ECMO Specialist Training Manual

Fourth Edition

Senior Editor:
Thomas V. Brogan, MD

Editors:
Gail Annich, MD, MS, FRCP(C)
W. Cory Ellis, CCP
Barb Haney, RN, MSN, RNC-NIC, CPNP-AC, FELSO
Micheal Heard, RN, FELSO
Roberto Lorusso, MD, PhD
Senior Editor: Thomas V. Brogan, MD
Editors:
Gail Annich, MD, MS, FRCP(C)
W. Cory Ellis, CCP
Barb Haney, RN, MSN, RNC-NIC, CPNP-AC, FELSO
Micheal Heard, RN, FELSO
Roberto Lorusso, MD, PhD
Manuscript Editor: Cindy Cooke, BA
Layout and Production: Peter Rycus, MPH, FELSO
Cover Design: Sara Frontalini, MD
Gail Annich, MD, MS, FRCP(C)

© 2018 Extracorporeal Life Support Organization, Ann Arbor, Michigan

Printed in the United States of America
ISBN 978-0-9656756-6-6

This material is made available through the Extracorporeal Life Support Organization (ELSO).
No endorsement of any product, service, or equipment should be inferred or is intended.
Dedication
ECMO Pioneer Ted Kolobow

Ted Kolobow MD was a pioneer in the development of ECMO.

He passed away in 2018. Dr Kolobow spent his career at the National
Institutes of Health where he studied artificial organs in acute illness.
In addition to major contributions in dialysis and nephrology, he was
instrumental in developing the devices that made ECMO possible.
His “spiral coil” membrane lung was used by all ECMO developers
from 1980 to 2008. His thin walled wire wound cannulas were the
first high flow/low resistance cannulas. His research was the first to
separate CO2 removal from oxygenation, making ECCO2R possible.
Ted trained over 40 fellows including Warren Zapol, Luciano Gat-
tinoni, and Antonio Pesenti. He is remembered with great respect by
the artificial organs and ECMO community.

Robert Bartlett, MD

ECMO Champion Richard Thomas (“Tad”) Fiser

Tad Fiser MD was a great champion of ECMO. He died in July 2017,

and is survived by his wife and three children. As a member of the
faculty at University of Arkansas College of Medicine & Arkansas
Children’s Hospital (ACH) he directed the ACH ECMO program
for most of his last decade and led development of the Adult ECMO
program at University Hospital. Under his leadership, ACH’s ECMO
program flourished, developing numerous high fidelity simulation
courses. Many children’s hospitals from the around the US came to
Arkansas to train and model their own programs based on the Arkan-
sas’ team work. Dr. Fiser became a recognized national authority in
pediatric ECMO. He authored national clinical practice guidelines
using this technology. Dr. Fiser mentored over 30 successful trainees,
and many pursued scholarly projects involving ECMO under his
guidance. He is remembered for his legacy of excellence and integrity.

Steve Schexnayder, MD

iii
List of Contributors

Cara Agerstrand, MD Jan Bělohlávek, MD, PhD

Division of Pulmonary, Allergy & Critical Care, General Teaching Hospital
Columbia University College of Physicians and Charles University
Surgeons Prague, Czech Republic
New York-Presbyterian Hospital, New York
Thomas V. Brogan, MD
Rocio Agliati, RN University of Washington School of Medicine
ECMO Unit Coordinator Seattle Children’s Hospital
Clínica Las Condes
Santiago, Chile L. Mikael Broman, MD, PhD
ECMO Centre Karolinska,
Francesco Alessandri, MD Karolinska University Hospital and Karolinska
Department of Anesthesia and Intensive Care Institutet
Medicine, Stockholm, Sweden
Sapienza University of Rome
Policlinico Umberto I, Rome, Italy Patrick D. Brophy, MD, MHCDS
University of Rochester School of Medicine and
Cory M. Alwardt, PhD, CCP Dentistry
Mayo Clinic College of Medicine The Golisano Children’s Hospital
Chief Perfusionist and ECMO Coordinator, Rochester, New York
Mayo Clinic, Phoenix, Arizona
Katie Butler, MSN, RN
Gail Annich, MD, FRCPC Children’s Hospital, Colorado,
The Hospital for Sick Children Aurora, Colorado
University of Toronto, Canada
Monika F. Cardona, RN, BSN
Jana Assy, MD ECMO and CRRT Nurse Manager
Director, Adult and Pediatric ECMO Program The Medical University of South Carolina
American University of Beirut Medical Center Charleston, South Carolina
Beirut, Lebanon
Titus Chan, MD, MS, MPP
Robert H. Bartlett, MD Pediatric Cardiac Critical Care Medicine
ECMO Experimental Laboratory Seattle Children’s Hospital
University of Michigan Medical Center Seattle, Washington
Ann Arbor, Michigan

v
Contributors

Ira M Cheifetz, MD, FCCM, FAARC Matteo Di Nardo, MD

Duke University School of Medicine Children’s Hospital Bambino Gesù
Duke Children’s Hospital IRCCS, Rome, Italy
Durham, North Carolina
Linda Edwards MB ChB, MRCPCH (UK)
Sertac Cicek, MD ECLS Program Director
Department of Cardiovascular Surgery British Columbia Children’s Hospital
Mayo Clinic Vancouver, Canada
Rochester, Minnesota
W. Cory Ellis, CCP
Alain Combes, MD, PhD Children’s Hospital Colorado
Groupe Hôpital de la Pitié Salpetrière, Aurora, Colorado
Assistance Publique-Hôpitaux de Paris
Université Pierre et Marie Curie, Paris Patricia English, MS, RRT-NPS
Massachusetts General Hospital
James T Connelly, BS, RRT-NPS Boston, Massachusetts
The Children's Hospital of Philadelphia
Philadelphia, Pennsylvania James D. Fortenberry, MD
Pediatrician-in-Chief
Steven A. Conrad, MD, PhD Emory University
Louisiana State University Health Sciences Center Atlanta, Georgia
Shreveport, Louisiana
John C. Greenwood, MD
Carl Davis, MD, FRCS Perelman School of Medicine at the University of
Royal Hospital for Children Pennsylvania
Glasgow, Scotland Philadelphia, Pennsylvania, USA

Joseph A. Dearani, MD Anne-Marie Guerguerian, MD, PhD

Department of Cardiovascular Surgery Hospital for Sick Children
Mayo Clinic Toronto, Ontario, Canada
Rochester, Minnesota
Kyle Gunnerson, MD
Archana Dhar, MD Massey Family Foundation Emergency Critical
UT Southwestern Care Center
Children's Health-Children's Medical Center University of Michigan Medical Center
Dallas Texas Ann Arbor, Michigan

Rodrigo Diaz, MD Barbara Haney, RN, MSN, RNC-NIC, CPNP-AC,

Department of Anesthesia FELSO
Clínica Las Condes Children’s Mercy Kansas City
Santiago, Chile Kansas City, Missouri

Robert DiGeronimo, MD Silvia M. Hartmann, MD

Seattle Children’s Hospital Seattle Children’s Hospital
University of Washington School of Medicine University of Washington School of Medicine

vi
 Contributors

Chris Harvey, MB ChB, MRCS Roberto Lorusso, MD, PhD

Glenfield Hospital ECLS Centrum
Leicester, UK. Maastricht University Medical Centre
Maastricht, Netherlands
Micheal L. Heard, RN, FELSO
ECMO & Advanced Technologies Coordinator William R. Lynch, MD
Children’s Healthcare of Atlanta Section of Thoracic Surgery
Atlanta, Georgia University of Michigan School of Medicine
Ann Arbor, MI
Daniel L. Herr, MD
Shock Trauma Center Graeme Maclaren, MBBS, FRACP, FRCP,
University of Maryland, School of Medicine FCICM, FCCM
Baltimore, Maryland National University Health System, Singapore
Royal Children’s Hospital
Aparna U. Hoskote, MBBS, MD, MRCP University of Melbourne, Australia
Consultant Pediatric Cardiac Intensivist Care
Great Ormond Street Hospital for Children George Makdisi, MD, MPH, MS
Lindon, England Tampa General Hospital
University of South Florida
Hanneke IJsselstijn, MD, PhD Tampa, Florida
Erasmus MC
Sophia Children’s Hospital Patti Massicotte, MD
Rotterdam, Netherlands Stollery Children’s Hospital
Edmonton, Alberta, Canada
Sarah Keene, MD
Emory University School of Medicine Ali McMichael, MD
Children’s Healthcare of Atlanta UT Southwestern School of Medicine
Atlanta, Georgia Dallas, Texas

Peter Chi Keung Lai, RN Tracy Morrison, MSQA, BSN, RN

Queen Mary Hospital Premier Health
Hong Kong Miami Valley Hospital
Dayton, Ohio
Christa Jefferis Kirk, PharmD
Heart Center Clinical Pharmacy Specialist Chirine Mossadegh, RN
Seattle Children’s Hospital Hôpital Pitié-Salpêtrière,
Seattle, Washington Assistance Publique-Hôpitaux de Paris, France

Paul Kratz, BSc, CPC, CCP Jessica Nicoll, MD, FRCPC

Cardiovascular Perfusion/ECLS Program The Research Institute
The Hospital for Sick Children Hospital for Sick Children
Toronto, Ontario, Canada Toronto, Ontario, Canada

Jan Hau Lee, MBBS, MRCPCH, MCI Mark T Ogino MD

KK Women’s and Children’s Hospital Nemours Alfred I. duPont Hospital for Children
Duke-NUS Medical School Wilmington, Delaware
Singapore

vii
Contributors

Matt Paden, MD Raisa M. Schiller, PhD

Emory University Erasmus MC
Children’s Healthcare of Atlanta, Sophia Children’s Hospital
Atlanta, Georgia Rotterdam, Netherlands

Bhavesh M. Patel, MD, FRCP(C), RDMS Jayne Sheldrake, RN

Mayo Clinic College of Medicine Intensive Care Unit
Mayo Clinic The Alfred Hospital
Phoenix, Arizona Melbourne, Australia

Vincent Pellegrino, MBBS, FRACP, CICM Farah Siddiqui, MBChB

Alfred Intensive Care Unit, Leicester Royal Infirmary
Alfred Hospital Leicester, United Kingdom
Melbourne, Australia
Rachel Sirignano, MD
Edoardo Piervincenzi, MD Emory University
Sapienza University of Rome Children’s Healthcare of Atlanta,
Policlinico Umberto I Atlanta, Georgia
Rome, Italy
Justyna Swol, MD
Francesco Pugliese, MD Paracelsus Medical University
Sapienza University of Rome Nuremberg, Germany
Policlinico Umberto I
Rome, Italy Fabio S. Taccone, MD, PhD
Hôpital Erasme
Monique Radman, MD, MAS Université Libre de Bruxelles
Seattle Children’s Hospital, Brussels, Belgium
University of Washington School of Medicine
Seattle, Washington Ravi R. Thiagarajan, MBBS, MPH
Children’s Hospital Boston
Kollengode Ramanathan, MD Harvard Medical School
Division of Cardiothoracic Intensive Care Boston, Massachusetts
National University Heart Centre
Singapore Mark Todd, HBSc, RRT
The Hospital for Sick Children
Marco V. Ranieri MD Toronto, Canada.
Sapienza University of Rome
Policlinico Umberto I Leen Vercaemst, BCP, ECCP, RN, RM
Rome, Italy ECMO Coordinator, Perfusion School Leuven
Gasthuisberg Campus
Kenneth A. Schenkman, MD, PhD UZ Leuven, Belgium
Seattle Children’s Hospital
University of Washington School of Medicine Deborah Wagner, PharmD
University of Michigan
Ann Arbor, MI

viii
 Contributors

Scott Wagoner MBA, RRT

Children's Healthcare of Atlanta at Egleston
Atlanta, Georgia

I-wen Wang MD, PhD

Indiana University School of Medicine
Indiana University Health, Methodist Hospital
Indianapolis, Indiana

E.D. Wildschut, MD, PhD

Erasmus MC
Sophia Children’s Hospital
Rotterdam, Netherlands

Susan B. Williams, MSN, RNC, CIT

Children’s Hospital of Philadelphia
Philadelphia, Pennsylvania

Anne Marie Winkler, MD, MSc

Director of Medical Affairs
Instrumentation Laboratory
Bedford, Massachusetts

Trisha E. Wong MD, MS

Oregon Health & Science University School of
Medicine
Portland, Oregon

Gillian Wylie, RN, RSCN, BSc

Royal Hospital of Children
Glasgow, Scotland

Larissa Yalon, BSN, RN, CCRN

Seattle Children’s Hospital
Seattle, Washington

Bishoy Zakhary MD
Oregon Health & Science University School of
Medicine
Portland, Oregon

ix
Preface to the 4th Edition

To follow knowledge like a sinking star,

Beyond the utmost bound of human thought
“Ulysses”
Alfred, Lord Tennyson

The scope of the application of extracorporeal life support (ECLS) has undergone a great expansion
if not a revolution in the eight years since the 3rd Edition of the ECMO Specialist Training Manual was
published. These changes have not only encompassed advances in technology and equipment as well as
growth in the populations served but also in the sheer imagination with which mechanical cardiopulmo-
nary support has been applied. We have aimed to expand this edition of the Manual to reflect the growth
and creativity that has occurred in the uses of ECMO. To do so, the editors have recruited an international
cadre of contributors who have brought their global perspective to production of this edition. We owe
them a large debt of gratitude.
The reader will note that there are a number of new chapters in this edition to address the specific
populations and technologies used in supporting patients with critical illness. The goal of this manuscript
is to provide the education that undergirds the training of the bedside ECMO Specialist. To that extent, in
addition to the core subject chapters, we have tried to provide chapters on the approach to emergencies,
troubleshooting, and simulation as well as questions and case scenarios. We trust that these chapters will
aid program directors to improve the education and training efforts.
As with all such training/educational manuals, this one will fall short and may not address specific
practices in some institutions. One of the loveliest aspects of ECMO is the broad community that is AL-
WAYS willing to aid colleagues near or far. Please reach out. The authors of this manual may be a good
place to start.
I would like to thank the amazing array of authors and the hardworking editors. It was my distinct
honor and pleasure to do so. The authors and the editors worked to meet rather stringent deadlines. I thank
you. I must also thank Peter Rycus and Cindy Cooke who made this book go. And finally, thank you to all
the clinicians who care for patients who receive this amazing support technology.

Thomas V. Brogan, MD
Seattle, WA

xi
 Table of Contents

Dedication.......................................................................................................................... iii
Preface to the 4th Edition................................................................................................. xi
1. History of ECMO Specialists...................................................................................... 19
Summary................................................................................................................22
References..............................................................................................................22
2. Respiratory Physiology and Gas Exchange............................................................... 23
Native Circulation Gas Exchange..........................................................................23
Oxygenation...........................................................................................................23
Carbon Dioxide......................................................................................................26
Physiology of ECLS..............................................................................................26
Summary .............................................................................................................. 28
References............................................................................................................. 29
3. Cardiac Physiology Relevant to Extracorporeal Life Support.................................31
Introduction............................................................................................................31
Oxygen Delivery....................................................................................................31
Hemodynamics: Pressure, Flow, Resistance and Compliance..............................35
Relevance to ECMO..............................................................................................36
Summary................................................................................................................36
References............................................................................................................. 38
4. Management of Anticoagulation and Blood Products on ECMO........................... 39
Objectives............................................................................................................. 39
Introduction .......................................................................................................... 39
ECLS Anticoagulation.......................................................................................... 39
Anticoagulation Monitoring (Table 4-4)............................................................... 40
Conclusion.............................................................................................................41
References..............................................................................................................42
5. Transfusion Management during Extracorporeal Life Support............................. 43
Transfusion Support during Extracorporeal Life Support (ECLS)........................43
Blood Group Review............................................................................................ 44
Red Blood Cell Transfusion.................................................................................. 44
Platelet Transfusion................................................................................................46
Plasma Transfusion................................................................................................46
Cryoprecipitate Transfusion...................................................................................47
Variability of Transfusion Practice in ECLS..........................................................47
Complications of Transfusion................................................................................47

xiii
Table of Contents

Adjunctive Therapy to Blood Product Transfusion.............................................. 49

References..............................................................................................................51
6. Medication Management in ECLS............................................................................. 55
Introduction............................................................................................................55
Pharmacokinetics...................................................................................................55
Pharmacokinetic Changes in Critical Illness and ECLS........................................56
Dosing Strategy......................................................................................................57
References............................................................................................................. 64
7. Respiratory Diseases Associated with ECMO............................................................67
Neonatal Respiratory ECMO ................................................................................67
Pediatric Respiratory ECMO................................................................................ 68
Adult Respiratory ECMO..................................................................................... 70
References .............................................................................................................73
8. Neonatal, Pediatric, and Adult Cardiac Extracorporeal Life Support................... 75
ECMO in Congenital Heart Diseases....................................................................75
Preoperative Support..............................................................................................75
Failure to Wean Off Cardiopulmonary Bypass......................................................76
Postoperative Low Cardiac Output State...............................................................77
Extracorporeal Cardiopulmonary Resuscitation (ECPR)......................................77
Glenn and Fontan Circulation on ECMO..............................................................77
Myocarditis and Cardiomyopathy........................................................................ 78
Cardiogenic Shock in Acute Myocardial Infarction and Chronic Heart Failure.. 79
Pulmonary Hypertension...................................................................................... 80
Arrhythmias............................................................................................................81
References..............................................................................................................82
9. ECMO for Cardiopulmonary Arrest and Resuscitation.......................................... 85
Introduction............................................................................................................85
Patient Selection....................................................................................................86
The Team................................................................................................................86
Equipment..............................................................................................................86
Vascular Access and Cannulation for ECPR......................................................... 89
Space and Location............................................................................................... 89
ECPR Algorithm, Protocols, Checklists, Cognitive Aids..................................... 90
Postcardiac Arrest Care (PCAC) Following ECPR ..............................................91
Ethics and Informed Consent.................................................................................92
ECPR Outcomes....................................................................................................93
References............................................................................................................. 94
10. Special Conditions.......................................................................................................97
ECLS in Trauma ...................................................................................................97
ECLS in Pregnancy............................................................................................... 98
ECLS in Sepsis................................................................................................... 100
ECLS in Oncology Patients.................................................................................101
Controversies.......................................................................................................102
References............................................................................................................103
11. Cannulation and Initiation of ECLS.......................................................................107
Objectives............................................................................................................107
The Team and Equipment....................................................................................107
Patient Preparation.............................................................................................. 108

xiv
 Table of Contents

Approach to Cannulation and Preparing ECMO................................................ 108

Vessel Cannulation.............................................................................................. 109
ECMO Initiation................................................................................................. 109
Preparation for Transportation.............................................................................110
References............................................................................................................112
12. Venoarterial ECMO in Adult and Pediatric Patients............................................113
Venoarterial ECMO in Adults..............................................................................113
V-A ECMO in Neonates and Children.................................................................116
13. Venovenous Extracorporeal Membrane Oxygenation.......................................... 123
Introduction..........................................................................................................123
Respiratory ECMO..............................................................................................123
Circuit Operation.................................................................................................123
Cannulation......................................................................................................... 124
Cannulae..............................................................................................................125
Weaning................................................................................................................126
Summary..............................................................................................................127
References .......................................................................................................... 128
14. Special Configurations in Adult ECLS....................................................................131
Definitions and Terminology...............................................................................131
Common Clinical Scenarios ...............................................................................131
References............................................................................................................137
15. Extracorporeal Carbon Dioxide Removal (ECCO2R).......................................... 139
Introduction......................................................................................................... 139
ECCO2R Technology.......................................................................................... 139
Potential Clinical Indications.............................................................................. 140
Complications......................................................................................................141
Future Perspective and Advances........................................................................142
References............................................................................................................143
16. ECMO Equipment and Circuit Design...................................................................147
Introduction..........................................................................................................147
ECMO Pump Console.........................................................................................147
ECMO Circuit Design........................................................................................ 148
17. ECMO Circuit Components.....................................................................................151
Objectives............................................................................................................151
Introduction .........................................................................................................151
Considerations for Cannula Selection.................................................................152
Blood Pumps-Rationale for Choosing a Blood Pump for ECLS........................ 154
Bladders...............................................................................................................157
Oxygenators.........................................................................................................157
Heat Exchangers................................................................................................. 159
References............................................................................................................161
18. ECLS Safety and Other Monitoring Devices......................................................... 163
Introduction..........................................................................................................163
Circuit Related Monitoring..................................................................................163
Patient Related Monitoring..................................................................................165
19. Management of the Neonate on ECMO..................................................................167
Introduction..........................................................................................................167
Initiation and Circuit Considerations...................................................................167

xv
Table of Contents

Daily Patient Management.................................................................................. 168

Conclusion...........................................................................................................173
References........................................................................................................... 174
20. Pediatric Respiratory ECLS: Patient Management..............................................177
Introduction..........................................................................................................177
Patient Management ...........................................................................................177
Associated Measures........................................................................................... 180
Nursing Care........................................................................................................181
References........................................................................................................... 184
21. Patient Management: Neonatal and Pediatric Cardiac ECLS.............................187
Introduction..........................................................................................................187
Indictions, Contraindications, and Cannulation ..................................................187
Daily Management of the Cardiac ECLS Patient............................................... 188
ECLS Emergencies..............................................................................................191
Weaning from ECLS............................................................................................193
References........................................................................................................... 194
22. Patient Management for the Adult Patient with Respiratory Failure on ECLS.195
General Principles of Care...................................................................................195
Systems Management and Goals.........................................................................195
References........................................................................................................... 200
Addendum............................................................................................................201
23. Adult Cardiac ECLS: Patient Management.......................................................... 203
Cannulation Strategies.........................................................................................203
Pain and Sedation.................................................................................................203
Preventing Complications................................................................................... 204
Hemodynamic Monitoring ................................................................................. 204
Limb and Hand Ischemia.................................................................................... 204
Differential Hypoxia............................................................................................205
Fluid Management...............................................................................................205
Ventilator Management........................................................................................206
References............................................................................................................207
24. Procedures on ECLS................................................................................................ 209
Objectives........................................................................................................... 209
Introduction......................................................................................................... 209
Preparation and Support ..................................................................................... 209
The Environment................................................................................................. 210
Common Noncardiac Surgical Procedures.......................................................... 211
Abdominal Surgery.............................................................................................. 211
References............................................................................................................213
25. Weaning and Trial Off............................................................................................. 215
Introduction..........................................................................................................215
When to Wean......................................................................................................215
How to Wean........................................................................................................216
Trial Off ECMO...................................................................................................217
References........................................................................................................... 219
26. Outcomes and Followup of Patients Receiving ECLS...........................................221
Introduction..........................................................................................................221
Long-term Outcomes...........................................................................................221

xvi
 Table of Contents

Long-term Followup............................................................................................222
References........................................................................................................... 224
27. Ambulation and ECLS............................................................................................ 225
Introduction..........................................................................................................225
Rationale for Ambulatory ECMO........................................................................225
A Programatic Approach .....................................................................................225
Long-term Outcomes.......................................................................................... 228
References........................................................................................................... 229
28. Acute Kidney Injury and Renal Support Therapy during ECLS........................231
Acute Kidney Injury in ECLS Patients................................................................231
CRRT during ECLS.............................................................................................232
References............................................................................................................236
29. Extracorporeal Support with Therapeutic Apheresis.......................................... 239
Introduction......................................................................................................... 239
Apheresis during ECLS....................................................................................... 239
Technical Aspects of Apheresis during ECLS.................................................... 240
Use and Outcomes of Apheresis during ECLS....................................................243
Summary of Apheresis during ECLS................................................................. 244
References............................................................................................................245
30. Mechanical Liver Support.......................................................................................247
Abstract................................................................................................................247
Introduction..........................................................................................................247
Approach to Treatment........................................................................................ 248
Selected Mechanical Extracorporeal Hepatic Support....................................... 248
Conclusions..........................................................................................................251
Appendix 1: (Adapted from the MARS Product Users Guideline).....................252
References............................................................................................................253
31. Staffing for the ECMO Patient............................................................................... 255
Introduction..........................................................................................................255
The Staffing Models and the Decision Process....................................................255
Choosing the Right Staffing Model for Your Program........................................256
Location of Patients and Staffing Impact.............................................................256
Linking Staffing, Education, and Quality............................................................257
Current State of ECMO Programs.......................................................................257
Conclusion.......................................................................................................... 258
References........................................................................................................... 260
32. ECMO Emergencies.................................................................................................261
Introduction..........................................................................................................261
The ECMO Caregiver..........................................................................................261
Off-Bypass Procedure..........................................................................................261
The ECMO Circuit Check....................................................................................262
Basic Emergency Management...........................................................................263
Thrombosis......................................................................................................... 264
Air Embolism and Circuit Pressures....................................................................265
Membrane Oxygenator Failure............................................................................266
Tubing Rupture....................................................................................................267
Centrifugal Pump Head Failure...........................................................................267
Cannula Problems............................................................................................... 270

xvii
Table of Contents

Accidental Decannulation................................................................................... 270

Equipment Failure................................................................................................271
Miscellaneous Circuit Components ....................................................................271
Summary..............................................................................................................271
References............................................................................................................273
33. Patient Troubleshooting.......................................................................................... 275
References............................................................................................................287
34. Case Simulations...................................................................................................... 289
Introduction......................................................................................................... 289
Simulation Scenario Format............................................................................... 289
References............................................................................................................291
35. Configuration Nomenclature for Extracorporeal Membrane Oxygenation...... 305
The ELSO Classification for Peripheral Cannulation Configurations.................305
Classification for Central Cannulation Configuration........................................ 308
Summary............................................................................................................. 308
36. Case Scenarios........................................................................................................... 311
37. Chapter Questions....................................................................................................337
Index................................................................................................................................ 363

xviii
1

History of ECMO Specialists

Robert H. Bartlett, MD, Jayne Sheldrake, RN, Roberto Lorusso, MD, PhD

The first clinical cases of extracorporeal mem- that study, research in ECMO stopped altogether
brane oxygenation (ECMO) were done in the early except for two centers, one in Milan Italy and one
1970s after a decade of development in the animal at University of California Irvine. At Irvine, we
laboratory. The teams who managed these cases had used the technology for newborn infants with
were the surgeons, fellows, perfusionists, and nurses respiratory failure with remarkably good results.
who developed the technique in the laboratory. The The lab team consisted of the surgeons, students,
lab ECMO circuit was a compilation of used operat- technicians, perfusionists, residents, and nurses
ing room (OR) pumps and heat exchangers, early who were studying the technique in the laboratory.
membrane oxygenators (often assembled on site), Looking back on it, that remarkable group
cannulas made from chest tubes, and some devices of nurses and perfusionists were the first ECMO
for control fabricated in the laboratory. The team for Specialists, although the technique was purely a
these early cases was the lab team and the circuit research enterprise at the time and the need for a
was cleaned up and brought to the intensive care full-time team was certainly not appreciated.
unit (ICU). The improvement in physiology was The UC Irvine team first reported and devel-
dramatic and the concept and practice of ECMO oped ECMO in newborn infants between 1975 and
was born. 1980. This was the first ECMO specialist team. It
In 1975, approximately 20 successful cases was headed by ICU nurses Nancy Wetmore and
of ECMO for respiratory failure were reported in Tammy Medley who were working on ECMO in
the literature. Based on this experience, the NIH the laboratory. The team was composed of ICU
empaneled nine centers to conduct a prospective nurses, perfusionists Nick Haiduc and John Tooma-
randomized controlled trial of ECMO for severe sian, surgical residents, and medical students. All
respiratory failure in adults. Only three of the nine were volunteers. The training was on the job in the
centers had clinical experience using a homemade laboratory where we were putting sheep on ECMO.
ECMO circuit. The other six centers developed We used the lab devices in the hospital. The concept
circuits which were all different. They tested cir- of the ECMO Specialist was described by Nancy
cuits in the laboratory and brought their lab team Wetmore in 1979.1
to the bedside. The first patients in these six centers In 1980, the UC Irvine project moved to the Uni-
were also included in the clinical trial, adding the versity of Michigan and there we established a full
learning curve to the ECMO arm. Only 10% of the ECMO service for respiratory failure in newborn
patients of both arms (ECMO and control) survived infants. This team was led by Robert Bartlett and
and it was declared that ECMO was of no value in perfusionist John Toomasian. As the need for ECMO
adult respiratory failure. In retrospect this was a in newborns increased, we established a designated
premature, poorly designed, and poorly conducted fulltime team to manage the ECMO patients. The
clinical trial. In fact, we use this trial as the proto- team was paid by the hospital as ECMO Specialists.
type to explain how not to conduct a clinical trial The team consisted of a coordinator, two nurses, two
of a new life support technique in a syndrome in respiratory therapists, and two perfusionists. The
which the endpoint is death. However, based on composition of this original team was an intended

19
Chapter 1

experiment to determine how an ECMO Specialist In 1991, Julie Vilardi described the nursing
could be created. An ECMO Specialist capable of based ECMO Specialist at Oakland Children’s
managing a patient on ECMO required the skills of Hospital.2 Specialists were selected from a group of
an experienced ICU nurse, a respiratory therapist, ICU nurses who applied for the position. The initial
and a perfusionist. The team had to learn the physiol- criteria were two years of ICU nursing experience,
ogy of extracorporeal circulation, the function and leadership ability, and outstanding performance.
malfunction of devices, and the management of Twenty-four nurses initiated the specialist team. The
patients on an extracorporeal system. Each specialist training included 24 hours of didactic instruction, 24
from the three disciplines had to learn the skills of hours of practice in the animal lab, and 8 hours of
the other disciplines. Under the direction of coordi- proctored bedside ECMO management. The respon-
nator John Toomasian, this team drilled in the animal sibilities included all aspects of ECMO from patient
laboratory until they were ready for actual patient selection and circuit preparation to daily manage-
care. The team could manage all aspects of ECMO ment and followup. The team held regular case
care and the experiment was successful. review meetings, skill drills, and reporting data to
The experience with neonatal ECMO at Michi- the ELSO Registry. Julie noted the overall improve-
gan attracted the attention of neonatologists from ment in understanding and physiologic management
other major academic centers. We established in the neonatal ICU which resulted from the ECMO
courses to teach other centers our practices, which team. This intangible benefit of an ECMO program
included the concept of ECMO Specialists and occurs in every ICU where ECMO is used.
ECMO teams, the training regimen, the actual ex- The background and role of ECMO Specialists
perience, and certification for bedside care. Soon, is subject to hospital policy, politics, and regulatory
ECMO Specialists and specialist teams were es- agencies. In the U.S. regulation of medical practice
tablished at Children’s National Medical Center is managed at the state level. Regulatory agencies
in Washington, Columbia University in New York, exist in perfusion, nursing, and respiratory therapy.
Children’s Hospital in Detroit, Boston Children’s These agencies decree what their certificants are
Hospital, University of Pittsburgh, and several other qualified to do (which may include ECMO). How-
major university centers. Many of these centers ever, these agencies cannot claim exclusivity of the
trained other centers, increasing the rate of train- technology and cannot decree what other profes-
ing of dedicated ECMO Specialists throughout the sions can do. Since specialists come from one of
country. Some centers based their specialist team these three backgrounds, there has been no specific
totally on perfusionists (Columbia), respiratory need to establish a board or certifying agency for
therapists (Detroit), or ICU nurses (Washington). “ECMO Specialists.” Rather, the specialists are certi-
The Michigan team continued to recruit specialists fied by their home discipline, with subspecialization
from all three disciplines, continuing the experiment. left to each hospital credentialing committee. Most
By 1989, there were more than 30 ECMO cen- ECMO centers develop ECMO Specialists from all
ters in North America, Europe, Japan, and Australia. three disciplines. In some countries, the ECMO Spe-
These centers joined together to form the Extracor- cialists are all physicians (China). In some countries,
poreal Life Support Organization, ELSO. ELSO the ECMO Specialists’ training and responsibility is
established a registry of ECMO cases, published included in ICU nursing (France). In some countries,
guidelines on patient care, published an inclusive the specialist is the perfusion team, particularly in
textbook (The Red Book), and published the first centers where cardiac ECMO is the primary activity.
ECMO Specialist Training Manual in 1993. The With the availability of ECMO specific devices
second edition was published in 1999 and the third around 2008, the technology has become much
edition in 2010. The current edition is the fourth simpler, safer, and easier. With some extra training,
edition (2018). This manual describes the role, re- any ICU nurse can manage ECMO and the ECMO
sponsibilities, and training of ECMO Specialists in circuit most of the time. The “primary care giver
detail. It is now used throughout the world to train model” of ECMO management evolved because
ECMO Specialists and teams. the devices were safer, simpler, and more reliable

20
 History of ECMO Specialists

than the homemade systems. Devices ran for days postoperative VA-ECMO cases. This responsibility
or weeks without complications and fatal circuit included equipment management, circuit configura-
complications were very rare, so a specialist at the tion and priming, patient selection, retrieval, can-
bedside 24/7 was no longer necessary.3 As this ICU nulation and commencement of ECMO, as well as
nurse based care is established, the specialist team the daily management and weaning of patients from
takes on a different but very important role. In nurse- ECMO support. Perfusionists remained involved,
centered ECMO practices the specialist team is but in a consultation and support role.
responsible for education, coordination throughout During the same transition period the ICU nurs-
the hospital, triage of patients, establishing standard ing staff, as the primary care givers at the bedside,
practice and protocols, case review, equipment, and also incorporated ECMO management into their
establishment of standard policies, and frequent scope of practice. Specific patient care roles were
bedside rounds including management of emergen- allocated to nursing staff. Initially, a small group of
cies, cannulation, decannulation, and transport as nurses were trained in immediate circuit manage-
well as supervising day to day management. The ment, but as ECMO patient days have increased, this
specialist team can be made up of nurses, respiratory responsibility has now grown to involve more than
therapists, or perfusionists. With ICU as the primary 50% of the permanent ICU nursing staff in the unit.
caregiver, the specialist team can be developed Australian ICUs do not use respiratory thera-
within the hospital perfusion service, as is done at pists, renal therapists, or ECMO Specialists (as
Columbia.4 Thus, it is often the case when the use described by the U.S. model). The bedside nurse
of ECMO is primarily for support of cardiac surgery. is responsible for all of these therapies, including
A recent survey of ECMO centers worldwide ECMO, in an integrated approach with the inten-
showed that the ICU nurse-based specialist was used sivist. This model of care allows for cost efficient,
as a 1:1 nurse patient ratio in 59% of centers, with timely ECMO support with little to no delay in
26% reporting a nurse:ECMO patient ratio from management decision making or action. The model
1:2 to 1:5. The ICU nurse specialist with support of a bedside nurse as the primary care giver provides
from perfusionists was a frequently reported model.5 streamlined, integrated patient care of all therapies
The specialist team at The Alfred hospital, required by the patient, including renal replacement,
Melbourne Australia is an excellent example of ventilation, nitric oxide, inotropic drugs in addition
a modern ECMO specialist team. In 1990, when to ECMO.
ECMO commenced at The Alfred Hospital, the Intensivists and ECMO nurses complete the
ECMO model of care was a traditional model using same training course prior to caring for this patient
bedside perfusionists 24 hours a day and intensiv- group. Medical and nursing staff also complete an-
ists managing the patient group. The nursing team nual competencies with both theoretical and prac-
cared for these patients on a 1:1 ratio, but at this time tical components. They undertake simulation and
ECMO management was solely the responsibility of wet lab training in a multidisciplinary forum. This
the ICU intensivists, cardiothoracic surgeons, and encourages team work and a shared understanding
perfusionists. In 2003, as ECMO devices became of mutual respect.
more reliable and exposure to this patient group Incorporating all aspects of patient management
grew, the model evolved from one that relied on into the role of the bedside nurse in collaboration
24/7 perfusionist presence to a collaborative model with medical and allied health team members has
between intensivists, allied health and the bedside led to the development of this specialist ECMO
nursing team. Perfusionists continued to attend team. This team has the ability to provide cost ef-
ward rounds and were available for circuit com- fective, streamlined ECMO care with flexibility and
mencement and intervention. In 2009, the ECMO adaptability to accommodate variations in ECMO
model of care evolved again. Intensivists became configurations and growing patient numbers. This
responsible for all aspects of ECMO management collaborative approach has been highly successful
within the ICU (VV and VA-ECMO) and worked in with enthusiastic staff uptake, enhanced team par-
conjunction with the cardiothoracic team managing

21
Chapter 1

ticipation and most importantly excellent patient References

outcomes.
1. Wetmore NE, Bartlett RH, Gazzaniga AB, Hai-
Summary duc NJ. Extracorporeal membrane oxygenator
(ECMO): a team approach in critical care and
The ECMO Specialist is a new paramedical life support research. Heart Lung 1979; 8:288-
profession which grew out of the need for unique 295.
specialists to manage the new technique of pro- 2. Vilardi J. Defining a new expanded role for staff
longed extracorporeal life support. The specialist nurses: role development and clinical practice
comes from a background in nursing, perfusion, or issues for the ECMO specialist. J Perinat Neo-
respiratory therapy. In recent years nursing based natal Nurs 1991;5(3):51-60.
specialist programs are the most widely used for 3. Freeman R, Naulat C, Mowry J, Baldridge P.
day to day bedside management. The specialist Expanded resources through utilization of a
team is responsible for all aspects of managing the primary care giver extracorporeal membrane
ECMO program. oxygenation model. Crit Care Nurs Q 2012;
35(1):39-49.
4. Mongero LB, Beck JR, Charette KA. Manag-
ing the extracorporeal membrane oxygenation
(ECMO) circuit integrity and safety utilizing the
perfusionist as the “ECMO Specialist.” Perfu-
sion 2013; 28(6):552-554.
5. Daly KJ, Camporota L, Barrett NA. An inter-
national survey: the role of specialist nurses
in adult respiratory extracorporeal membrane
oxygenation. Nurs Crit Care 2017; 22(5):305-
311.

22
2

Respiratory Physiology and Gas Exchange

Cara Agerstrand, MD, Thomas V. Brogan, MD

Management of patients receiving extracor- air, 159 mmHg (760 mmHg of atmospheric pressure
poreal life support (ECLS) requires a thorough x 21% oxygen in atmospheric air = 159 mmHg O2).
understanding of respiratory and cardiovascular Oxygen travels through the conducting airways (tra-
physiology, tissue oxygenation, and metabolic chea, bronchi, and bronchioles), known collectively
rate.1 Application of these physiologic principles as the anatomic dead space, prior to reaching the al-
to the care of patients receiving ECLS, coupled veoli. The air becomes humidified in the conducting
with comprehension of circuit gas exchange and airways, but gas exchange does not occur. Due to the
oxygen delivery is central to the successful delivery increased presence of water vapor and expired CO2,
of ECLS to neonate, pediatric, and adult patients. the partial pressure of oxygen (PAO2) in the alveoli
is lower than that of ambient air, at approximately
Native Circulation Gas Exchange 100 mmHg. Oxygen diffuses into the pulmonary
capillary blood where the partial pressure of arte-
Understanding the physiology of gas exchange rial oxygen (PaO2) reaches 90-95 mmHg, slightly
during ECLS requires comprehension of the factors decreased from PAO2 due to venous admixture from
that determine oxygen delivery and carbon dioxide the returning systemic circulation, the difference
regulation in native circulation.2 The movement of of which is known as the “A-a gradient.” The A-a
any gas from the atmosphere to the alveoli, tissues, gradient may be elevated in disease states.
and back again is determined by the partial pres- Blood flows through the arterial circulation until
sure of gases in various compartments of the body, it reaches the tissue capillaries where, under normal
with the movement of gas occurring down a gradi- homeostatic conditions, aerobic metabolism occurs.
ent, from an area of higher concentration to lower Following aerobic metabolism by body tissues, the
concentration, in search of equilibrium. partial pressure of oxygen in the venous system
The partial pressure of gas is measured in milli- (PvO2) is approximately 40 mmHg, equivalent to a
meters of mercury (mmHg) or torr (Torr). The origin hemoglobin saturation of ~75%. Since gas exchange
of these units is based on the pressure exerted at the does not occur in venules or veins, this is equivalent
base of a 1 millimeter column of mercury, the ele- to the oxygen in the pre-alveoli capillary bed of the
ment historically used for measuring pressure. The lungs and can be measured from a pulmonary artery
atmosphere exerts a total pressure of 760 mmHg catheter as the mixed venous oxygen (SvO2) or from
at sea level; 1 Torr equals 1/760 atms. For practi- an internal jugular or subclavian artery central line
cal purposes, 1 mm Hg and 1 Torr are equivalent as the central venous oxygen (ScvO2). As the ScvO2
and are the units of pressure measurement used in only represents the amount of oxygen returning from
medicine today. upper body circulation, this is about 5-10% higher
than the SvO2.3 Venous blood may drain from the
Oxygenation inferior vena cava or superior vena cava into the
ECLS circuit, therefore pre-oxygenator saturation
Oxygen enters the body through the oropharynx represents a coarse estimation of the central venous
at a partial pressure equivalent to that of atmospheric saturation, provided there is minimal recirculation.

23
Chapter 2

Oxygen Content and Hemoglobin by the solubility coefficient of oxygen in plasma

of 0.003 ml/dL. PaO2 is used to assess lung injury
Oxygen content refers to the amount of oxygen severity and oxygen requirements, however, as most
in arterial (CaO2) or venous (CvO2) blood, an ag- oxygen is bound to hemoglobin and not dissolved
gregate of the oxygen bound to hemoglobin (98.5%) in plasma, its value is also limited in assessing total
and dissolved in plasma (1.5%). Oxygen content is oxygen body content.
the quantity of oxygen per unit of blood (typically Taken together, the total oxygen content of
measured in mL/dL), and though often overlooked blood is detailed in Equation 1:
in clinical applications, is crucial to understanding
tissue oxygen delivery in normal circulation and Equation 1: Oxygen content = (1.34 x hemoglo-
during ECLS. Since the vast majority of oxygen is bin x oxygen saturation) + (PO2 x 0.003)
bound to hemoglobin molecules of red blood cells,
hemoglobin concentration (measured in g/dL) is the
primary determinant of oxygen content (Figure 2-1). At a normal hemoglobin concentration of 15 g/
Each gram of hemoglobin, when fully saturated, dL with a saturation of 100% and a normal physi-
binds 1.34 ml of oxygen. The fraction of oxygen- ologic PaO2 of 90 mmHg, arterial oxygen content
saturated hemoglobin (oxyhemoglobin) to total (CaO2) = 20 g/dL. A normal oxygen content of ve-
hemoglobin is known as oxygen saturation SpO2 nous blood (CvO2) at the same hemoglobin with a
when measured peripherally. While a frequently PvO2 of 40 mmHg and a saturation of 75% = 15 g/dL.
monitored clinical parameter, SpO2 itself reveals The difference between CaO2 and CvO2 is know
nothing about the body’s actual oxygen content un- as the A-V oxygen difference and is an important
less the hemoglobin level is known as well. component to calculating tissue oxygen consump-
In addition to oxyhemoglobin, a small amount tion. Similarly, during ECLS, pre and post-oxy-
of oxygen is delivered to the tissues dissolved in genator gases can be used to calculate the pre/post
plasma. This amount is directly measured by blood oxygenator content of the blood and thereby ECLS
gas analysis and known as the PaO2. The total dis- oxygenator efficiency.
solved arterial oxygen content is the PaO2 multiplied
Oxygen Delivery and Consumption

Oxygen delivery (DO2) is the amount of oxygen

delivered to the tissues per unit of time (typically
in mL/min) and determined by the oxygen con-
tent of the blood multiplied by the cardiac output
(Equation 2).

Equation 2: DO2 = CaO2 x cardiac output

With a normal CaO2 of 20 ml/dL and at a normal

cardiac output of 5 LPM, DO2 equals 1000 mL/min.
Within the limits of normal physiology, CaO2 is
relatively constant, so cardiac output must increase
to meet increased physiologic demands, by either
increasing heart rate or stroke volume.
Figure 2-1. The oxygen content of the blood (CO2) The DO2 required is based on the oxygen con-
is determined by hemoglobin, oxyhemoglobin
saturation, and the partial pressure of oxygen (PO2). sumption of the tissues (VO2) and maintained under
Hemoglobin is the most important determinant of normal physiologic conditions at a ratio of 4:1–5:1,
oxygen content. As depicted in this figure, O2 con- so the oxygen extraction ratio (O2ER) is 20-25%.
tent varies significantly based on the hemoglobin
concentration even at the same PO2 or saturation. Based on the Fick Principle (Equation 3), which

24
 Respiratory Physiology and Gas Exchange

states that the amount of oxygen absorbed across dissociation curve and an increased A-V O2 differ-
the lung is equal to oxygen consumption (VO2) by ence (Figure 2-5).
peripheral tissues (Figure 2-2). Since normal O2ER is 20-25%, normal SvO2 is
75-80%. Therefore, SvO2 = 1 – O2ER. A low SvO2
Equation 3: VO2 = cardiac output x (A-V O2 indicates that overall oxygen delivery is inadequate
difference within the lung). and that tissue oxygen utilization exceeds the com-
pensatory mechanisms of cardiac output.4 This holds
true in both native circulation and during ECLS, so
Therefore when VO2 increases, such as during that if the pre-oxygenator saturation is low, oxygen
exercise, infection, hyperthermia, catecholamine consumption is disproportionate to delivery.
release, or with excessive thyroid hormone, DO2 is
the first parameter to increase to match increased
metabolic demands (Figure 2-3).3 If DO2 cannot
match VO2, the O2ER increases to a point of ~50%
(maximally ~70% in a conditioned athlete) in order
to maintain aerobic metabolism; beyond that critical
DO2:VO2 threshold of approximately 2:1, anaerobic
metabolism is triggered and results in tissue hy-
poxia and elevated lactate (Figure 2-4). The VO2
is approximately 3-5 mL/kg/min in adults, 4-6 ml/
kg/min in children, and 5-8 ml/kg/min in infants.
Increases in temperature, PCO2, hydrogen ions
(lower pH), and 2,3 bisphosphoglycerate (such as in
hyperthyroidism and chronic anemia) will facilitate
oxygen unloading of hemoglobin and delivery to
tissues, with a right-shift of oxygen-hemoglobin Figure 2-3. The relationship between DO2 and VO2,
with percent change in various conditions. The
DO2:VO2 ratio is maintained at 5:1, allowing for
continued aerobic metabolism.

Figure 2-2. An example of resting oxygen kinetics

is depicted in this figure. The product of the arterial
oxygen content (CaO2) and cardiac output equals Figure 2-4. DO2:VO2 relationships are depicted
oxygen delivery (DO2) to tissues. Under normal at normal conditions (ratio of 4:1–5:1) and during
homeostatic conditions, DO2 is maintained at a 5:1 physiologic stress, with decreased DO2:VO2. As the
to oxygen consumption (VO2). Based on a respira- DO2:VO2 ratio drops, oxygen extraction increases
tory exchange ratio (RER) of 1, CO2 production resulting in a low venous saturation. Below a ratio
(VCO2) equals oxygen consumption. of 2:1, anaerobic metabolism ensues.

25
Chapter 2

Carbon Dioxide admixture and the ease of CO2 diffusion, PACO2 is

equivalent to PaCO2.
CO2 is produced during tissue metabolism at a
rate closely related to VO2. This rate is known as Physiology of ECLS
the respiratory exchange ratio (RER) and varies
based on diet, ranging from 0.7 (100% fat diet) to ECLS support for patients with respiratory fail-
1.0 (100% carbohydrate diet). At an RER of 1, CO2 ure is based on the same principles of gas exchange
production is equal to O2 consumed (3 mL/kg/min in the native circulation, with the movement of gas
of CO2 will be produced for every 3 mL/kg/min of determined by the gradient of partial pressures
O2 consumed, for example). The RER is defined in between various chambers.5 In this method of sup-
Equation 4. port, known as venovenous (VV) ECLS, the circuit
functions as an artificial lung, acting in series with a
Equation 4: Respiratory Exchange Ratio (RER) patient’s native circulation. As the lungs are, at best,
= VCO2/VO2 only partially functional during VV-ECLS, the bulk
of gas exchange occurs within the membrane oxy-
As with oxygen, movement of CO2 from the ve- genator. Venovenous ECLS is achieved by draining
nous system to the atmosphere is driven by the gra- venous blood from the patient, pumping it though a
dient in partial pressures. CO2 is transported in the membrane oxygenator then returning this blood to
blood in 3 ways, dissolved in plasma (10%), bound the venous system.
to hemoglobin (carbaminohemoglobin, 30%), and As with all types of ECLS, gas exchange occurs
carried in the form of bicarbonate (60%). Normal within the oxygenator by diffusion of oxygen and
PaCO2 is 40 mmHg, while PvCO2 in venous blood carbon dioxide across the oxygenator membrane.
is 46 mmHg. PACO2 is 40 mmHg, which is easily Blood flows through one chamber of the oxygenator
exhaled given a PCO2 of 0.3 mmHg in ambient air). and a continuous supply of gas, known as sweep
Exhaled CO2 may be measured directly by capnog- gas and typically comprised of 100% oxygen, flows
raphy as end-tidal CO2 (ETCO2, with a normal value through the other chamber. Modern membranes are
ranging between 35-45 mmHg). Because of venous composed of polymethylpentene-coated hollow
fibers across which gas exchange occurs down a
partial pressure gradient. As in the alveoli of the
lung, the patient’s venous blood entering the oxy-
genator has a low PO2 and a high PCO2 (PaO2 ~40
mmHg, CO2 ~45 mmHg) oxygen will diffuse into
the blood from the gas chamber (PO2 ~600 mmHg,
CO2 0 mmHg) and carbon dioxide will diffuse out,
after which it is expelled from a gas outlet. Blood
exiting the oxygenator therefore has a high PO2 and
a low PCO2. (Figure 6). The pre and post membrane
PO2, PCO2, and saturations can be measured by
analyzing pre and post membrane blood gasses or
by non-invasive circuit monitors.

Oxygen Delivery during ECLS

As DO2 in the native circulation is primarily

Figure 2-5. Oxygen – hemoglobin dissociation determined by the product of cardiac output and
curve. Acidosis (increased H+ ios), PCO2, tempera- CaO2. Oxygenation in ECLS is determined by the
ture, and 2,3 bisphosphoglycerate facilitate oxygen blood flow rate of the ECLS circuit multiplied by
unloading to tissues.
the oxygen content of the post-oxygenator gas

26
 Respiratory Physiology and Gas Exchange

(Equation 5), provided there is no recirculation (an Thus total oxygen delivery includes pump flow
important caveat). times the post-oxygenator O2 content plus the na-
tive cardiac output plus the venous oxygen content:
Equation 5: ECLS DO2 = C(post-oxygenator gas)O2 x
circuit blood flow
Equation 6: DO2 = CvO2 x cardiac output +
As such, at stable oxygen content, the primary C(post-oxygenator gas)O2 x circuit blood flow
determinant of DO2 during ECLS is circuit blood
flow. And since the ECLS and native circulations In addition to circuit blood flow, other factors
operate in series, a higher circuit blood flow rate that affect oxygenation during ECLS include the
means that a greater proportion of the patient’s car- patient’s cardiac output, tissue VO2, and degree of
diac output is being oxygenated by the ECLS circuit, lung function. Other circuit-related factors that af-
as opposed to the native circulation (Figure 2-7). For fect oxygenation during ECLS include the amount
example, if a patient’s cardiac output is 6 LPM and of blood recirculation and the efficiency of the
ECLS circuit blood flow rate is 4 LPM, two-thirds oxygenator itself. Recirculation can render oxygen
of the patient’s blood is being oxygenated maximally exchange rather inefficient when the recirculation
(in a well-functioning circuit), while one-third is be- fraction is high. The oxygenator itself is flow-rated,
ing oxygenated to the degree the lungs are able to do meaning that at the “rated flow” the oxygenator is
so. Because the patient’s actual cardiac output likely functioning to maximize DO2. At flows exceeding
exceeds ECLS circuit blood flow, total systemic the rated flow, the blood will flow through the oxy-
oxygen delivery will be a mixture of the ECLS and genator faster than the diffusion properties of the
native blood flows (Figure 2-7). oxygenator can match, so that blood at the center
of the tubing will not be optimally oxygenated by
the time it exits the oxygenator.

Carbon Dioxide Clearance during ECLS

While circuit blood flow rate is the primary de-

terminant of oxygenation during ECLS, the sweep
gas flow rate is the primary determinant of CO2

Figure 2-7. Venovenous ECLS showing oxygen

content, saturation, and partial pressure (PO2) in
various compartments of the body. Total oxygen
Figure 2-6. Gas transfer within the membrane content and delivery are dependent on the mixing
oxygenator down a partial pressure gradient. of circuit and native circulations.

27
Chapter 2

clearance, akin to ventilation. Due to its high solu-

bility, CO2, transfers across the membrane occurs 6
times faster than does oxygen. Given this relative
ease of CO2 transfer, clearance will occur provided
an adequate CO2 gradient is maintained across the
oxygenator membrane. Maintenance of this gradient
is achieved by constantly delivering a fresh sup-
ply of oxygen-rich, carbon dioxide-poor or absent
sweep gas to the gas chamber of the oxygenator.
The rate of sweep gas flow is controlled by the
circuit flow meter and percentage of that gas which
is oxygen (fraction of delivered oxygen, FDO2) is
controlled by the blender. With the exception of
ECLS weaning, the FDO2 is often maintained at
100%, although this may vary by center. Carbon di-
oxide clearance is independent of circuit blood flow
so that even at low blood flow rates adequate CO2
diffusion occurs provided sweep gas flows within
the gas chamber at a high enough rate, typically
ranging between 1-11 LPM. Patients with a higher
PvCO2 will therefore require greater sweep gas flow
to maintain adequate CO2 elimination. In addition
to the PCO2 gradient across the membrane, other
factors that determine the amount of CO2 clearance
include the membrane surface area and oxygenator
efficiency (condensation within the gas chamber
will gas impair CO2 transfer).

Summary

A clear understanding of the physiologic

principles detailed in this chapter are essential
to managing all patients with respiratory failure,
particularly those receiving ECLS. Incorporation
of these concepts of oxygen content, delivery, and
consumption, as well as the fundamentals of gas
transfer will allow for optimal management of these
critically ill patients. Further details can be found
in Chapters 7, 13, 19, 20 and 22.

28
 Respiratory Physiology and Gas Exchange

References

1. Brogan T, Lequier L, Lorusso R, MacLaren

G, Peek G. Extracorporeal Life Support: The
ELSO Red Book, Fifth Edition. Ann Arbor,
Extracorporeal Life Support Organization 2017.
2. Leach RM, Treacher DF. The Pulmonary
Physician in Critical Care 2: Oxygen Delivery
and Consumption in the Critically Ill. Thorax
2002;57:150-57.
3. Vignati C, Cattadori G. Measuring Cardiac Out-
put During Cardiopulmonary Exercise Testing.
Am J Resp Crit Care Med 2017;14(S1):S48-52.
4. Walley KR. Use of Central Venous Oxygen
Saturation to Guide Therapy. Am J Resp Crit
Care Med 2011;184(5):514-20.
5. Short BL, Williams L. ECMO Specialist Train-
ing Manual, Third Edition. Ann Arbor, Extra-
corporeal Life Support Organization 2010.

29
3

Cardiac Physiology Relevant to Extracorporeal Life Support

Monique Radman, MD, MAS, Cory M. Alwardt, PhD, CCP, Bhavesh M. Patel, MD

Introduction following concepts, oxygen content values can be

calculated quite easily (Equation 1).
The cardiovascular system delivers oxygen
and nutrients to the tissues and removes carbon Equation 1: Oxygen content = oxygen bound
dioxide and metabolic byproducts. In addition, va- to hemoglobin + oxygen dissolved in plasma
soactive substances regulate myocardial function
and vascular tone. Cardiac activity is impacted by Hemoglobin can carry 1.34 mL of oxygen per
cardiovascular disease, including congenital heart gram of hemoglobin. Therefore, a patient with a
disease, acquired heart disease, and cardiac surgery. hemoglobin level of 12.0 g/deciliter has a total of
An understanding of the basic principles underlying 16.08 mL of oxygen bound to hemoglobin in each
normal cardiac physiology is essential when con- deciliter (dL) of fully saturated blood. Because a
sidering the use of ECMO in the neonatal, pediatric, dL is one-tenth of a liter, the patient would there-
and adult patient with cardiac disease. fore have 160.8 mL in one liter of fully saturated
blood. However, even arterial blood is not usually
Oxygen Delivery fully saturated. Normal arterial oxygen saturations
in a healthy individual are 97-100%, and this must
Because a primary job of the cardiovascular be accounted for when calculating oxygen content
system is to deliver oxygen to the tissues, a thorough (Equation 2). The same patient described above
understanding of determinants of oxygen delivery with 160.8 mL of oxygen in fully saturated blood
is crucial in comprehending normal cardiovascular would only have 156.0 mL of oxygen if the arterial
physiology and ECMO. To understand how much oxygen saturation was 97%.
oxygen is being delivered to the tissues, one must
first know how much oxygen is in a given amount of Equation 2: Oxygen bound to hemoglobin =
blood (oxygen content), and also how much blood [1.34 g/dL] x [Hemoglobin (g/dL)] x [% satura-
is being pumped (cardiac output). These concepts tion]
can also be applied to artificial circulation.
While the calculation above represents the
Oxygen Content vast majority of oxygen in the blood, we must also
account for the oxygen dissolved in the plasma
Oxygen content refers to the amount of oxygen (Equation 3). For each mm Hg of partial pressure
(in mL) in a given volume of blood. To determine of oxygen (PO2), there is 0.003 mL of oxygen dis-
the quantity of oxygen in a given volume of blood, solved in each deciliter of blood. A patient whose
one must understand how oxygen is transported in PaO2 is 100 mmHg would therefore have 0.3 mL of
the blood. The vast majority of oxygen is bound to oxygen dissolved in each deciliter of blood. Again,
hemoglobin, while a smaller amount is dissolved in remember that 10 deciliters make a liter, so that
the plasma. With a thorough understanding of the same patient would have 3 mL of oxygen dissolved
in each liter of blood.

31
Chapter 3

Equation 3: Oxygen dissolved in plasma = Starling mechanism--the method by which stretched

[0.003 mL/dL] x PO2 or elongated heart muscle can change its force of
contraction. Within a normal range, increases in
As mentioned above, total oxygen content is preload typically result in increased force of con-
the sum of bound and dissolved oxygen. By adding traction and therefore higher SV. Chronic excessive
the hemoglobin-bound oxygen in the patient above preload can result in pathologic morphologic adap-
(156.0 mL) to the amount of oxygen dissolved in tation from the left ventricle resulting in a dilated
the plasma (3 mL), we see that the patient has an cardiomyopathy. Disease states such as aortic valve
oxygen content of 159 mL of oxygen per liter of regurgitation can cause unwanted changes in the left
blood. However, oxygen content does not reflect ventricle (i.e. dilated cardiomyopathy) due to chron-
how fast this oxygen is delivered to the tissues--it ic elevations in preload. Centrifugal ECMO pumps
is only a measure of oxygen in a given volume of also depend on preload--they need an adequate
blood. In order to determine how much oxygen is amount of blood surrounding the drainage cannula.
delivered, we must understand the contribution of Patients with low preload on ECMO can have the
cardiac output. tissues surrounding the drainage cannula sucked up
against the cannula, preventing blood drainage and
Cardiac Output causing dramatic decreases in blood flow and even
causing tissue and red blood cell damage.
Cardiac output is a measure of the total amount In contrast, afterload refers to the amount of
of blood flow pumped from the heart per minute. resistance to ejection of blood from the heart. Af-
Cardiac function is determined by the contractile terload can be thought of as the load after the heart.
state of the heart, valvular competence, diastolic A high afterload causes increased wall stress on the
properties, ventricular interdependence and heart heart muscle, and SV decreases. Ejection of blood
rhythm as well as the work it must perform. Nu- from the heart is sensitive to afterload. Disease
merically, cardiac output equals the amount of states such as high systemic blood pressure or aortic
blood ejected from the heart with each beat (stroke valve stenosis cause increases in afterload on the
volume) multiplied by the heart rate (Equation 4). left heart, and can cause problematic changes in the
structure and function of the chambers of the heart
Equation 4: Cardiac output = stroke volume (i.e. hypertrophic cardiomyopathy). Similarly, right
x heart rate heart function can also be detrimentally affected
by excessive afterload from primary or secondary
Several factors influence stroke volume (SV): causes of pulmonary arterial hypertension.4 Cen-
preload, afterload, and contractility. Each of these trifugal ECMO pumps are also sensitive to afterload.
factors can dramatically affect SV, and therefore A high blood pressure, for example, can decrease
cardiac output and oxygen delivery.1 ECMO flow.
Preload refers to the amount of blood in the heart The last primary determinant of SV is contractil-
at the end of relaxation, or diastole. High preload ity of the myocardium, or heart muscle. Contractility
refers to a heart that contains a high volume of blood refers to the ability of the muscle fibers to contract.
at end diastole, meaning that may or may not be Assuming heart valves are functioning normally,
ejected with each beat. A dehydrated patient often high contractility results in more ejection of blood
has low preload, and therefore lower SV. Stroke from the heart at a given preload and afterload. Con-
volume depends on a certain minimal amount of tractility depends on the overall myocardial health
preload. In addition to hydration levels (i.e. blood including its ability to respond to circulating factors
volume), venous tone can affect preload. Higher such as catecholamines. Catecholamines (such as
venous tone produces increased mean systemic ve- epinephrine or norepinephrine) released from adre-
nous pressure, resulting in more blood being pushed nal gland or sympathetic nervous system stimulation
into the heart and higher preload. The relationship act on alpha and beta receptors in cardiac and vas-
between SV and preload is referred to as the Frank- cular tissue. Their clinical effect partially depends

32
 Cardiac Physiology Relevant to Extracorporeal Life Support

upon which receptors are predominantly stimulated. technique provides an estimate of cardiac output
Alpha-receptor (α) stimulation result in arterial va- based on the temperature change of blood as it
soconstriction and venoconstriction, which increase travels through the right heart after injecting a small
systemic vascular resistance and venous return amount of dye (i.e. cold saline). Echocardiography,
respectively. Beta-receptor (β) stimulation can or an ultrasound of the heart, can also provide an
be divided into 3 broad clinical effects: increased estimate of cardiac output by estimating SV based
force of myocardial contraction (inotropy – β1 ef- on the change in size of the ventricles during systole
fect), increased heart rate (chronotropy - β1 effect) versus diastole or by Doppler velocity time integral
and arterial vasodilation (β2 effect). The relative (VTI) of blood flow through the left ventricular
influence of various catecholamines on the α− and outflow tract.2 This, however, simply represents
β-receptors is shown in Figure 3-1 with epinephrine a snapshot and does not provide continuous data.
being equipotent on the α− and β-receptor effects. Newer techniques attempt to estimate cardiac output
Also shown is milrinone, a non-catecholaminergic based on the arterial blood pressure waveforms, but
agent causing similar cardiovascular effects as β1 the accuracy in various clinical conditions is still
and β2 stimulation in addition to facilitating myo- being investigated.
cardial relaxation (lusitropy). Vasopressin, a non- ECMO support (particularly VA-ECMO), ren-
catecholaminergic vasoconstrictor, causes selective ders the interpretation of cardiac output data from
vasoconstriction of systemic vasculature with less these techniques difficult. Intuitively, however,
effect on the renal and pulmonary vasculature at during ECMO support the pulse pressure (systolic
low doses. In practice, dobutamine and milrinone pressure–diastolic pressure) can also provide infor-
are generally accepted first line inotropic drugs to mation about the amount of native cardiac output.
increase cardiac contractility while norepinephrine Because ECMO pumps are nonpulsatile, patients
and vasopressin are first line vasoconstrictors. can have no pulse pressure and a flat arterial blood
Clinical estimation of cardiac output can be pressure waveform. However, when the heart ejects
done using a number of techniques. The gold stan- it can generate a pulse pressure despite high levels
dard is the thermodilution technique performed of nonpulsatile flow from the circuit. Most clinicians
using a Swan Ganz pulmonary artery catheter. This agree that a small amount of pulse pressure (5-15
mmHg) is preferred in order to ensure that the heart
ejects at least a small amount of blood to prevent
it from becoming over distended and help prevent
thrombus formation in the heart and ascending aorta.

Calculation of Oxygen Delivery

In the above section, we described how to cal-

culate oxygen content. In order to calculate oxygen
delivery (Equation 5), we must now multiply oxygen
content and the rate at which it is delivered (cardiac
output) (Figure 3-2).

Equation 5: Oxygen delivery = oxygen content

x cardiac output
Figure 3-1. Vasopressor and Inotropic Cardiovas-
cular Medications. Alpha catecholamine receptor
activating spectrum, α; Beta catecholamine recep- In the patient above with a hemoglobin of 12 g/
tor activating spectrum, β. Italics medications are dL, an arterial oxygen saturation of 97%, and a PaO2
non-catecholamine receptor activating medications: of 100 mmHg, we calculated the oxygen content
milrinone, phosphodiesterase (PDE)3 inhibitor;
vasopressin, V1a receptor activator. to be 159 mL of oxygen per liter of blood. If the
patient had a cardiac output of 5 liters per minute,

33
Chapter 3

the oxygen delivery to the tissues would be 795 mL The most common and convenient estimate of
of oxygen per minute. adequate oxygen delivery is systemic venous oxy-
For patients on ECLS, the calculation of oxygen genation, or SvO2. While individual vascular beds
delivery provided by the circuit is done by measur- extract variable amounts of oxygen (pulmonary,
ing the oxygen and hemoglobin values from the cerebral, renal, etc.), the cardiovascular system, as
arterial side of the circuit and multiplying by the a whole, normally delivers 3-5 times more oxygen
ECLS pump flow as the “cardiac output.” You can than the tissues extract for metabolism, resulting
find information on the relative contributions of in a SvO2 of approximately 25-30% below arterial
ECLS flow versus native circulation elsewhere in saturation. Lower SvO2 values could potentially sig-
this manual. nal inadequate oxygen delivery. The Fick Equation
can also be used to calculate the rate at which the
Adequacy of Oxygen Delivery body is utilizing oxygen, or the VO2. This equation
takes into account cardiac output and the difference
While it is important to understand the concepts in oxygen levels between arterial and venous blood.
and calculations associated with oxygen delivery, a During ECLS, use of SvO2 and the Fick Equa-
specific oxygen delivery is not usually targeted. The tion can be problematic because it is often impos-
microcirculation of each organ has multiple signal- sible to measure true mixed venous oxygen satura-
ing pathways to ensure adequate nutrient delivery tion. In this scenario, clinicians must rely on trends
to meet metabolic demands. The microcirculatory in factors such as serum lactate as well as tissue
determinants of blood flow include the tissue perfu- oximetry to understand adequacy of perfusion. As
sion pressure, regional arteriolar tone, capillary pa- a byproduct of anaerobic metabolism, increases in
tency and hemorheology.3 Normal values of global serum lactate can signal inadequate oxygen delivery.
oxygen delivery range from 650 mL to 1400 mL Regardless of how tissue perfusion is being deter-
per minute (3-20 mL/kg/min for pediatric patients), mined, a thorough understanding of the concepts
largely based on size, gender and body composi- and calculations related to oxygen delivery are vital
tion. Clinicians must understand clinical markers of in determining the next appropriate steps in patient
end-organ perfusion to determine whether oxygen
delivery is adequate (Figure 3-3).

Figure 3-2. Determinants of oxygen delivery: The Figure 3-3. Balance of Oxygen Delivery and Con-
product of arterial blood oxygen content times car- sumption: serum lactate, mixed venous oxygen
diac output. In turn, arterial blood oxygen content saturation and tissue oximetry are used in clinical
is directly associated with hemoglobin concentra- context to estimate adequacy of the balance be-
tion (HgB), arterial oxygen saturation (O2 Sat) and tween global oxygen delivery and global oxygen
arterial oxygen tension (PaO2). Cardiac output is consumption. Oxygen, O2; Content of arterial
directly related to heart rate and stroke volume. oxygen, CaO2; Hemoglobin, Hgb; Arterial oxygen
Stroke volume is dependent on preload, afterload saturation, SaO2; Partial pressure of arterial oxygen,
and myocardial contractility. PaO2; Adenosine triphosphate, ATP, Krebs cycle,
Krebs; Cori cycle, Cori.

34
 Cardiac Physiology Relevant to Extracorporeal Life Support

management. For example, increasing the rate of

blood flow or increasing hemoglobin levels using Equation 6: Flow through a vessel ∝(Pr4) / ηL
banked blood can have a much larger impact than
increasing the PaO2. The impact of such maneuvers Notably, since resistance changes with the
can be calculated prior to making patient manage- fourth power of the radius, vascular tone or cannula
ment decisions. size has a magnified effect on flow. This can lead to
dramatic changes in both preload and afterload, and
Hemodynamics: Pressure, Flow, Resistance and in turn, cardiac output and oxygen delivery.
Compliance The relationship between changes in intravas-
cular volume (∆V) and the concomitant change in
When considering ECMO, it is important to pressure (∆P) determines the compliance (or elastic-
consider the factors that affect flow through the ity) of a vessel (Equation 7):
circuit. These factors include vascular function,
cardiac function and, lastly, the interaction between Equation 7: Vessel compliance = ∆V / ∆P
the two, or cardiovascular coupling.1,4
The venous system has more compliance and
Vascular Function larger vessels than the arterial system, which is why
the majority of vascular volume remains in the ve-
With laminar flow, the primary factors influenc- nous system at any one point in time. Accordingly,
ing the amount of flow through a vessel or ECMO fluid resuscitation leads to larger increases in venous
cannula include the difference between the inflow volume than arterial volume (Figure 3-4). Although
and outflow pressures (P) and the resistance of the overlooked at times, hematocrit has a significant
vessel, which is directly proportional to its length effect on blood viscosity, which is an important
(L) and inversely proportional to its radius (r). Poi- determinant of flow (hemorheology) in both native
seuille’s Equation describes this relationship where and ECMO circulations.3
η is blood viscosity (Equation 6):
Cardiac Function

Graphically, cardiac function can be thought of

as the opposite of the vascular function relationship
shown above. In other words, alterations in central
venous pressure (CVP) directly affect cardiac output
(Q). This is a representation of the Frank Starling
relationship in which myocardial stretch, as a result
of increased CVP, augments Q (Figure 3-5).5 This
relationship can be affected by the other determi-
nants of cardiac output that have previously been
discussed in this chapter. Specifically, increased
afterload [e.g. systemic vascular resistance (SVR)]
due to, for example, peripheral vasoconstriction,
will decrease cardiac output, shifting the curve
Figure 3-4. Central venous pressure (CVP) de- downward. In turn, decreased afterload will shift
creased as cardiac output (Q) increases. The criti- the curve upward.
cal closing pressure of the venous system (Pcc) is
the point where the highest Q and lowest CVP are
achieved for a certain blood volume. Decreases Cardiovascular Coupling
and increases in total blood volume are associated
with decreases and increases in the vascular func- Theoretically, the cardiovascular system oper-
tion curve. ates at a point of intersection between the vascular

35
Chapter 3

and cardiac function curves. Increased SVR shifts ing of cardiopulmonary physiology and ECMO
both curves downward, moving their point of inter- circuit function. Venoarterial ECMO allows blood
section downward (Figure 3-6). to bypass the heart and lungs, draining the venous
Cardiac output results from cardiac function blood and sending it through an oxygenator and into
and venous return acting simultaneously. The ve- the systemic circulation. Venovenous ECMO does
nous tone, intravascular volume, and resistance to not provide direct hemodynamic support, but deliv-
venous flow determine venous return to the right ers oxygenated blood to the right heart. Not only can
heart. Right heart dysfunction increases resistance ECMO restore a normal ratio of oxygen delivery to
to venous return and commonly results from a oxygen consumption, but it also permits weaning
sudden elevation in pulmonary vascular resistance of deleterious ventilator settings and, potentially
(PVR). Some examples of rapidly reversible causes toxic cardiac medications. Most importantly, ECMO
of elevated PVR include hypoxia, decreased lung allows time for cardiac and/or respiratory function
volumes (e.g. compressive pleural effusions) and to recover. VA ECMO brings the disadvantage of
excessive intrathoracic pressures (e.g. pneumo- increased left ventricular afterload related to higher
thorax, high mechanical ventilation tidal volumes systemic arterial blood pressure generated by the
or PEEP).6 Venous return and cardiac output are artificial circulation.8 However, this is offset by
interdependent and must be considered together maintenance of coronary perfusion and decreased
when contemplating any therapeutic intervention, preload, time for myocardial recovery and/or car-
including ECMO.7 diac transplant.

Relevance to ECMO Summary

ECMO is used to ensure adequate oxygen deliv- The cardiovascular system provides delivery
ery when heart and/or lung failure occurs. Managing of oxygen and nutrients to the tissues, as well as
a patient on ECMO requires a thorough understand- removes carbon dioxide and metabolic by-products.
Oxygen delivery equals the product of oxygen

Figure 3-5. Increased central venous pressure Figure 3-6. Theoretical interaction between the car-
(CVP) leads to increased ventricular contractility diac function and vascular function curves shown
and, as a result, cardiac output (Q). Increase and above. The cardiovascular system, or ECMO cir-
decreases in afterload (systemic vascular resis- cuit, functions at the point of intersection.
tance) shift the cardiac function curve down and
up, respectively.

36
 Cardiac Physiology Relevant to Extracorporeal Life Support

content (i.e. hemoglobin, pO2, oxygen saturation)

and the rate of blood flow (i.e. cardiac output and/
or ECMO blood flow). Oxygen delivery can be eas-
ily calculated by understanding the principles and
calculations reviewed in this chapter. Calculation of
oxygen delivery allows clinicians to make informed
decisions on how to improve patient care. Under-
standing markers of perfusion can be used to deter-
mine whether or not oxygen delivery is adequate to
the tissues. In addition, an understanding of vessel
hemodynamics is important to comprehending the
relationship between blood pressure and blood flow
in both the arterial and venous systems. A thorough
understanding of cardiovascular physiology is im-
perative to proper treatment of patients both with
and without ECMO support.

37
Chapter 3

References

1. Nichols DG. Critical Heart Disease in Infants

and Children. Mosby Incorporated; 2006.
2. Huntsman LL, Stewart DK, Barnes SR,
Franklin SB, Colocousis JS, Hessel EA. Non-
invasive Doppler determination of cardiac
output in man. Clinical validation. Circulation.
1983;67(3):593–602.
3. Baskurt OK, Meiselman HJ. Blood rheology
and hemodynamics. Semin Thromb Hemost.
2003;29(5):435–450.
4. Shaffner DH, Nichols DG. Rogers’ Textbook of
Pediatric Intensive Care. Lippincott Williams &
Wilkins; 2015.
5. Berlin DA, Bakker J. Starling curves and central
venous pressure. Crit Care. 2015;19:55..
6. Harjola V-P, Mebazaa A, Čelutkienė J, et al.
Contemporary management of acute right
ventricular failure: a statement from the Heart
Failure Association and the Working Group
on Pulmonary Circulation and Right Ven-
tricular Function of the European Society of
Cardiology. European Journal of Heart Failure.
2016;18(3):226–241.
7. Sakamoto K, Saku K, Kishi T, et al. Predic-
tion of the impact of venoarterial extracorpo-
real membrane oxygenation on hemodynam-
ics. AJP: Heart and Circulatory Physiology.
2015;308(8):H921–H930.
8. Femoral Cannulation and Veno-Arterial ECMO
- Mayo Clinic [Video]. YouTube Https://Youtu.
Be/NGGA-8zXVGE. Published January 29,
2007. Accessed April 30, 2010.;2013.

38
4

Management of Anticoagulation and Blood Products on ECMO

Jessica Nicoll, MD, Patti Massicotte, MD, Gail Annich, MD

Objectives understanding of their mechanism of action, me-

tabolism, dosing range, and monitoring (Table 4-2).
After completion of this chapter, the participant In addition anticoagulant response varies according
should be able to: 1) discuss common risks associ- to patient age, diagnosis, and metabolic condition.
ated with ECLS; 2) outline routine anticoagulation
strategies for a patient on ECLS; 3) discuss antico- Unfractionated Heparin (UNFH)
agulation monitoring while on ECLS; and 4) discuss
strategies for minimizing risk of thrombotic and Unfractionated heparin remains the most
hemorrhagic complications. widely used anticoagulation therapy for patients
on ECLS support. UNHF primarily acts by form-
Introduction ing a complex with a plasma protein, antithrombin
(AT).2,3 Together, they inhibit factor Xa and throm-
The most frequent of the complications on ex- bin activity, thereby preventing the formation of a
tracorporeal life support (ECLS) involve either clots cross-linked fibrin clot.2,3 UNFH also increases the
that form within the circuit, or clots and/or bleeding release of tissue factor protein inhibitor (TFPI) from
within the patient.1 The etiology is multifactorial the endothelium, which inhibits activation of hemo-
and related to patient diagnosis, physiology, and stasis through TF-FVIIa.4 AT deficiency can affect
response to artificial circuitry. (Table 4-1) the amount of UNFH effect and its level should be
assessed in the setting of increasing heparin require-
ECLS Anticoagulation ments.6 Furthermore, developmental hemostasis of
the neonate must be considered for such patients
The most common anticoagulant is unfraction-
ated heparin (UNFH), although some centers have
Anticoagulant Mechanism Metabolism and Half Life Dose Monitoring
converted to the use of direct thrombin inhibitors of Action excretion (minutes) Range
Metabolism:liver 30-602,3
for anticoagulation during ECLS instead of only Unfractionated Inhibits
heparin Factor Xa, Excretion:kidney
2,3
bolus 50 to
100
Anti-Factor Xa
assay
thrombin units/kg7 0.3 – 0.7
using them when heparin induced thrombocy- and Maintenanc IU/mL
increases e ACT
topenia (HIT) is suspected/diagnosed. For each tissue factor 10-60 aPTT7
protein units/kg/h7
anticoagulant used during ECLS, there must be an inhibitor2,3, 4
Bivalirudin Synthetic Excretion: 25-35 8,9 Bolus 0.05- aPTT 50-60s7
direct proteolytic 0.5mg/kg7 **
thrombin enzymes (80%), Maintenanc
Complication % of Total % of Total % of Total inhibitor 9 kidneys (20%)10 e
Complications Complications in Complications 0.03-
in Adults Pediatrics in Neonates 0.5mg/kg/h
7
VV/ VA-ECMO VV/ VA-ECMO VV/ VA-ECMO
Cannula hemorrhage 13.2/18.5 18.3/ 15.6 7.9/ 10.7 Argatroban Synthetic Hepatic 40-50 8,9 Maintenanc aPTT 50-60s7*
direct metabolism9 e
Surgical hemorrhage 10.5/ 20.2 12.6/ 28.9 6.3/ 29.3 thrombin 0.5-2
Renal failure 9.3/12.3 12.9/7.2 7.8/12.3 inhibitor9
µg/kg/min7
Infection 17.5/ 13 16.8/ 11 5.8/7.1 *The ranges for aPTT when using DTIs are derived from the literature. However, there is
variability in the use of these agents for example in pediatrics, frequently no bolus dose is
given for bivalirudin and is started at a higher level, 0.3mg/kg/hr, if not transitioning from
Table 4-1. Frequent ECLS complications (Ex- UFH. The aPTT is checked in 2 hours to achieve a range between 60-90 sec.
tracorporeal Life Support Organization Registry
International Report 20161). Table 4-2. Anticoagulants.

39
Chapter 4

on ECLS as their responses vary more widely than ACT target range depends on device and should be
those of adults, requiring customization. (Table 4-3) adjusted based on patient bleeding, circuit clotting,
In addition to the risk of bleeding, patients are or the measured anti-factor Xa level. It remains the
also at risk for heparin induced thrombocytopenia simplest whole blood bedside test for hemostatic
(HIT), which is rare. In adults, HIT typically pres- monitoring taking into account all factors that in-
ents with at least a 50% decrease in platelet count fluence clot formation beyond UNFH infusion, but
that occurs within 5-14 days after initiating UNFH it uses contact activation rather than tissue factor
and is associated with venous and arterial thrombi. 8 activation. Recommendation is that it be paired with
In the case of HIT, heparin resistance or significant a heparin monitoring test such as aPTT or antiXa
thrombocytopenia, UNFH may be replaced by a assay. There is no correlation between the tests but
direct thrombin inhibitor for ECLS anticoagulation. the combination helps to separate heparin effect
of anticoagulation versus individual patient physi-
Direct thrombin inhibitors (DTI) ologic hemostatic conditions and responses to their
illness and the artificial circuitry.
Bivalirudin and argatroban directly inhibit
thrombin without binding AT. DTIs should be initi- Activated Partial Thromboplastin Time (aPTT)
ated at the low end of the maintenance range and
increased to target activated partial thromboplastin Activated partial thromboplastin time (aPTT)
time (aPTT) of 50-60 seconds however this range has classically been used to titrate UNFH and is
comes from the literature and the range used clini- the preferred monitoring parameter for titrating
cally through experience is more closely aligned DTIs because it is easy to use while other specific
between 60-90 seconds.7 tests for monitoring the effects of hirudin analogs
and arginine derivatives are rare and often need
Anticoagulation Monitoring (Table 4-4) to be sent out.7,16 The aPTT is the time measured
between calcium addition and clot formation.16
Activated Clotting Time (ACT) However, aPTT can be prolonged at baseline in
pediatric patients (developmental hemostasis) and
Activated clotting time (ACT) is a rapid, bed- falsely shortened by acute phase reactants.17 This
side test that requires a drop of blood to measure limits its applicability for anticoagulation titration
clotting of whole blood.5 It is used by almost all in critically ill patients.5, 7, 16 As such, it has poor cor-
ECLS centers for UNFH dosage.7,12 Limitations relation with anti-factor Xa in children and patients
include a lack of precision and inability to differen- on ECLS support. 16
tiate between causes of impaired clotting.13-15 The

Step 1: Consider blood prime 50-100 units may be added

anticoagulation to each full unit of blood ACT Q1h – Q2h
prime aPTT Q6h – Q12
Step 2: Once the vessels are Bolus 50 to 100 units/kg
exposed and before ECLS
Anti–factor Xa Assay Q6h
cannula insertion Platelets Q6h - Q12h
Step 3: ACT ≤ 300 Initiate infusion at 7.5-20 INR Q6h - Q12h
units/kg/h Fibrinogen Q12h - Q24h
Adults, transthoracic CBC Q6h - Q12h
cannulation, severe Antithrombin Level Daily - PRN
coagulopathy, or active
bleeding start at lower end
Thromboelastography Daily – PRN
of range for bleeding
Step 4: Titrate infusion to Maximum dose 40-60 or clotting
effect units/kg/h complications
Table 4-3. ELSO Anticoagulation Guidelines 2014: Table 4-4. ELSO Anticoagulation monitoring
Protocol for UNFH anticoagulation. laboratory schedule.

40
 Management of Anticoagulation and Blood Products on ECMO

Anti-factor Xa Assay fibrinolysis in whole blood. It is useful in the case of

complications from bleeding or clotting. It may help
Anti-factor Xa assay measures UNFH anticoag- to determine which blood products to administer to
ulation activity and depends on AT levels. Although control the bleeding.
the most reliable heparin assay, it can be falsely
low with hemolysis (elevated plasma hemoglobin) Thromboelastometry
and hyperbilirubinemia. 5 Knowledge of how the
test is performed is imperative as some assays add Thromboelastometry is a similar method but
exogenous AT or dextran sulfate, which can falsely instead of the cup rotating around a pin, the pin ro-
overestimate the heparin effect. The recommenda- tates inside the cup. Knowledge of this technology
tion is to perform the assay without exogenous addi- is required as is a baseline graph in order to use it
tion to AT for accuracy of heparin effect. (Table 4-5) effectively in ECLS patients. It is similar to ACT
in that it is a whole blood test assessing hemostasis
Antithrombin in the patient taking into account all influences and
not just heparin effect.
Antithrombin is an essential cofactor of UNFH
and may be replaced with either FFP or AT con- Conclusion
centrate in the case of low AT levels with heparin
resistance (e.g. low anti-factor Xa assay levels).7 Management of anticoagulation for ECLS re-
AT concentrate delivers a higher AT concentration quires an understanding of the anticoagulant agents
in a small volume but is expensive when compared used and the tests for anticoagulation monitoring.
to FFP.19-20 However, it has been shown to reduce Therapeutic anticoagulation dosing goal is to de-
UNFH dose requirements.19 FFP is a poor means to crease the risk of both clotting and bleeding. Targets
increase AT, however when it is the only agent avail- for platelets, INR, and fibrinogen have also been
able it does aid in optimizing heparin functionality. recommended to further decrease risk of bleeding
In two single center studies it was not associated (Table 4-6).
with increased bleeding or transfusion,6,19 but in a
multicenter, retrospective analysis of an administra-
tive database its use was associated with increased Platelets >80,000/μL to 100,000/μL
clotting and bleeding.21 INR FFP transfusion to maintain INR <2.0
Fibrinogen Cryoprecipitate to maintain
Thromboelastography (TEG) fibrinogen > 100 mg/dL
OR
> 150 mg/dL if bleeding or prior
Thromboelastography (TEG) is the measure of surgical intervention
time needed to form a fibrin clot, clot strength, and Hematocrit PRBCs to maintain hematocrit > 30%
(consider higher goal for neonates
and children with cyanotic congenital
heart disease or lower goal for stable,
Anti-factor Xa Anti-factor UNFH Rate ACT Goal
adult patients)
Goal Range Xa Level Change Range
Antithrombin >50%-80% (>0.5-0.8u/mL), consider
(units/mL) (units/mL) (seconds)
AT replacement if on maximum dose
< 0.3 ↑10 – 20% ↑10 – 20 of UNFH and unable to obtain
0.3 – 0.5 0.3 - 0 .5 No Change No Change therapeutic anti-factor Xa assay
> 0.5 ↓10 – 20% ↓10 – 20
< 0.4 ↑10 – 20% ↑10 – 20 Table 4-6. Blood product and factor replacement.
0.4 – 0.6 0.4 – 0.6 No Change No Change
> 0.6 ↓10 – 20% ↓10 – 20
< 0.5 ↑10 – 20% ↑10 – 20
0.5 – 0.7 0.5 – 0.7 No Change No Change
> 0.7 ↓10 – 20% ↓10 – 20
Table 4-5. ELSO UNFH titration and ACT goal
range based on anti-factor Xa levels.

41
Chapter 4

References extracorporeal membrane oxygenation: an

international survey. PediatrCrit Care Med
1. Thiagarajan R, Barbaro R, Rycus PT, et al. Ex- 2013;14(2):e77-84.
tracorporeal Life support organization registry 13. Uden DL, Payne NR, Kriesmer P, Cipolle RJ.
international report 2016. ASAIO J 2017; 63 Procedural variables which affect activated
(1): 60-67. clotting time test results during extracorporeal
2. Verstraete M. Pharmacotherapeutic aspects membrane oxygenation therapy. Crit Care Med.
of unfractionated and low molecular weight 1989;17(10):1048-1051.
heparins. Drugs. 1990;40(4):498-530. 14. Seay RE, Uden DL, Kriesmer PJ, Payne NR.
3. Sinauridze EI, Panteleev MA, Ataullakhanov FI. Predictive performance of three methods of
Anticoagulant therapy: basic principles, classic activated clotting time measurement in neonatal
approaches and recent developments. Blood ECMO patients. ASAIO J 1993;39(1):39-42
Coagul Fibrinolysis. 2012;23(6):482-493. 15. Bosch YP, Ganushchak YM, de Jong DS.
4. Lupu C, Poulsen E, Roquefeuil S, Westmuck- Comparison of ACT point-of-care measure-
ett AD, Kakkar, VV, Lupu F. Cellular effect of ments: repeatability and agreement. Perfusion.
heparin on the production and release of tissue 2006;21(1):27-31.
factor pathway inhibitor in human endothelial 16. Annich, GM, Zaulan O, Neufeld M, Wagner
cells in culture. Arterioscler Thromb Vasc Biol. D, Reynolds MM. Thromboprophylaxis in
1999; 19: 2251-2262 extracorporeal circuits: current pharmacologic
5. Ryerson LM, Lequier LL. Anticoagulation strategies and future directions. Am J Cardio-
management and monitoring during pediatric vasc Drugs. 2017; 17(6): 425-429.
extracorporeal life support: a review of current 17. Teruya J. Coagulation Tests Affected by Acute
issues. Front Pediatr 2016; 4: 67 Phase Reactants Such as CRP and Factor VIII.
6. Niebler RA, Christensen M, Berens R, Wellner Paper presented at: International Conference on
H, Mikhailov T, Tweddell JS. Antithrombin Hematology and Blood Disorders; September
replacement during extracorporeal membrane 23 - 25, 2013; Research Triangle Park, NC USA.
oxygenation. Artific Organs 2011; 35(11):1024- 18. Winkler AM Managing the Precarious Hemo-
1028 static Balance during Extracorporeal Life Sup-
7. ELSO anticoagulation guidelines 2014 port: Implication for Coagulation Laboratories.
8. Warkentin TE, Greinacher A. Heparin-induced Semin Thromb Hemost 2017;43:291-299.
thrombocytopenia: recognition, treatment, and 19. Ryerson LM, Bruce AK, Lequier L, Kuhle S,
prevention: the Seventh ACCP Conference Massicotte MP, Bauman ME. Administration
on Antithrombotic and Thrombolytic Therapy. of Antithrombin Concentrate in Infants and
Chest. 2004;126(3 Suppl):311S-337S. Children on ECLS Improves Anticoagulation
9. Hirsh J, O’Donnell M, Weitz JI. New antico- Efficacy. ASAIO J 2014; 60(5):559-563.
agulants. Blood. 2005;105(2):453-463. 20. Mintz PD, Blatt PM, Kuhns WJ, Roberts HR.
10. Gladwell TD. Bivalirudin: a direct thrombin Antithrombin III in fresh frozen plasma, cryo-
inhibitor. Clinical Therapeutics. 2002;24(1):38- precipitate, and cryoprecipitate-depleted plasma.
58. Transfusion. 1979;19(5):597-598.
11. Phillips MR, Khoury AI, Ashton RF, Cairns BA, 21. Wong TE, Huang YS, Weiser J, Brogan TV,
Charles AG. The dosing and monitoring of arg- Shah SS, Witmer CM. Antithrombin concen-
atroban for heparin-induced thrombocytopenia trate use in children: a multicenter cohort study.
during extracorporeal membrane oxygenation: J Pediatrics. 2013;163(5):1329-1334.e1321.
a word of caution. Anaesth Intensive Care.
2014;42(1):97-98.
12. Bembea MM, Annich G, Rycus P, Oldenburg
G, Berkowitz I, Pronovost P. Variability in
anticoagulation management of patients on

42
5

Transfusion Management during Extracorporeal Life Support

Anne Marie Winkler, MD, MSc, Trisha E. Wong, MD, MS

Transfusion Support during Extracorporeal Life As a result, transfusions of red blood cells (RBC),
Support (ECLS) platelets, fresh frozen plasma (FFP), and cryopre-
cipitate are typically given prophylactically during
A blood transfusion can loosely be defined as ECLS even though there have been no prospective
administration of blood or any blood component studies investigating their use.
to a recipient. Reasons to transfuse include restora- Currently, the most common strategy adopted
tion of oxygen carrying capacity, maintenance of by hospitals worldwide is to transfuse blood derived
hemostatic balance, and/or management of hemor- from the separation of whole blood into multiple
rhagic complications. Another indication specific to components (RBC, platelets, FFP, cryoprecipitate)
extracorporeal applications includes priming of the via centrifugation after collection from a single do-
circuit at initiation. nor. While platelets can be derived from this process,
Patients on ECLS have some of the highest most platelet units currently used for transfusion are
transfusion needs of any hospitalized patient for obtained by apheresis. Regardless of the collection
many reasons including poor production and/or method, each product must meet stringent regulatory
consumption due to underlying illness, activation requirements, be stored in controlled environments,
induced by the circuit, frequent phlebotomy for and have well-defined expiration dates (Table 5-1).
laboratory testing, and higher transfusion thresholds. After being passé for several decades, there is re-
Product Description Storage Expiration
Temp (additive solution)
Unseparated blood that contains 21 days (CPD)
RBC, platelets, plasma and possibly
Whole Blood 1-6ºC
WBC (if not filtered) from one 35 days (CPD-A)
donor.
RBC separated from plasma and 21 days (CPD)
Packed Red platelets, suspended in a 35 days (CPD-A)
1-6ºC
Blood Cells preservative solution. Used to
42 days (AS-1, 3, or 5)
increase oxygen-carrying capacity.
24h after thawing when
The aqueous part of blood stored at 1-6ºC; can be
remaining after removal of RBC, converted to thawed
Plasma ≤18ºC
WBC, and platelets. Used mostly to plasma for up to 5 days
replete clotting factors after thaw when stored
at 1-6ºC
The precipitate that falls out of 4h after thawing when
solution when plasma is thawed at pooled, 6 hr after
Cryoprecipitate ≤18ºC
1-6ºC. Used mostly to replete thawing for a single
fibrinogen. unit
Platelets suspended in plasma from
one donor. One apheresis-derived
Platelets 20-24ºC 5-7 days
unit is equivalent to 4-6 platelet
concentrates.
Granulocytes suspended in RBC and
plasma from one donor. Indicated
Granulocytes 20-24ºC 24hr
for treatment of refractory infection
in a severely neutropenic patient.
RBC=red blood cell(s); WBC=white blood cell(s);hr=hour

Table 5-1. Description, storage temperature, and expiration date for blood components.

43
Chapter 5

newed enthusiasm about transfusing whole blood, fetus who inherits the RhD antigen from the father.
the starting point for blood product manufacturing. Transfusion services make reasonable attempts to
This chapter will briefly describe the various blood match RhD for all recipients but priority is always
components, transfusion practices as they apply to given to females who are of childbearing potential.
adult and pediatric ECLS, complications of transfu- When a “type and screen” is ordered, the
sion, and alternatives to transfusion. “type” instructs the hospital transfusion service to
determine the patient’s ABO and RhD type and
Blood Group Review the “screen” instructs them to perform an antibody
screen to identify any non-ABO antibodies such
Hundreds of blood groups have been identified as those formed against the blood groups listed in
within 36 blood systems (Table 5-2).1 However, not Table 5-2.
all are considered clinically significant as they are
not implicated in hemolytic transfusion reactions Red Blood Cell Transfusion
or hemolytic disease of the newborn. This chapter
reviews only ABO and RhD blood groups because The factors list above (low production due to
they are the most relevant. ABO is the most im- underlying disease, frequent phlebotomy, bleeding,
portant blood group for transfusion because, by 6 sheering from circuit pressure gradients, and in vitro
months of age, nearly all individuals have naturally hemolysis) contribute to anemia in ECLS patients
occurring antibodies against ABO which are capable and RBC transfusion requirements are high. A RBC
of hemolyzing incompatible red cells. The RBC sur- unit is composed of approximately 130-240 mL
face contains A antigens (type A), B antigens (type of packed RBC, 50-80 grams of hemoglobin, and
B), both (type AB), or neither (type O) (Table 5-3). 150-250 mg of iron.2 The total volume of a RBC
As a result, Group O donors are the universal ABO unit ranges from 225 to 350 mL with a hematocrit
donor for RBC because they neither express the A typically between 55 and 65%, depending on the
or B antigen. Furthermore, a person will only make anticoagulant preservative solution used. Citrate-
antibodies to ABO antigens they lack. Therefore, phosphate-dextrose (CPD) serves as the base anti-
Group AB donors are the universal ABO donor for coagulant and preservative solution. Adenine can be
all plasma-based products (FFP, platelets, and cryo- added (CPD-A) to extend the shelf life. Proprietary
precipitate), because their plasma does not contain additive solutions (AS-1, -3, or -5) have the longest
anti-A or anti-B. shelf life and represent the largest inventory of RBC
In contrast, Rh is a blood groups system for products in a hospital transfusion service (Table 5-1).
which antibodies are only formed once an individual Unless otherwise stated, most transfusion ser-
is exposed through transfusion, transplant, or preg- vices issue units that are closest to their expiration
nancy. The Rh system contains over 50 antigens date as not to waste an expensive, limited resource.
but RhD is the most immunogenic. If a female who Throughout storage, RBC undergo metabolic and
lacks RhD (“Rh-negative”) is exposed to RhD posi- structural changes often referred to as the “storage
tive RBC and makes an anti-D, these antibodies can lesion,” which includes development of a pH less
cross the placenta and cause hemolysis in any future than 6.5, increased lactate concentrations, depletion
of 2,3-diphosphoglycerate, impairment of sodium
ABO (ABO) Lewis (LE) Scianna (SC) Kx (XK) Raph (RAPH) FORS and potassium exchange, and alteration to the RBC
MNS (MNS) Duffy (FY) Dombrock (DO) Gerbich John Milton
(FORS)
JR (JR) membrane.3 Several large, randomized, controlled
(GE) Hagen (JMH)
P1PK (P1PK) Kidd (JK) Colton (CO) Cromer I (I) LAN
(CROM) (LAN)
Rh (RH) Diego (DI) Landsteiner- Knops (KN) Globoside Vel (VEL) Patient ABO Antigen(s) Antibodies in Compatible Compatible
Wiener (LW) (GLOB) Blood Type on RBC Surface Plasma RBC Plasma Products
Lutheran (LU) Yt (YT) Chido/Rogers Indian (IN) Gill (GIL) CD59 A A Anti-B A, O A, AB
(CH/RG) (CD59)
B B Anti-A B, O B, AB
Kell (KEL) Xg (XG) H (H) Ok (OK) Rh-associated Augustine
glycoprotein (AUG)
AB A, B None AB, A, B, O AB
(RHAG) O None Anti-A, Anti-B O O, A, B, AB
RBC=Red Blood Cells
Table 5-2. International Society of Blood Transfu-
sion defined blood group systems (system symbol). Table 5-3. ABO compatibility.

44
 Transfusion Management during Extracorporeal Life Support

clinical trials have failed to demonstrate a differ- Since RBCs are a cellular product, it can be
ence in clinical outcomes following transfusion of modified to best suit the patient’s needs. Modifica-
fresh versus old RBC across a variety of clinical tions can be determined by the transfusion service,
settings.4-8 Each hospital should have a policy or pro- hospital policy, or by the ordering provider. All of
cedure describing when fresh RBC units are issued. the modifications in Table 5-4 apply to RBC units.
Well-established indications for fresh units include RBC transfusions are indicated for decreased
intrauterine transfusions and high volume transfu- oxygen carrying capacity and/or tissue hypoxia
sions to a neonate. Patients on ECLS may or may as a result of inadequate circulating red cell mass
not be an indication depending on the institution. or ongoing hemorrhage. Importantly, hemoglobin
concentration alone is a poor measure of circulating

Modification Process End Goal Possible Indication

Leukocyte reduction Pre-storage filtration Reduces: Recurrent febrile non-hemolytic
(synonyms: “LR”, of cellular products to - CMV transfusion reactions
leukodepleted, insure < 5 x 106 - HLA alloimmunization Prevention of primary HLA
leukofiltered, CMV- WBC/unit. Some - Febrile transfusion reactions alloimmuization
safe) blood banks only Decrease transfusion-transmitted
issue leukocyte CMV infections in
reduced cellular units, immunocompromised patients
while others issue it
at discretion of
ordering provider
CMV-Negative Specific donors are Immunocompromised recipients Immunocompromised patients
tested and negative cannot contract CMV from a donor including:
for both IgG and IgM if they have never had CMV - Fetuses and young children
antibodies to CMV - Congenital or acquired
immunodeficiency
Irradiation Breaks DNA of Leukocytes are unable to - HSCT patients
leukocytes proliferate, thereby preventing - Patients with a hematological
transfusion-associated GVHD malignancy
- Intrauterine transfusions
- Related directed donors
- Populations with little HLA
diversity
- Congenital immunodeficiency
Term Neonate
Volume reduced Cellular products are To either: - Patient at risk for circulatory
spun and supernatant (1) Decrease the overall volume of overload
is removed unit - Incompatible plasma
(2) Reduce amount of a harmful - Recurrent, moderate allergic
substance in the supernatant transfusion reactions
Washed Cellular products are To remove a harmful substance - IgA deficient recipients
spun, supernatant - Recurrent severe allergic
removed, transfusion reactions
resuspended in saline - Removal of potassium
or albumin and
repeated several
times
Pathogen reduction In general, to induce Prevent pathogen survival and/or - All blood products given to any
damage to a vital replication without altering the population can be treated with
aspect of viruses and blood product pathogen reduction
bacteria, such as
lipids or DNA
Hemoglobin S negative SickleDex testing Donors are confirmed not to carry - Sickle cell patients
sickle cell trait - Pediatric cardiopulmonary bypass
patients
WBC=white blood cell; CMV=Cytomegalovirus; HLA=human leukocyte antigen;
DNA= deoxyribonucleic acid; GVHD=graft vs host disease; HSCT=hematopoietic stem cell transplant

Table 5-4. Available blood component modifications.

45
Chapter 5

RBC mass because of the physiologic compensatory ECLS-induced platelet dysfunction may lead to a
mechanisms that preserve oxygen transport such propensity for thrombosis.
as reduced blood viscosity to increase blood flow Like RBC, platelet products are manufactured
to tissues, redistribution of blood flow, increased from whole blood (also known as, whole blood
unloading of oxygen to tissues, and maintenance of derived platelets, random donor platelets, platelet
blood volume due to expansion of plasma volume. concentrates) or apheresis (apheresis platelets,
Because of these concepts and observations made in single donor platelets, plateletpheresis) donations.
Jehovah’s Witness patients who decline transfusion Four to six units of whole blood-derived platelets
based on religious beliefs, and in underdeveloped must be pooled to equal an apheresis platelet. To
countries where RBCs were unavailable or limited, maximize their effectiveness, platelets must be
readjustment of transfusion practice to a lower he- stored at room temperature. However, this increases
moglobin threshold has been investigated.9-14 In sev- the risk of bacterial growth and therefore platelets
eral large randomized, controlled trials conducted expire after 5-7 days. In otherwise healthy people
to investigate restrictive versus liberal transfusion with thrombocytopenia, increased bleeding is not
thresholds in adult and pediatric patients and in a typically evident until platelet counts fall below 5-10 
recent metaanalysis, no difference was detected in x 109/L.32 Except for Hgb S negative, all modifica-
mortality or other adverse events when a restrictive tions described in Table 5-4 can apply to platelets.
transfusion strategy was utilized.15-21 To date, no pro- The donor plasma type should be ABO compatible
spective studies investigating hemoglobin thresh- with the recipient’s RBC to avoid hemolysis (Table
olds in patients receiving extracorporeal support 5-3), but compatibility testing is unnecessary for
have been performed; however, two retrospective platelet transfusions. Transfusion of 4-6 units of
studies have demonstrated a reduction in RBC trans- pooled whole-blood derived platelets or one unit
fusion using a restrictive transfusion threshold.22,23 of apheresis platelets to an adult or 10 mL/kg of
Current ELSO guidelines recommend maintenance platelets to a pediatric patient should increase the
of the hematocrit over 40% to optimize oxygen de- platelet count by approximately 30,000/uL in the
livery while allowing the lowest reasonable blood absence of bleeding, clotting, splenomegaly, dis-
flow. While many centers target this transfusion seminated intravascular coagulopathy, or immuno-
threshold, there is a lack of evidence at the current logic refractoriness.
time to guide transfusion thresholds in ECLS and
each case should be considered individually. Plasma Transfusion

Platelet Transfusion Exposure of patient blood to the artificial

surface of an ECLS circuit results in initiation of
Platelets acquire both quantitative and qualita- coagulation, consumption of clotting factors and
tive defects during ECLS. Specifically, the under- deposition of fibrinogen throughout the circuit, par-
lying illness, sheer stress, and indwelling cannulas ticularly in areas of slow or turbulent flow. To offset
result in exposed collagen and activation of the this propensity for thrombosis, an anticoagulant
clotting cascade. Subsequently, platelets bind to (typically unfractionated heparin) is given which can
fibrinogen and platelet counts fall to less than 40% contribute to activation and clearance of antithrom-
of normal within the first few hours of ECLS.24 bin and other clotting factors. Simultaneously, the
Platelets may not aggregate properly during ECLS, natural fibrinolytic pathway attempts to lyse clots in
so taken all together, ECLS-induced platelet dys- an attempt to balance hemostasis. Furthermore, all
function may lead to a propensity to bleed.25-29 clotting factors are made in the liver so underlying
On the other hand, heparin induced throm- liver disease can contribute to poor production.33-40
bocytopenia (HIT) is a described complication of Similar to RBC and platelet products, plasma
ECLS.30-31 In HIT, platelets are activated by an IgG can be manufactured from either whole blood or
antibody against PF4-heparin complex and therefore apheresis donations. Although fresh frozen plasma
(FFP) is a term commonly used throughout the

46
 Transfusion Management during Extracorporeal Life Support

hospital to describe all plasma products, it is just temperatures, which forms a precipitate that is
one of several types of plasma. All plasma products collected, suspended in a small amount of plasma
are considered clinically equivalent but differ in the and then refrozen. Once thawed, cryoprecipitate
collection method (whole blood vs. apheresis), the must be stored at room temperature or else it will
time from collection to when it is frozen (e.g. never reprecipitate, and therefore expires 4-6 hours after
frozen, 8 hours, 24 hours), and how long each can thawing depending if pooled or not. One unit of
be thawed and stored at refrigeration temperatures cryoprecipitate should contain at least 150 mg of
(e.g. never frozen, 24 hours, 5 days).2 The latter two fibrinogen and 80 IU of factor VIII suspended in
factors largely determine the subtle differences in 5-20 mL of plasma.2 Typically 5-6 units are pooled
the concentration of procoagulant and anticoagulant together to make one adult dose. Transfusion of one
factors that are functional at time of transfusion. At a unit of cryoprecipitate per 10 kg of body weight will
dose of 10-15 mL/kg, clotting factor activity should typically increase the fibrinogen concentration by
increase by approximately 30% in the absence of 50-75 mg/dL and 5-6 units pooled together will raise
activation or loss. the fibrinogen in a 70-kg person by approximately
As with platelets, donor plasma should be com- 35 mg/dL.
patible with recipient RBC to avoid hemolysis and
as such, Group AB plasma is the universal plasma Variability of Transfusion Practice in ECLS
donor. In an attempt to decrease rates of transfusion
related acute lung injury (TRALI), female plasma Transfusion practices vary widely among ECLS
donors are now discouraged in many developed centers and are not currently evidence based. Rather,
countries, thereby making the availability of Group they are based on clinical experience, historical
AB plasma even more limited. Therefore, many literature, and extrapolation from other patient
transfusion services are using Group A plasma populations.43,44 While each ECLS center’s protocol
from donors who have a low concentration of anti- varies in how and how often to monitor and when
B as their routine, emergency issue plasma. Even and how to transfuse, ECLS centers invariably have
with low levels of anti-B present, Group A plasma a policy or procedure that describes their transfu-
is considered safe in limited amounts even when sion approach for a typical patient. For example,
transfused to Group B recipients.41 the hematocrit at which RBC are transfused to a
The most common indications for plasma typical ECLS patient ranges from 25 to 40% from
transfusion include reversal of coagulopathy in two surveys.43,45 The platelet count threshold for
bleeding patients or for prophylactic use in patients platelet transfusion in an otherwise uncomplicated
undergoing procedures; however, there is a paucity ECLS run ranges from 50,000 to 200,000/µL for
of well-designed studies to define plasma transfu- pediatric only programs, as compared to 20,000-
sion practice.42 Besides hemostatic factors, plasma 100,000/µ for adult only and mixed adult/pediatric
also contains hundreds of other molecules, many programs. 43 Fibrinogen threshold to transfuse
of which survive freezing and thawing and might plasma or cryoprecipitate ranges from 50 to 200 mg/
be active at time of transfusion. It is unknown what dL. The ELSO website contains general guidelines
other biologic effects plasma transfusion may have, regarding optimal replacement of blood products.44
especially in the setting of extracorporeal support.
Complications of Transfusion
Cryoprecipitate Transfusion
To ensure blood transfusion safety, the Food
As described above, ECLS contributes to activa- and Drug Administration and AABB (formerly
tion of clotting factors and deposition of fibrinogen. known as American Association of Blood Banks)
Cryoprecipitate is rich in fibrinogen, factor VIII, have established strict criteria for donation and
factor XIII, von Willebrand factor (VWF), and fibro- testing of donated blood including screening for
nectin, and is typically given to replete fibrinogen. potentially transfusion transmitted infections (TTI).
It is manufactured by thawing FFP at refrigerated In the United States, rates for human immunode-

47
Chapter 5

ficiency virus (HIV) and hepatitis C virus (HCV) rare, transfusion reactions can be fatal. As a result,
are less than 1 in 1 million.46 Infectious risks of clinicians should be aware of the risks, signs and
transfusion are at an all-time low, and therefore symptoms, and management of transfusion related
noninfectious serious hazards of transfusion have adverse events. If a transfusion reaction is suspected,
become the primary transfusion complications. The the transfusion should be stopped immediately, the
rate of these noninfectious adverse reactions, com- patient assessed by a provider, and the reaction
monly termed “transfusion reactions,” is 660 per should be reported to the transfusion service. A
100,000 individuals in an international registry.47 summary of the prevalence, signs and symptoms,
In the United States, 239.5 adverse reactions were and management of transfusion adverse reactions
reported per 100,000 units transfused. Although is presented in Table 5-5.

Prevalence (per
Reaction 100,000 units Signs and Symptoms Management
transfused)60
Allergic transfusion reaction 112.2 Urticaria, rash, skin itching, swelling Antihistamines
(throat, eye, etc.)
Anaphylactic transfusion 8 Bronchospasm, dyspnea, angioedema, Epinephrine, Corticosteroids,
reaction hypotension, tachycardia Antihistamines, Fluid Bolus
Acute hemolytic transfusion 2.5-7.0 Fever, chills, dyspnea, hypotension, Symptomatic treatment,
reaction tachycardia, back pain, nausea, Diuretics and fluid
vomiting, oliguria/anuria, administration
hemoglobinuria, positive DAT For future transfusion
antigen negative RBC will be
provided
Delayed hemolytic 40 Occurs 2-14 days after transfusion; Symptomatic treatment
transfusion reaction jaundice, anemia, elevated bilirubin,
reticulocytosis, spherocytosis, For future transfusion
increased LDH, positive antibody antigen negative RBC will be
screen, positive DAT provided
Delayed serologic 48.9-75.7 Occurs 2-14 days after transfusion; Symptomatic treatment
transfusion reaction positive antibody screen, positive DAT For future transfusion
antigen negative RBC will be
provided
Febrile non-hemolytic 1000-3000 Occurs within 4 hours of transfusion; Antipyretic or close
transfusion reaction temperature of 38oC and/or increase of observation
1oC from pre-transfusion value with or
without chills and rigors
Post-transfusion purpura Unknown (varies by Occurs 2-14 days after transfusion; Self-limiting, IVIG with or
component) severe thrombocytopenia, petechiae, without corticosteroids
purpura, identification of platelet For future transfusion
antibodies antigen negative platelets
will be provided
Septic transfusion reaction 0.03-3.3 Fever, chills, hypotension, tachycardia Antipyretic, empiric,
antibiotics

Culture blood product

Transfusion associated 10.9 Occurs within 2 hours of transfusion; Diuretic administration,
circulatory overload dyspnea, tachycardia, hypertension, reduce fluid intake
headache, jugular venous distension
Transfusion associated graft Extremely rare Occurs 1-4 weeks after transfusion; Supportive care, no treatment
versus host disease pancytopenia, maculopapular rash, except bone marrow
vomiting, diarrhea transplantation
Transfusion related acute 0.4-1.0 (varies by Occurs within 6 hours of transfusion; Symptomatic treatment,
lung injury component) dyspnea, hypoxemia, fever, Intubation and mechanical
hypotension, copious frothy pink- ventilation as required
tinged fluid from endotracheal tube,
bilateral pulmonary infiltrates
DAT=direct antiglobulin test; RBC= red blood cell; IVIG= intravenous immune globulin
Table 5-5. Adverse reactions to transfusion.

48
 Transfusion Management during Extracorporeal Life Support

Adjunctive Therapy to Blood Product Transfu- reported use of rFVIIa to help manage hemorrhage
sion in ECLS patients.43

At present, there are no commercially available Recombinant Factor VIIa

blood substitutes, limiting alternatives to blood
product transfusion. Recombinant factor VIIa (rF- No randomized controlled trial in ECLS patients
VIIa), prothrombin complex concentrates (PCC), has been conducted, but rFVIIa in doses ranging
and fibrinogen concentrate have been used off-label from 30 to 171 mg/kg have been used during ECLS
in ECLS and are summarized in Table 5-6. In addi- and outcomes described in case reports and case
tion, tranexamic acid (TXA) and e-aminocaproic series. Off-label use of rFVIIa during ECLS has
acid, medications that prevent the breakdown of been shown to reduce blood loss, most commonly
fibrin clots, have also been used during ECLS. De- as measured by chest tube output, and transfusion
spite limited evidence, 63/94 (67%) ELSO centers volumes for a defined period of time post-admin-

Medication Mechanism of Action On-Label Use Results of Off-Label Use in ECLS

Recombinant factor VIIa • When complexed with • Treatment of bleeding episodes • Reduced blood loss, most
(NovoSeven®RT, Novo Nordisk, tissue factor activates factor and perioperative management commonly as measured by chest
Bagsværd, Denmark) X to factor Xa, as well as, in adults and children with tube output, and reduced
factor IX to factor IXa hemophilia A or B with transfusion volumes for a
• Tenase complex converts inhibitors defined period of time post-
prothrombin to thrombin • Congenital factor VII administration48-53
• Thrombin generation occurs deficiency • Fatal thrombosis has been
on the activated platelet • Glanzmann’s thrombasthenia reported54-57
with or without anti-platelet
antibodies who are refractory to
platelet transfusion
• Treatment of bleeding episodes
and perioperative management
in adults with acquired
hemophilia
Activated PCC • Contains non-activated • Control or prevention of • Efficacy not established
(FEIBA, Baxalta, Westlake Village, factors II, IX, and X, and bleeding, perioperative • Fatal thrombosis has been
California) activated factor VII management, or routine reported54
prophylaxis to prevent or
reduce the frequency of
bleeding in patients with
hemophilia A or B
3F-PCC • Contain nonactivated factors • Prevention and control of • Efficacy and safety not
Bebulin® (Baxalta, Westlake Village, II, IX, X, and low levels of bleeding episodes in patients established
California) FVII (typically not in excess with hemophilia B
of 35 IU/mL)
Profilnine® (Grifols Biologicals Inc.,
Los Angeles, California)
4F-PCC • Contain vitamin K • Urgent reversal of acquired • Efficacy and safety not
Kcentra® or Beriplex®P/N (CSL dependent coagulation coagulation factor deficiency established
Behring GmbH, Marburg, Germany) factors, II, VII, IX, X and induced by VKA therapy in
proteins C and S adult patients with acute major
Octaplex® (Octapharma, Lachen , bleeding or need for an urgent
Switzerland), surgery or invasive procedure
• Treatment of bleeding and
perioperative prophylaxis of
bleeding in congenital
deficiency of any of the vitamin
K dependent coagulation
factors only if a purified
specific coagulation factor
product is unavailable, in
addition to, in cardiac surgery,
trauma, and liver disease
Fibrinogen concentrate • Contains plasma-derived • Treatment of acute bleeding in • Efficacy and safety not
RiaSTAP® or Fibryga® (CSL human fibrinogen patient with congenital established
Behring GmbH, Marburg Germany) fibrinogen deficiency • When given as part of a
prophylactic coagulation
protocol, may improve incidence
and severity of intracranial
hemorrhage61

Table 5-6. Adjunctive therapy to blood product transfusion.

49
Chapter 5

istration.48-53 In clinical trials, thrombotic adverse Antifibrinolytic Therapy

events occurred in 0.2% of patients treated with
rFVIIa with hemophilia A or B and 4% of patients Antifibrinolytic therapy has been shown to
with acquired hemophilia; the risk of thromboem- reduce bleeding associated with activation and dys-
bolic complications in patients receiving rFVIIa for regulation of fibrinolysis in major surgery (cardiac,
off-label use is unknown. Moreover, the thrombo- liver, neurosurgery, and obstetric hemorrhage) and
embolic risk of rFVIIa administration has not been trauma. The synthetic lysine analogues, tranexamic
established in the setting of ECLS; however, case acid (TXA) and ε-aminocaproic acid, are the most
reports of fatal thrombosis after administration of widely used antifibrinolytic agents. Both drugs block
rFVIIa have been published.54-57 As a result, care- the lysine binding sites of plasminogen, preventing
ful consideration must be given to administration activation to plasmin and lysis of polymerized fi-
of rFVIIa in ECLS patients who have multiple risk brin. Data to support routine use of antifibrinolytic
factors for thrombosis. therapy during ECLS are lacking; however, prophy-
lactic perioperative use or as an adjunct therapy to
Prothrombin Complex Concentrates blood product and coagulation factor therapy may
be successful to reduce bleeding.58 While the risk
Prothrombin Complex concentrates contain of thrombosis may not be significantly increased,
vitamin K-dependent clotting factors, II (prothrom- fatal thrombosis has been reported in a neonate with
bin), VII, IX and X. Two types of prothrombin congenital diaphragmatic hernia (CDH) treated with
complex concentrates are currently available for use, ε-aminocaproic acid.59 As a result, larger studies
nonactivated (PCC) and activated (aPCC). Similar are needed to determine the efficacy and safety of
to rFVIIa, thromboembolic adverse events have antifibrinolytic therapy during ECLS.
been reported with aPCC especially at high doses
or in patients with thrombotic risk factors. To date,
the efficacy of PCC or aPCC in patients receiving
ECLS has not been established and there has been
a single case report of massive intracardiac and
circuit thromboses in a 56 year old man placed on
ECMO support following lung retransplantation
who received two doses of rFVIIa followed by aPCC
for refractory bleeding.54
Nonactivated PCCs are classified as either 3
factor PCC (3F-PCC) or 4 factor PCC (4-PCC)
based on the concentration of FVII. While 3F-
PCC has been used off-label mainly for vitamin
K antagonist (VKA) reversal, FVII concentrations
had to be supplemented with either FFP or rFVIIa
and their use has mostly been replaced by 4F-PCC.
Currently, there are two available 4F-PCCs. How-
ever, the efficacy and safety of PCC in the setting
of extracorporeal support has not been established.
As a result, careful consideration must be given to
administration of either aPCC or PCC in patients
requiring extracorporeal support as these patients
have multiple risk factors for thrombosis until
the efficacy and safety can be established in this
population.

50
 Transfusion Management during Extracorporeal Life Support

References 13. Lackritz EM, Hightower AW, Zucker JR, et

al. Longitudinal evaluation of severely anemic
1. ISBT Blood Group Systems, v5. Feb 2017; children in Kenya: the effect of transfusion
http://www.isbtweb.org/fileadmin/user_upload/ on mortality and hematologic recovery. AIDS.
Working_parties/WP_on_Red_Cell_Immu- 1997;11(12):1487-1494.
nogenetics_and/Table_of_blood_group_sys- 14. Shander A, Javidroozi M, Naqvi S, et al. An
tems_v5.0_170205.pdf. Accessed 11/7/2017. update on mortality and morbidity in patients
2. Circular of Information For The Use of Human with very low postoperative hemoglobin levels
Blood and Blood Components. 2013. http:// who decline blood transfusion (CME). Transfu-
www.aabb.org/tm/coi/Documents/coi1113.pdf. sion. 2014;54(10 Pt 2):2688-2695.
3. Orlov D, Karkouti K. The pathophysiology and 15. Carson JL, Terrin ML, Noveck H, et al. Lib-
consequences of red blood cell storage. Anaes- eral or restrictive transfusion in high-risk
thesia. 2015;70 Suppl 1:29-37, e29-12. patients after hip surgery. N Engl J Med.
4. Fergusson DA, Hebert P, Hogan DL, et al. Effect 2011;365(26):2453-2462.
of fresh red blood cell transfusions on clinical 16. Hajjar LA, Vincent JL, Galas FR, et al. Trans-
outcomes in premature, very low-birth-weight fusion requirements after cardiac surgery: the
infants: the ARIPI randomized trial. JAMA. TRACS randomized controlled trial. JAMA.
2012;308(14):1443-1451. 2010;304(14):1559-1567.
5. Heddle NM, Cook RJ, Arnold DM, et al. Effect 17. Hebert PC, Wells G, Blajchman MA, et al. A
of Short-Term vs. Long-Term Blood Storage on multicenter, randomized, controlled clinical
Mortality after Transfusion. The N Engl J Med. trial of transfusion requirements in critical care.
2016;375(20):1937-1945. Transfusion Requirements in Critical Care In-
6. Lacroix J, Hebert PC, Fergusson DA, et al. Age vestigators, Canadian Critical Care Trials Group.
of transfused blood in critically ill adults. N N Engl J Med. 1999;340(6):409-417.
Engl J Med. 2015;372(15):1410-1418. 18. Holst LB, Haase N, Wetterslev J, et al. Low-
7. Steiner ME, Ness PM, Assmann SF, et al. Ef- er versus higher hemoglobin threshold for
fects of red-cell storage duration on patients transfusion in septic shock. N Engl J Med.
undergoing cardiac surgery. N Engl J Med. 2014;371(15):1381-1391.
2015;372(15):1419-1429. 19. Lacroix J, Hebert PC, Hutchison JS, et al.
8. Cooper DJ, McQuilten ZK, Nichol A, et al. Transfusion strategies for patients in pe-
Age of Red Cells for Transfusion and Out- diatric intensive care units. N Engl J Med.
comes in Critically Ill Adults. N Engl J Med. 2007;356(16):1609-1619.
2017;377(19):1858-1867. 20. Villanueva C, Colomo A, Bosch A, et al. Trans-
9. Carson JL, Noveck H, Berlin JA, Gould SA. fusion strategies for acute upper gastrointestinal
Mortality and morbidity in patients with very bleeding. N Engl J Med. 2013;368(1):11-21.
low postoperative Hb levels who decline blood 21. Holst LB, Petersen MW, Haase N, Perner A,
transfusion. Transfusion. 2002;42(7):812-818. Wetterslev J. Restrictive versus liberal trans-
10. Carson JL, Patel MS. Red blood cell transfusion fusion strategy for red blood cell transfusion:
thresholds: can we go even lower? Transfusion. systematic review of randomised trials with
2014;54(10 Pt 2):2593-2594. meta-analysis and trial sequential analysis. BMJ.
11. English M, Ahmed M, Ngando C, Berkley J, 2015;350:h1354.
Ross A. Blood transfusion for severe anae- 22. Cahill CM, Blumberg N, Schmidt AE, et al.
mia in children in a Kenyan hospital. Lancet. Implementation of a Standardized Transfusion
2002;359(9305):494-495. Protocol for Cardiac Patients Treated With
12. Lackritz EM, Campbell CC, Ruebush TK, 2nd, Venoarterial Extracorporeal Membrane Oxy-
et al. Effect of blood transfusion on survival genation Is Associated With Decreased Blood
among children in a Kenyan hospital. Lancet. Component Utilization and May Improve Clini-
1992;340(8818):524-528. cal Outcome. Anesth Analg. 2017.

51
Chapter 5

23. Sawyer AA, Wise L, Ghosh S, Bhatia J, Stans- sion policy for patients with acute leukaemia.
field BK. Comparison of transfusion thresholds Lancet. 1991;338(8777):1223-1226.
during neonatal extracorporeal membrane oxy- 33. Annich GM. Extracorporeal life support: the
genation. Transfusion. 2017;57(9):2115-2120. precarious balance of hemostasis. J Thromb
24. Robinson TM, Kickler TS, Walker LK, Ness Haemost. 2015;13 Suppl 1:S336-342.
P, Bell W. Effect of extracorporeal membrane 34. Urlesberger B, Zobel G, Zenz W, et al. Activa-
oxygenation on platelets in newborns. Crit Care tion of the clotting system during extracorporeal
Med. 1993;21(7):1029-1034. membrane oxygenation in term newborn infants.
25. Nair P, Hoechter DJ, Buscher H, et al. Pro- J Pediatr. 1996;129(2):264-268.
spective observational study of hemostatic 35. Arnold P, Jackson S, Wallis J, Smith J, Bolton
alterations during adult extracorporeal mem- D, Haynes S. Coagulation factor activity during
brane oxygenation (ECMO) using point-of-care neonatal extra-corporeal membrane oxygen-
thromboelastometry and platelet aggregometry. ation. Intensive Care Med. 2001;27(8):1395-
J Cardiothorac Vasc Anesth. 2015;29(2):288- 1400.
296. 36. McManus ML, Kevy SV, Bower LK, Hickey PR.
26. Cheung PY, Sawicki G, Salas E, Etches PC, Coagulation factor deficiencies during initiation
Schulz R, Radomski MW. The mechanisms of extracorporeal membrane oxygenation. J
of platelet dysfunction during extracorporeal Pediatr. 1995;126(6):900-904.
membrane oxygenation in critically ill neonates. 37. Zavadil DP, Stammers AH, Willett LD, Deptula
Crit Care Med. 2000;28(7):2584-2590. JJ, Christensen KA, Sydzyik RT. Hematological
27. Mutlak H, Reyher C, Meybohm P, et al. Multiple abnormalities in neonatal patients treated with
electrode aggregometry for the assessment of extracorporeal membrane oxygenation (ECMO).
acquired platelet dysfunctions during extra- J Extra Corpor Technol. 1998;30(2):83-90.
corporeal circulation. Thorac Cardiovasc Surg. 38. Hundalani SG, Nguyen KT, Soundar E, et al.
2015;63(1):21-27. Age-based difference in activation markers of
28. Saini A, Hartman ME, Gage BF, et al. Incidence coagulation and fibrinolysis in extracorporeal
of Platelet Dysfunction by Thromboelastogra- membrane oxygenation. Pediatr Crit Care Med.
phy-Platelet Mapping in Children Supported 2014;15(5):e198-205.
with ECMO: A Pilot Retrospective Study. Front 39. Peek GJ, Firmin RK. The inflammatory and
Pediatr. 2015;3:116. coagulative response to prolonged extra-
29. Kestin AS, Valeri CR, Khuri SF, et al. The plate- corporeal membrane oxygenation. ASAIO.
let function defect of cardiopulmonary bypass. 1999;45(4):250-263.
Blood. 1993;82(1):107-117. 40. Gorbet MB, Sefton MV. Biomaterial-associated
30. Sokolovic M, Pratt AK, Vukicevic V, Sarumi M, thrombosis: roles of coagulation factors, com-
Johnson LS, Shah NS. Platelet Count Trends plement, platelets and leukocytes. Biomaterials.
and Prevalence of Heparin-Induced Throm- 2004;25(26):5681-5703.
bocytopenia in a Cohort of Extracorporeal 41. Dunbar NM, Yazer MH. Safety of the use of
Membrane Oxygenator Patients. Crit Care Med. group A plasma in trauma: the STAT study.
2016;44(11):e1031-e1037. Transfusion. 2017;57(8):1879-1884.
31. Pollak U, Yacobobich J, Tamary H, Dagan O, 42. Karam O, Tucci M, Combescure C, Lacroix J,
Manor-Shulman O. Heparin-induced throm- Rimensberger PC. Plasma transfusion strategies
bocytopenia and extracorporeal membrane for critically ill patients. Cochrane Database
oxygenation: a case report and review of the lit- Syst Rev. 2013(12):Cd010654.
erature. J Extra Corpor Technol. 2011;43(1):5- 43. Bembea MM, Annich G, Rycus P, Oldenburg
12. G, Berkowitz I, Pronovost P. Variability in
32. Gmur J, Burger J, Schanz U, Fehr J, Schaffner A. anticoagulation management of patients on
Safety of stringent prophylactic platelet transfu- extracorporeal membrane oxygenation: an

52
 Transfusion Management during Extracorporeal Life Support

international survey. Pediatr Crit Care Med. following open heart surgery. Pediatr Crit Care
2013;14(2):e77-84. Med. 2005;6(4):473-476.
44. Lequier L, Annich G, Al-Ibrahim O, et al. ELSO 54. Bui JD, Despotis GD, Trulock EP, Patterson
Anticoagulation Guideline. 2014; https://www. GA, Goodnough LT. Fatal thrombosis after
elso.org/Portals/0/Files/elsoanticoagulation- administration of activated prothrombin com-
guideline8-2014-table-contents.pdf. Accessed plex concentrates in a patient supported by
11/7/2017. extracorporeal membrane oxygenation who
45. Esper SA, Welsby IJ, Subramaniam K, et had received activated recombinant factor VII. J
al. Adult extracorporeal membrane oxygen- Thorac Cardiovasc Surg. 2002;124(4):852-854.
ation: an international survey of transfusion 55. Chalwin RP, Tiruvoipati R, Peek GJ. Fatal
and anticoagulation techniques. Vox sang. thrombosis with activated factor VII in a
2017;112(5):443-452. paediatric patient on extracorporeal mem-
46. Carson JL, Triulzi DJ, Ness PM. Indications for brane oxygenation. Eur J Cardiothorac Surg.
and Adverse Effects of Red-Cell Transfusion. N 2008;34(3):685-686.
Engl J Med. 2017;377(13):1261-1272. 56. Swaminathan M, Shaw AD, Greenfield RA,
47. Frazier SK, Higgins J, Bugajski A, Jones AR, Grichnik KP. Fatal Thrombosis After Factor
Brown MR. Adverse Reactions to Transfu- VII Administration During Extracorporeal
sion of Blood Products and Best Practices for Membrane Oxygenation. J Cardiothorac Vasc
Prevention. Crit Care Nurs Clin North Am. Anesth. 2008;22(2):259-260.
2017;29(3):271-290. 57. Syburra T, Lachat M, Genoni M, Wilhelm
48. Agarwal HS, Bennett JE, Churchwell KB, et al. MJ. Fatal outcome of recombinant factor
Recombinant factor seven therapy for postoper- VIIa in heart transplantation with extracorpo-
ative bleeding in neonatal and pediatric cardiac real membrane oxygenation. Ann Thorac Surg.
surgery. Ann Thorac Surg. 2007;84(1):161-168. 2010;89(5):1643-1645.
49. Long MT, Wagner D, Maslach-Hubbard A, 58. Buckley LF, Reardon DP, Camp PC, et al. Ami-
Pasko DA, Baldridge P, Annich GM. Safety and nocaproic acid for the management of bleeding
efficacy of recombinant activated factor VII for in patients on extracorporeal membrane oxygen-
refractory hemorrhage in pediatric patients on ation: Four adult case reports and a review of
extracorporeal membrane oxygenation: a single the literature. Heart Lung. 2016;45(3):232-236.
center review. Perfusion. 2014;29(2):163-170. 59. Hocker JR, Saving KL. Fatal aortic thrombosis
50. Niebler RA, Punzalan RC, Marchan M, Lankie- in a neonate during infusion of epsilon-amino-
wicz MW. Activated recombinant factor VII caproic acid. J Pediatr Surg. 1995;30(10):1490-
for refractory bleeding during extracorporeal 1492.
membrane oxygenation. Pediatr Crit Care Med.. 60. Delaney M, Wendel S, Bercovitz RS, et al.
2010;11(1):98-102. Transfusion reactions: prevention, diagnosis,
51. Veldman A, Neuhaeuser C, Akintuerk H, et al. and treatment. Lancet. 2016;388(10061):2825-
rFVIIa in the treatment of persistent hemor- 2836.
rhage in pediatric patients on ECMO following 61. Kalbhenn J, Wittau N, Schmutz A, Zieger B,
surgery for congenital heart disease. Paediatr Schmidt R. Identification of acquired coagula-
Anaesth. 2007;17(12):1176-1181. tion disorders and effects of target-controlled
52. Walker A, Davidson M, Chalmers E. Use of coagulation factor substitution on the incidence
activated recombinant factor VII in pediatric and severity of spontaneous intracranial bleed-
extracorporeal membrane oxygenation. Pediatr ing during veno-venous ECMO therapy. Perfu-
Crit Care Med. 2010;11(4):537-538. sion. 2015;30(8):675-682.
53. Wittenstein B, Ng C, Ravn H, Goldman A.
Recombinant factor VII for severe bleeding
during extracorporeal membrane oxygenation

53
6

Medication Management in ECLS

Christa Kirk, PharmD, Gail Annich, MD, Deborah Wagner, PharmD, E.D. Wildschut, MD, PhD

Introduction their concentrations fluctuate due to changes in

volume status. Highly lipophilic drugs distribute
Managing medication therapy in critically ill into tissues leaving lower concentrations in the
patients is challenging due to rapidly changing blood, increasing Vd. Lipophilicity, which can be
physiology associated with the severe underlying ill- predicted by using the partition coefficient or log P
ness. The addition of an extracorporeal circuit adds value of a medication, may also influence whether
another layer of complexity due to the interactions a medication has a high affinity for circuit surfaces.
between the biochemical properties of both patient Critical illness and ECLS can also alter the Vd of a
and circuit. Therefore, understanding how altera- drug and is discussed later.
tions in pharmacokinetics and pharmacodynamics
in ECLS patients can affect the dose-response re- Protein Binding
lationship of medications may create opportunities
to individualize dosing strategies that maximize Alterations in plasma proteins also affect the Vd
therapeutic effect while minimizing toxicities. of certain medications. Drugs that are significantly
bound to plasma proteins have a lower Vd. There-
Pharmacokinetics fore, significant changes in plasma protein concen-
trations due to sepsis, hypoproteinemia, and ECLS
Pharmacokinetics describe drug disposition may impact the bioavailability of protein bound
over time. The plasma concentration of a drug is medications. One must also consider the effect of
responsible for most of its effects as it determines the inflammation on various protein subtypes as well.
amount of uptake into the tissues. Pharmacokinetics
help predict drug concentrations within the plasma Clearance
allowing providers to select therapeutic drug doses
and dosing intervals as well as make adjustments Medication elimination depends upon Vd as
when needed. well as both renal and hepatic clearance. Organ
impairment is often associated with critical illness
Volume of Distribution and the addition of the ECLS circuit can lead to or
complicate renal or liver dysfunction. Both VV and
Understanding volume of distribution (Vd) VA-ECLS are associated with renal dysfunction and
is essential to designing medication regimens for misdistribution of blood flow to the liver, which
patients on ECLS. Vd represents the volume of can affect clearance as well.1 The addition of CRRT,
fluid needed to dilute a medication at steady state hemodialysis, or peritoneal dialysis also increases
to a desired concentration. Physiological character- clearance depending on the drug.
istics and specific biochemical properties of each
medication such as protein binding, lipophilicity,
and molecular weight affect Vd. In general, hy-
drophilic (water-soluble) drugs have a low Vd and

55
Chapter 6

Sequestration If doses are not adjusted to overcome sequestration,

treatment failure may ensue. Although data remain
The ECLS circuit alters patient-specific pharma- inconclusive, leaching of drugs from the tubing
cokinetics. Medications may react to both plastics back into the circulation may occur during weaning
and silicone in the ECLS circuit; however, data processes, a potential cause of toxicity.5,7
concerning drug sequestration within ECLS circuits
are limited. Early studies in membrane oxygenators Pharmacokinetic Changes in Critical Illness and
were done with polypropylene membranes and not ECLS
the current polymethylpentene products, which
affects the applicability of these data to current Severe illness and ECLS can have dramatic and,
circuits. Binding of drugs to circuit components somewhat unpredictable, effects on pharmacokinet-
depends upon the chemical characteristics of the ics. Figure 6-1 illustrates the complex changes in
drugs, design of the oxygenator, and type of priming pharmacokinetics associated with critical illness
fluid. More highly lipophilic medications gener- while Figure 6-2 demonstrates the additive pharma-
ally have a strong affinity for circuit components. cokinetic challenges created by the ECLS circuit.8
When studied in ECLS circuits, medications with Both Vd and clearance can be altered by the addition
higher log P values (lipophilic) were less likely to be of ECLS secondary to hypervolemia or sequestra-
recovered post-oxygenator, strongly suggesting se- tion of drugs due to protein binding. Hypervolemia
questration in the circuit.2-4 The large surface area of develops due to fluid retention secondary to inflam-
the silicone and plastic ECLS components increases mation as well as circuit volume, blood products,
sequestration which, in turn, can lead to decreased and bolus volume needed to maintain flows. Edema
bioavailability of the drugs as they become trapped and added fluid volume increases Vd, which is more
in the circuit.5 Polyvinyl chloride used in tubing prominent in smaller patients due to relative size.
often contains a plasticizer to maintain flexibility Substantial protein losses affect concentrations of
that increases adsorption of drugs to the tubing. In protein-bound medications and impacts Vd. Highly
addition to the physical properties of the tubing, both protein bound or hydrophilic medications are affect-
the flow rate and infusion time of the medication ed by fluid shifts associated with ECLS necessitating
independently contribute to the rate of adsorption.6 careful medication management. Therefore, changes

Figure 6-1. Pharmacokinetic changes in critical illness.10

56
 Medication Management in ECLS

to Vd or clearance impact the expected half-life of sequestration. Appendices I and II highlight the
the drug, which is the basis for determination of pharmacokinetic properties of commonly used
dosing and frequency. medications in critically ill patients and provide
recommendations for dosing adjustment. Finally,
Dosing Strategy renal or hepatic dysfunction also affects medication
elimination. Manufacturer-recommended dosage
Pharmacodynamics adjustments should be considered; however, poten-
tial alterations in pharmacokinetics created by the
Pharmacodynamics describes the relationship ECLS circuit as well as large interpatient variability
between the concentration of a drug and patient may necessitate clinically appropriate deviation
response. Table 6-1 highlights the pharmacokinetic from their guidance.
and pharmacodynamic changes associated with the
addition of the ECLS circuit and dosing strategies Assessment of Risk
for overcoming these obstacles. Utilizing pharma-
cokinetic properties of drugs and patient–specific Aggressive dosing with close surveillance for
factors allows for estimation in choosing an initial the emergence of adverse drug reactions should
dosing regimen. Steady state concentrations are direct management strategies. Whenever possible,
determined by the clearance of a drug and affect therapeutic drug monitoring should be used to
the maintenance dose. In general, standard dosing is ensure targeted concentrations are achieved while
appropriate for hydrophilic medications with lower avoiding toxicity. It is critical that the components
Vd while medications with higher Vd will be more of the ECLS system and patient specific factors be
dramatically affected by fluctuations in fluid status. taken into consideration to allow for appropriate
Higher loading doses are recommended for patients medication administration. Though there may never
with hypervolemia to achieve therapeutic levels be a perfect strategy for ECLS medication man-
more rapidly. An increase in protein binding and agement, continuous assessment of clinical status
lipophilicity will necessitate higher doses. Highly combined with the known biochemical properties of
lipophilic or protein bound medications should be drugs gives providers a guide for thoughtful initia-
avoided, when possible, to prevent prolonged under tion and adjustment of medications in the setting of
dosing. If unavoidable, significantly higher and life-threatening critical illness.
more frequent dosing may be required at initiation
and after circuit changes to decrease the effects of

Figure 6-2. Pharmacokinetic changes in ECLS.10

*Hydrophilic medications **Lipophilic and highly protein bound medications 57

Chapter 6

ECLS PK Physiologic PK Affected Therapeutic Implication

Change Changes Medications
Priming/ Hemodilution ↑ Vd Hydrophilic Consider loading dose.
Transfusions May need higher
maintenance doses.
Drug inactivation ↓ Bioavailability Lipophilic Consider loading dose.
by priming fluids Highly protein May need higher
bound maintenance doses.
Circuit factors Sequestration ↓ Bioavailability Lipophilic Consider loading dose.
(Effect more Highly protein May need higher
significant with bound maintenance doses.
newer circuits)
Patient factors Systemic ↑ Vd Hydrophilic Consider loading dose.
inflammation ↑ Clearance May need higher
Sepsis maintenance doses.
Consider increased
frequency or continuous
infusion.
Organ failure ↑ Vd Renally or Avoid agents cleared by
↓ Clearance hepatically affected organ system if
cleared possible. May need higher
maintenance doses but
decreased frequency of
administration
Dehydration ↓ Vd Renally or May require increased
hepatically frequency with standard
cleared dosing

Table 6-1. Pharmacodynamic Changes and Medication Dosing in ECLS (Adapted from:
Jefferis Kirk et al. ECLS 5th Ed)

58
 Medication Management in ECLS

Appendix I: Pharmacokinetic Properties of Common ECLS Medications and Dosing Strategy

Analgesic/Sedative Agents (IV)

Goal: Maintain sedation to ensure circuit safety while minimizing pain and potential toxicity
Dosing strategy: Consider standard starting doses except where indicated with increased bolus doses and
more substantial rate changes based on clinical response.

Medication Vd (L/kg)* Protein X Log P** Dosing Strategy

Low: ≤1
Binding Hydrophilic: <1
Mod: 1-5 Low: <30% Lipophilic: 1-2
Initiation Maintenance
High: >5 Mod: 30-70% Highly
High: >70% lipophilic: >2
Acetaminophen 1 Low 0.5 Standard Standard
Dexmedetomidine11 1.7-2.2 High 3.1 ↑ 50-100% ↑ 50-100%
Consider loading
dose 100 -
500mcg/kg/dose
if severe
agitation
Fentanyl12 4-6 High 4 Consider other
agents a
Hydromorphone13 2.9 Low 1.8 Standard ↑ 50-100%
Ketamine14 2.4 Moderate 2.2 Standard ↑ 50-100%
Lorazepam15 1.4 High 2.4 Standard ↑ 50-100%
Midazolam16,12 Adult: 1-3.1 High 3.9 Consider other
Ped: 1.24-2 agents a
Morphine16 Adult: 1-4.7 Low 0.8 Standard ↑ 25-50%
Infant: 2.8-5.5
Pentobarbital17 Adult: 1 High 2.1 Standard ↑25-50%
Ped: 0.8
Propofol16 Adult: up to 60 High 3.8 Consider other
Ped: 5-10 agents a
a If
necessary to use medication - significantly higher loading and maintenance doses will be needed to overcome sequestration
when initiating ECLS or at circuit change – consider reducing doses by 25-50% when coming off of ECLS and monitor patients
closely

59
Chapter 6

Antimicrobial Agents (IV)

Goal: Quickly attain appropriate concentrations to treat infection while minimizing toxicity.
Dosing strategy: Consider higher doses based on MIC data for more serious infections. Adjust frequency
based on clearance recognizing that more frequent dosing may be necessary. Use close therapeutic drug
monitoring when available.

Medication Vd (L/kg)* Protein X Log P** Dosing Strategy

Low: ≤1
Binding Hydrophilic: <1
Mod: 1-5 Low: <30% Lipophilic: 1-2
Initial Dose Clearance
High: >5 Mod: 30-70% Highly
High: >70% lipophilic: >2
Ampicillin 0.28 Low 0.67 Consider Renal
alternative agents
Ampicillin- Adult: 0.25 Moderate -1.1 / -1 Standard dosing Renal
sulbactam Ped: 0.35 (amp/sulb)
Azithromycin19 0.45 Moderate 4 Standard dosing Hepatic
Cefotaxime19 Standard dosing
Cefazolin8 0.2 High -0.4 Higher end of Renal
dosing/frequency
Cefepime Adult: 0.26 Moderate -0.1 Standard dosing Renal
Ped: 0.3
Infant: 0.5
Ceftaroline 0.29 Low 1.6 Standard dosing Renal
Ceftriaxone 0.4 High -1.3 Standard dosing Hepatic
Ciprofloxacin 1.2-2.7 Moderate 1.5 Standard dosing Renal
Gentamicin19 Adult: 0.3 Low -4.1 Standard dosing – Renal
Ped: 0.4 follow levels
Neonate: 0.5
Levofloxacin19 1.3-1.6 Moderate -0.4 Standard dosing Renal
Linezolid21 0.6 Moderate 0.5 Standard dosing Renal
MRSA MIC ≥1:
higher end of
dosing range19
Meropenem22 0.2-0.3 Low -2.4 Standard dosing Renal
MIC > 2: higher
end of dosing
range19
Neonate: consider
loading dose and
continuous
infusion for serious
infections23
Metronidazole 0.51-1.1 Low 0 Standard dosing Hepatic
Piperacillin- 0.15-0.22 Moderate 0.3 / -2 Standard dosing Renal
tazobactam19 (pip/tazo)
Polymyxin B Central: 0.09 High -2.5 Higher end of Renal
Peripheral/tissues: dosing range
0.33
Rifampin19 1.6 High 1.5 Higher end of Hepatic
dosing range

60
 Medication Management in ECLS

Tobramycin Adult: 0.2 Low -6.2 Standard dosing – Renal

Ped: 0.3 follow levels
Infant: 0.4
Trimethoprim- Adult:1.3 High 0.9/0.9 Standard dosing Renal
sulfamethoxazole Ped:0.8-1 (trim/sulfa)
Infant:1.5
Vancomycin19 Adult: 0.9 Moderate -3.1 Standard dosing – Renal
Ped: 0.6 follow levels
Neonate: 0.8 closely
Neonates/children:
20mg/kg IV X1
then standard
dosing with
levels19

Antifungal/Antiviral Agents (IV)

Goal: Quickly attain appropriate concentrations to treat infection while minimizing toxicity.
Dosing strategy: Consider higher doses based on MIC data for more serious infections. Adjust frequency
based on clearance recognizing that more frequent dosing may be necessary. Use close therapeutic drug
monitoring when available.

Medication Vd (L/kg)* Protein X Log P** Dosing Strategy

Low: ≤1
Binding Hydrophilic: <1
Mod: 1-5 Low: <30% Lipophilic: 1-2
Initial Dose Clearance
High: >5 Mod: 30-70% Highly
High: >70% lipophilic: >2
Acyclovir18 Adult: 26 Low -1.6 Higher end of Renal
Ped: 26 dosing range
Neonate: 13 Neonatal HSV:
Consider
continuous
infusion18
Amphotericin B19 0.7-3.99 High 0 Standard dosing Renal
Amphotericin B 0.16 High 0 Standard dosing Renal
liposomal19
Caspofungin19 0.14 High 0.3 Standard dosing Hepatic
Neonates: may
need higher doses
– follow levels if
available20
Fluconazole19 Adult: 0.6-0.8 Low 0.4 Higher end of Renal
Ped: 0.8-1.1 dosing range
Infant: 0.8-2.6 Neonatal/Infant
prophylaxis:
12mg/kg IV x1
then 6mg/kg/day
Neonatal/Infant
treatment:

61
Chapter 6

25mg/kg IV x1
then
12mg/kg/day19
Ganciclovir19 Adult: 0.74 Low -2 Standard dosing Renal
Ped: 0.64 Neonatal CMV:
6mg/kg IV q12h19
Micafungin19 Adult: 0.39 High -1.6 Standard dosing Hepatic
Ped: 0.3
Infant: 0.7-1.5
Oseltamivir 0.4 Moderate 1.16 Standard dosing Renal
Voriconazole19 Adult: 4.6 Moderate 1.5 Significantly higher Hepatic
Ped: 1.9 doses required –
avoid if possible.
Neonates/children:
14mg/kg IV q12h
Follow levels if
available19

Miscellaneous Agents (IV)

Goal: Achieve desired effect
Dosing strategy: Adjust per levels or titrate to desired effect

Medication Vd (L/kg)* Protein X Log P** Dosing Strategy

Low: ≤1
Binding Hydrophilic: <1
Mod: 1-5 Low: <30% Lipophilic: 1-2
High: >5 Mod: 30-70% Highly
High: >70% lipophilic: >2
Argatroban24,25 0.18 Low 1.3 Higher bolus doses and rate
changes may be required
Titrate to goal
Amiodarone 26,27 40-80 High 7.6 Higher bolus doses may be
required
Infusion rates as high as
20mcg/kg/min have been reported
Titrate to effect
Bivalirudin27,28 0.2 Low -7.1 Standard dosing
Titrate to goal
Cisatracurium10 0.1-0.15 Unknown unknown Titrate to effect
Fosphenytoin7 0.06-0.15 High 0.6 Higher doses may be needed
Follow levels if available or
consider other agents
Heparin 0.07 High -10.4 Higher bolus doses and rate
changes may be required
Titrate to goal
Levetiracetam10,29 0.6 Low -0.3 Standard dosing
Follow levels if available
Phenobarbital16,7 Ped: 0.6-0.8 Moderate 1.5 Standard dosing
Neonate: 0.8-1 Follow levels if available
Neonate: 5mg/kg loading dose

62
 Medication Management in ECLS

recommended
Rocuronium10 0.2-0.3 Moderate 5 Avoid if possible
Succinylcholine10 0.05 Unknown 0.6 Titrate to effect
Vecuronium10 0.3-0.4 Moderate to unknown Titrate to effect
High
*Vd reflects adult data unless otherwise noted **log P values are computed log P values which may differ from
experimental log P values and are to be used as reference only.

Appendix II: Risk of Sequestration Based on X Log P Values and Protein Binding30

Protein Binding X Log P Risk Dosing strategy

Low <1 Low Standard dosing likely appropriate
(<30%) 1-2 Low to moderate
>2 Moderate Consider higher doses or increased
frequency when appropriate.
Moderate <1 Low to moderate Standard dosing likely appropriate
(30-70%) 1-2 Moderate Consider higher doses or increased
frequency when appropriate.
>2 Moderate to High Use higher loading and/or maintenance
doses and/or increased frequency when
appropriate.
High <1 Moderate Consider higher doses or increased
(>70%) frequency when appropriate.
1-2 Moderate to High Use higher loading and/or maintenance
doses and/or increased frequency when
appropriate
>2 High Avoid combination if possible. If
required – use higher loading doses
and/or maintenance doses and increased
frequency if appropriate. Follow
medication levels if available.

63
Chapter 6

References Extracorporeal Life Support Organization;

2017:795-803.
1. Shekar K, Fraser JF, Smith MT, et al. Pharma- 11. Wagner D, Pasko D, Phillips K, Waldvogel
cokinetic changes in patients receiving extra- J,Annich G. In vitro clearance of dexmedetomi-
corporeal membrane oxygenation. J Crit Care. dine in extracorporeal membrane oxygenation.
2012; 27(741): e9-74. Perfusion. 2013; 28(1): 40-46.
2. Wildschut ED, Ahsman MJ, Allegaert K, Mathot 12. Fung Y, Barnett A, Fisquet S, et al. Sequestration
RA, Tibboel D. Determinants of drug absorption of drugs in the circuit may lead to therapeutic
in different ECMO circuits. Intensive Care Med. failure during extracorporeal membrane oxy-
2010; 36(12): 2109-16. genation. Critical Care. 2012; 16(5): 1-7.
3. Shekar K, Roberts JA, Mcdonald CI, et al. 13. Reiter PD, Ng J, Dobyns EL. Continuous hy-
Protein-bound drugs are prone to sequestration dromorphone for pain and sedation in mechani-
in the extracorporeal membrane oxygenation cally ventilated infants and children. J Opioid
circuit: Results from an ex vivo study. Crit Care. Management. 2012; 8(2): 99-104.
2015; 19(1): 164. 14. Tellor B, Avidan M. Ketamine infusion for
4. Wildschut ED, de Hoog M, Ahsman MJ, et al. patients receiving extracorporeal membrane
Plasma concentrations of oseltamivir and osel- oxygenation support. J of Heart Lung Transpl.
tamivir carboxylate in critically ill children on 2015; 34(4): S144.
extracorporeal membrane oxygenation support. 15. Varsha Bhatt-Mehta GA. Sedative clearance
PLoS ONE. 2010; 5(6): e10938. during extracorporeal membrane oxygenation.
5. Preston TJ, Hodge AB, Riley JB, Leib-Sargel C, Perfusion. 2005; 20(6): 309-315.
Nicol KK. In vitro drug adsorption and plasma 16. Wildschut ED, van Saet A, Pokorna P, Ahsman
free hemoglobin levels associated with hollow MJ, et al. The impact of extracorporeal life
fiber oxygenators in the extracorporeal life support and hypothermia on drug disposition in
support (ECLS) circuit. J Extra Corpor. 2007; critically ill infants and children. Pediatr Clin
39(4): 234-7. N Am. 2012; 59(5): 1183-1204.
6. Preston TJ, Ratliff TM, Gomez D, et al. Modi- 17. Ahsman MJ. Determinants of pharmacokinetic
fied surface coatings and their effect on drug variability during extracorporeal membrane
adsorption within the extracorporeal life support oxygenation: A roadmap to rational pharmaco-
circuit. J Extra Corpor. 2010; 42(3): 199-202. therapy in children. 2010.
7. Mehta NM, Halwick DR, Dodson BL, et al. Po- 18. Cies JJ, Moore WS, Miller K, Small C, et al.
tential drug sequestration during extracorporeal Therapeutic drug monitoring of continuous-in-
membrane oxygenation: Results from an ex fusion acylovir for disseminated herpes simplex
vivo experiment. Intensive Care Med. 2007; virus infection in a neonate receiving concurrent
33(6): 1018-1024. extracorporeal life support and continuous renal
8. Roberts JA . Pharmacokinetic issues for antibiot- replacement therapy. Pharmacotherapy. 2015;
ics in the critically ill patient. Crit Care Med. 35(2): 229-233.
2009; 37(3): 840-51. 19. Sherwin J, Heath T,Watt K. Pharmacokinetics
9. Shekar K, Roberts JA, Mullany DV, et al. In- and dosing of anti-infective drugs in patients
creased sedation requirements in patients re- on extracorporeal membrane oxygenation: A
ceiving extracorporeal membrane oxygenation review of the current literature. Clin Therapeu-
for respiratory and cardiorespiratory failure. tics. 2016; 38(9): 1976-1994.
Anaesth Intensive Care. 2012; 40(4): 648-55. 20. Koch BCP, Wildschut ED, Goede AL. de Hoog
10. Jefferis Kirk C, Abel EE, Muir J, et al. Strate- M, de Brüggemann RJM. Insufficient serum
gies for Medication Management in ECLS. IN: caspofungin levels in a paediatric patient on
Brogan TV, Lequier L, Lorusso R, MacLaren G, ECMO. Medical Mycology Case Reports 2013;
Peek GJ (eds). Extracorporeal life support:The 2: 23-24.
ELSO Red Book. 5th ed. Ann Arbor, Michigan:

64
 Medication Management in ECLS

21. De Rosa FG, Corcione S, Baietto L, et al. Phar- 30. Ha MA, Sieg AC. Evaluation of altered drug
macokinetics of linezolid during extracorpo- pharmacokinetics in critically ill adults receiv-
real membrane oxygenation. Int J Antimicrob ing extracorporeal membrane oxygenation.
Agents. 2013; 41(6): 590-1. Pharmacotherapy. 2017; 37(2): 221-235.
22. Shekar K, Fraser JF, Taccone FS, et al. The
combined effects of extracorporeal membrane
oxygenation and renal replacement therapy
on meropenem pharmacokinetics: A matched
cohort study. Crit Care. 2014; 18(6).
23. Cies JJ, Moore WS 2nd, Dickerman MJ, et
al. Pharmacokinetics of continuous-infusion
meropenem in a pediatric patient receiving
extracorporeal life support. Pharmacotherapy.
2014; 34(10): 175-9.
24. Beiderlinden M, Treschan T, Görlinger K, Pe-
ters J. Argatroban in extracorporeal membrane
oxygenation. Artificial Organs. 2007; 31(6):
461-465.
25. Latham GJ, Jefferis Kirk C, Falconer A, Dickey
R, Albers EL, McMullan DM. Challenging
argatroban management of a child on extracor-
poreal support and subsequent heart transplant.
Semin Cardiothorac Vasc Anesth. 2016; 20(2):
168-174.
26. Watt K, Li JS, Benjamin DK Jr, Cohen-Wolkow-
iez M. Pediatric cardiovascular drug dosing in
critically ill children and extracorporeal mem-
brane oxygenation. J Cardiovasc Pharmacol.
2011; 58(2): 126-32.
27. Sanfilippo F, Asmussen S, Maybauer DM, et
al. Bivalirudin for alternative anticoagulation
in extracorporeal membrane oxygenation: A
systematic review. J Int Care Med. 2016; 32(5):
321-319.
28. Ranucci M, Ballotta A, Kandil H, Isgrò G,
Carlucci C, Baryshnikova E, Pistuddi V. Bi-
valirudin-based versus conventional heparin
anticoagulation for postcardiotomy extracor-
poreal membrane oxygenation. Crit Care. 2011;
15(6): R275.
29. Nei SD, Wittwer ED, Kashani, KB, Frazee EN.
Levetiracetam pharmacokinetics in a patient
receiving continuous venovenous hemofiltra-
tion and venoarterial extracorporeal membrane
oxygenation. Pharmacotherapy. 2015; 35(8):
e127-e130.

65
7

Respiratory Diseases Associated with ECMO

Sarah Keene, MD, Silvia M. Hartmann, MD, William R. Lynch, MD

Neonatal Respiratory ECMO Indications

Background Neonatal candidates for ECMO have severe

respiratory failure and have failed extensive medical
Neonatal respiratory cases provided the first therapy (Table 7-1).3 The goal of ECMO is to
substantial population of ECMO patients with improve both morbidity and mortality; delays in
good outcomes, which allowed ECMO to flourish implementation, as well as inappropriate use, in-
as a therapy. Neonates are no longer the most com- crease mortality.4,5 Oxygenation index (mean airway
mon ECMO patients. In the last few years, ELSO pressure x FiO2/PaO2) is the most commonly used
Registry has recorded 850-900 neonatal respiratory numerical criteria for ECMO, but many other patient
ECMO runs, decreased from more than 1500 in factors contribute to the decision to cannulate.
1992.1,2 ECMO remains a lifesaving therapy for a
specific subset of gravely ill infants, with increasing Contraindications
frequency of those with CDH and other congenital
lung abnormalities, and less of reversible diseases, Table 7-2 lists traditional contraindications to
such as MAS (meconium aspiration syndrome) or ECMO; however, substantial variability occurs
classic RDS (respiratory distress syndrome). between centers and even among providers in

CANDIDATES Pathophysiologic findings that should drive

consideration for ECMO
Late preterm or Term Infants with: 1. Oxygenation Index > 40 for >4 hours
• Severe respiratory or cardiac 2. Oxygenation Index >25 on 100% oxygen
failure despite prolonged (>24-48h) maximal therapy
• Severe pulmonary hypertension 3. Severe hypoxic respiratory failure with
recurrent acute decompensations
Candidates should: 4. Pulmonary hypertension with progressive
• be refractory to maximal right ventricular dysfunction
medical management 5. Persistent air leak or other ongoing lung
• have a high predicted morbidity damage
and mortality, and
• have a potentially reversible
etiology
Table 7-1. Indications for ECMO in neonatal respiratory patients.

ABSOLUTE CONTRAINDICATIONS RELATIVE CONTRAINDICATIONS

Lethal congenital anomalies Size (<2kg, <1.6kg)
Lethal genetic syndromes Gestational age <34 weeks
Severe irreversible brain damage Coagulopathy with bleeding
Grade III or IV IVH Irreversible organ injury
Active, uncontrolled bleeding Mechanical ventilation for >10-14 days
Table 7-2. Contraindications for ECMO in neonates.

67
Chapter 7

decisions surrounding ECMO cannulation.5 Some previously excluded populations, including oncol-
physicians consider hypoxic ischemic encephalopa- ogy patients and hematopoietic stem cell transplant
thy (HIE), even of moderate severity, to be a con- recipients. Different centers have varying inclusion/
traindication to ECMO.6 Similarly, gestational age exclusion criteria. Some common exclusion criteria
criteria are applied differently, with some physicians are progressive neurodegenerative disorders, life-
advocating for younger patients due to reasonable limiting diagnoses with poor prognosis despite best
outcomes.7 Although a lethal genetic anomaly is available treatment, and ongoing hemorrhagic shock.
classically considered a contraindication, it is often As of 2015, the ELSO Registry recorded almost
unclear at the time of consideration what diagnoses 7,000 children supported with ECMO for respiratory
may prove lethal. indications.16,17 Both venoarterial (VA) and venove-
nous (VV) ECLS are used to support patients with
Physiology respiratory failure. The use of VV-ECLS has been
increasing and accounts for approximately 50% of
The normal in utero status is, essentially, one patients.18
of physiologic pulmonary hypertension. Pressures The three most common diagnoses in a review
in the pulmonary arterial system exceed those of of the ELSO Registry were: 1) respiratory failure
the systemic, shunting blood away from the lungs without meeting criteria for acute respiratory dis-
toward the brain and body, thus the historical name tress syndrome (ARDS), 2) Respiratory Syncytial
for pulmonary hypertension (PPHN) in neonates, virus (RSV) pneumonia, and 3) bacterial pneumonia.
“persistent fetal circulation.” At birth this changes Almost 1 in 5 pediatric patients were described by
in a matter of a few minutes, and the entire system the category of “respiratory failure, non-ARDS”
must work for this to happen quickly and success- which is a term to capture a heterogeneous group
fully.8 The infant must start breathing, expanding of patients without more specific diagnoses, such
and bringing oxygen to the lungs, stop production as influenza or pulmonary hemorrhage recognized
of in-utero vasoconstrictors, increase that of vaso- in the Registry reporting form. Table 7-4 describes
dilators and the pulmonary arteries must respond. etiologies of respiratory failure captured by the
Anything that interferes with this process—infec- ELSO Registry by prevalence.
tion, meconium, structural lung changes, hypoxia The underlying cause of respiratory failure is
and acidosis due to any cause---will interfere with helpful in prognostication as the survival to dis-
this transition and can cause pathologic pulmonary charge varies significantly depending on the diag-
hypertension. nosis. Patients with status asthmaticus have the best
survival (>80%).19 Patients with RSV pneumonia
Common Disease Processes and aspiration pneumonia treated with ECLS also
have good survival (>70%).19 On the other hand,
Table 7-3 shows ELSO average data 2012- fungal pneumonia has the worst outcome with less
20161,3 for common disease processes. than 25% survival. Pertussis and ARDS related to
sepsis also have poor outcomes with approximately
Pediatric Respiratory ECMO 40% survival. Pediatric respiratory failure patients
had an overall survival of 57%.19
Children age 31 days to 18 years receive extra- Increasingly, pediatric patients treated with
corporeal life support (ECLS) for respiratory failure ECLS have associated comorbid conditions. About
as rescue therapy for refractory hypoxemia or severe 1 out of every 3 patients supported with ECLS has
respiratory acidosis or as an alternative modality to an underlying condition, acute kidney injury (AKI)
provide oxygenation and ventilation when mechani- and chronic lung disease (CLD) being the most
cal ventilator settings have a high risk of causing common.19 The presence of AKI is associated with
ongoing lung injury. Many different diagnoses significantly worse survival of only 33%; whereas
may lead to these scenarios. Contraindications to CLD or biventricular congenital heart disease did
ECLS for children are evolving with inclusion of

68
 Respiratory Diseases Associated with ECMO

DIAGNOSIS PHYSIOLOGY/RUN LENGTH/SURVIVAL/EXPECTATIONS

Meconium Aspiration Physiology: physical obstruction to ventilation by meconium, surfactant
Syndrome inactivation, hypoxia and acidosis leading to pulmonary hypertension
Length of Run: 5-6 days
Survival: 91-94%
Expectations: Short runs with excellent survival, mortality usually related to
complications of ECMO, good respiratory outcomes but higher risk for
asthma9
Pneumonia/sepsis Physiology: physical obstruction due to infection, structural lung damage,
failure of pulmonary vasorelaxation due to ongoing inflammation, hypoxia
Length of Run: 7-14 days
Survival: 42-48% (low number of cases in last 5 years)
Expectations: survival and risk of long term lung sequelae depends on
primary disease process
Hypoxic Ischemic Physiology: Acidosis and lack of ventilation prior to and at delivery leading
Encephalopathy (HIE) to PPHN
Length of Run: 5-7 days
Survival: 70-85%10
Expectations: Cooling may continue on ECMO, potentially increased risk
for bleeding, ICH due to coagulopathy and reperfusion injury11
Congenital Physiology: pulmonary hypoplasia causes hypercarbia and hypoxia, poorly
Diaphragmatic Hernia developed and over muscularized pulmonary arteries + hypoxia/acidosis
(CDH) cause PPHN12
Length of Run: 12-14 days but many cases longer, chance of survival
decreases substantially after the 3rd and 4th week of ECMO13
Survival: 45-55%
Expectations: Prolonged runs and recalcitrant PPHN are common,
additional medications for PPHN may be tried. CDH repair may occur
before, during, or after ECMO. Chronic respiratory disease is common in
survivors and may be lifelong.14

“Idiopathic” pulmonary Physiology: Classic “black lung” PPHN is usually due to failure of
hypertension (PPHN) vasorelaxation due to variety of causes of abnormal transition: fetal stress,
acidosis, maternal medication use and is usually reversible
Length of Run: 6-8 days
Survival: 72-76%
Expectations: Low flow requirements and low ventilator settings are
typical, further evaluation if no improvement

“Other” Physiology: Includes various causes of pulmonary hypoplasia, lymphatic

disease, congenital lung disease [alveolar capillary dysplasia/acinar
dysplasia/surfactant protein deficiencies], others
Length of Run: Variable, average 6-8 days, longer runs for other etiologies
Survival: 53-69%
Expectations: Patients with “idiopathic” respiratory failure and PPHN who
have not shown improvement by a week of ECMO should be investigated
for other causes --usually genetic testing or a lung biopsy.15 Depending on
diagnosis and comorbidities, a limited number are candidates for lung
transplant.

Table 7-3. Indications for ECMO support for neonatal respiratory failure.

69
Chapter 7

not impact mortality.19 Table 7-5 lists such comor- of patients do not respond to this approach. Such
bidities. patients experience an ongoing risk of ventilator
A small percent, estimated at 12%, of children induced lung injury along with the consequences of
supported with ECLS have a prolonged course long duration sedation. ECMO should be considered
of greater than 21 days.20 Survival does decline for this group of patients.
steadily with the duration of ECLS therapy but
does so gradually with no dramatic decreases that Indications
might represent a reasonable discontinuation point.
In this study, there was a survival rate of 25% after Despite the growth in ECMO for adult respi-
45 days of ECLS.20 ratory failure, the indications for its use have not
changed. ECMO should be considered for adults
Adult Respiratory ECMO with hypoxic respiratory failure when the mortality
risk is 50% or greater and is firmly indicated when
The first reported ECMO survivor was an adult the mortality risk exceeds 80%. Mortality risk is
in respiratory failure21 but the majority of ECMO challenging to quantify but a variety of indicators
cases during the following three decades were neo- help to suggest it. PaO2/FiO2 can be calculated as
nates. The ELSO Registry reveals dramatic growth can Adult Oxygenation Index (AOI),23 Murray Lung
in adult ECMO cases in the past 10-15 years with Injury Score,24 and “APPS” score25 can help guide
2670 cases in 2016, twice the number of combined decisions on the appropriate use of ECMO.
neonatal and pediatric respiratory cases in 2016 ECMO for hypercarbic respiratory failure is
and comparable to all adult respiratory cases from indicated for persistent CO2 retention with venti-
1990 -2010 (2793 cases).22 Adult respiratory failure lator plateau pressures >30 cm H2O.26 Severe air
is hypoxic, hypercarbic, or a combination of both. leak syndrome can be a consequence of positive
The initial approach for these patients includes ventilation,26 and worsens hypercarbia. Air leaks
mechanical ventilation and sedation but a number

Most Common Etiologies More Common Etiologies Least Common Etiologies

(>10% of cohort) (5-10% of cohort) (<5% of cohort)
Acute respiratory failure, ARDS related to sepsis Severe Pertussis
non-ARDS Aspiration pneumonia Status asthmaticus
Respiratory syncytial virus ARDS related to trauma or Pulmonary hemorrhage
pneumonia post-operative state Influenza
Bacterial pneumonia Submersion injury
Viral pneumonia (non-RSV) Pneumocystis pneumonia
Fungal pneumonia
Table 7-4. Prevalence of specific diagnoses leading to pediatric respiratory failure in the
ELSO Registry. Adapted from Zabrocki et al.4

More Common Comorbidities Least Common Comorbidities

(≥5% of cohort) (<5% of cohort)
Renal failure Cancer
Chronic lung disease Bone marrow transplant
Biventricular congenital heart Solid organ transplant
disease Primary immunodeficiency
Single ventricle congenital heart
disease
Cardiomyopathy/myocarditis
Liver failure
Table 7-5. Prevalence of comorbidities among pediatric patients with
respiratory failure treated with ECLS. Adapted from Zabrocki, et al.4

70
 Respiratory Diseases Associated with ECMO

persist until positive pressure ventilation is reduced A relatively new ECMO application is in the
or discontinued. This cannot happen without ECMO. setting of endstage chronic lung disease leading
The overwhelming majority of adult respiratory to lung transplantation. Chronic lung diseases
failure patients have Acute Respiratory Distress commonly considered for lung transplantation are
Syndrome (ARDS). In 2011, a consensus descrip- shown in Table 7-7. Most lung transplant programs
tion of ARDS was revised and is referred to as the now take advantage of ECMO to support patients to
“Berlin Definition” of ARDS. The criteria require a transplant. The concept is to utilize ECMO so that
chest x-ray with bilateral opacities (not explained a patient can remain active and vigorous prior to
by effusions, atelectasis, consolidation, or edema transplantation (see Chapter 27). Successful “bridge”
secondary to heart failure), hypoxia, and a clinical to transplant typically occurs with a few weeks of
course of 1 week or less from the inciting insult. Se- ECMO support.27,28 Unfortunately, it is not possible
verity of hypoxia is defined by PaO2/FiO2 ratio with to know when a lung will become available. As
mild being 200-300, moderate 100-200 and severe ECMO support continues, complications accrue
≤100.26 The pathophysiology of ARDS includes (infections, blood transfusion associated antibod-
alveolar epithelial and capillary endothelial damage ies, deconditioning) that can result in removing
that results in failure of the alveolar capillary barrier. a patient from the transplant list. These patients
ARDS is often considered as primary and second- might be awake, alert, and adequately supported
ary (or pulmonary and extrapulmonary),27 as shown with ECMO but they will die on ECMO. Realizing
in Table 7-6. These different etiologies for ARDS there are barely 2000 lung transplants per year in
have variations in incidence and survival; however, the U.S., this should remain a relatively small group
the fundamental pathophysiology in the adult lung of patients requiring ECMO.
appears to be very similar.

PRIMARY (PULMONARY) SECONDARY (EXTRAPULMONARY)

Pneumonia Non pulmonary sepsis
Bacterial Trauma
Viral Pancreatitis
Fungal Blood transfusions
Aspiration Cardiopulmonary bypass
Inhalation injury Thoracic surgery
Ventilator Injury Burns
Vasculitis
Embolism
Table 7-6. Etiologies by category of ARDS.

DISEASES
Idiopathic pulmonary fibrosis
Suppurative lung diseases
Cystic fibrosis
Bronchiectasis
Emphysema
α1-Antitrypsin Deficiency
Pulmonary Hypertension
Connective Tissue Disorders
Scleroderma
Rheumatoid arthritis
Lupus
Sjögren syndrome
Polymyositis
Dermatomyositis

Table 7-7. Chronic lung diseases considered for transplantation.

71
Chapter 7

Contraindications

There are no absolute contraindications to

ECMO for respiratory failure. Cannulation can be
done safely and quickly. The technology (pumps,
oxygenators) is reliable, allowing prolonged ECMO
support to be offered for weeks and months. There
are a group of relative contraindications associated
with worse outcome on ECMO.27,29,30,31,32

• High risk of systemic bleeding, recent or ex-

panding CNS hemorrhage, other contraindica-
tions to anticoagulation
• Comorbidities with short expected survival like
terminal malignancies, advanced CNS injuries
or other significant changes in quality of life;
immune suppression
• Older age recognizing that there is not a defined
upper age limit
• Mechanical ventilation with high Pplat >30 cm
H2O with FiO2 >90% for longer than 7 days.

72
 Respiratory Diseases Associated with ECMO

References poreal membrane oxygenation. The Journal of

pediatrics. 2010;157(3):499-501, 501 e491.
1. ECLS Registry Report: International Summary 12. Harting MT, Davis,C.F., Lally, K.P. Extracor-
January 2017. 2017. Accessed June 7, 2017. poreal Life Support: The ELSO Red Book. 5th
2. Thiagarajan RR, Barbaro RP, Rycus PT, et al. ed. Ann Arbor, Michigan: Extracorporeal Life
Extracorporeal Life Support Organization Reg- Support Organization; 2017.
istry International Report 2016. ASAIO journal. 13. Prodhan P, Stroud M, El-Hassan N, et al. Pro-
2017;63(1):60-67. longed extracorporeal membrane oxygenator
3. Short BL, Soghier, L. Extracorporeal Life Sup- support among neonates with acute respiratory
port: The ELSO Red Book. In: Brogan TV, ed. failure: a review of the Extracorporeal Life
5th ed. Ann Arbor, Michigan: Extracorporeal Support Organization registry. ASAIO journal.
Life Support Organization; 2017. 2014;60(1):63-69.
4. Smith KM, McMullan DM, Bratton SL, Rycus 14. Spoel M, Laas R, Gischler SJ, et al. Diagnosis-
P, Kinsella JP, Brogan TV. Is age at initiation related deterioration of lung function after
of extracorporeal life support associated with extracorporeal membrane oxygenation. The
mortality and intraventricular hemorrhage in European respiratory journal. 2012;40(6):1531-
neonates with respiratory failure? Journal of 1537.
perinatology: official journal of the California 15. Gupta A, Zheng SL. Genetic disorders of sur-
Perinatal Association. 2014;34(5):386-391. factant protein dysfunction: when to consider
5. Suttner D. Extracorporeal Life Support: The and how to investigate. Archives of disease in
ELSO Red Book. 5th ed. Ann Arbor, Michigan: childhood. 2017;102(1):84-90.
Extracorporeal Life Support Organization; 2017. 16. Paden ML, Rycus PT, Thiagarajan RR, Reg-
6. Chapman RL, Peterec SM, Bizzarro MJ, Mercu- istry E. Update and outcomes in extracorpo-
rio MR. Patient selection for neonatal extracor- real life support. Seminars in perinatology.
poreal membrane oxygenation: beyond sever- 2014;38(2):65-70.
ity of illness. Journal of perinatology: official 17. www.elso.org. 12/18/2015.
journal of the California Perinatal Association. 18. Gray BW, Haft JW, Hirsch JC, Annich GM,
2009;29(9):606-611. Hirschl RB, Bartlett RH. Extracorporeal life
7. Church JT, Kim AC, Erickson KM, et al. Push- support: experience with 2,000 patients. ASAIO
ing the boundaries of ECLS: Outcomes in <34 journal. 2015;61(1):2-7.
week EGA neonates. Journal of pediatric sur- 19. Zabrocki LA, Brogan TV, Statler KD, Poss
gery. 2017. WB, Rollins MD, Bratton SL. Extracorporeal
8. Fuloria M, Aschner JL. Persistent pulmonary membrane oxygenation for pediatric respiratory
hypertension of the newborn. Seminars in fetal failure: Survival and predictors of mortality.
& neonatal medicine. 2017;22(4):220-226. Critical care medicine. 2011;39(2):364-370.
9. Hamutcu R, Nield TA, Garg M, Keens TG, 20. Brogan TV, Zabrocki L, Thiagarajan RR, Rycus
Platzker AC. Long-term pulmonary sequelae in PT, Bratton SL. Prolonged extracorporeal mem-
children who were treated with extracorporeal brane oxygenation for children with respiratory
membrane oxygenation for neonatal respiratory failure. Pediatric critical care medicine: a jour-
failure. Pediatrics. 2004;114(5):1292-1296. nal of the Society of Critical Care Medicine and
10. Cuevas Guaman M, Lucke AM, Hagan JL, Kai- the World Federation of Pediatric Intensive and
ser JR. Bleeding Complications and Mortality in Critical Care Societies. 2012;13(4):e249-254.
Neonates Receiving Therapeutic Hypothermia 21. Hill JD, O’Brien TG, Murray JJ et all. Pro-
and Extracorporeal Membrane Oxygenation. longed extracorporeal oxygenation for acute
American journal of perinatology. 2017. post-traumatic respiratory failure (shock lung
11. Massaro A, Rais-Bahrami K, Chang T, Glass P, syndrome): use of the Bramson Membrane
Short BL, Baumgart S. Therapeutic hypother- Lung. N Engld J Med. 1972;286(12):629-634.
mia for neonatal encephalopathy and extracor-

73
Chapter 7

22. ECLS Registry Report: International Summary

January 2017. 2017. Accessed June 7, 2017.
23. Murray JF, Matthay MA, Luce JM, Flick MR:
An expanded definition of the adult respiratory
distress syndrome [published erratum appears
in Am Rev Respir Dis 1989 Apr; 139(4):1065].
American Review of Respiratory Disease 1988,
138:720-723.
24. Dechert RE, Park PK, Bartlett RH: Evaluation
of the oxygenation index in adult respira-
tory failure. J Trauma Acute Care Surg 2014;
76:469–473
25. Villar J, Ambrós A, Soler JA, et al; Stratifica-
tion and Outcome of Acute Respiratory Distress
Syndrome (STANDARDS) Network: Age,
Pao2/ Fio2, and Plateau Pressure Score: A Pro-
posal for a Simple Outcome Score in Patients
with the Acute Respiratory Distress Syndrome.
Crit Care Med 2016; 44:1361–1369
26. Ranieri VM, Rubenfeld GD, Thompson BT,
Ferguson ND, Caldwell E et al. ARDS Defi-
nition Task Force. Acute respiratory distress
syndrome: the Berlin Definition. JAMA. 2012
Jun 20. 307 (23):2526-33.
27. Rezoagli E, Fumagelli R, Bellani G. Definition
and epidemiology of acute respiratory distress
syndrome. Ann Transl Med 2017; 5(14):282.
28. ELSO. Extracorporeal Life Support Organiza-
tion Guidelines for Adult Respiratory Failure.
August 2017.
29. Biscotti M, Gannon WD, Agerstrand C, Abrams
D, Sonett J, Brodie D, Bacchetta M. Awake ex-
tracorporeal membrane oxygenation as bridge to
lung transplantation: A 9 year experience. Ann
Thorac Surg. 2017 Aug;104(2):412-419.
30. Brodie D, Bacchetta M. Extracorporeal mem-
brane oxygenation for ARDS in adults. N Engl
J Med. 2011; 365:1905-1914.
31. Richard C, Arguad L, Blet A, et al. Extracorpo-
real life support for patients with acute respira-
tory distress syndrome: report of a Consensus
Conference. Ann Intensive Care 2014; May 24;
4:15.
32. Schmidt M, Brechot N, Combes A. Ten situa-
tions in which ECMO is unlikely to be success-
ful. Intensive Care Med. 2016; 42(5):750-752

74
8

Neonatal, Pediatric, and Adult Cardiac Extracorporeal Life Support

Sertac Cicek, MD, Titus Chan, MD, MS, MPP, Joseph Dearani, MD

ECMO in Congenital Heart Diseases septostomy in patients with d-transposition of the

great arteries. When used to stabilize patients with
Extracorporeal membrane oxygenation (ECMO) evidence of end-organ failure, ECMO enhances
is the most commonly used form of mechanical end-organ perfusion, reverses acidosis, and fa-
support in children and adult patients who have cilitates recovery of cardiac function making them
undergone cardiac surgery and cannot be success- better surgical candidates and potentially improving
fully separated from cardiopulmonary bypass (CPB). outcomes.1,2
ECMO support can provide cardiac and pulmonary Although most neonates with ductal-dependent
support and can be instituted via central cannulation cyanotic heart disease or isolated parallel circula-
or peripheral cannulation. The duration of ECMO tions are well managed with prostaglandin infusion
support can range from days to months depending to establish or maintain ductal patency or a balloon
on the initial indications for implementation. In the atrial septostomy to increase mixing at the atrial
post-cardiotomy setting, the goal is sufficient myo- level, ECMO may be necessary to stabilize patients
cardial recovery to allow successful separation from who present late with hypoxic shock, refractory to
ECMO. When separation is not feasible, conversion pharmacological restoration of pulmonary blood
to a more long-term form of mechanical support flow only. CHDs such as d-transposition of the
(e.g. left or right ventricular assist device) is gener- great arteries with refractory pulmonary hyperten-
ally performed. This chapter reviews the role and sion,3 pulmonary atresia with intact ventricular
outcomes of ECMO in the setting of heart disease. septum, hypoplastic left heart syndrome, and severe
Ebstein’s malformation may require ECMO for
Preoperative Support preoperative stabilization.
Another group of patients are those with insuf-
Despite its common use in children postcardi- ficient pulmonary blood flow (acute aortopulmonary
otomy, ECMO is employed much less commonly for shunt occlusion or refractory tetralogy of Fallot
preoperative stabilization and as bridge to surgery. spells, etc.) These patients can often undergo a
Preoperative ECMO support provides stabilization straightforward repair with immediate discontinu-
and/or recovery of cardiopulmonary function as well ation of ECMO after surgery.4 ECMO is a vital ad-
as recovery of other end-organ functions prior to junct for the preoperative stabilization of patients
palliation or complete repair. Gupta et al. reported with congenital diaphragmatic hernia with severe
that 14% of perioperative ECMO cases (494/3498) pulmonary hypoplasia and associated pulmonary
from the pediatric health information system (PHIS) hypertension.5
database represented patients receiving preoperative The cannulation strategy for preoperative
support.1 Bautista-Hernandez et al. reported 26 pa- ECMO depends on patient anatomy, ECMO indica-
tients with congenital heart disease (CHD) bridged tion, expected duration, surgical timing, and patent
to surgery with ECMO. Of these 62 survived to vasculature. Cervical cannulation with the right
hospital discharge.2 The most common indication internal jugular vein and the right common carotid
for ECMO was hypoxemia despite balloon atrial artery provides good exposure and easy insertion

75
Chapter 8

and is the most common approach for preoperative Castenada’s famous quote “what went wrong with
ECMO cannulation. With current cannulas, per- the operation?” should be kept in mind and residual
manent ligation of the carotid artery distal to the structural lesions should be actively sought using
arterial cannula is unnecessary and vascular repair all diagnostic modalities.
is feasible at decannulation. Femoral cannulation Postcardiotomy patients with single-ventricle
for larger children, although easier, has the poten- physiology and depressed myocardial function or
tial for leg ischemia and vascular complications, reactive pulmonary vasculature are reported to have
and may lead to suboptimal oxygen delivery to the improved survival on ECMO.7 Transient myocardial
cerebral and coronary circulation. Central cannula- dysfunction after CPB in the single-ventricle neona-
tion through the right atrium and ascending aorta tal heart results mostly from systemic inflammatory
provides initiation of support with larger cannula response syndrome (SIRS), ischemia-reperfusion
sizes and higher flows, better left heart decompres- injury and pulmonary vascular reactivity and can
sion with placement of additional cannula if needed. benefit significantly from the rest during ECMO
However, bleeding and infection are important risks. support. Patients with biventricular physiology
ECMO is indicated only for the temporary who receive ECMO are reported to have high risk
provision of tissue oxygen delivery in patients for early death after repair of complex intracardiac
with reversible cardiac and/or respiratory failure in lesions.1 Data show that the overall hospital sur-
whom conventional medical strategies do not suffice. vival for patients who fail to separate from CPB
Neonates who receive ECMO for preoperative sta- is approximately 47% with a range from 35% to
bilization experience greater risk for perioperative 61%.8 Use of ECMO following repair of CHD
morbidity and death; however, their survival is better requires special attention to the unique physiol-
than other cardiac ECMO populations.1 ogy being supported. Single-ventricle physiology,
once considered a relative contraindication, is
Failure to Wean Off Cardiopulmonary Bypass now the most common underlying diagnosis in
neonates receiving ECMO.9 These patients require
Despite advances in CPB, myocardial protec- complex management. Balancing the systemic and
tion, and surgical techniques, myocardial dysfunc- pulmonary circulations to prevent low coronary
tion following repair of CHD occurs commonly. perfusion and pulmonary edema remains a chal-
Fortunately, in most cases myocardial dysfunction lenge. Management of systemic to pulmonary artery
is mild to moderate and easily managed with ino- shunt flow is controversial and varies at different
tropic support, afterload reduction, and pulmonary institutions. Some teams leave the shunt open and
vasodilators. However, patients with severe myo- compensate for pulmonary runoff with increased
cardial dysfunction following long crossclamp and flow (200 ml/kg), at the risk of pulmonary over
CPB times and those with preexisting ventricular circulation;10 whereas, others partially occlude the
dysfunction might not respond to conventional shunt to improve the systemic perfusion.11 We prefer
agents and fail separation from CPB. Initiation of to partially occlude the shunt with a clip to maintain
ECMO in a timely manner in these patients creates systemic perfusion with some pulmonary blood flow.
a favorable environment for recovery by improving A persistent elevated lactate level in patients with
systemic perfusion and facilitates myocardial recov- a patent systemic-to-pulmonary shunt most likely
ery by increasing oxygen delivery and preventing represents inadequate oxygen delivery to the tissue
ventricular distension. Timing of initiation of me- with pulmonary overcirculation and should prompt
chanical support directly correlates with improved immediate modification of ECMO configuration.
postoperative survival. However, key feature for Intrathoracic cannulation of the right atrium or
favorable outcomes is the absence of residual le- the superior and inferior vena cava and the ascend-
sions which are common and may be difficult to ing aorta is the most appropriate approach for pa-
identify. In a large group of patients who receive tients who require intraoperative support for failure
ECMO after cardiac surgery, 70 % were reported to to wean from CPB.12 Left ventricular (LV) disten-
have a technically inadequate operation.6 Dr. Aldo sion must be avoided during ECMO, especially in

76
 Neonatal, Pediatric, and Adult Cardiac Extracorporeal Life Support

patients with preexisting LV dysfunction because it echocardiographic demonstration of poor ventricu-

may prevent LV recovery or worsen injury due to lar function. Adequacy of organ perfusion must be
inadequate coronary perfusion pressure. Left-sided actively and repeatedly assessed once ECMO has
distension can be treated by increasing ECMO flow been started and any abnormalities must be ad-
rates to decrease the pulmonary venous return to the dressed.
left heart. However, when this is unsuccessful, left Our policy calls for central thoracic cannulation
heart venting should be performed. This is easily in patients requiring ECMO support within the first
accomplished in patients with an open chest and cen- 5-7 days following cardiac surgery, while neck can-
tral cannulation, placement of a vent into the right nulation is often used if support is required more
superior pulmonary vein, dome of the left atrium, than one week postoperatively. Central cannula-
or left atrial and provides excellent decompression tion allows placement of larger cannulas, multiple
of the left heart. For patients cannulated via periph- cannulas, or decompression of the LV as required
eral vessels, LV decompression can be achieved by by patient anatomy and ECMO physiology. The
transcatheter or surgical atrial septostomy. Adequate chest remains routinely open with sterile draping
ECMO circuit flow depends on proper cannula po- to prevent tamponade.
sitioning, so confirmation of appropriate cannula
placement is extremely important. Extracorporeal Cardiopulmonary Resuscitation
(ECPR)
Postoperative Low Cardiac Output State
Cardiac arrest may complicate the postopera-
Postoperative myocardial dysfunction and tive course following surgical repair of CHD with
clinical deterioration can follow initial successful reported survival rates ranging between 14 and 44%.
separation from CPB. The multifactorial etiology Duration of CPR strongly impacts survival in con-
includes SIRS, ischemia-reperfusion injury, arrhyth- ventional CPR. CPR lasting more than 30 minutes
mias, residual lesions and pulmonary hypertension. yields survival rates of 0% to 5%.14 Rapid deploy-
Most patients require pharmacologic and ventilator ment of ECMO to resuscitate patients during cardiac
support. However, if hemodynamic instability and arrest appears to improve survival with hospital dis-
poor tissue oxygenation persists despite escalating charge rates ranging from 34% to 73%.15 Improved
doses of vasoactive medications, ECMO support survival to hospital discharge makes cardiac arrest
should be considered. Again, residual surgically cor- a common indication for ECMO, comprising nearly
rectable or treatable cardiac lesions must be detected half of all cannulations in pediatric cardiac centers.16
as reintervention improves clinical outcomes.13 An effective ECPR program requires an organi-
Echocardiographic imaging can be challenging and zation which enables rapid cannulation; minimizing
suboptimal in this clinical setting, necessitating a the time delay between cardiac arrest and actual
low threshold for cardiac catheterization in case of starting time of ECMO. Continuous “24/7” avail-
inconclusive echocardiograms and ECMO provides ability of a credentialed surgeon and ECMO Spe-
only support not treatment. cialist, a pre-primed ECMO circuit, and equipment
Defining at which level pharmacologic and ven- readily available save critical time. Development
tilator support becomes insufficient is arbitrary and of protocols, delineation of roles, and simulation
varies between institutions. As cardiac failure and reduces the time to initiating ECMO support.17
organ complications due to low cardiac output are
the most common cause of mortality after repair of Glenn and Fontan Circulation on ECMO
congenital heart defects, potential consideration for
more timely/earlier support may improve survival. Patients following cavopulmonary connection
We prefer early institution of ECMO support rather operations (Glenn and Fontan) constitute a com-
than using it as a last resort and consider initiation plicated group due to technical difficulties and the
in patients with maximal dose of epinephrine (0.15 requirement for placement of multiple venous can-
mcg/kg/min) and signs of poor tissue perfusion or nulas to adequately drain the disconnected venous

77
Chapter 8

circulation. The literature is sparse with specific data nulation in the patients with bidirectional Glenn
on patients with cavopulmonary connections sup- is more challenging because of the separation of
ported with ECMO and mostly includes single case systemic venous drainage, SVC to the pulmonary
reports, descriptions of institutional experiences artery, and IVC to the heart.18 Although a central
with cardiac patients on ECMO that contain small venous cannula placed in the common atrium may
subsets of patients with Fontan or BDG circulations. provide adequate ECMO flows it is often insufficient
In a single center retrospective report, Booth et al. even when the ventricle is ejecting well. If separate
reported 20 patients (14 Fontan and 6 bidirectional cannulas for SVC and IVC are used, the SVC must
Glenn) supported with ECMO.18 Of the 14 Fontan first be decompressed to prevent neurologic injury.
patients, 7 were withdrawn from ECMO or died Patients with the Fontan circulation pose unique
within 48 hours of decannulation, whereas 5 of 6 and difficult challenges for ECMO support that may
Glenn patients died before hospital discharge and decrease survival in these patients. Fortunately, sur-
the lone survivor had neurologic injury at followup. gical outcomes for patients who have undergone the
Patients with an acute and reversible cause for their bidirectional Glenn or Fontan procedures are typi-
severe cardiac dysfunction are more likely to benefit cally very good and ECMO use in this population
from ECMO than those with severe long standing remains relatively uncommon.
myocardial dysfunction or patients who require
resuscitation or ECPR. Myocarditis and Cardiomyopathy
A clear understanding of patient anatomy, physi-
ology, and factors contributing to clinical deteriora- The myocardial inflammation seen in myocardi-
tion is necessary for selecting the most appropriate tis typically results from infection (most commonly
cannulation strategy in patients palliated with Glenn viral), autoimmune disease, or exposure to medica-
and Fontan operations. Cannulation is more difficult tions. Cardiomyopathies are diseases of the heart
due to complex anatomy, previous operations, and myocardium and can be broadly differentiated into
lack of proper vascular access sites due to preexist- three categories: hypertrophic, dilated, and restric-
ing occlusions. The anatomy of the cavopulmonary tive cardiomyopathy. While the pathophysiology
connection may limit venous drainage and preclude for each type of cardiomyopathy and myocarditis
adequate decompression of the heart. Multiple is unique, these diseases can produce cardiac dys-
venous sites may need to be cannulated, compli- function producing pump failure necessitating VA-
cating the circuit and patient management. High ECMO for cardiac support (Table 8-1). Additionally,
systemic venous pressure despite adequate ECMO patients with hypertrophic cardiomyopathy and
support may limit organ perfusion and can increase myocarditis are at increased risk of arrhythmias
susceptibility to end-organ injury. The distal tip of and may require extracorporeal support as a result
the cannula should be located as centrally as pos- of cardiac arrest. Patients with restrictive cardiomy-
sible toward the superior vena cava (SVC)-Fontan opathy will also develop pulmonary hypertension
baffle-inferior vena cava junction (IVC). The can- and may require ECLS during pulmonary hyperten-

MYOCARDITIS DILATED HYPERTROPHIC RESTRICTIVE

CARDIOMYOPATHY CARDIOMYOPATHY CARDIOMYOPATHY
Systolic ↓↓ ↓↓ ↑↑ N/↓
Function
Diastolic N/↓ N ↓↓ ↓↓
Relaxation
Pulmonary - - - +
Hypertension
Arrhythmias +/- +/- + -
Disease Improvement Progressive Progressive Progressive
Course over weeks
Table 8-1. Characteristics of myocarditis and cardiomyopathies.

78
 Neonatal, Pediatric, and Adult Cardiac Extracorporeal Life Support

sive crises. Typical indications for ECLS support in In patients who are cannulated through the fem-
myocarditis and cardiomyopathy patients include oral vessels, monitoring for differential oxygenation
low cardiac output states, cardiac arrest/arrhythmia, should be considered. In those with no native cardiac
and pulmonary hypertension. output, oxygen saturations should be uniform in all
When considering mechanical circulatory extremities as the ECLS circuit provides the only
support for these patients, the time course of each oxygen delivery. However, as these patients regain
disease should also be considered. In children cardiac function, they may begin to eject blood from
with myocarditis, eventual full or partial recovery the LV which may be poorly oxygenated if they
of myocardial function occurs in the majority of have concurrent lung injury or inadequate ventila-
cases. In these patients, mechanical circulatory tor settings. These patients then eject deoxygenated
support (MCS) is likely to be temporary until the blood into the aorta and preferentially into the head
myocardium recovers and survival for myocarditis and neck arteries. This could lead to the delivery
patients on ECLS is typically excellent.19-21 In con- of deoxygenated blood to the brain. If this occurs,
trast, the majority of patients with cardiomyopathy efforts should be made to improve the oxygenation
have a progressive disease that is unlikely to show in the lungs or add a cannula to deliver ECLS to the
significant recovery and ECLS will likely serve RA to ensure adequate oxygen delivery to the brain.
as a bridge to a ventricular assist device (VAD) or
cardiac transplant. Cardiomyopathy patients who Cardiogenic Shock in Acute Myocardial
would not be a candidate for cardiac transplantation Infarction and Chronic Heart Failure
or VAD are poor ECLS candidates.
Cannulation for ECLS typically occurs via the Adult patients with acute myocardial infarction
neck or the femoral vessels, depending on patient (AMI) and decompensated heart failure have been
size and age. As a rule, patients who are old enough supported with ECLS. In this patient population,
to be ambulatory can likely be cannulated through ECLS is used in the setting of refractory cardiogenic
the femoral vessels. A distal perfusion cannula shock (primary pump failure) unresponsive to mul-
should strongly be considered in the leg in which tiple inotropes and intraaortic balloon pump.22-28 In
the arterial cannula was placed. In patients with patients with AMI, ECLS is used for hemodynamic
both myocarditis and cardiomyopathy, ECLS typi- support before, after, or during percutaneous coro-
cally provides the majority of cardiac output and nary intervention.23,25,28 Among patients with acute
flows should be adjusted to achieve adequate tissue decompensated heart failure, ECLS may be used to
oxygen delivery. bridge these patients to long-term MCS such as a
In patients with decreased systolic function, VAD or cardiac transplantation.29 Patients with an
the left heart must be adequately decompressed in acute disease that precipitates cardiogenic shock
order to optimize coronary perfusion and prevent (such as AMI or peripartum cardiomyopathy) are
pulmonary edema, pulmonary hypertension, and more likely weaned off ECLS while patients with a
thrombus formation. Evidence that there is none chronic causes of cardiogenic shock (chronic heart
or inadequate left ventricular emptying, including failure) are more likely to require an alternate form
echocardiogram showing no aortic valve opening, of MCS or transplant.29,30
LV thrombi, or pulmonary hypertension or the Most patients receive cannulation through the
arterial catheter reveals no pulsatility, or evidence femoral vessels, using venoarterial ECLS. Some
of pulmonary edema, the patient should undergo centers favor placing the arterial and venous cannula
further investigation to evaluate and address left on opposite femoral sites to improve venous drain-
heart decompression. This can be accomplished with age of the leg with the arterial cannula. A distal per-
creating an atrial communication, implanting a left fusion cannula may need to be placed in the leg with
atrial drainage cannula, a transaortic pigtail drain, the arterial cannula. As in patients with myocarditis
implantation of an Impella VAD, or increasing LV and cardiomyopathy, efforts must be undertaken to
contraction with inotropes. ensure that the left heart is decompressed. Similar
to patients with myocarditis and cardiomyopathy,

79
Chapter 8

differential hypoxemia must be assessed to ensure ECLS may be used to support patients with pulmo-
that deoxygenated blood is not preferentially sup- nary hypertensive episodes unresponsive to medical
plied to the head. management. This can occur preoperatively and
postoperatively, such as in patients with TAPVR
Pulmonary Hypertension who may present in extremis preoperatively and
who may also have significant increased PVR
Pulmonary hypertension is a term that is used postoperatively. In these patients, there is a high
practically to refer to increased pulmonary vascu- likelihood that the PVR will decrease with medical
lar resistance (PVR). Increased PVR can lead to management over time, permitting weaning from
decreased pulmonary blood flow (PBF) as the in- ECLS. In contrast, patients with severe idiopathic
creased resistance increases the afterload on the right pulmonary hypertension unresponsive to medical
ventricle (RV), leading to decreased RV ejection. In management have a progressive disease with a lower
patients with biventricular pathology, decreased RV likelihood of recovery and clinical improvement. In
ejection leads to decreased PBF and then subsequent these patients, the reasons behind the current pul-
decreased LV preload. Increased PVR also leads to monary hypertensive crises should be understood.
increased RV work which, in turn, can cause RV If the episode results from a transient cause such
dysfunction that further decreases PBF. The end as a viral respiratory infection, the likelihood of
result of this cascade is decreased cardiac output, for weaning from ECLS is probably higher than if the
which patients may require cardiac ECLS. The wide episode is a manifestation of worsening disease. As
variety of causes of pulmonary hypertension include such, consideration should be given as to whether
idiopathic pulmonary hypertension, medication- these patients are also candidates for possible lung
induced pulmonary hypertension, pulmonary hy- assist devices and heart-lung or lung transplant when
pertension associated with congenital heart disease, considering ECLS.
pulmonary hypertension associated with left heart Cannulation and ECLS support type depend
disease, and pulmonary hypertension associated upon the circumstances surrounding clinical de-
with other systemic diseases (such as connective terioration. Patients with CHD and postoperative
tissue disorders). Congenital heart disease patients pulmonary hypertension may be cannulated through
with long-standing, unrepaired left-to-right shunts the chest for VA-ECLS. In contrast, preoperative
(such as tetralogy of Fallot, ventricular septal de- patients with CHD and those with idiopathic pul-
fects),31 lesions involving the pulmonary veins or monary hypertension may be cannulated through
left atrial hypertension (such as total anomalous the neck or femoral vessels. Some patients may
pulmonary venous return [TAPVR]) and single receive adequate support with venovenous ECLS
ventricle physiology (such as some instances of as oxygenated blood, in addition to pulmonary
double outlet right ventricle), who have unrestricted hypertension medications, may decrease the PVR
PBF, can develop increased PVR that may require enough to allow sufficient PBF, providing adequate
ECLS support. Patients with CHD may require preload to the LV. However, if a trial of VV-ECLS
ECLS during pulmonary hypertensive episodes in does not improve the cardiac output, conversion to
the perioperative period or during other intercur- VA is necessary to provide adequate oxygen deliv-
rent illnesses. Patients with idiopathic pulmonary ery. In pulmonary hypertension patients, titrating
hypertension typically have progressive disease the ECLS flows to allow for some RV ejection into
where their PVR increases while their RV function the pulmonary vascular tree may allow oxygenated
worsens. These patients may have severe episodes blood and pulmonary hypertension medications to
of pulmonary hypertension due to intercurrent ill- be delivered to the pulmonary vasculature, possibly
nesses or as a progression of their disease. improving chances of recovery.
As with all diseases requiring ECLS, special
attention must be paid to the estimated time course
and likelihood of recovery when considering ECLS
for pulmonary hypertension. In children with CHD,

80
 Neonatal, Pediatric, and Adult Cardiac Extracorporeal Life Support

Arrhythmias supported with VA-ECLS. It should be noted that

these patients may alternate between normal sinus
As stated previously, patients with myocarditis rhythm and arrhythmia, which may translate into
and cardiomyopathy can develop arrhythmias and alternating periods of normal and reduced cardiac
subsequently require cardiac ECLS to maintain sys- output. ECLS flows may need to be adjusted during
temic perfusion. In many such cases, ECLS may be periods of normal sinus rhythm that may produce
initiated as extracorporeal cardiopulmonary resusci- hypertension.
tation (ECPR) where these patients are cannulated
while they are actively receiving chest compres-
sions in response to a cardiac arrest. Additionally,
other patients with malignant arrhythmias, such as
patients with atrial tachycardias, junctional ectopic
tachycardia, and ventricular tachycardia may also
require ECLS to maintain systemic perfusion.32-35
These arrhythmias can occur in the setting of post-
cardiac surgery patients (both adult and congenital),
adult patients with advanced heart failure, ischemic
cardiac disease, as well as in patients without other
heart disease. In all types of patients, the malignant
arrhythmia is usually associated with decreased
cardiac output. In tachyarrhythmias, the excessive
heart rate and loss of atrial-ventricular synchrony
may prevent adequate ventricular filling, resulting in
inadequate preload and cardiac output. Additionally,
an excessively elevated heart rate also decreases the
diastolic time, which decreases coronary perfusion,
resulting in ischemia of the myocardium. Over a pro-
longed period of time, chronic ischemia may result
in ventricular dysfunction and decreased cardiac
output. Patients with malignant arrhythmias and
marginal hemodynamics are at risk of decompensa-
tion, while medical management (including medi-
cations and catheter ablation) is instituted. In these
instances, ECLS is used to support the cardiac output
until the arrhythmias can be adequately controlled.
Cannulation strategy depends on the type of
patient (ECPR, postoperative, or nonsurgical heart
disease) and any planned arrhythmia interventions.
Postoperative patients are likely to be cannu-
lated centrally while others are usually cannulated
through the neck or femoral vessels. ECPR patients
can be cannulated in all three regions. If ablation
is a therapeutic option, cannulation considerations
should include future venous access that will be
required for an electrophysiologic study and abla-
tion (as this may occur while on ECLS).36 Given
that patients with arrhythmias have insufficient
cardiac output, these patients are almost universally

81
Chapter 8

References in infants with systemic-pulmonary shunts. J

Thorac Cardiovasc Surg. 2016;152:912-918.
1. Gupta P, Robertson MJ, Beam BW, Rettiganti M. 11. Allan CK, Thiagarajan RR, del Nido P, et al.
Outcomes associated with preoperative use of Indication for initiation of mechanical circula-
extracorporeal membrane oxygenation in chil- tory support impacts survival of infants with
dren undergoing heart operation for congenital shunted single-ventricle circulation supported
heart disease: A multi-institutional analysis. with extracorporeal membrane oxygenation. J
Clin Cardiol. 2014;38(2):99-105 . Thorac Cardiovasc Surg. 2007, 133:660-667.
2. Bautista-Hernandez V, Thiagarajan RR, Fynn- 12. Stulak J, Dearani JA, Burkhart HN, et al. ECMO
Thompson F, et al. Preoperative extracorporeal cannulation controversies and complications.
membrane oxygenation as a bridge to cardiac Seminars in Cardiothoracic and Vascular An-
surgery in children with congenital heart disease. esthesia. 2009;13:176-182.
Ann Thorac Surg. 2009;88(4):1306–1311. 13. Agarwal HS, Hardison DC, Saville BR, et al.
3. Luciani GB., Chang AC, Starnes VA. Surgical Residual lesions in postoperative pediatric
repair of transposition of the great arteries in cardiac surgery patients receiving extracorpo-
neonates with persistent pulmonary hyperten- real membrane oxygenation support. J Thorac
sion. Ann Thorac Surg. 1996; 61: 800–805. Cardiovasc Surg. 2014;147: 434-441.
4. Hunkeler NM, Canter CE, Donze A, Spray T. 14. Morris MC, Wernovsky G, Nadkarni, VM.
Extracorporeal life support in cyanotic congeni- Survival outcomes after extracorporeal car-
tal heart disease before cardiovascular operation. diopulmonary resuscitation instituted during
Am J Cardiol. 1992;69 (8):790–793. active chest compressions following refractory
5. Harrington KP, Goldman AP. The role of extra- in-hospital pediatric cardiac arrest. Pediatr Crit
corporeal membrane oxygenation in congenital Care Med. 2004; 5: 440–446.
diaphragmatic hernia. Semin Pediatr Surg 15. Alsoufi B, Al-Radi OO, Nazer RI, et al. Survival
2005;14:72e6. outcomes after rescue extracorporeal membrane
6. Howard TS, Kalish BT, Wigmore D, et al. As- oxygenation for the resuscitation of pediatric
sociation of extracorporeal membrane oxygen- patients with heart disease after cardiac arrest.
ation support adequacy and residual lesions with J Thorac Cardiovasc Surg 2007;134:952-959.
outcomes in neonates supported after cardiac 16. Philip J, Burgman C, Bavare A, et al. Nature
surgery. Pediatr Crit Care Med. 2016;17: 1045- of the underlying heart disease affects survival
1054. in pediatric patients undergoing extracorporeal
7. Aharon AS, Drinkwater DC, Churchwell KB, cardiopulmonary resuscitation. J Thorac Car-
et al. Extracorporeal membrane oxygenation diovasc Surg 201; 148: 2367-2372.
in children after repair of congenital cardiac 17. Su L, Spaeder MC, Jones MB, et al. Imple-
lesions. Ann Thorac Surg, 2001;72:2095-2101. mentation of an extracorporeal cardiopulmo-
8. Chrysostomou C, Morell VO, Kursch BA, et al. nary resuscitation simulation program reduces
Short and intermediate-term survival after ex- extracorporeal cardiopulmonary rtesuscitation
tracorporeal membrane oxygenation in children times in real patients. Pediatr Crit Care Med
with cardiac disease. J Thorac Cardiovasc Surg, 2014; 15:856-860.
2013; 146:317-325. 18. Booth KL, Roth SJ, Thiagarajan RR, et al. Ex-
9. Alsoufi B, Wolf M, Botha P, et al. Late out- tracorporeal membrane oxygenation support of
comes of infants supported by extracorporeal the Fontan and bidirectional Glenn circulation.
membrane oxygenation following the Norwood Ann Thorac Surg. 2004; 77: 1341–1348.
operation. World Journal for Pediatric and Con- 19. Xiong H, Xia B, Zhu J, Li B, Huang W. Clinical
genital Heart Surgery 2015; 6: 9-17. outcomes in pediatric patients hospitalized with
10. Botha P, Deshpande SR, Wolf M, et al. Ex- fulminant myocarditis requiring extracorporeal
tracorporeal membrane oxygenator support membrane oxygenation: A meta-analysis. Pedi-
atr Cardiol. 2017;38:209-214.

82
 Neonatal, Pediatric, and Adult Cardiac Extracorporeal Life Support

20. Lorusso R, Centofanti P, Gelsomino S, et al. 29. Dangers L, Brechot N, Schmidt M, et al. Ex-
Venoarterial extracorporeal membrane oxygen- tracorporeal membrane oxygenation for acute
ation for acute fulminant myocarditis in adult decompensated heart failure. Crit Care Med.
patients: A 5-year multi-institutional experience. 2017;45:1359-1366.
Ann Thorac Surg. 2016;101:919-926. 30. Tarzia V, Bortolussi G, Bianco R, et al. Extracor-
21. Diddle JW, Almodovar MC, Rajagopal SK, poreal life support in cardiogenic shock: Impact
Rycus PT, Thiagarajan RR. Extracorpo- of acute versus chronic etiology on outcome. J
real membrane oxygenation for the support of Thorac Cardiovasc Surg. 2015;150:333-340.
adults with acute myocarditis. Crit Care Med. 31. Cullen M, Splittgerber F, Sweezer W, Hakimi
2015;43:1016-1025. M, Arciniegas E, Klein M. Pulmonary hyperten-
22. Bermudez CA, Rocha RV, Toyoda Y, et al. Extra- sion postventricular septal defect repair treated
corporeal membrane oxygenation for advanced by extracorporeal membrane oxygenation. J
refractory shock in acute and chronic cardiomy- Pediatr Surg. 1986;21:675-677.
opathy. Ann Thorac Surg. 2011;92:2125-2131. 32. Baratto F, Pappalardo F, Oloriz T, et al. Extracor-
23. Chung S, Tong M, Sheu J, et al. Short-term and poreal membrane oxygenation for hemodynam-
long-term prognostic outcomes of patients with ic support of ventricular tachycardia ablation.
ST-segment elevation myocardial infarction Circulation Arrhythmia and electrophysiology.
complicated by profound cardiogenic shock 2016;9.
undergoing early extracorporeal membrane oxy- 33. Le Pennec-Prigent S, Flecher E, Auffret V, et
genator-assisted primary percutaneous coronary al. Effectiveness of extracorporeal life sup-
intervention. Int J Cardiol. 2016;223:412-417. port for patients with cardiogenic shock due to
24. Hryniewicz K, Sandoval Y, Samara M, et al. intractable arrhythmic storm. Crit Care Med.
Percutaneous venoarterial extracorporeal mem- 2017;45:e281-e289.
brane oxygenation for refractory cardiogenic 34. Salerno JC, Seslar SP, Chun TUH, et al. Pre-
shock is associated with improved short- and dictors of ECMO support in infants with
long-term survival. ASAIO J. 2016;62:397-402. tachycardia-induced cardiomyopathy. Pediatr
25. Lee W, Fang C, Chen H, et al. Associations Cardiol. 2011;32:754-758.
with 30-day survival following extracorporeal 35. Scherrer V, Lasgi C, Hariri S, et al. Radiofre-
membrane oxygenation in patients with acute quency ablation under extracorporeal mem-
ST segment elevation myocardial infarction and brane oxygenation for atrial tachycardia in
profound cardiogenic shock. Heart & lung : the postpartum. J Card Surg. 2012;27:647-649.
journal of critical care. 2016;45:532-537. 36. Ucer E, Fredersdorf S, Jungbauer C, et al. A
26. Negi SI, Sokolovic M, Koifman E, et al. Con- unique access for the ablation catheter to treat
temporary use of veno-arterial extracorporeal electrical storm in a patient with extracorpo-
membrane oxygenation for refractory cardio- real life support. Europace : European pacing,
genic shock in acute coronary syndrome. J arrhythmias, and cardiac electrophysiology :
Invasive Cardiol. 2016;28:52-57. journal of the working groups on cardiac pacing,
27. Sattler S, Khaladj N, Zaruba M-, et al. Ex- arrhythmias, and cardiac cellular electrophysi-
tracorporal life support (ECLS) in acute ology of the European Society of Cardiology.
ischaemic cardiogenic shock. Int J Clin Pract. 2014;16(2):299-302.
2014;68:529-531.
28. Tsao N, Shih C, Yeh J, et al. Extracorporeal
membrane oxygenation-assisted primary per-
cutaneous coronary intervention may improve
survival of patients with acute myocardial in-
farction complicated by profound cardiogenic
shock. J Crit Care. 2012;27:530.e1-530.11.

83
9

Extracorporeal Membrane Oxygenation for Cardiopulmonary Arrest and Resuscitation

Anne-Marie Guerguerian, MD, PhD, Kyle Gunnerson, MD, Ravi R. Thiagarajan, MBBS, MPH

Introduction bridge to decision that can lead to a bridge to organ

recovery (or transplant) and delivery of therapy (in-
Survival with good neurologic outcomes fol- cluding another ECLS device), or that can transition
lowing standard cardiopulmonary resuscitation to a bridge to palliative care.
(CPR) for cardiopulmonary arrest (CPA) either in The use of ECPR is increasing around the world
the in-hospital (IH) or out-of-hospital (OOH) set- (Figure 9-1) but requires proper systems, resources,
tings has improved but remains poor in children1-4 expertise, team training, and performance assess-
and in adults.5,6 Prolonged CPR without return of ment. Teams in many institutions can deploy ECMO
circulation (ROC) is associated with poor survival rapidly and successfully during CPR. However,
and neurological outcomes. ECPR is the rapid de- standards have evolved over several decades, and
ployment of venoarterial (VA) extracorporeal mem- ECPR should not be applied as an ad hoc procedure.
brane oxygenation (ECMO) or cardiopulmonary It should be considered only in select populations
bypass to provide immediate cardiovascular support when delivered by organized systems.7 Applying
for patients in the context of CPA.7 The objective ECPR is different than emergency preservation and
of ECMO instituted as a resuscitation measure in resuscitation (EPR) for cardiac arrest from trauma;
patients in CPA is to provide circulation and gas the later involves rapid intraaortic retrograde
exchange (delivery of oxygen and removal of CO2). flushing of ice cold preservation solution, venous
During the early postcardiac arrest care (PCAC) drainage and deep hypothermia (<10°C) within
period, ECMO support may allow for time to iden- minutes developed in animals10-12 and piloted in
tify the cause of the event, to facilitate diagnostic trauma trials.13
procedures (e.g. interventional cardiac catheteriza- Rapid deployment of ECMO for patients in
tion), or to facilitate intra or interhospital transport CPA was pioneered systematically in patients with
to undertake specific therapies, surgeries, or inter-
ventions. The ELSO Registry collects information
TOTAL ECMO % ECPR USE
on children undergoing ECMO cannulation during
cardiac compressions since 1992 and with a separate 7000 25

addendum since 2011.8 An updated terminology is 6000

under consideration to classify patients supported 5000
20

with ECPR when ECMO flows are instituted dur- 15

4000
ing CPR (manual or mechanical) and up to within
20 minutes of return spontaneous of circulation
3000
10

(ROSC) (without ongoing compressions). Patients 2000

5
cannulated after 20 min of sustained circulation 1000

would not be classified as ECPR.9 0 0

Before the initiation or during deployment, the

1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015

purpose of ECMO initiation is defined by the medi- Figure 9-1. Trends in ECPR use among ELSO
cal team to guide the care plan. ECPR can serve as a Centers 1992 to 2015.

85
Chapter 9

primary cardiac diseases14 in the hospital settings tions before and during ECMO deployment. This
(cardiac intensive care units, operating rooms and resuscitation is coordinated by a physician leader
cardiac catheterization laboratories). The success who ensures the delivery of high quality CPR and
and experience gained in using this technology has launches the ECMO team. A second group of indi-
led to the expansive utilization of ECMO in patients viduals includes surgeons (or physicians) dedicated
with a wider variety of diagnoses, who suffer cardiac to rapid ECMO cannulation and others who prepare
arrest in a variety of settings including in the OOH the ECMO circuit, and diagnose and manage un-
settings.15,16 expected complications. The process of instituting
ECPR requires the coordination of a large number
Patient Selection of skilled providers and thus should always have
an event manager. The individual serving as event
Despite its rapid expansion, patient selection manager must also assure that adequate ECMO
remains critically important to provide optimal support is provided once on ECMO. In addition to
cardio-cerebral CPR.17 It is unknown whether CPR the event manager, a health care provider who can
measures or techniques should be modified before, mobilize resources beyond ECMO equipment is
during, and after ECMO initiation, to optimize also necessary. The team must apply appropriate
survival and best functional outcomes in survivors. communication and crisis resource management
Different criteria are used in different institutions principles.20 Finally, a calling system (e.g. team
(e.g. >2 kg) or regional organizations (e.g. adult paging system) to efficiently mobilize the entire
cardiac arrest with shockable-witnessed-with by- ECPR team is crucial.
stander CPR in adults less than a certain age),18 and ECPR teams vary in composition across in-
these criteria evolve as new evidence in resuscita- stitutions and within in-or out-of-hospital settings.
tion science becomes available. Initiation of ECPR Examples of tasks and team members are shown
to support the organ donor involves a different set in Table 9-1. In some organizations, ECPR teams
of protocols (e.g. organ donation in uncontrolled are available to deploy ECMO during CPR 24/7.
donation)19 and is not included herein. Whether an ECPR team should be available on-site
Because maintaining a sustainable ECPR sys- 24/7 remains controversial and is not universal. Be-
tem and expert team requires considerable resources, cause ECPR is low volume high risk event, system
more information on operational issues are urgently and team training and testing, using simulation is
needed. This chapter focuses on current basic necessary to maintain skilled and high performing
knowledge and practical pointers for launching teams.
the team, rapid preparation, and use of an ECMO
system. The legal and ethical framework related Equipment
to ECMO in the context of the cardiopulmonary
resuscitation are covered elsewhere. (See related ECMO equipment including pumps and oxy-
chapters in ELSO Red Book, 5th edition) genators to provide ECPR varies widely. ECPR
equipment should be stored so that they can be
The Team mobilized rapidly on rolling carts. Most centers
use a version of a preassembled and/or wet prime
ECPR requires a well-coordinated team of ECMO circuits that can be rapidly deployed. In ad-
expertly trained professionals to ensure rapid and dition, many centers have ECMO carts that contain
successful deployment. A general approach involves a wide range of cannulas and connectors that can be
having multiple concurrent tasks performed by mobilized to the bedside or the site of ECMO can-
individuals who know their role and their responsi- nulation. Finally, surgical instruments for ECMO
bilities (Tables 9-1 and 9-2). Standard resuscitation cannulation should also be readily available and
measures are initiated and delivered by a first group brought to the bedside.
of individuals, often led by an ICU physician, who ECMO circuits can be configured with either
focus on high quality CPR with minimal interrup- roller or centrifugal pumps. ECMO circuits us-

86
 Extracorporeal Membrane Oxygenation for Cardiopulmonary Arrest and Resuscitation

ing centrifugal pumps are generally compact and to support one target temperature vs. another,21 a
have improved mobility compared to roller pumps. set target temperature should be defined by the
Some centers configure ECMO circuits with a heat responsible physician based on patient factors and/
exchanger. Heat exchanger can be used to provide or institutional protocols, and hyperthermia should
targeted temperature management (TTM) set to be avoided. Equal to blood flow and to blood pres-
hypothermia on initiation for neurological protec- sure targets, core temperature targets should not be
tion. There are no comparative trials that suggest random; temperature should be measured centrally
an optimal target temperature; however, in Boston especially when no heat exchanger is used.
Children’s Hospital and The Hospital for Sick Some centers prime their ECPR circuits with
Children, the heat exchanger is set to 34°C. Fol- crystalloid solutions (e.g. lactated ringers or plas-
lowing cannulation, temperature management is malyte) while other use blood prime circuit. No evi-
ordered, including target temperature. While there dence supports one priming method above another.
is no evidence based on randomized control trials However, delays ECMO deployment while awaiting

Task Team Member

Patient Type Pediatric Adult
Location of event* BCH ICU SickKids** ICU UM ED
Stat ECMO group Charge nurse ICU physician ED physician 1
page faculty or delegate
Event manager ICU physician ICU faculty ED physician 2
physician
CPR team leader ICU physician ICU fellow ED physician 2
ECMO cannulation Surgeon CVS surgeon 1 ECPR physicians 1
(in-house) (fellow in-house) and 2
CVS surgeon 2
ECMO circuit ECMO specialist ECMO specialist 1 ECMO specialist 1
(in-house) (in-house)
Perfusion specialist ECMO specialist 2
or ECMO specialist
-2
CPR medications Beside nurse Bedside & Bedside nurses
medication nurses (1, 2)
(1, 2, 3)
Compressions 2 assigned 2 assigned Manual or
mechanical device
Heparin bolus Bedside nurse ICU staff physician Bedside nurse at the
& ECMO specialist direction of ECPR
double check physician 1
Resources Charge nurse Charge or clinical Charge nurse or
support nurse clinical support
nurse
Airway Respiratory Respiratory Respiratory
therapist therapist therapist with ED
physician
Documentation Assigned Documentation Documentation
nurse or RT nurse
BCH=Boston Children’s Hospital, SickKids: Hospital for Sick Children;
UM=University of Michigan.
*Institutions may apply different versions of this role and responsibilities based on the
location of CPA event and the predefined location for cannulation.
**Electronic Centralized Team Page & Members who do not call back but go directly
to cannulation location: CVS Surgeon (fellow and faculty), ECMO Specialist,
Perfusion Specialist

Table 9-1. Tasks and team members with preassigned roles and responsibilities.

87

1
Chapter 9

Role Responsibility
ICU Physician 1 (ICU Fellow) Code Team Leader focused on high quality CPR
Calls ICU Faculty if not bedside
ICU Physician 2 (ICU Faculty) Event Manager
Launches E-CPR team page weight and cannulation location
Registered Nurse (RN) 1 Charge/clinical support nurse
Assigns roles
Brings defibrillator and resuscitation cart
Enforces crowd control
Ensures all equipment is available
RN 2 Assistant to cannulating surgeons
Ensures patient is in proper position with CPR board under
patient
Provide sutures, dressings and other necessary surgical
supplies
Cardiothoracic Surgeons 1 and 2 Places roll under neck or hips
(Fellow and Faculty) Opens all ECMO trays
Assigns CPR compressors with sterile gloves (2 people)
Prepares surgical sites for cannulation
ECMO Specialist/Perfusion Call blood bank and requests blood products
Specialist (2 people) Assemble ECMO circuit, cannula, and ACT machine carts
Prepare ECMO circuit for initiation
Assist surgeons with selecting cannula and connectors
Add medications and albumin to circuit only when cannulas
are being inserted
Double checks Heparin patient bolus (50U/kg) dose with
ICU physician faculty
Blood primes circuit if indicated and time allows
Respiratory Therapist (RT) 1 Responsible for airway & ventilator
Continuous end tidal CO2
Prepares equipment for intubation
Runs blood gas samples
Bedside RN 1 Ensures bedside set up complete
Connects medication delivery vascular access line to patient
Administers medications and fluids to patient
Communicates clearly to Recorder what and when
medications are given
Takes patient blood samples
Completes electronic medical record documentation (vitals,
medications, fluids)
Medication RN 1 and RN 2 Prepare medications and fluids
Label and double checks all medications and fluids
Recorder 1 (RN or RT) Positions self to oversee resuscitation, read monitor, hear
ordering physician and confirm that medications and fluids
have been administered by Bedside RN, using stool if
necessary
Keeps track of time of events
Completes documentation of resuscitation and ensures
record is signed
Social Worker Supports family
Table 9-2. The Hospital for Sick Children roles & responsibilities for ECPR cannulation location in ICU.
(Note: Cannulations in the Operating Room or in the Catheterization Laboratory involve some adaptation to
this general list.)

88
 Extracorporeal Membrane Oxygenation for Cardiopulmonary Arrest and Resuscitation

availability of blood for priming is unacceptable. Institutions should deliver antibiotics for surgical
While centers have been using ECPR for several prophylaxis based on their local usual ECMO can-
decades with local protocols, there is no published nulation protocols (for peripheral or open chest
evidence to guide oxygenator gas management dur- cannulation).
ing ECPR deployment, and the practice has been
adopted from the cardiopulmonary bypass (CPB) Space and Location
clinical practice. For the IH setting, many centers
use blended gases on their standard ECMO circuits The patient, equipment, and team members
(O2 & Air) to titrate FdO2 and attempt to minimize involved in ECPR require a large amount space.
hyperoxia as it may worsen ischemia related reper- Either for IH or OOH initiations, most organiza-
fusion injury. Similarly, carbon dioxide management tions use a space map to organize team members
in ECPR has not been studied and is extrapolated and equipment around the bed that anticipates the
from the neonatal respiratory ECMO, anesthesia, requirements of cannulation (e.g. location of the
CPB literature;22 early measurement of arterial blood cannulation surgeon and the ECMO machine and
gases on initiation will help avoid hypocapnia and cart). Adult and pediatric space map examples are
hypercapnia which alter cerebral blood flow and included (Figure 9-2, and Table 9-3). Each system
may influence neurologic outcome. should designate locations suitable and predefined
for cannulation (e.g. ICU or catheterization labora-
Vascular Access and Cannulation for ECPR tory) and have a protocol to transport the patient to
these locations with ongoing CPR. These locations
Cannulation for ECPR must be achieved rap- are assigned based on patients at risk, size, and fea-
idly. Either open surgical cutdown or percutaneous tures of the location necessary to accommodate the
techniques can be used; percutaneous cannulation is
being increasingly used in adult ECPR. Institutional Team Members (#)
Primary ED team (2-3)
Role (s)
Code Leader—Direct ACLS,
Position
Code Leader-foot of bed
protocols should set the default approach in advance direct flow of resuscitation.
Airway—secure airway, then
Airway-head of bed

to expedite the procedure. Choice of cannulation assist with echo, IVC

ultrasound, and place right
site is determined by the desired target blood flow ECPR physicians (2)
radial arterial line
ECPR 1—Obtain femoral ECPR 1—at right groin if
estimated by patient size and the need based on the venous/arterial access, final
review of inclusion/exclusion
right-handed, left groin if
left-handed
disease (witnessed ventricular fibrillation arrest in criteria, venous and arterial
cannulation and connection to
ECPR 2—opposite ECPR
1
OOH setting vs. postcardiotomy arrest in-intensive circuit.
ECPR 2—Assist ECPR 1 with
care unit). Chest or central cannulation may be used equipment, wire control during
cannulation, flushing during
in patients who have undergone a recent sternotomy connection to circuit
Registered Nurse RN (3-4) Recording RN—document Recording RN—to one
for cardiac surgery. In these cases, the right atrium interventions side
Bedside RN1—IV access, RN1 and RN2—on either
and aorta may be cannulated directly. In young medications, fluids, side of patient

children the internal jugular vein and carotid artery mechanical CPR devices,
arterial line set up
may be used whereas the femoral veins and arteries Bedside RN2—IV access,
medications, fluids,
may be used in older children and adults. Neck can- mechanical CPR devices,
arterial line set up
nulation was shown to be associated with reduced Optional: RN Team Leader for
extra coordination/assistance
mortality in children in one series.23 Peripheral com- Respiratory Therapy (1) Control ventilation, set up
waveform capnography
Head of bed

pared to central cannulation may be associated with Tech/Paramedic (1-2) IV access, lab draws, cardiac
monitors, defibrillation
To either side of patient

less frequent interruptions to cardiac compressions Pharmacist Prepare code medications,

vasoactive infusions, and
At the left foot of the
patient
during ECPR, however there no evidence exists to heparin
ECMO specialists (2) Prepare and operate ECMO At the right foot of the
support one approach over the other. In children and circuit patient -
Social work Obtain patient information, Doorway
adults who have had vessels accessed previously, including known comorbidities
and social history, from family
the vascular integrity may be compromised; it is and/or EMS

important to have knowledge of occluded vessels

Table 9-3. University of Michigan Department of
to prevent delay in connection to ECMO by wast- Emergency Medicine adult ECPR roles related to
ing time accessing narrow or thrombosed vessels. Figure 9-2A Space Map 1

89
Chapter 9

proper performance of the team and the equipment. core protocol includes general patient selection and
Transporting a patient while undergoing active initiation criteria, tasks, roles and responsibilities,
CPR is complex and may compromise the quality location of cannulation if not in the location of the
of CPR; however, completing a cannulation and cardiac arrest, instructions for transport and CPR
initiating proper postcardiac arrest care, should be measures. Figure 9-3 shows an example of ECPR
orchestrated in the environment that will optimize process care flow diagram.
patient outcomes. Institutions should establish internal process
measures to review their performance and detect
ECPR Algorithm, Protocols, Checklists, Cogni- vulnerable areas in the algorithm. Prolonged times
tive Aids to attaining ECPR blood flow are associated with
worse outcomes although no universal time-to-
Most organizations have a written protocol, a cannulation benchmarks for IH or for OOH events
list of roles, responsibilities, ordered tasks, indi- exist. For in-hospital cardiopulmonary arrests, at the
vidual or shared checklists, a process flow diagram Hospital for Sick Children and Boston Children’s
or algorithm, to enhance process performance and Hospital, we aim to have processes in place to
minimize disruptive variability. Because multiple achieve ROC within 30 minutes, in any CPA, either
tasks are accomplished simultaneously, the ECPR by return of extracorporeal circulation (ROEC) or
algorithm should specify tasks, order of comple- ROSC. If ROSC occurs within 30 min, the respon-
tion, and person responsible for completion of sible physician either decides to pursue extracor-
task during ECPR. The algorithm can also contain poreal circulation or not based on patient factors
“time-outs” to ensure safety and should target a (e.g. low cardiac output or profound hypoxemia).24
time to full ECMO flows during CPR. No refer- Understanding the process lag times and the inter-
ence documents or cognitive aids can ensure high vals required to launch the ECPR team members
performance; fundamentally, ECPR does not lend and have them arrive at bedside (approximately
itself to improvisation, rather predetermined tasks 5 minutes), conduct the cannulation (approximately
should be assigned before the event as described 15-17 minutes) while providing high quality CPR,
above (e.g. racecar pit teams). A written protocol and prepare the circuit, helps the event manager
describing predefined roles and responsibilities of during the procedure to direct care (ECPR decision
team members must be followed. It includes all steps and request by the 7th minute of CPR). These bench-
to operationalize the efficient conduct of ECPR. The marks are used for performance review, and for

Figure 9-2. Examples of space maps used at (A) the University of Michigan Emergency Department and (B)
the Hospital for Sick Children.

90
 Extracorporeal Membrane Oxygenation for Cardiopulmonary Arrest and Resuscitation

team and system training (these are never used as a Postcardiac Arrest Care (PCAC) Following
clock times to stop resuscitation measures). Mature ECPR
regional out-of-hospital systems in Europe and Asia
use well-coordinated ambulance retrieval systems Postcardiac arrest care in patients who receive
to minimize delays when integrating ECPR. Each ECPR begins immediately after ROC and gas ex-
geographic setting that uses ECPR predetermines change have been established. ECMO flow must be
the option of “stay and play” to cannulate in the field optimized and end-organ perfusion restored in order
or “pack and go” to transport to the ECPR centers. to optimize oxygen delivery. Adequate ECMO flow
In the adult ECPR out-of-hospital setting, outcomes may be indicated by improved clinical perfusion and
are generally best in witnessed cardiac arrest events neurologic function, improvement and resolution
with intervals from CPA to ROEC than span less of laboratory markers of ischemia (e.g. metabolic
than 60 minutes (see outcomes below). The general acidosis or lactate), and restoration of urine output.
process algorithm referred to above1 is helpful to Inotropic support if used prior to the arrest should be
indicate when the patient should be transported weaned as much as tolerated. Although recent stud-
(either IH or OOH) to the cannulation location if ies have shown no difference in survival with good
different from the location of the CPA event. neurological outcomes in children who received
therapeutic hypothermia, ECPR programs provide
temperature targeted management (TTM) and some
maintain patients in with hypothermia (33–34o C)
for 24 to 48 hours post arrest to optimize neurologic
outcomes. In these institutions, duration of TTM and
Cardiac Arrest
CPR Initiated rewarming protocols are applied. TTM should not
Ice Packs to the Head
interfere with established ECMO care plans.
5 min CPR Diagnostic or therapeutic procedures may
No ROSC
include radiological imaging and interventional
STAT ECMO page* cardiac catheterization. These should be planned and
safely undertaken without delay. Patients with signs
Optimize CPR of left atrial hypertension and pulmonary edema/
Team Members Assembled
Blood Bank Notification
hemorrhage may require emergent left ventricle
ECMO/Surgical Equipment to Bedside (LV) decompression. This can be achieved in the
catheterization laboratory by creating a communi-
cation between the right and left atria.25 In adults,
TIME OUT #1 (Surgeon) adequate coronary perfusion is a key target and clini-
Patient Identification
Cannulation Plan** cians aim to maintain diastolic blood pressures >35
mmHg to help achieve target coronary perfusion
ECMO Cannulation completed pressure of >25 mmHg (assuming a central venous
pressure of 10 mmHg).26 Before using interven-
TIME OUT #2
tions or devices to decompress the LV, one strategy
(Surgeon and ECMO Specialist) to assist LV emptying involves decreasing ECMO
Identify correct Arterial &
Venous connections flow slightly to allow for the atrial-ventricular valve
to open and the ventricle to eject, with or without
inotropic support. Intraaortic balloon pump (IABP)
or a percutaneous Impella device placement across
ECMO FLOW
Goal 100ml/kg/min or 2L/min the aortic valve via the femoral artery may provide
Post-resuscitation Management
left sided decompression.27 Patients cannulated via
* STAT ECMO pages Cannulating Surgeon and ECMO specialist the femoral artery require a reperfusion cannula to
** Peripheral Vessel Cannulation: Check Vascular Occlusion Sheet perfuse the leg distal to the ECMO cannula should
Figure 9-3. be undertaken without delay. Because neurological

91
Chapter 9

injury occurs commonly in ECPR patients, assess- Ethics and Informed Consent
ment and monitoring of neurologic status is para-
mount. Neurologic assessment may include awak- Because of the rapid nature of ECPR deploy-
ening with clinical assessment, neuromonitoring ment obtaining informed consent for ECPR is im-
(cerebral oximetry, continuous EEG), neuroimaging, practical. In high volume centers (cardiac, transplant,
and consultation with a neurologist. Neurologic surgical) information should be given and consent
assessment for future prognostication is beyond secured at the time of the anesthesia and/or surgi-
the scope of this chapter, but is an integral part of cal consent. However, in unexpected ECPR events
ECPR postcardiac arrest care, and individualized guardians (or surrogate decision makers) should
care based on pediatric or adult evidence should be informed soon after stable vital signs have been
guide practice accordingly. achieved. These conversations should include po-

Author Year Diagnosis Institution Total Survival

Pediatric Patients
del Nido28 1992 Cardiac Pittsburg 11 64%
Dalton29 1993 Cardiac Pittsburg 29 45%
Duncan30 1998 Cardiac Boston 11 54%
Morris31 2004 All Philadelphia 64 33%
Thiagarajan14 2007 All ELSO-R 682 38%
Alsoufi32 2007 All Toronto 80 34%
Huang33 2008 All Taiwan 27 41%
Tajik34 2008 All Metaanalysis 288 40%
Chan23 2008 Cardiac ELSO-R 492 42%
Kane35 2010 Cardiac Boston 172 51%
Raymond36 2010 All GWTG-R 199 44%
Wolf37 2012 Cardiac Atlanta 150 56%
Lasa38 2016 All GWTG-R 591 40%
Adult Studies
Younger39 1999 Cardiac Ann Arbor 25 36%
Chen40 2008 All Taiwan 59 24%
Thiagarajan41 2009 All ELSO 297 27%
Fagnoul42 2013 IHCA, Brussels 24 25%
OHCA
Chou43 2014 IHCA Taiwan 43 35%
Sawamoto44 2014 Hypothermia Sapporo 26 39%
Sakamoto45 2014 OHCA Japan 260 12.3%*
Stub46 2013 IHCA,OHCA Melbourne 24 50%
Yannopoulos47 2017 OHCA Minnesota 50 45%**
All=includes patients with cardiac and noncardiac diagnosis; ELSO-
R=Extracorporeal Life Support Organization Registry; GWTG-R=Get
With The Guidelines Registry. IHCA=in-hospital cardiac arrest; OHCA:
out-of-hospital cardiac arrest.
*Outcome rate of favorable neurological outcome (Cerebral
performance categories 1 & 2) at one month.
**Survivors are reported among the prospectively enrolled cohort
28/62; a subgroup of 47 underwent percutaneous coronary intervention
(PCI) after OHCA. 62 OHCA prospectively enrolled, where 50 were
placed on ECMO, 5 had ROSC and 7 died; 8 of 50 ECMO died early.
42 ECMO and 5 with ROSC underwent PCI, of which 28 survived

Table 9-4. Survival to hospital discharge in children and adults

supported with ECPR.

92
 Extracorporeal Membrane Oxygenation for Cardiopulmonary Arrest and Resuscitation

tential reasons for cardiac arrest, a plan of care that

involves the possibilities of survival and of death,
and prognostic information based on patient relevant
data that may be available. These conversations are
best conducted on a regular basis, which allows
families to be aware of changes in condition or
prognostic information.

ECPR Outcomes

Survival following ECPR reported in adult and

pediatric studies are shown in Table 9-4. Rare are
studies comparing ECPR vs. standard CPR. Some
studies comparing survival with good neurologic
function at hospital discharge between patients
receiving ECPR and those managed with standard
CPR show improved outcomes for ECPR patients.
However, these studies are retrospective analyses
(with residual selection bias and confounders)
and thus limit inferential conclusions. In general
survival to hospital discharge for children is higher
compared to adults. Better outcomes for children
may be related to the fact that ECPR is largely of-
fered to children who are inpatients and often are in
intensive care unit environments, with significantly
shorter times to return of extracorporeal circulation
(ROEC), and cannulation strategies that more often
exploit neck or central vessels compared to femoral
access. In adults, ECPR used for witnessed OOH
cardiac arrest is an emerging approach in systems
with protocols and resources in place, most often
using a femoral VA cannulation approach. Neuro-
logic injury results commonly and these patients
often die prior to hospital discharge. Long-term
outcomes for patients who suffered cardiac arrest
and were exposed to ECPR including functional
and cognitive outcomes remain undefined. Early
crude neurological assessment at the time of hospital
discharge among survivors suggests that they have
good neurological outcomes.

93
Chapter 9

References Maastricht Treaty for Nomenclature in Extra-

corporeal Life Support. A Position Paper of the
1. Thiagarajan R, Brediger S, Larsen S, Flynn- Extracorporeal Life Support Organization. Am
Thompson F, Alexander P. Extracorporeal J Respir Crit Care Med. 2018.
Cardiopulmonary Resuscitation in Children. 10. Safar P, Tisherman SA, Behringer W, Capone A,
In: TV B, L L, R L, G M, G P, editors. ELSO Prueckner S, Radovsky A, et al. Suspended ani-
Red Book 5th Edition ed. Ann Arbor, Michigan, mation for delayed resuscitation from prolonged
USA: Extracorporeal Life Support Organiza- cardiac arrest that is unresuscitable by standard
tion; 2017. p. 321-330. cardiopulmonary-cerebral resuscitation. Crit
2. Tijssen JA, Prince DK, Morrison LJ, et al. Care Med. 2000;28(11 Suppl):N214-218.
Time on the scene and interventions are as- 11. Kutcher ME, Forsythe RM, Tisherman SA.
sociated with improved survival in pediatric Emergency preservation and resuscitation
out-of-hospital cardiac arrest. Resuscitation. for cardiac arrest from trauma. Int J Surg.
2015;94:1-7. 2016;33(Pt B):209-212.
3. Fink EL, Prince DK, Kaltman JR, et al. Un- 12. Moffatt SE, Mitchell SJB, Walke JL. Deep and
changed pediatric out-of-hospital cardiac ar- profound hypothermia in haemorrhagic shock,
rest incidence and survival rates with regional friend or foe? A systematic review. J R Army
variation in North America. Resuscitation. Med Corps. 2017.
2016;107:121-128. 13. NCT01042015 Cg. Emergency Preservation and
4. Girotra S, Spertus JA, Li Y, et al. Survival trends Resuscitation for Cardiac Arrest From Trauma.
in pediatric in-hospital cardiac arrests: an analy- 2017.
sis from Get With the Guidelines-Resuscitation. 14. Thiagarajan RR, Laussen PC, Rycus PT, Bartlett
Circ Cardiovasc Qual Outcomes. 2013;6(1):42- RH, Bratton SL. Extracorporeal membrane
49. oxygenation to aid cardiopulmonary resus-
5. Girotra S, Chan PS. Trends in survival af- citation in infants and children. Circulation.
ter in-hospital cardiac arrest. N Engl J Med. 2007;116(15):1693-1700.
2013;368(7):680-681. 15. Wang CH, Chou NK, Becker LB, et al. Im-
6. Grunau B, Reynolds JC, Scheuermeyer FX, proved outcome of extracorporeal cardiopul-
et al. Comparing the prognosis of those with monary resuscitation for out-of-hospital cardiac
initial shockable and non-shockable rhythms arrest--a comparison with that for extracorpo-
with increasing durations of CPR: Informing real rescue for in-hospital cardiac arrest. Resus-
minimum durations of resuscitation. Resuscita- citation. 2014;85(9):1219-1124.
tion. 2016;101:50-56. 16. de Chambrun MP, Brechot N, Lebreton G, et
7. de Caen AR, Kleinman ME, Chameides L, et al. Venoarterial extracorporeal membrane
al. Part 10: Paediatric basic and advanced life oxygenation for refractory cardiogenic shock
support: 2010 International Consensus on Car- post-cardiac arrest. Intensive Care Med.
diopulmonary Resuscitation and Emergency 2016;42(12):1999-2007.
Cardiovascular Care Science with Treatment 17. Safar P. Cerebral Resuscitation from Temporary
Recommendations. Resuscitation. 2010;81 Complete Global Brain Ischemia In: Pinksy
Suppl 1:e213-259. CM, Vincent JL, editors. Cerebral Blood Flow
8. Barbaro R, Rycus P, Conrad S, Thiagarajan R, Mechanisms of Ischemia, Diagnosis, and
Paden M. The Registry of the Extracorporeal Therapy 2002.
Life Support Organization. In: TV B, L L, R L, 18. Belohlavek J, Chen Y-S, Morimura N. Extra-
G M, G P, editors. ELSO RedBook 5th Edition. corporeal Cardiopulmonary Resuscitation in
Ann Arbor, Michigan, USA: Extracorporeal Adults. In: TV B, L L, R L, G M, G P, editors.
Life Support Organization; 2017. p. 809-814. ELSO RedBook 5th Edition ed. Ann Arbor,
9. Conrad SA, Broman LM, Taccone FS, et al. Michigan, USA: Extracorporeal Life Support
The Extracorporeal Life Support Organization Organization; 2017. p. 501-16.

94
 Extracorporeal Membrane Oxygenation for Cardiopulmonary Arrest and Resuscitation

19. Rodriguez-Arias D, Deballon IO. Protocols for after cardiac surgery. Circulation. 1992;86(5
uncontrolled donation after circulatory death. Suppl):II300-304.
Lancet. 2012;379(9823):1275-1276. 29. Dalton HJ, Siewers RD, Fuhrman BP, et al.
20. Allan CK, Thiagarajan RR, Beke D, et al. Sim- Extracorporeal membrane oxygenation for car-
ulation-based training delivered directly to the diac rescue in children with severe myocardial
pediatric cardiac intensive care unit engenders dysfunction. Crit Care Med. 1993;21(7):1020-
preparedness, comfort, and decreased anxiety 1028.
among multidisciplinary resuscitation teams. J 30. Duncan BW, Ibrahim AE, Hraska V, et al. Use
Thorac Cardiovasc Surg. 2010;140(3):646-652. of rapid-deployment extracorporeal membrane
21. Moler FW, Silverstein FS, Holubkov R, et oxygenation for the resuscitation of pediatric
al. Therapeutic Hypothermia after In-Hospital patients with heart disease after cardiac arrest.
Cardiac Arrest in Children. N Engl J Med. J Thorac Cardiovasc Surg. 1998;116(2):305-311.
2017;376(4):318-329. 31. Morris MC, Wernovsky G, Nadkarni VM.
22. Cashen K, Reeder R, Dalton HJ, et al. Hyper- Survival outcomes after extracorporeal car-
oxia and Hypocapnia During Pediatric Extracor- diopulmonary resuscitation instituted during
poreal Membrane Oxygenation: Associations active chest compressions following refractory
With Complications, Mortality, and Functional in-hospital pediatric cardiac arrest. Pediatr Crit
Status Among Survivors. Pediatr Crit Care Med Care Med 2004;5(5):440-446.
2018;19(3):245-253. 32. Alsoufi B, Al-Radi OO, Nazer RI, et al. Survival
23. Chan T, Thiagarajan RR, Frank D, Bratton outcomes after rescue extracorporeal cardiopul-
SL. Survival after extracorporeal cardiopul- monary resuscitation in pediatric patients with
monary resuscitation in infants and children refractory cardiac arrest. J Thorac Cardiovasc
with heart disease. J Thorac Cardiovasc Surg. Surg. 2007;134(4):952-959 e2.
2008;136(4):984-992. 33. Huang SC, Wu ET, Chen YS, et al. Extracor-
24. de Mos N, van Litsenburg RR, McCrindle poreal membrane oxygenation rescue for car-
B, Bohn DJ, Parshuram CS. Pediatric in- diopulmonary resuscitation in pediatric patients.
intensive-care-unit cardiac arrest: incidence, Crit Care Med. 2008;36(5):1607-1613.
survival, and predictive factors. Crit Care Med. 34. Tajik M, Cardarelli MG. Extracorporeal mem-
2006;34(4):1209-1215. brane oxygenation after cardiac arrest in chil-
25. Kotani Y, Chetan D, Rodrigues W, et al. Left dren: what do we know? Eur J Cardiothorac
atrial decompression during venoarterial extra- Surg. 2008;33(3):409-417.
corporeal membrane oxygenation for left ven- 35. Kane DA, Thiagarajan RR, Wypij D, et al.
tricular failure in children: current strategy and Rapid-response extracorporeal membrane oxy-
clinical outcomes. Artif Organs. 2013;37(1):29- genation to support cardiopulmonary resuscita-
36. tion in children with cardiac disease. Circulation.
26. Paradis NA, Martin GB, Rivers EP, et al. Coro- 2010;122(11 Suppl):S241-8.
nary perfusion pressure and the return of spon- 36. Raymond TT, Cunnyngham CB, Thompson MT,
taneous circulation in human cardiopulmonary et al. Outcomes among neonates, infants, and
resuscitation. JAMA. 1990;263(8):1106-1113. children after extracorporeal cardiopulmonary
27. Parekh D, Jeewa A, Tume SC, et al. Percu- resuscitation for refractory inhospital pediatric
taneous Mechanical Circulatory Support Us- cardiac arrest: a report from the National Reg-
ing Impella(R) Devices for Decompensated istry of Cardiopulmonary Resuscitation. Pediatr
Cardiogenic Shock: A Pediatric Heart Center Crit Care Med 2010;11(3):362-371.
Experience. ASAIO J. 2018:64(1):98-104. 37. Wolf MJ, Kanter KR, Kirshbom PM, Kogon BE,
28. del Nido PJ, Dalton HJ, Thompson AE, Siew- Wagoner SF. Extracorporeal cardiopulmonary
ers RD. Extracorporeal membrane oxygen- resuscitation for pediatric cardiac patients. Ann
ator rescue in children during cardiac arrest Thorac Surg. 2012;94(3):874-879; discussion
879-880.

95
Chapter 9

38. Lasa JJ, Rogers RS, Localio R, et al. Extracor- 47. Yannopoulos D, Bartos JA, Raveendran G, et
poreal Cardiopulmonary Resuscitation (E-CPR) al. Coronary Artery Disease in Patients With
During Pediatric In-Hospital Cardiopulmonary Out-of-Hospital Refractory Ventricular Fi-
Arrest Is Associated With Improved Survival brillation Cardiac Arrest. J Am Coll Cardiol.
to Discharge: A Report from the American 2017;70(9):1109-1117.
Heart Association’s Get With The Guidelines-
Resuscitation (GWTG-R) Registry. Circulation.
2016;133(2):165-176.
39. Younger JG, Schreiner RJ, Swaniker F, Hirschl
RB, Chapman RA, Bartlett RH. Extracorporeal
resuscitation of cardiac arrest. Acad Emerg
Medi:1999;6(7):700-707.
40. Chen YS, Yu HY, Huang SC, et al. Extracorpo-
real membrane oxygenation support can extend
the duration of cardiopulmonary resuscitation.
Crit Care Med. 2008;36(9):2529-2535.
41. Thiagarajan RR, Brogan TV, Scheurer MA,
Laussen PC, Rycus PT, Bratton SL. Extra-
corporeal membrane oxygenation to support
cardiopulmonary resuscitation in adults. Ann
Thorac Surg. 2009;87(3):778-785.
42. Fagnoul D, Taccone FS, Belhaj A, et al. Extra-
corporeal life support associated with hypother-
mia and normoxemia in refractory cardiac arrest.
Resuscitation. 2013;84(11):1519-1524.
43. Chou TH, Fang CC, Yen ZS, et al. An obser-
vational study of extracorporeal CPR for in-
hospital cardiac arrest secondary to myocardial
infarction. Emerg Med J. 2014;31(6):441-447.
44. Sawamoto K, Bird SB, Katayama Y, et al. Out-
come from severe accidental hypothermia with
cardiac arrest resuscitated with extracorporeal
cardiopulmonary resuscitation. Am J Emerg
Med. 2014;32(4):320-324.
45. Sakamoto T, Morimura N, Nagao K, et al.
Extracorporeal cardiopulmonary resuscitation
versus conventional cardiopulmonary resuscita-
tion in adults with out-of-hospital cardiac arrest:
a prospective observational study. Resuscitation.
2014;85(6):762-768.
46. Stub D, Bernard S, Pellegrino V, Smith K,
Walker T, Sheldrake J, et al. Refractory cardiac
arrest treated with mechanical CPR, hypother-
mia, ECMO and early reperfusion (the CHEER
trial). Resuscitation. 2015;86:88-94.

96
10

Special Conditions

Justyna Swol, MD, Farah Siddiqui, MBChB, Graeme MacLaren, MBBS, FRACP, FRCP, Matthew L.
Paden, MD

ECLS in Trauma which may be overlooked or underestimated, poten-

tially resulting in severe respiratory failure. Clinical
The majority of traumatic pulmonary and chest findings are subtle and frequently missed due to
injuries can be managed without ECLS. Conven- multiple injuries that direct attention elsewhere.
tional modalities developed to prevent or treat Chest trauma needs to be treated on an individual
severe posttraumatic complications (hypothermia, basis, depending on the location and type of thoracic
metabolic acidosis, bleeding, and acute lung injury) injury or co-injuries.
include lung-protective ventilation, volume infusion, The following are clinical scenarios in which
blood component therapy, and rewarming tech- venovenous ECLS can be considered in trauma
niques to avoid or correct hypothermia. However, patients: trauma-associated acute severe respiratory
some patients die of their complications despite failure due to posttraumatic ARDS (e.g. impair-
having potentially survivable injuries. ECLS offers ments in gas exchange related to mass transfusion,
supplemental capacity in the early resuscitation and blunt chest trauma, or pulmonary contusion) and/
treatment of secondary complications of patients or injury of the trachea-bronchial tree, resulting
with major injuries when their primary injuries are in inadequate gas exchange.1,2 Other indications
being evaluated. ECLS devices have the capacity to include near drowning3 and inhalation trauma with
oxygenate and rewarm blood, infuse fluid volumes, or without cutaneous burns.4 In cases of damage to
correct hypercarbia, and provide circulatory support, the tracheal-bronchial tree, profound hypoventila-
offering an advanced technology for life support tion can result, and ECLS cannulation is lifesaving,
and resuscitation. acting as a bridge to further reconstructive surgery.5,6
In general, ECLS indications depend on several ECLS improves oxygenation in patients with pa-
factors: renchymal disruption and/or bleeding. Furthermore,
• the nature of the injury, ECLS helps to restore hemodynamics, improve
• the prognosis of the injury or combination of metabolic acidosis, and rewarm patients in cases of
injuries, hypothermia. The indication for venoarterial ECLS
• possibilities for definitive therapy, cannulation in trauma patients is either acute trau-
• comorbidities and the patient´s biological age, matic cardiac failure (myocardial contusion or rup-
• severity of the dysfunction, ture) resulting in inadequate circulation (cardiogenic
• status of the central organ, such as the brain, shock or cardiac arrest) or hemorrhagic shock.7,8
kidney, and liver.
Controversies
Clinical Scenarios
Trauma patients are often young and have a
Chest injuries cause many deaths each year, and high risk of complications as well as a complex pat-
blunt thoracic trauma is an especially critical injury tern of life-threatening injuries. They may develop
mechanism in major trauma patients. Blunt force hypothermia, metabolic acidosis, and coagulopathy,
injury typically produces a pulmonary contusion causing hemorrhage and hypoxemia from severe

97
Chapter 10

underlying injuries. Major trauma patients suffer Indications for ECMO

systemic inflammatory response syndrome (SIRS)
in the early posttraumatic phase and are hemody- Severe Respiratory Failure
namically unstable with a high risk of multiorgan The confidential enquiry in maternal deaths dur-
failure.9,10 The risk of bleeding, limited feasibility ing the H1N1 influenza outbreak in 2011 highlighted
of prone positioning in brain, spine, and pelvic inju- early consideration of ECMO if the pregnant patient
ries, and the need for further operative intervention with severe intrinsic lung disease does not respond
during ECLS for brain injury make extracorporeal to conventional ventilation.17,18 The indications for
therapy of trauma patients challenging. ECLS sup- therapy are similar to those in other adults.
port becomes a balancing act between the need for
anticoagulation and the risk of bleeding. Second-hit Acute Heart Failure in Pregnancy
complications and organ failure later in the treat-
ment course can also make extracorporeal support ECMO is increasingly being used for short
necessary. term support for patient in acute heart failure ei-
Traumatic brain injury is not an absolute con- ther following cardiac surgery or post arrest. In
19

traindication, but the risk of lethal intracerebral pregnancy, the plasma volume is increased and the
bleeding has to be considered. A heparin free ECLS oncotic pressure is decreased, rendering the patient,
course is possible.11 Jugular venous cannulation may particularly with complex cardiac disease, more
obstruct venous outflow; optional femoro-femoral susceptible to pulmonary edema and heart failure.
cannulation for VV-ECMO may be considered. Earlier delivery may be beneficial to reduce the
Successful treatment with a beneficial neurological cardiac output, plasma volume, and oxygen require-
outcome suggests that extracorporeal devices should ments of pregnancy.
not be withheld from major trauma patients with TBI
who present with life-threatening hypoxemia.12,13 Contraindications
Decision making in the treatment of coexist-
ing injuries requires an interdisciplinary consensus There are no specific contraindications in preg-
between an ECLS-specialized intensivist, trauma nant patients that do not apply to other adults. These
surgeon, general surgeon, neurosurgeon, cardio- women are often younger and have no other major
thoracic surgeon, orthopedic surgeon, anesthesia comorbidities and are therefore likely to benefit
specialist, and all other physicians involved. from instituting ECMO early. A multidisciplinary
approach including the obstetric teams is useful to
ECLS in Pregnancy discuss whether delivery before starting ECMO is
in mother's and baby's best interest.
Adult respiratory distress syndrome (ARDS) or
cardiac failure poses specific challenges in pregnant Cannulation Strategies
women and in the puerperium (up to one month
after delivery) due to the need to consider, not The augmented cardiac output associated with
only the effects of any treatment on the developing pregnancy necessitates higher flows on VV-ECMO
fetus, but also the physiological changes that oc- to ensure adequate oxygenation. Consequently,
cur in normal pregnancy.15 VV-ECMO is now an cannulation should be performed with the largest
established modality of short term support when a bore cannula that can be safely used. Appropriate
pregnant woman does not respond to conventional sizing with ultrasound measurement of vessels may
ventilation support.16 be useful.
Caval compression by the gravid uterus limits
effective venous drainage from the inferior vena
cava and therefore dual lumen jugular cannula-
tion, or at least jugular drainage and femoral return,
may be advisable for VV-ECMO. With VA-ECMO,

98
 Special Conditions

femoral cannulation is generally performed although Fetal Monitoring

venous drainage may still limit adequate pump flows.
The inability to achieve adequate flows may require Fetal assessment is required as part of the early
the insertion of a second venous drainage cannula care pathways with plans in place should delivery
both for VV- and VA-ECMO. be required. At a gestation less than 24 weeks, the
neonate is not usually viable; delivery essential
Position terminates the pregnancy.21 It is sufficient to aus-
cultate the fetal heart with a Sonicaid (Huntleigh
Since venous drainage to maintain adequate Ltd, Cardiff, UK) to ensure the fetus is alive.22 As
circuit flow can be difficult, left lateral positioning delivery is based on maternal indications to improve
(with a wedge under the right hip) to a 30-degree her health, emergency delivery due to concerns with
inclination can reduce the aortocaval compression fetal health are likely to be detrimental to the mother.
by the gravid uterus, thus improving hypotension
and uterine blood flow. Timing and Mode of Delivery

Respiratory and Hemodynamic Targets Women who receive ECMO early in gestations
who improve and are successfully weaned from
The oxygen targets on ECMO are somewhat ECMO and mechanical ventilation continue to
higher in this patient group. Fetal oxygenation re- term with good maternal and neonatal outcomes.23
quires a maternal arterial oxygen tension (PaO2) >70 Conversely, in late pregnancy when considering
mmHg or an oxygen saturation >=95%. The PaCO2 a trial off ECMO or the patient remains difficult
should be maintained at 30-32 mmHg, considered to oxygenate, consideration should be made to
the normal level in pregnancy. deliver the fetus. Any decision to deliver should
be discussed as a multidisciplinary team including
Anticoagulation the ECMO team, adult intensivists, obstetric anes-
thetists, obstetricians, and the neonatal teams and
Heparin does not cross the placental barrier and documented clearly.24 Antenatal steroid treatment
therefore can be administered using conventional (e.g. dexamethasone 12 mg IM, 2 doses at least 12
targets for maintenance of the ECMO circuit. Direct hours apart) should be considered when planning
thrombin inhibitors are not recommended in preg- delivery between 24-36 weeks gestation, to reduce
nancy. These agents can easily cross the placental the risk of severe neonatal respiratory distress.25
barrier due to their smaller size. An infusion of magnesium sulphate at gestations
before 32 weeks has been shown to reduce the risk
Drug Therapy of neurodisability.26
A filtered non-heparinized circuit can be run up
With regards to administration of drugs a risk- to 6 hours. Once the plan for delivery has been made,
benefit decision making process is needed with the then consideration regarding the mode of delivery
input of obstetricians and pharmacists with knowl- becomes crucial. Case reports demonstrate good
edge of the effects of the medication on the mother outcomes with spontaneous vaginal births while on
and the fetus. Drugs should be chosen according to ECMO. Similarly, in cases of fetal death, vaginal
the least risk for maternal and fetal morbidity. In life- birth may be appropriate. Provision should be made
saving situations, potentially fetotoxic medication for emergency operative delivery if complications
or medication where the fetus effects are not known arise. Most units would consider elective caesarean
can be justified if there is no safe alternative.20 section at a time when the obstetric, cardiothoracic,
ECMO, anaesthetic (both cardiovascular and ob-
stetric anesthetists) and neonatal teams are present.

99
Chapter 10

Specific Complications of Pregnancy to Consider Postpartum Hemorrhage

during ECLS
Once the placenta has delivered, significant
Preeclampsia bleeding may occur due to atony. Uterotonics such
as boluses of Syntocinon, Syntocinon infusion, car-
Preeclampsia is a relatively common complica- boprost, or misoprostol should be considered. Cell
tion of pregnancy, is associated with a high risk of salvage should be available. Careful consideration
intracranial bleeds, eclamptic seizures, and mater- should be made to hemostasis and drain place-
nal death.27 Preeclampsia may be diagnosed if the ment to estimate any following ongoing bleeding.
blood pressure is difficult to control with evidence Tranexamic acid, protamine, and recombinant factor
of new onset proteinuria. The risk of intracerebral VII should be available.
hemorrhage increases exponentially once the blood
pressure exceeds 160/100 (mean arterial pressure of ECLS in Sepsis
125) and therefore blood pressure control is critical.
Initial management involves treatment controlling Sepsis is defined as ‘life-threatening organ dys-
blood pressure with intravenous hydralazine or la- function caused by a dysregulated host response to
betalol infusion particularly in a fully anticoagulated infection.’28 However, the complete definition was
patient. Magnesium sulphate is used to reduce cere- developed for adult patients and different criteria
bral irritability and the risk of eclampsia (seizures).14 should probably be applied to children, although
Preeclampsia is due to abnormal placentation; the the exact details of this are the subject of ongoing
conditions improve postdelivery; thus early multi- research.29 Many patients with sepsis may be placed
disciplinary discussion is required to determine the on ECLS to help treat their underlying illnesses;
timing of delivery. in particular, acute respiratory distress syndrome
Chorioamnionitis (ARDS). However, ECLS is rarely needed in pedi-
atric septic shock and is almost never needed or ap-
If there is evidence of ruptured membranes, propriate for hemodynamic support in adult patients.
then ascending infection will result in infection and This section covers important issues when ECLS is
inflammation. used specifically as mechanical circulatory support
for septic shock refractory to all other treatment.
Abruption Sepsis can induce a number of different hemo-
Abruption is defined as bleeding from the pla- dynamic responses, which in turn affects potential
centa bed. If severe it can result fetal death, maternal cannulation strategies. These responses include
coagulopathy, and DIC. Therefore, management right heart failure and pulmonary hypertension
with broad spectrum antibiotics and delivery of the (neonatal response), isolated left ventricular dys-
fetus should be considered. function (young children), and distributive shock
(adolescents and adults), with an increase in cardiac
Fetal Death output and reduction in vasomotor tone. However,
these responses cannot be reliably predicted from
The risk of fetal demise is high in women with the age of the patient alone, may coexist with other
profound shock or severe systemic disease. The hemodynamic patterns (e.g. biventricular failure),
diagnosis of fetal death can be made by ultrasound. and must be diagnosed individually in each patient
The retained fetus may cause a significant coagu- using a combination of clinical examination and
lopathy in the mother. Therefore, once stabilized, objective diagnostic tools such as echocardiography
consideration should be made to deliver the fetus. A and/or cardiac output monitoring devices.
vaginal delivery is a reasonable option. Due to the If ECLS is used for mechanical circulatory
risk of catastrophic bleeding, once developing signs support, then venoarterial ECLS should be chosen.
of labor, temporary cessation of heparin should be Cannulation can be peripheral or central. Central
considered cannulation avoids the risk of differential hypoxia

100
 Special Conditions

and allows higher flow rates but increases the risk function, and overall prognosis of their underlying
of bleeding and requires the involvement of a car- cancer. However, the ECLS community has looked
diac surgeon. In general, using ECLS for distribu- closer at the cancer population and related immu-
tive shock is not recommended because the native nosuppressed states, and found specific populations
cardiac output is already high. This may explain where ECLS could be useful.
the high mortality rates in some studies of patients
receiving peripheral ECLS for distributive shock.30,31 Populations and Outcomes
Only one study has demonstrated survival rates over
50% using ECLS for adult septic shock in which The initial population that was evaluated was
the patients had developed acute left ventricular childhood cancer, because 5-year survival in the
failure and were cannulated peripherally.32 Stud- United States for hematologic and solid tumors
ies in children have demonstrated better outcomes, exceeded 80%.36 Gow et al. reviewed the ELSO
likely reflecting the greater utility of ECLS in non- Registry experience with children with oncologic
distributive patterns of shock.33-35 diseases and found 35% survival of 107 patients
Particularly high risk groups of septic shock between 1994 and 2007.37 In a survey of active
patients in whom ECLS should be used with cau- pediatric ELSO centers, they found that 95% did
tion and in consultation with (or at) an experienced not consider oncologic diseases to be an absolute
center include febrile neutropenic patients, neonates contraindication to ECLS. Since that publication, we
with disseminated herpes simplex virus, and young have seen increasing use of ECLS in this pediatric
infants with Bordetella pertussis pneumonia. population. A recent update to that paper showed
Septic patients on ECLS have a high incidence that in the 4 years subsequent to its publication, 178
of multiorgan dysfunction and disseminated intra- patients with oncologic diagnoses were supported
vascular coagulation. Renal replacement therapy with 56% survival.38 Little has been described re-
is frequently required for acute kidney injury or to garding the use of ECLS in the adult oncology
prevent fluid overload. Bleeding should be managed population, but findings were similar to the pediatric
with aggressive use of blood products and factor population in that 72 patients received support from
replacement and may require surgical intervention. 1992-2008 with 32% survival.39 More recent stud-
Patients on ECLS for septic shock generally ies demonstrate improving survival (7/14, 50%) in
have short runs, 3-5 days is normal. Occasionally, hematologic malignacies.40
some patients require conversion from VA- to VV- Several points should be emphasized regarding
ECLS because the circulation often recovers before these data. First, while a growing indication, patients
the lungs do in particularly severe cases. with oncologic conditions are still rare users of
Finally, it must be highlighted that ECLS is not a ECLS representing only 1.03% of all ECLS use.38
treatment for sepsis, but rather a form of life support. In the last 5 years, reported survival in the oncology
Every attempt must be made to adequately diagnose ECLS population is only 5-10% worse than other
and aggressively treat the underlying infection. patients supported with ECLS who do not have on-
cologic diagnoses. However, this enthusiasm should
ECLS in Oncology Patients be tempered by the small numbers of patients, and
very likely selection bias occurs in the oncology
Historically, the presence of active cancer was patients who are chosen for ECLS. Hence, these
an absolute contraindication to ECLS support. This outcomes should not be expected if this technique
decision was made due multiple risk factors for is applied to a wider group of oncologic patients.
complications and poor outcome of these patients, Additionally, little is known about long-term (>5
which often include combinations of an immuno- year) survival and morbidity rates in this population.
suppressed state, increased risk of bleeding from
thrombocytopenia, increased risk of thrombosis as-
sociated with malignancy, reduced rate of infectious
clearance due to low white blood cell numbers and

101
Chapter 10

Clinical Scenarios

The bedside practice of ECLS in patients

with cancer differs only slightly from other ECLS
patients; however, patient selection is often more
involved. Offering ECLS is not appropriate for all
patients with cancer and must be individualized. A
multidisciplinary review of the clinical case by the
critical care, surgery, oncology, and ECLS teams
should occur. Discussion of the potential risks and
benefits of ECLS therapy must be weighed in the
setting of long-term cancer prognosis, expected
survival, remission status, other organ involvement,
remaining anticancer therapeutic options, and likely
reversibility of the acute problem for which ECLS is
being considered. Discussion with the family should
include clear descriptions of the goals of support,
how success and failure are defined, higher risk of
ECLS related complications, as well as incorporat-
ing palliative care principles from the beginning
of ECLS.

Controversies

Hematopoietic stem cell transplant patients

have received ECLS; however, their survival re-
mains poor at 10-20% in both the adult and pediatric
populations.41 Survival is even worse in patients
supported peri transplant (~100 days), and those
who have not completed engraftment and immune
reconstitution. Many consider ECLS use in this
setting an absolute contraindication.

102
 Special Conditions

References 12. The Antenatal Magnesium Sulphate for Neu-

roprotection Guideline Development Panel.
1. Reece A, Hobbins J. Clinical Obstetrics: The Antenatal magnesium sulphate prior to preterm
Fetus and the Mother. 3rd ed. Wiley-Blackwell birth for neuroprotection of the fetus, infant
Publishing, 2007. and child: national clinical practice guidelines.
2. Sharma N, Wille R, Bellot S, Diaz-Guzman E. The Australian Research Centre for Health of
Modern use of extracorporeal life support in Women and Babies, The University of Adelaide;
pregnancy and postpartum. ASAIO 61(1):110- 2010. http://www.sahealth.sa.gov.au/wps/
114. wcm/connect/86f3f2804ee4f45294189dd150
3. Modder J. Review of maternal deaths in the UK ce4f37/magnesium+sulphate+neuroprotect+f
related to A H1N1 2009 influenza (CMACE). etus+risk+preterm_27042016.pdf?MOD=AJ
https://www.rcm.org.uk/sites/default/files/ PERES&CACHEID=ROOTWORKSPACE-
Review-of-Maternal-Deaths~he-United-King- 86f3f2804ee4f45294189dd150ce4f37-lm-tQIa
dom-re.pdf Accessed April 24, 2018. Accessed April 24, 2018.
4. Pollock W, Rose L. Pregnant and postpartum ad- 13. NICE. Hypertension in pregnancy. Clinical
missions to the intensive care unit: a systematic guideline 107. 2010 https://www.rcpch.ac.uk/
review. Intensive care med 2010; 36(9):1465-74. sites/default/files/asset_library/Research/Clini-
5. Koster, A. A., Pappalardo, F., Silvetti, S., cal%20Effectiveness/Endorsed%20guidelines/
Schirmer, U., Lueth, J. U., Dummler, R. Sandica, Hypertension%20in%20Pregnancy/NICE%20
E. Cesarean section in a patient with non-com- GuidelineI.pdf Accessed April 24, 2018.
paction cardiomyopathy managed with ECMO 14. Eclampsia Trial Collaborative Group. Which
. Heart, Lung and Vessels, 2013;5(3):183–186. anticonvulsant for women with eclampsia?
6. Laegreid L Olegard R et al. Teratogenic effects Evidence from the Collaborative Eclampsia
of benzodiazepine use during pregnancy. J Pe- Trial. Lancet 1995, 345:1455–63.
diatr. 1989;114:126-1131. 15. Incagnoli P, Blaise H, Mathey C, Vinclair M,
7. Sauer PM. Maternal-fetal assessment of the Albaladejo P. Pulmonary resection and ECMO:
critically ill parturient: Decisions related to a salvage therapy for penetrating lung trauma.
delivery. AACN Clin Issues 1997;8:564-573. Ann Fr Anesth Reanim. 2012;31(7-8):641-643.
8. Daniel E. Cole, MD; Tara L. Taylor, MD; Deir- 16. Yuan KC, Fang JF, Chen MF. Treatment of en-
dre M. McCullough, MD; Catherine T. Shoff, dobronchial hemorrhage after blunt chest trau-
DO; Stephen Derdak, DO. Acute respiratory ma with extracorporeal membrane oxygenation
distress syndrome in pregnancy. Crit Care Med (ECMO). J Trauma. 2008;65(5):1151-1154.
2005; 33[Suppl.]:S269 –S278. 17. Champigneulle B, Bellenfant-Zegdi F, Follin
9. Kristine Madsen, Ditte Gry Strange, Morten A, et al. Extracorporeal life support (ECLS)
Hedegaard Maternal and fetal recovery after for refractory cardiac arrest after drowning: an
severe respiratory failure due to influenza: a 11-year experience. Resuscitation. 2015;88:126-
case report. BMC Research Notes, 2013;6(1):1. 131.
10. Costeloe Kate L, Hennessy Enid M Haider 18. Nelson J, Cairns B, Charles A. Early extracor-
Sadia et al. Short term outcomes after extreme poreal life support as rescue therapy for severe
preterm birth in England: comparison of two acute respiratory distress syndrome after inhala-
birth cohorts in 1995 and 2006 (the EPICure tion injury. J Burn Care Res. 2009;30(6):1035-
studies) BMJ 2012;345:e7976. 1038.
11. Antenatal corticosteroids to reduce neonatal 19. Walker JL, Wiersch J, Benson C, Young HA,
morbidity and mortality. Green Top guideline Dearmond DT, Johnson SB. The successful use
2010. RCOG https://www.glowm.com/pdf/ of cardiopulmonary support for a transected
Antenatal%20Corticosteroids%20to%20Re- bronchus. Perfusion. 2012;27(1):34-38.
duce%20Neonatal%20Morbidity.pdf Accessed 20. Zhou R, Liu B, Lin K, et al. ECMO support for
April 24, 2018. right main bronchial disruption in multiple trau-

103
Chapter 10

ma patient with brain injury--a case report and 30. Huang CT, Tsai YJ, Tsai PR, Ko WJ. Extracor-
literature review. Perfusion. 2015;30(5):403- poreal membrane oxygenation resuscitation in
406. adult patients with refractory septic shock. J
21. Kim DW, Lee KS, Na KJ, Oh SG, Jung YH, Thorac Cardiovasc Surg 2013; 146:1041-1046.
Jeong IS. Traumatic rupture of the coronary 31. Park TK, Yang JH, Jeon K, et al. Extracorporeal
sinus following blunt chest trauma: a case report. membrane oxygenation for refractory septic
J Cardiothorac Surg. 2014;9:164. shock in adults. Eur J Cardiothorac Surg 2015;
22. Weber SU, Hammerstingl C, Mellert F, 47:e68-74.
Baumgarten G, Putensen C, Knuefermann P. 32. Brechot N, Luyt CE, Schmidt M, et al. Veno-
Traumatic tricuspid valve insufficiency with arterial extracorporeal membrane oxygenation
right-to-left shunt: bridging using extracor- support for refractory cardiovascular dysfunc-
poreal venovenous membrane oxygenation. tion during severe bacterial septic shock. Crit
Anaesthesist. 2012;61(1):41-46. Care Med 2013; 41:1616-26.
23. Hill JD, O’Brien TG, Murray JJ, et al. Pro- 33. McCune S, Short BL, Miller MK, Lotze A,
longed extracorporeal oxygenation for acute Anderson KD. Extracorporeal membrane oxy-
post-traumatic respiratory failure (shock-lung genation therapy in neonates with septic shock.
syndrome). Use of the Bramson membrane lung. J Pediatr Surg 1990;25:479-482.
N Engl J Med. 1972;286(12):629-634. 34. Hocker JR, Simpson PM, Rabalais GP, Stewart
24. Schupp M, Swanevelder JL, Peek GJ, Sosnows- DL, Cook LN. Extracorporeal membrane oxy-
ki AW, Spyt TJ. Postoperative extracorporeal genation and early-onset group B streptococcal
membrane oxygenation for severe intraopera- sepsis. Pediatrics 1992;89:1-4.
tive SIRS 10 h after multiple trauma. Br J An- 35. MacLaren G, Butt W, Best D, Donath S. Cen-
aesth. 2003;90(1):91-94. tral extracorporeal membrane oxygenation for
25. Muellenbach RM, Kredel M, Kunze E, et al. refractory pediatric septic shock. Pediatr Crit
Prolonged heparin-free extracorporeal mem- Care Med 2011; 12:133-136.
brane oxygenation in multiple injured acute 36. Murphy SL, Xu J, Kochanek KD: Deaths: Final
respiratory distress syndrome patients with data for 2010. National Vital Statistics Reports
traumatic brain injury. J Trauma Acute Care 2013; 61:1-117.
Surg. 2012;72(5):1444-1447. 37. Gow KW, Heiss KF, Wulkan ML,et al. Extracor-
26. Muellenbach RM, Redel A, Kustermann J, et poreal life support for support of children with
al. Extracorporeal membrane oxygenation and malignancy and respiratory or cardiac failure:
severe traumatic brain injury. Is the ECMO- The extracorporeal life support experience. Crit
therapy in traumatic lung failure and severe Care Med 2009;37:1308-16.
traumatic brain injury really contraindicated?. 38. Armijo-Garcia V, Froehlich CD, Carrillo S,
Anaesthesist. 2011;60(7):647-652. Gelfond J, Meyer AD, Paden ML. Outcomes
27. Young N, Rhodes JK, Mascia L, Andrews of extracorporeal life support for children with
PJ. Ventilatory strategies for patients with malignancy: A report from the ELSO registry.
acute brain injury. Curr Opin Crit Care. Proceedings of the 24th Annual ELSO Meeting,
2010;16(1):45-52. 2012, Philadelphia, PA.
28. Singer M, Deutschman CS, Seymour CW, et al. 39. Gow KW, Lao OB, Leong T, et al. Extracor-
The Third International Consensus definitions poreal life support for adults with malignancy
for sepsis and septic shock (Sepsis-3). JAMA and respiratory or cardiac failure: the extra-
2016; 315:801-810. corporeal life support experience. Am J Surg
29. Schlapbach LJ, MacLaren G, Festa M, et al. 2010;199:669-75.
Prediction of pediatric sepsis mortality within 40. Wohlfarth P, Ullrich R, Staudinger T, et al.
1 h of intensive care admission. Intensive Care Extracorporeal membrane oxygenation in
Med 2017; 43:1085-1096. adult patients with hematologic malignancies

104
 Special Conditions

and severe acute respiratory failure. Crit Care

2014;18:R20.
41. Di Nardo M, Locatelli F, Palmer K, et al. Extra-
corporeal membrane oxygenation in pediatric
recipients of hematopoietic stem cell trans-
plantation: the Extracorporeal Life Support
Organization experience. Intensive Care Med.
2014; 40(5):754-6.

105
11

Cannulation and Initiation of ECLS

James T. Connelly, RRT-NPS, Susan B. Williams, MSN, RNC, CIT, Alain Combes, MD, PhD

Objectives assessment for competency to perform circuit as-

sembly and priming.
After completion of this chapter, the participant ECLS cannulation varies among patient popula-
should be able to: tions, disease processes, and institutions. Cannula-
tion procedures vary according to technical aspects
• Describe personnel required for activation and of the surgical procedure and individual anatomy
timely initiation of ECLS. (see Chapters 12-14). The purpose of this chapter
• Detail the interprofessional roles and depart- is to describe the team and the requirements related
ment requirements related to the technical to the technical provision of ECLS from decision to
provision of the ECLS circuit. initiation of ECMO support.
• Describe the system of Primer workflow from Assembly and priming of the ECLS circuit
activation, through circuit prime preparation, to varies by center. Several ECMO circuits may be
‘handing off the lines’ at the time of connection available within each institution to provide neonatal,
to the cannulas. pediatric and adult support. The mode of support
• Identify potential complications immediately (pulmonary vs. cardiac) may also influence circuit
after initiation of VA- and VV-ECMO flow. choice.3, 4 The equipment necessary for cannulation,
including the extracorporeal circuit, is usually main-
The Team and Equipment tained and provided by the ECMO specialist primer
and the surgical team. All equipment (consumables
When a patient is deemed a candidate for and ECMO console) must always be ready and
ECMO, prompt notification of the multidisciplinary available, usually preassembled dry or crystalloid-
team is essential. Ideally, the team consists of a primed. The circuit and other supplies are typically
critical care specialist, the nurse responsible for stored in a secure area in close proximity to the
the patient, a senior surgeon, a resident, an operat- intensive care units (ICU). Supplies include steril-
ing room nurse, an ECMO specialist primer, and ized instruments, sterile drapes, surgical gowns,
in most pediatric and neonatal centers, an ECMO cannulas, introducers, radiopaque dressings, and
Specialist.1, 2 Blood bank, pharmacy, radiology, and surgical consumables.5 A checklist may help to as-
clinical laboratory departments also require notifica- sure rapid deployment of the equipment.
tion of an ECMO initiation to prioritize the urgency Medical orders must be entered to obtain nec-
of ECMO orders. The ECMO primer assembles the essary blood products, medications, and infusions
circuit and primes it with blood in preparation for in a timely manner. An ECMO cannulation order
neonates and children < 30 kg before ‘handing off set in the electronic health record (EHR), permits
the lines’ for connection to the cannulas. Priming in the ordering clinician to select appropriate options
larger patients can be bloodless. The ECMO primer from a “Menu.”
may be a perfusionist, nurse, or respiratory therapist For patients who require a blood prime, the
institutionally certified via specialized training and blood bank must be notified to urgently prepare
blood products.6, 7 Many centers have a predefined

107
Chapter 11

requirement for packed red blood cells (PRBC) tion, both sides of the groin are prepared. This has
used for cannulation related to freshness of product. the advantage of varying cannulation sites (artery
If no suitable PRBC are available to meet institu- and vein from opposite sides to reduce the risk of
tional criteria, the blood products can be ultrafiltered ischemia). The electrocautery pad should be placed
washed, although the latter may delay by 30-60 away from any metal prosthesis and on healthy skin,
minutes per unit. Additionally, most centers and the followed by a sterile bath according to unit protocol,
European Union require that all blood products be from the ear and covering to the knees bilaterally.
leukoreduced to minimize passive antibody trans- Any dressings in those areas must first be removed.
mission and transfusion reactions. This process may To avoid the risk of an electric arc between the
remove other potentially harmful viral (cytomega- electrocautery and the presence of alcohol vapor,
lovirus, Epstein Barr virus) and/or bacterial agents which could cause severe burns, do not use alcohol
from donor blood.8 Immediate notifications must antiseptic. The resuscitator bag (flow-inflating or
also include clinical pharmacy, radiology, and the self-inflating) cannot be located under or near the
operating room (OR). Pharmacy prepares needed draped patient to minimize the possibility of a surgi-
infusions, including heparin and sedatives for con- cal fire related to an oxygen source.
nection to the circuit. Pharmacy may also provide
the medications used for the circuit blood prime, Approach to Cannulation and Preparing ECMO
such as sodium bicarbonate (NaHCO3), calcium glu-
conate, and heparin; whereas, some hospitals may Once prepared, the cannulation site can be
have a prepared available in each ICU that allows dressed.9 A pack of surgical drapes is positioned
for immediate preparation of these necessary circuit on a side table and the positioning of the sterile
additives. Radiology may deliver and position an drapes must meet the standards of sterility. When
x-ray plate prior to sterile draping. Cardiology must a surgical approach is used, a sterile suction line
be notified when echocardiography is essential. The is provided, connected to a strong suction with a
OR may require additional time to mobilize to the collection bucket. The nurse/resident installs the
ICU if the cannulation procedure is to occur at the instrumentation table with all the necessary instru-
bedside. ments for cannulation and prepares local anesthesia
if necessary.
Patient Preparation At this point, the ECMO primer ensures that
the ECMO pump is in the room. The console is
The room must contain an operating light, connected to power, air, and oxygen supplies. The
electrocautery unit, OR tables, and enough space console includes an air/oxygen gas mixer, an emer-
for the surgeons to move around the patient in the gency crank, a centrifugal or roller pump, a clamp to
presence of the ECMO circuit. In order to facilitate hold the oxygenator, and two Vorse tube press clips
rapid cannulation, the equipment should be prepared, (usually called Weiss clamps). A heater warms the
the patient positioned and draped in a sterile fashion prime, and then regulates the patient’s temperature.
prior to cannulation.6 Once in the room, the primer de-airs the circuit.
The patient is positioned supine with a block, Before starting, the primer checks circuit in-
usually made of rolled sheets, under the pelvis or tegrity and expiration dates, removes bubbles, and
shoulder to improve exposure of the surgical site. primes the circuit in a sterile manner. Blood priming
Multiple staff are required to protect indwelling must be performed carefully. “VA-ECMO is the
lines and tubes, position the ventilator, and ensure most forgiving of an imperfect blood prime". Even
a secure airway. ICU nurse teams provide ongoing if a patient is initiated with a crystalloid prime mix,
critical care monitoring and medication/infusion a low hematocrit, or unbalanced electrolytes, most
administration during the procedure. patients will overcome the prime and maintain ap-
In addition to sterile preparation of the surgical propriate vital signs. VV-ECMO however, requires
site, a backup site should always be prepared in case a well-balanced circuit prime at bypass initiation.10
of cannulation failure. Usually, for femoral cannula- While the circuit warms, the primer gathers the

108
 Cannulation and Initiation of ECLS

necessary additives for blood priming a crystalloid gravity to expel air from the entire circuit. Once
circuit. Every program has its own version, but a complete, air is removed using the blood pump,
typical neonatal/pediatric blood prime procedure initially at low speed, then gradually increasing
can be: the speed until no air bubbles remain. The ease and
speed of de-airing depends on the circuit model.
1. Add 50 ml of 25% albumin to the crystal- Once bubble removal is complete, the circuit is
loid solution and circulate to allow protein secured: all valves are closed and checked for pos-
coating sible leaks. Sensors and ultrasonic flow meter(s) are
2. Clamp and drain the priming reservoir of placed as required. The de-airing lines should be
crystalloid kept until ECMO initiation in case they are needed
3. Add one or more units of preferably < 7 further.11
day old PRBC
4. Add 100 units heparin per unit of blood Vessel Cannulation
5. Add 100 ml or more of FFP (if desired)
6. Add 10-15 mEq of NaHCO3 per unit of Specific cannulation configurations for re-
blood spiratory and cardiac ECMO are detailed in
7. Unclamp the venous connection of the Chapters 12-14. Often, a separate cannula cart is
reservoir, open the drain line, and restart brought to the scene, containing a large variety
the pump to “chase” the crystalloid into of cannulas and connectors, capable of providing
a collection bottle by displacing the fluid the size and type of cannula requested without
with blood from the reservoir. Stop the delay. Since the percutaneous approach using the
pump again when the blood nears the drain Seldinger technique under ultrasound guidance is
line; close the drain line and open clamps the rule for VV-ECMO and is increasingly used for
to the reservoir on the return side to allow VA-ECMO, an ultrasound device should be avail-
for blood to circulate through the reservoir. able. Also, fluoroscopy is often used when inserting
8. With blood circulating, add 500-600 mg the Avalon cannula. Before cannulation, a heparin
calcium gluconate to the primed blood bolus should be administered to prevent cannula
through the reservoir. and circuit thrombosis.12 Patients usually receive
9. Initiate sweep flow to the circulating blood. a bolus of 5000 IU (or 50-100 IU/kg in children)
FiO2 varies from 21% to 100%, and dura- standard dose of unfractionated heparin.13 Once
tion from at least one minute. Flow rates inserted, ECMO cannulas are purged with saline
are usually 1 liter per minute or greater. solution or back filled with blood. With femoral
The goal is to assure gas exchange across VA-ECMO a reperfusion catheter is also placed to
the oxygenator, historically called the “pink prevent limb ischemia.
test,” while attempting to approximate the
patient’s blood gas CO2 level. ECMO Initiation
10. Circuit blood gases should be drawn and
corrected as necessary. A watchful eye for Once cannulation is complete, sterile lines are
appropriate pH, PCO2, PO2 and ionized handed to the surgeon. The connection lines are
calcium (iCa2+) should guide additive or cleansed of residual air using a prefilled saline sy-
sweep adjustments. It is helpful to have ringe. The ECMO initiation checklist is reviewed
a POCT device for quickly results (e.g. by the surgeon, anesthetist, and specialist primer. It
iSTAT). should include:

Taps, plugs, and fittings are checked manually. • Initiation of the gas supply.
Each ECMO manufacturer publishes recommenda- • Verification of the priming temperature: it must
tions for de-airing that must be followed (regularly be warm to avoid rhythm disorders due to po-
updated). The priming is initially performed by tential cold shock at initiation.

109
Chapter 11

• Verification that the reinjection line is clamped. ECMO flow begins. Surgical braided sutures fix
• Verification of the rotation speed of the blood lines and cannulas at three points. The cannula
pump, which must be set at a minimum thresh- must be secured in the axis of the vessel to prevent
old preventing backflow when ECMO begins. decannulation. Meanwhile, the ECMO Specialist
manages the balance between patient hemodynam-
The lines and cannula position must be managed ics and ECMO support with the help of the physi-
carefully until sutures are secured. This represents a cians (hypovolemia, vascular resistance, control
precarious time, due to continuing responsibilities of ventilation). When cannulated surgically, the
of the surgical team and the primer. Often, the sur- surgeon closes the incision after local hemostasis
gical team is awaiting radiographic verification of verification. The occlusive and sterile dressing is
cannulation while suturing and dressing the incision. placed by the nurse or resident. It can be transpar-
Simultaneously, the primer is advancing flow, start- ent if the puncture sites are clean or opaque in case
ing the sweep gas, verifying adequate drainage and of bleeding (Tegaderm, Cicaplaie, Mepor). The
pressures, and initiating the heparin infusion. Thus, reperfusion line to prevent limb ischemia is covered
one individual should be responsible to support the with a transparent dressing to detect thrombosis of
weight of the tubing on the cannulas. This is a key the tubing and prevent kinking (see Chapter 23).
time as inadvertent decannulation may occur. Once on ECMO, patient ventilation should be
The initiation of ECMO flow depends spe- weaned.15 The patient ventilator should provide
cifically on the type of support. One caveat is to minimal support as the patient advances to full
commence ECMO support slowly. Increase flow ECMO support. Desired ‘rest’ or ‘open lung’ ven-
as tolerated depending upon patient circumstances tilator settings are established and achieved, often
as some patients may be at risk for cardiac stun. A within the first few hours of ECMO support. Care
properly blood primed circuit (i.e. without hyperka- should be exercised when weaning the ventilator
lemia), slow advancement of flow, and prophylactic because rapid decreases in support may cause rapid
administration of a dose of 100 mg/kg of calcium atelectasis with acute increases in PVR and acute
gluconate minimizes this potentially catastrophic right heart dysfunction. These settings usually vary
complication. This may require additional manage- by type of ECMO support provided and discussed
ment of blood pressure with either volume or possi- daily for the duration of the ECMO run. An ad-
bly calcium gluconate administration. Normalizing ditional order for ‘emergency ventilator settings’
the ionized calcium can prevent cardiac instability. is also established in case of an emergent need to
Volume expanders should be available during the clamp off ECMO to troubleshoot a complication.
transition to full ECMO support. Many pediatric programs transfuse platelets
After removing sterile drapes, secure the con- after stabilization of ECMO flow. Of note, most
nections of the circuit tubing to the cannula with centers do not include platelets in their prime recipe,
tystraps as soon as possible. When the patient is and the hemodilution of native platelets with the
optimally positioned, the tubing must be anchored platelet poor prime may produce low platelet counts
securely to prevent accidental decannulation. in infants and children but less commonly so in
Sweep flow is initiated along with circuit blood adults. Adult centers often can avoid administration
flow. The initial rate may vary per institution, but of platelets after cannulation.
a good starting point is to begin the sweep flow at
or below the blood flow and adjust slowly over 12 Preparation for Transportation
hours or more in order to avoid rapid changes in
cerebral blood flow and the potential for intracranial This step follows cannulation at a referring
hemorrhage. Sweep flow should be initiated with hospital.16 After the surgical dressings are removed,
this goal in mind and monitored closely with patient the perfusionist secures the tubing connection with
arterial blood gases. links and a gasket-holder. The perfusionist performs
In the context of extracorporeal cardiopulmo- a dissociated fixing of the lines at the thigh using
nary resuscitation (eCPR),14 CPR is stopped once a nonwoven extensible tape. Ideally, a horizontal

110
 Cannulation and Initiation of ECLS

fixing system for drains and probes (Hollister) can

optimize the binding of the lines on the patient.
The heater is set according to patient needs and
the advice of the medical team. The final patient
preparation, performed by the perfusionist, includes
keeping the patient in the supine position during
transportation. The patient may have the head of
the bed raised to a maximum of 30°: Beyond this,
the risk of an internal kink or bleeding from the scar
increases. ECMO lines are secured in the bed and
positioned along the leg: a loop can be made around
the patient’s foot. The lines must never be in contact
with the ground. The console is placed on a carriage
(preferably with a brake system) positioned at the
bedside; the ECMO control panel must be visible
from the outside, as well as the patient’s monitoring
device. The crank handle is positioned proximate
to the centrifugal head to facilitate any manual
relay. A panel with the crank handle explaining the
emergency procedure is left for every person who
cares for the patient. The 24/7 phone number of the
ECMO team is written on one side of the console.

111
Chapter 11

References (ECMO) circuit integrity and safety utilizing the

perfusionist as the «ECMO Specialist». Perfu-
1. Combes A, Brodie D, Bartlett R, et al. Position sion. 2013; 28(6):552-554.
paper for the organization of extracorporeal 12. Brodie D, Bacchetta M. Extracorporeal mem-
membrane oxygenation programs for acute re- brane oxygenation for ARDS in adults. N Engl
spiratory failure in adult patients. Am J Respir J Med. 2012;365(20):1905-1914.
Crit Care Med. 2014;190(5):488-496. 13. Sklar MC, Sy E, Lequier L, Fan E, Kanji HD.
2. Daly KJ, Camporota L, Barrett NA. An inter- Anticoagulation Practices during Venovenous
national survey: the role of specialist nurses Extracorporeal Membrane Oxygenation for
in adult respiratory extracorporeal membrane Respiratory Failure. A Systematic Review. Ann
oxygenation. Nurs Crit Care. 2017; 22(5):305- Am Thorac Soc. 2016;13(12):2242-2250.
311. 14. Ouweneel DM, Schotborgh JV, Limpens J, et
3. Bosarge PL, Raff LA, McGwin J, et al. Early al. Extracorporeal life support during cardiac
initiation of extracorporeal membrane oxygen- arrest and cardiogenic shock: a systematic re-
ation improves survival in adult trauma patients view and meta-analysis. Intensive Care Med.
with severe adult respiratory distress syndrome. 2016;42(12):1922-1934.
J Trauma Acute Care Surg. 2016; 81(2):236-243. 15. Serpa Neto A, Schmidt M, Azevedo LC, et al.
4. Frenckner B. Extracorporeal membrane oxygen- Associations between ventilator settings dur-
ation: a breakthrough for respiratory failure. J ing extracorporeal membrane oxygenation for
Int Med. 2015; 278(6):586-598. refractory hypoxemia and outcome in patients
5. ELSO guidelines, https://www.elso.org/Re- with acute respiratory distress syndrome: a
sources/Guidelines.aspx, Last accessed August pooled individual patient data analysis: Me-
27th 2017. chanical ventilation during ECMO. Intensive
6. Van Kiersbilck C, Gordon E, Morris D. Ten Care Medicine. 2016;42(11):1672-1684
things that nurses should know about ECMO. 16. Beurtheret S, Mordant P, Paoletti X, et al.
Intensive Care Med. 2016;42(5):753-755. Emergency circulatory support in refractory car-
7. Ang AL. Predictors of increased transfusion diogenic shock patients in remote institutions:
requirements and optimizing transfusional sup- a pilot study (the cardiac-RESCUE program).
port in patients on extracorporeal membrane Eur Heart J. 2013;34(2):112-120.
oxygenation (ECMO). ISBT Science Series.
2012; 7(1):89-91.
8. Thiele T, Krüger W, Zimmermann K, et al.
Transmission of cytomegalovirus (CMV) infec-
tion by leukoreduced blood products not tested
for CMV antibodies: a single‐center prospective
study in high‐risk patients undergoing alloge-
neic hematopoietic stem cell transplantation
(CME). Transfusion. 2011; 51(12):2620-2626.
9. MacLaren G, Combes A, Bartlett RH. Contem-
porary extracorporeal membrane oxygenation
for adult respiratory failure: life support in the
new era. Intensive Care Med. 2012;38(2):210-
220.
10. Short, BL & Williams, L. ECMO Specialist
Training Manual. (3rd ed.). Ann Arbor, MI:
Extracorporeal Life Support Organization; 2010.
11. Mongero LB, Beck JR, Charette KA. Manag-
ing the extracorporeal membrane oxygenation

112
12

Venoarterial ECMO in Adult and Pediatric Patients

Roberto Lorusso, MD, PhD, Matteo Di Nardo, MD, I-wen Wang, MD, PhD, George Makdisi, MD, MPH, MS

Venoarterial ECMO in Adults Despite its versatility and ease of use, ECMO
is not suitable for all patients. One fundamental
Venoarterial (V-A) ECMO indications and usage consideration prior to initiating ECMO is the consid-
have progressed over the last 20 years becoming eration of an exit strategy so as not to initiate ECMO
an essential tool in the care of adults and children in a futile situation or simple prolong the dying
with severe cardiopulmonary failure refractory to process. Contraindications are listed in Table 12-2.
conventional management.1-5 ECMO support has
become more reliable with improved outcomes, with V-A ECMO Technique
improvement in technology, and increased clinical
experience. Furthermore, it is typically instituted in Blood in V-A ECMO drains from the right
a time-sensitive manner as definitive therapy to sta- atrium or vena cava and returns to the arterial system.
bilize critically ill patients rather with a salvage only Conceptually, the V-A ECMO circuit is connected in
intent. In the contemporary era of cardiothoracic
surgery, VA-ECMO is also emerging as a powerful Emergent Use of V-A ECMO
tool in nonemergent applications.6,7 This chapter Cardiogenic shock Severe cardiac failure due to almost any other
cause:
will focus on this powerful life-saving technology. Acute coronary syndrome (consider ECMO support
before perfusion in cases of refractory cardiac arrest)
Refractory cardiogenic shock after PCI for AMI
Indications of V-A ECMO Cardiac arrhythmic storm refractory to other measures
Sepsis with profound cardiac depression
Myocarditis
The indications for V-A ECMO extend from Drug overdose⁄toxicity with profound cardiac depression
emergent use for cardiogenic shock to more pro- Pulmonary embolism
Isolated cardiac trauma
longed use in intensive care units, such as bridge- Acute anaphylaxis
to-transplant for both cardiac and lung transplant Postcardiotomy: Inability to wean from cardiopulmonary bypass
(Table 12-1).1-12 ECMO can also be used as an after cardiac surgery
Post heart transplant: Primary graft failure after heart or heart-
alternative to conventional cardiopulmonary bypass lung transplantation
(CPB) in patients during lung transplantation. V-A Acute on chronic cardiomyopathy ECMO is used as:
Bridge to longer term VAD support
ECMO is also used posttransplant to support pri- Bridge to decision
mary graft dysfunction.13- 15 For patients undergoing Bridge to transplant
Pulmonary failure:
high risk coronary artery interventions or transcath- V-V ECMO is not able to achieve enough support
eter aortic valve replacement, V-A ECMO has been Combine heart and lung failure
Nonemergency Use of ECMO as Intraoperative Support in:
used emergently to support unstable patients, and Periprocedural support for high-risk percutaneous cardiac
prophylactically to prevent hemodynamic instabil- interventions including TAVI, Stenting, and other procedures
ity.16-21 V-A ECMO has been used to support patients Major thoracic resections (tracheal resection)
V-A=Venoarterial; PCI=percutaneous coronary intervention;
with poor pulmonary function during lung resections AMI=acute myocardial infarction; VAD=ventricular assist
and for management during complex airway repair/ device; V-V=venovenous; TAVI=trans-catheter aortic valve
implantation
reconstruction.
Table 12-1. Indications of V-A ECMO use.

113
Chapter 12

parallel to the heart and lungs--in contradistinction can be performed during cardiopulmonary resusci-
to venovenous (V-V) ECMO where the circuit is tation (CPR).
connected in series to the heart and lungs. The
outflow cannula connects to the arterial system Complications
peripherally via the femoral, axillary, or carotid
arteries (Figure 12-1). Alternatively, the outflow V-A ECMO is an invasive surgical procedure,
cannula may be placed centrally with direct access requiring complex management, significant per-
of the ascending aorta (Figure 12-2). sonnel and medical resources, and often associated
In cases with right ventricle (RV) failure and with serious complications. Peripheral insertion
hypoxemia, special ECMO configuration is used of V-A ECMO has its own spectrum of complica-
to support the RV and oxygenation. In this sce- tions, including vessel perforation with hemorrhage,
nario, oxygenated blood is delivered to the pul- arterial dissection, distal ischemia, and incorrect
monary artery (PA), bypassing only the right heart cannula location (e.g. venous cannula within the
(Figure 12-3a). A double lumen single cannula is artery) or development of pseudoaneurysm at the
now commercially available for this purpose and is site of insertion. Formation of lymphocele can also
usually introduced through the right internal jugular occur. In total, these complications occur uncom-
(IJ) into the PA (blood drains from the right atrium monly, (<5%).
[RA] and superior vena cava into the circuit then Three complications of V-A ECMO can have
inflowed directly into the PA). The same principle major impact on recovery:
could be achieved by using 2 venous cannulae; the 1) Left ventricle (LV) distention and thrombus
first (femoral vein) drains blood from the RA into formation. Poor LV unloading combined with in-
the circuit while the second cannula inserted in the creased afterload can cause ventricular distension
right IJ into the main PA (Figure 12-3b). In open with a high risk of myocardial ischemia and pulmo-
chest cases, the PA can be cannulated directly. nary edema. Ironically, high flow through the circuit
Femoral access (percutaneous or cutdown) is may result in low flow through the native cardiac
preferred for V-A ECMO in case of emergency or circulation, resulting in stasis in the left atrium
cardiogenic shock as the insertion is less invasive (LA) and LV. The resulting development of LA/
and faster and can be performed at bedside. It also LV thrombus increases mortality considerably.31-35
Ideally, stasis of flow in the LV should be avoided
ABSOLUTE CONTRAINDICATIONS because thrombus inevitably develops despite
Among these futile treatments without exit strategy in case of: full anticoagulation. One approach is to maintain
Unrecoverable heart and not a candidate for transplant or
destination therapy of VAD support the patient on inotropic support to encourage LV
Disseminated malignancy contraction. Also, a LV vent can be inserted. With
Known severe brain injury
Severe chronic organ dysfunction (emphysema, cirrhosis, peripheral V-A ECMO options to decompress the LV
renal failure) include percutaneous atrial septostomy, transapical
Unwitnessed cardiac arrest
Prolonged CPR without adequate tissue perfusion LV vent through left thoracotomy, and placing an
Unrepaired aortic dissection Impella.36-38
Severe aortic regurgitation
Compliance (financial, cognitive, psychiatric, or social
2) Upper body hypoxemia (also known by
limitations in patient without social support) Harlequin, or North-South syndrome) can occur
Pregnancy: if gestation age is less than 34 weeks
RELATIVE CONTRAINDICATIONS
with femoral V-A cannulation.39-42 Fully saturated
Contraindication for anticoagulation blood returns from the circuit into the femoral artery
Bloodless treatment patients (who refuse blood transfusion providing retrograde perfusion of the distal aorta.
products)
Obesity Blood ejected from the heart selectively perfuses
Advanced age the coronaries, brain, and upper extremities in an
Major peripheral vascular disease is contraindication for
peripheral V-A ECMO not for central antegrade manner. The two blood streams form a
V-A=Venoarterial; VAD=ventricular assist device “mixing cloud” or a watershed area. Depending on
native lung function, volume status, and contractility,
Table 12-2. Contraindications for V-A ECMO use. blood ejected from the LV may be poorly oxyge-

114
 Venoarterial ECMO in Adult and Pediatric Patients

Figure 12-1. Peripheral venoarterial ECMO cannulation approach: femoral vein (for drainage), (A) femoral,
(B) axillary, (C) carotid, artery used for perfusion.

Figure 12-2. Central venoarterial ECMO cannulation approach.

Figure 12-3. Special configuration to support RV with ECMO: (A) Double lumen single cannula (IJ to PA);
(B) Two cannula technique (femoral vein, and IJ to PA).

115
Chapter 12

nated, resulting in hypoxia of the coronaries and V-A ECMO in Neonates and Children.
cerebral circulation, worsening as the heart recovers
(ejection of blood from the LV). The cardiac and ce- V-A ECMO in Neonates
rebral hypoxia remain unrecognized if oxygenation
is monitored only from the lower extremities or even V-A ECMO replaces both cardiac and pulmo-
the left upper extremity. Detection of differential nary function. During partial V-A ECMO blood
hypoxemia requires meticulous attention through from the circuit mixes with LV blood which has
frequent blood gas monitoring from a right upper
extremity arterial line. Upper body hypoxemia does Parameter Score
not generally exist with central or subclavian arterial Acute cardiogenic shock diagnosis group (select one or more)
cannulation as oxygenated blood perfuses the arch Myocarditis 3
Refractory VT/VF 2
vessels in an antegrade fashion. Post heart or lung transplantation 3
3) Ischemia of the ipsilateral lower extremity Congenital heart disease 3
is another major complication. Cannula size and Other diagnoses leading to cardiogenic 0
shock requiring V-A ECMO
positioning, as well as the presence of underlying Age (years)
peripheral vascular disease, play a major role. Isch- 18–38 7
emia may lead to compartment syndrome resulting 39–52 4
in muscle necrosis, metabolic acidosis, and even 53–62 3
≥63 0
lower extremity amputation. This complication can Weight (kg)
be avoided by using a vascular conduit end-to-side ≤65 1
graft into the superficial femoral artery to avoid 65–89 2
the cannula obstruction of the distal arterial blood 0
Acute pre-ECMO organ failures (select one or more if required)
flow. The use of a retrograde or antegrade distal Liver failure -3
perfusion catheter (DPC) to perfuse the leg distally Central nervous system dysfunction -3
to the ECMO cannula also decreases this complica- Renal failure -3
tion significantly. DPC should be inserted at ECMO Chronic renal failure -6
Duration of intubation prior to initiation of
cannulation or shortly after. 43-47
ECMO (h)
≤10 0
Outcomes 11–29 -2
≥30 -4
Peak inspiratory pressure ≤20 cmH2O 3
According to the International Extracorporeal Pre-ECMO cardiac arrest -2
Life Support Organization (ELSO) Registry reports Diastolic blood pressure before ECMO ≥40 3
through July 2017,48 over 83,000 patients received mmHg
Pulse pressure before ECMO ≤20 mmHg -2
ECLS, of whom 69% were weaned off ECMO and HCO3 before ECMO ≤15 mmol/L -3
55% were discharged or transferred from the hos- Constant value to add to all calculations of SAVE-score
pital. Approximately 28,500 patients have received Total Score -35 to 17
ECLS for cardiac support with survival to discharge
Total SAVE Score Risk Class Survival %
rates of 41%, 51%, and 41% for neonatal, pediatric, Hospital survival by risk class
and adult populations, respectively.48 ECMO was >5 I 75
used for extracorporeal cardiopulmonary resuscita- 1-5 II 58
tion (eCPR) in over 9,000 patients with survival to -4 to 0 III 42
discharge rates of 41%, 41%, and 28% for neonates, -9 to -5 IV 30
<10 V 18
children, and adults, respectively. VT=ventricular tachycardia; VF=ventricular fibrillation:
Risk Scores have been developed to help pre- V-A=Venoarterial.. From Schmidt et al. Eur Heart J,
dictions of in-hospital outcome. The SAVE Score, 2015;36:2246-2256
derived from the ELSO Registry, might be helpful Table 12-3. Parameter Score (SAVE Score) to
to assess early prognosis in adults receiving V-A predict survival after ECMO for refractory cardio-
ECMO for cardiogenic shock (Table 12-3). genic shock.1

116
 Venoarterial ECMO in Adult and Pediatric Patients

traversed the lungs. Arterial oxygen and CO2 con- diaphragm and to insert the arterial cannula 3-3.5
tent represent mixtures of these two sources (native cm into the common carotid artery.
lung and artificial lung). Total V-A ECMO occurs
only in the operating theatre (CPB) when the heart Indication and Contraindication
is completely drained. In neonates and children
V-A ECMO is primarily used to manage refractory The major indication for V-A ECMO in chil-
cardiac failure. It is also used to manage respiratory dren is cardiac failure which can be divided in two
failure when V-V ECMO cannot be performed for groups: cases involving cardiac surgery and those
technical reasons but V-V ECMO is preferred for not involving cardiac surgery.53 With the former
respiratory failure.49 V-A ECMO may be employed group, the indications include preoperative stabi-
in smaller infants where the right IJ vein cannot ac- lization (neonates with total anomalous pulmonary
commodate the smallest V-V dual lumen cannula venous connection, tetralogy of Fallot with absent
(13F),50 and in neonates with severe pulmonary pulmonary valve syndrome; refractory pulmonary
hypertension (especially neonates with congenital hypertension in d-transposition of great arteries;
diaphragmatic hernia although this may depend on Ebstein’s anomaly and functional pulmonary atresia
institutional preferences),51 in neonates with low with ductal dependent flow and high pulmonary
cardiac output or with cardiac disease requiring vascular resistance), failure to wean from CPB, low
hemodynamic stabilization prior to surgery or after cardiac output syndrome, and cardiac arrest. The
surgery or those in septic shock. indications unrelated to cardiac surgery are: cardiac
arrest, myocarditis, cardiomyopathy, pulmonary
V-A ECMO Cannulation in Neonates and Infants hypertension, and intractable arrhythmias. 53-56
Contraindications have evolved. In recent years,
Each center should have a chart for cannula absolute contraindications include incurable malig-
selection based on weight and required flow. We nancy, multiple organ failure, severe central nervous
have a selection of cannulae immediately available system damage and patients who are not transplant
and optional cannulae once the vessels have been candidates. Patients with functional univentricular
visualized.52 Single lumen cannulae suitable for physiology no longer represent a contraindication
V-A ECMO are wire reinforced to avoid kinking. to ECMO, even though their survival to hospital
The arterial cannulae suitable for neonates have discharge remains poor.57
only an end hole, while venous cannulae have side
holes for better drainage. Arterial cannulae include Cannulation
8F and 10F (Bio-Medicus, Medtronic Minneapolis,
MN) while venous cannula include 8,10,12,14 Fr Cannulation technique should be based on type
(Bio-Medicus) for neonates between approximately of support needed, patient size (generally sizes >20
2-5 kg. In neonates, cannulation occurs preferably kg allow peripheral access, but vascular ultrasound
through the right IJ vein and common carotid artery. has become mandatory to evaluate vessel size be-
Left sided cervical cannulation is also possible but fore cannulation). For patients with postcardiotomy
technically challenging; therefore, we recommend failure, central cannulation is typically used. In case
the use of central cannulation (sternotomy) if right immediate support is needed such as ECPR, cardiac
sided vessels are unavailable. With cervical cannula- arrest in patients with myocarditis or failing cardio-
tion, the technique can be percutaneous, semiopen myopathy, peripheral access using the Seldinger
or open. It is practice of the authors to use the open technique or open arteriotomy and venotomy may
approach for V-A cannulation in neonates using be performed.52 Pump flow can be set according to
a 2 cm transverse incision over the right sterno- patient needs in order to maintain an aerobic me-
cleidomastoid muscle approximately 1 cm above tabolism (Table 12-4).
the clavicle. We also suggest the tip of the venous
cannula (which is radiopaque) sit 1 cm above the

117
Chapter 12

Venting majority of whom have recently undergone congeni-

tal heart surgery. Other cardiac conditions in which
Sometime heart function cannot recover if ECPR may be utilized include unrepaired congenital
left-sided venting is not performed. Venting allows heart disease, myocarditis and cardiomyopathy, and
myocardial perfusion which is compromised when intoxication-induced cardiovascular failure.58
the LV does not eject. During normal physiology,
myocardial perfusion occurs in diastole. When the Myocarditis and Cardiomyopathy
LV does not eject, it fills until its cavity pressure
reaches the myocardial perfusion pressure, halting Children with fulminant myocarditis can ben-
coronary blood flow. In these cases, LV venting efit from ECMO as a bridge to recovery, a bridge
is mandatory to allow recovery of the heart and to ventricular assist device (VAD), or a bridge to
to avoid pulmonary hemorrhage and myocardial transplant.59
ischemia.52 Venting can be performed in neonates Children with persistent low cardiac output
and infants using a percutaneous septostomy or state unresponsive to medical therapy can benefit
positioning a venting cannula in the right superior from ECMO. The ELSO Registry reports that 60%
pulmonary vein during surgery. of children with myocarditis have successfully
weaned from ECMO. ECMO should be considered
Postcardiotomy ECMO a temporary device to bridge to heart recovery or
VAD placement. It remains the support of choice
ECMO is often used both in neonates and chil- in patients experiencing cardiac arrest or severe
dren to provide postcardiotomy support. Sometimes decompensated state unable to tolerate sternotomy
ECMO is requested to allow the weaning from CPB for VAD placement. ECMO can also serve as a
without using high doses of inotropic drugs. In other bridge to transplantation in children with irrevers-
patients, it may be used to support a progressive low ible myocardial dysfunction in patients with di-
cardiac output state due to ventricular dysfunction, lated cardiomyopathy, restrictive cardiomyopathy,
pulmonary hypertension, or intractable arrhyth- endstage congenital cardiac disease, and chronic
mias.55,57 Patients requiring ECMO after CPB re- graft dysfunction after transplantation. In this case,
main cannulated centrally with a venous cannula in the urgency to stabilize the patient and the ability
the right atrium, an arterial cannula in the aorta and to tolerate sternotomy underscores the choice of
a vent in the left atrium. Patients requiring support ECMO. However, it is recommended to transit to a
later in their postoperative course are more likely more durable support (e.g. VAD) if heart recovery
to be cannulated with right cervical cannulation to does not occur within 7-10 days. Prolonged ECMO
preserve mediastinal integrity. runs have been associated with lower posttransplant
survival.
ECPR
Pulmonary Hypertension
Another increasing indication for ECMO is
extracorporeal cardiopulmonary support (ECPR) The recently published American Heart Asso-
in patients with heart disease (see Chapter 9) the ciation/American Thoracic Society 2015 Pediatric
Pulmonary Hypertension guidelines recommend
Patient Weight (kg) Blood Flow Rate the use of ECMO for the management of refractory
1.8-3 100-200 ml/kg/min postoperative pulmonary hypertension in children
3-10 100-150 ml/kg/min after cardiac surgery. Mechanical support is con-
10-15 80-125 ml/kg/min troversial for patients with irreversible pulmonary
15-30 50-100 ml/kg/min hypertension such as primary pulmonary hyperten-
>30 < 50 50-75 ml/kg/min
sion, as the likelihood of survival as a bridge to lung
>50 60-65 ml/kg/min
transplantation is minimal.60
Table 12-4. Blood flow rate by weight (kg).

118
 Venoarterial ECMO in Adult and Pediatric Patients

Arrhythmias with Hemodynamic Compromise

In selected patients with severe tachyarrhyth-

mias and bradyarrhythmias secondary to myo-
carditis or cardiomyopathy or junctional ectopic
tachycardia or ventricular tachycardia after cardiac
surgery, ECMO may prevent circulatory collapse.61

119
Chapter 12

References 10. Bartlett RH, Gattinoni L. Current status of

extracorporeal life support (ECMO) for car-
1. Makdisi G, Wang IW. Extra corporeal mem- diopulmonary failure. Minerva Anestesiol
brane oxygenation (ECMO) review of a lifesav- 2010;76(7):534-540.
ing technology. J Thorac Dis 2015;7(11):E166- 11. Aubron C, Cheng AC, Pilcher D, et al. Factors
167. associated with outcomes of patients on extra-
2. MacLaren G, Combes A, Bartlett RH. Contem- corporeal membrane oxygenation support: a
porary extracorporeal membrane oxygenation 5-year cohort study. Crit Care 2013;17(2):R73.
for adult respiratory failure: life support in the 12. Bréchot N, Luyt CE, Schmidt M, et al. Veno-
new era. Intensive Care Med 2012;38(2):210- arterial extracorporeal membrane oxygenation
220. support for refractory cardiovascular dysfunc-
3. Peek GJ, Mugford M, Tiruvoipati R, et al. Ef- tion during severe bacterial septic shock. Crit
ficacy and economic assessment of conventional Care Med 2013;41(7):1616-1626.
ventilatory support versus extracorporeal mem- 13. Bittner HB, Binner C, Lehmann S, et al.
brane oxygenation for severe adult respiratory Replacing cardiopulmonary bypass with ex-
failure (CESAR): a multicentre randomised tracorporeal membrane oxygenation in lung
controlled trial. Lancet 2009;374(9698):1351- transplantation operations. Eur J Cardiothorac
1363. Surg 2007;31(3):462-467; discussion 467.
4. Cooper DS, Jacobs JP, Moore L, et al. Cardiac 14. Gulack BC, Hirji SA, Hartwig MG. Bridge to
extracorporeal life support: state of the art in lung transplantation and rescue post-transplant:
2007. Cardiol Young 2007;17(Suppl 2):104-115. the expanding role of extracorporeal membrane
5. Combes A, Leprince P, Luyt CE, et al. Outcomes oxygenation. J Thorac Dis 2014;6(8):1070-
and long-term quality-of-life of patients sup- 1079.
ported by extracorporeal membrane oxygen- 15. Bermudez CA, Shiose A, Esper SA, et al. Out-
ation for refractory cardiogenic shock. Crit Care comes of intraoperative venoarterial extracor-
Med 2008;36(5):1404-1411. poreal membrane oxygenation versus cardiopul-
6. Smedira NG, Blackstone EH. Postcardiotomy monary bypass during lung transplantation. Ann
mechanical support: risk factors and outcomes. Thorac Surg 2014;98(6):1936-1942; discussion
Ann Thorac Surg 2001;71(3 Suppl):S60-66; 1942-3.
discussion S82-85. 16. Seco M, Forrest P, Jackson SA, et al. Extracor-
7. Chen YS, Lin JW, Yu HY, et al. Cardiopulmo- poreal membrane oxygenation for very high-risk
nary resuscitation with assisted extracorporeal transcatheter aortic valve implantation. Heart
life-support versus conventional cardiopulmo- Lung Circ 2014;23(10):957-62.
nary resuscitation in adults with in-hospital 17. Husser O, Holzamer A, Philipp A, et al. Emer-
cardiac arrest: an observational study and pro- gency and prophylactic use of miniaturized
pensity analysis. Lancet 2008;372(9638):554- veno-arterial extracorporeal membrane oxygen-
561. ation in transcatheter aortic valve implantation.
8. Shin TG, Choi JH, Jo IJ, et al. Extracorporeal Catheter Cardiovasc Interv 2013;82(4):E542-
cardiopulmonary resuscitation in patients with 551.
inhospital cardiac arrest: A comparison with 18. Banjac I, Petrovic M, Akay MH, et al. Extracor-
conventional cardiopulmonary resuscitation. poreal membrane oxygenation as a procedural
Crit Care Med 2011;39(1):1-7. rescue strategy for transcatheter aortic valve
9. Elsharkawy HA, Li L, Esa WA, et al. Outcome replacement cardiac complications. ASAIO J
in patients who require venoarterial extracor- 2016;62(1):e1-4.
poreal membrane oxygenation support after 19. Makdisi G, Makdisi PB, Wang IW. Use of ex-
cardiac surgery. J Cardiothorac Vasc Anesth tracorporeal membranous oxygenator in trans-
2010;24(6):946-951. catheter aortic valve replacement. Ann Transl
Med 2016;4(16):306.

120
 Venoarterial ECMO in Adult and Pediatric Patients

20. Arlt M, Philipp A, Voelkel S, et al. Early expe- 31. Hlavacek AM, Atz AM, Bradley SM, Bandisode
riences with miniaturized extracorporeal life- VM. Left atrial decompression by percutaneous
support in the catheterization laboratory. Eur J cannula placement while on extracorporeal
Cardiothorac Surg 2012;42(5):858-863. membrane oxygenation. J Thorac Cardiovasc
21. Alkhouli M, Al Mustafa A, Chaker Z, Algahtani Surg 2005;130(2):595–596.
F, Aljohani S, Holmes DR. Mechanical circula- 32. Aiyagari RM, Rocchini AP, Remenapp RT,
tory support in patients with severe aortic steno- Graziano JN. Decompression of the left atrium
sis and left ventricular dysfunction undergoing during extracorporeal membrane oxygenation
percutaneous coronary intervention. J Card using a transseptal cannula incorporated into the
Surg. 201732(4):245-249. circuit. Crit Care Med 2006;34(10):2603–2606.
22. Rinieri P, Peillon C, Bessou JP, et al. National 33. Guirgis M, Kumar K, Menkis AH, Freed D.
review of use of extracorporeal membrane Minimally invasive left-heart decompression
oxygenation as respiratory support in thoracic during venoarterial extracorporeal membrane
surgery excluding lung transplantation. Eur J oxygenation: an alternative to a percutaneous
Cardiothorac Surg 2015;47(1):87-94. approach. Interact Cardiovasc Thorac Surg
23. Makdisi G, Makdisi PB, Wang IW. New Ho- 2010;10(5):672–674.
rizons of non-emergent use of extracorporeal 34. Ogawa S, Richardson JE, Sakai T, Ide M, Tanaka
membranous oxygenator support. Ann Transl KA. High mortality associated with intracardiac
Med 2016;4(4):76. and intrapulmonary thromboses after cardiopul-
24. Woods FM, Nepture WB, Palatchi A, et al. monary bypass. J Anesth 2012;26(1):9–19.
Resection of the carina and main-stem bronchi 35. Makdisi G, Hashmi ZA, Wozniak TC, Wang
with the use of extracorporeal circulation. N I. Left ventricular thrombus associated with
Engl J Med 1961;264:492-494. arteriovenous extra corporeal membrane oxy-
25. Lei J, Su K, Li XF, et al. ECMO-assisted carinal genation. J Thorac Dis. 2015;7(11):E552-E554.
resection and reconstruction after left pneumo- 36. Avalli L, Maggioni E, Sangalli F, et al. Percuta-
nectomy. J Cardiothorac Surg 2010;5:89. neous left-heart decompression during extracor-
26. Spaggiari L, Rusca M, Carbognani P, Contini poreal membrane oxygenation: an alternative to
S, Barboso G, Bobbio P. Segmentectomy on a surgical and transeptal venting in adult patients.
single lung by femorofemoral cardiopulmonary ASAIO J 2011;57(1):38-40.
bypass. Ann Thorac Surg 1997;64(5):1519. 37. Aiyagari RM, Rocchini AP, Remenapp RT, et
27. Felten ML, Michel-Cherqui M, Puyo P, Fischler al. Decompression of the left atrium during
M. Extracorporeal membrane oxygenation use extracorporeal membrane oxygenation using a
for mediastinal tumor resection. Ann Thorac transseptal cannula incorporated into the circuit.
Surg 2010;89(3):1012. Crit Care Med 2006;34(10):2603-2606.
28. Brenner M, O’Connor JV, Scalea TM. Use of 38. Kawashima D, Gojo S, Nishimura T, et al. Left
ECMO for resection of post-traumatic ruptured ventricular mechanical support with Impella
lung abscess with empyema. Ann Thorac Surg provides more ventricular unloading in heart
2010;90(6):2039-2041. failure than extracorporeal membrane oxyge-
29. Souilamas R, Souilamas JI, Alkhamees K, et al. nation. ASAIO J 2011;57(3):169-176.
Extra corporal membrane oxygenation in gen- 39. Chung M, Shiloh AL, Carlese A. Monitoring of
eral thoracic surgery: a new single veno-venous the Adult Patient on Venoarterial Extracorporeal
cannulation. J Cardiothorac Surg 2011;6:52. Membrane Oxygenation. Scientific World Jour-
30. Oey IF, Peek GJ, Firmin RK, Waller DA. nal. 2014;2014:393258.
Post-pneumonectomy video-assisted thora- 40. Cove ME. Disrupting differential hypoxia in pe-
coscopic bullectomy using extra-corporeal ripheral veno-arterial extracorporeal membrane
membrane oxygenation. Eur J Cardiothorac oxygenation. Crit Care. 2015;19(1):280.
Surg2001;20(4):874-876. 41. Kinsella JP, Gerstmann DR, Rosenberg AA. The
effect of extracorporeal membrane oxygenation

121
Chapter 12

on coronary perfusion and regional blood flow 53. Di Nardo M, MacLaren G, Marano M, Cecchetti
distribution. Pediatr Res. 1992;31(1):80–84. C, Bernachi P, Amodeo A. ECLS in Pediatric
42. Hoeper MM, Tudorache I, Kühn C, et al. Ex- Cardiac Patients. Front Pediatr. 2016;4:109.
tracorporeal membrane oxygenation watershed. eCollection 2016.
Circulation. 2014;130(10):864–865. 54. Dalton HJ, Siewers RD, Fuhrman BP, et al.
43. Makdisi G, Makdisi T, Wang I-W. Use of Extracorporeal membrane oxygenation for car-
distal perfusion in peripheral extracorporeal diac rescue in children with severe myocardial
membrane oxygenation. Ann Transl Med. dysfunction. Crit Care Med 1993; 21(7):1020-
2017;5(5):103. 1028.
44. Aziz F, Brehm CE, El-Banyosy A, et al. Arterial 55. Aharon AS, Drinkwater DC, Churchwell KB,
complications in patients undergoing extracor- et al. Extracorporeal membrane oxygenation
poreal membrane oxygenation via femoral can- in children after repair of congenital cardiac
nulation. Ann Vasc Surg 2014;28(1):178-183. lesions. Ann Thorac Surg 2001; 72(6): 2095-
45. Cheng R, Hachamovitch R, Kittleson M, et al. 2101.
Complications of extracorporeal membrane 56. Booth KL, Roth SJ, Thiagarajan RR, Almodovar
oxygenation for treatment of cardiogenic shock MC, del Nido PJ, Laussen PC. Extracorporeal
and cardiac arrest: a meta-analysis of 1,866 adult membrane oxygenation support of the Fontan
patients. Ann Thorac Surg 2014;97(2):610-616. and bidirectional Glenn circulations. Ann Tho-
46. von Segesser L, Marinakis S, Berdajs D, et rac Surg 2004; 77(4): 1341-1348.
al. Prevention and therapy of leg ischaemia in 57. Chang AC, Wernovsky G, Kulik T, Jonas RA,
extracorporeal life support and extracorporeal Wessel DL. Management of the neonate with
membrane oxygenation with peripheral cannu- transposition of the great arteries and persistent
lation. Swiss Med Wkly2016;146:w14304. pulmonary hypertension. Am J Cardiol 1991;
47. Lafçı G, Budak AB, Yener AU, Cicek OF. Use 68(11):1253-1255.
of extracorporeal membrane oxygenation in 58. Chan T, Thiagarajan RR, Frank D, et al. Survival
adults. Heart Lung Circ 2014;23(1):10-23. after extracorporeal cardiopulmonary resuscita-
48. Extracorporeal Life Support Registry Report tion in infants and children with heart disease. J
international summary. Available online: https:// Thorac Cardiovasc Surg 2008; 136(4): 984-992.
www.elso.org/Registry/Statistics/International- 59. Asaumi Y, Yasuda S, Morii I, et al. Favourable
Summary.aspx . Accessed on October 8, 2017. clinical outcome in patients with cardiogenic
49. Brogan T, Lequier L, Lorusso R, MacLaren G, shock due to fulminant myocarditis supported
Peek G. Extracorporeal life support: the ELSO by percutaneous extracorporeal membrane oxy-
Red Book 5th Edition. genation. Eur Heart J 2005; 26(20): 2185-2192.
50. Frenckner B, Palmer K, Linden V. Neonates 60. Abman SH, Hansmann G, Archer SL, et al.
with congenital diaphragmatic hernia have Pediatric pulmonary hypertension: Guidelines
smaller neck veins than other neonates –an from the American Heart Association and Ame-
alterbative route for cannulation. J Pediatr Surg rican Thoracic Society. Circulation 2015; 132
2002; 37(6):906-908. (21): 2037-2099.
51. Fisher J, Jefferson RA, Kuenzler KA, Stolar 61. Walker GM, McLeod K, Brown KL, et al. Ex-
CJ, Arkovitz MS. Challenges to cannulation tracorporeal life support as treatment of supra-
for extracorporeal support in neonates with ventricular tachycardia in infants. Pediatr Crit
right-sided congenital diaphragmatic hernia. J Care Med 2003; 4(1):52-54.
Pediatr Surg 2007; 42(12):2123-2128.
52. Peek GJ, Hammond I. Neonatal/pediatric
cardiac cannulation, chapter 30 in: Brogan T,
Lequier L, Lorusso R, MacLaren G, Peek G.
Extracorporeal life support: the ELSO Red
Book 5th Edition.

122
13

Venovenous Extracorporeal Membrane Oxygenation

L. Mikael Broman, MD, PhD, Scott Wagoner, MBA, RRT

Introduction Respiratory ECMO

The two most common cannulation approaches While V-V ECMO may be the preferred mode
for extracorporeal membrane oxygenation (ECMO) to support respiratory failure patients, some may be
are venoarterial (V-A) ECMO and venovenous (V- at risk for increased pulmonary vascular resistance
V) ECMO. Venoarterial ECMO is offered to patients (e.g. primary or secondary cute respiratory distress
with heart, or combined heart and lung failure as it syndrome [ARDS]), leading to right ventricular
provides direct circulatory support since the oxygen- failure. Biventricular failure may also be seen from
ated blood is reinfused into the arterial circulation. the cytokine storm in septic shock. When assessing
V-A ECMO provides a portion or complete cardiac these patients before acceptance for cannulation
output support and is described in Chapter 12. and initiation of V-V ECMO, echocardiography is a
Venovenous ECMO lacks direct circulatory valuable tool. V-A ECMO is the better mode for the
support, and is employed for patients with refrac- sizeable subset of patients who require circulatory
tory respiratory failure with adequate native cardiac support. Septic shock patients commenced on V-V
output.1-3 The deoxygenated blood is drained from ECMO may have an increased risk for mortality
the right atrium or adjacent superior and/or infe- as do patients that later need to be converted from
rior caval veins (SVC and IVC, respectively), via V-V to V-A ECMO (unpublished data). Centers that
venous access(es), pumped though the membrane exclusively perform V-V ECMO should establish
lung (ML), oxygenated and then reinfused into the a rapport with a center who can accept these com-
venous circulation. The oxygenated ECMO blood plicated patients.
and systemic venous blood mixes in the major
veins, the right atrium (RA), and right ventricle. Circuit Operation
This results in reliance on native cardiac output for
oxygen delivery, (DaO2). Circuit and venous blood The circuits used for V-V ECMO are generally
mix before the lung; oxygen saturations should be less complex than for V-A ECMO, and may not in-
uniform no matter the site of measurement. In two clude a bridge, or a less sized loop which is an A-V
randomized controlled studies ECMO treatment shunt closer to the ML providing access.7
for refractory severe respiratory failure have shown The ECMO blood flow (QEC) required for ad-
increased survival compared to conventional inva- equate oxygenation support is generally regarded to
sive ventilation treatment in adults and newborns.4-5 be 100-150 mL/kg per minute in the newborn, 80-
At the ELSO webpage6 guidelines concerning 120 ml/kg in pediatric patients and >2 L/min (50-80
Respiratory ECMO for the different age groups are mL/kg per minute) in the adult, aiming for a SaO2 of
published and frequently updated. ELSO guidelines 75-85%.8-9 Thus, the requirements for cannula size
related to this chapter are Identification and manage- differ substantially from that needed for the flows
ment of recirculation on VV ECMO, and Ultrasound of 10 mL/kg per minute needed for extracorporeal
guidance for extracorporeal membrane oxygenation carbon dioxide removal (ECCO2R, see Chapter 15) .
venovenous ECMO.

123
Chapter 13

Table 13-1 shows oxygen saturation profiles Percutaneous cannulation is employed more
during changes in physiology and flow patterns, and commonly with V-V than with V-A ECMO. Still,
how to interpret and manage these. centers that developed from pediatric or neonatal
intensive care, or cardiac surgical units may have
Cannulation cannulating surgeons who are accustomed to the
semi or fully invasive approaches for cannulation.
Peripheral cannulation for V-V ECMO adheres Since the H1N1 pandemic and the CESAR trial,4
to the same Seldinger or semi-percutaneous tech- V-V ECMO treatment has been used in a growing
nique used for V-A ECMO.10 In semi-percutaneous adult population.12 Adult respiratory patients are
cannulations the vessel is visualized via a skin inci- often treated in intensive care units (ICU) where
sion after which the procedure follows the Seldinger providers are familiar with percutaneous techniques.
technique, or via direct surgical cannulation. The Thus, the adjustment from placing a larger central
percutaneous technique is preferred since both can- venous catheter compared to an ECMO cannula can
nulation and decannulation require fewer resources be easily accomplished.11 The goal of all cannulation
and exposes the patient to less risk.11 configurations for ECMO is to extract the blood with

Arterial Cephalad Venous Interpretation Management

Saturation Saturation Saturation
Increasing Increasing or Increasing or Patient is improving Wean ECMO flow
stable stable
Decreasing Decreasing or Decreasing or Patient is worse. Check catheter position
stable stable and pump flow. Try to
increase flow.
Decreasing Stable but with Increasing Increased Evaluate catheter position,
decreased flow recirculation most adjust head position, add
likely due to loss of or subtract shoulder roll,
cephalad catheter apply gentle traction to
flow catheter
Decreasing Decreasing or Increasing Increased Evaluate cardiac output
stable with good recirculation most and catheter position,
flow likely due to consider giving volume.
changing catheter
position
Decreasing or Increasing with Stable Poor ventilation. Check ABG. Adjust sweep
stable increased flow PaCO2 is increased gas flow or mix.
If off ECMO* adjust
ventilator support.
Stable Decreasing with Increasing Over ventilation, Check ABG. Adjust sweep
decreased flow PaCO2 is decreased gas flow or mix.
If off ECMO* adjust
ventilator support.
Decreasing and Decreasing and Increasing Worry about Consider pressure
decreased blood decreased flow compromised pneumothorax, or
pressure, cardiac output pericardial tamponade if
increasing heart pulse pressure is decreased
rate.
ABG=arterial blood gas
*In the weaning phase performing a trial-off with the sweep gas turned off.

Table 13-1. V-V ECMO Scenario Chart which shows oxygen saturation profiles during changes in physiology
and flow patterns, interpretation and management during V-V ECMO. Saturations obtained are patient arterial,
venous (SPREO2, draining limb of ECMO circuit), and from a cephalad catheter (venous flow and saturation
from a catheter placed in the internal jugular vein cranially to the venotomy for the major venous draining can-
nula). Modification of original table adapted from Children's Healthcare of Atlanta; ECMO training manual.

124
 Venovenous Extracorporeal Membrane Oxygenation

the lowest oxygen content possible. That would be directed towards the TV. The common cavoatrial
the lever arm for maximum oxygen delivery that DLC is placed via the RIJ but designs may promote
then depends on the performance of the ML, hemo- placement via either femoral vein as well. Recently
globin content, ECMO flow, and recirculation issues a cannula designed for atriopulmonary artery (RA-
(V-V). The reinfusion cannula should be placed to PA) configuration was introduced but should in
minimize recirculation, not to return into the most its essence be regarded as a dual lumen cannula
desaturated blood for respiratory and right ventricular V-A support.
Configurations for SLCs may be femoro-femoral,
Cannulae femoro-jugular, SVC(/RA)-femoral, or RA-femoral.
Some of these are illustrated in Figure 13-1.
To provide maximal oxygen delivery, V-V Hybrids: Under circumstances with draining
ECMO should extract blood with the lowest pos- problems, two draining cannulae may be used, i.e.
sible oxygen content and minimize recirculation. VV-V ECMO. In this configuration, ‘femoro and
This can be accomplished with either two or more jugular to RA’ would be the most efficient configu-
single lumen cannulae (SLC) or one dual lumen ration concerning recirculation which mimics the
cannula (DLC). The most common cannulation flow dynamics of a bicaval DLC. Some centers
sites are the right internal jugular vein (RIJ) and the use a cephalad cannula cranial to the venotomy for
femoral veins. DLCs are typically placed in the RIJ the primary jugular ECMO cannula. The cephalad
with fluoroscopic or echocardiographic guidance to cannula is an additional drainage to prevent venous
provide safe insertion and positioning. The DLCs stasis or thrombosis of cerebral veins.13
designs vary and may be placed only in the RA or
have a bicaval design that drains venous blood from Recirculation
the IVC and SVC and returns arterialized blood to
the RA directed towards the tricuspid valve (TV) to Because V-V ECMO drains venous blood and
minimize recirculation (Figure 13-1). The cavoatrial reinfuses arterialized blood in the venous circulation
cannula design drains venous blood from both the risk of recirculation exists. Recirculation is the
caval veins via proximal drainage sites and from portion of oxygenated blood that returns directly to
the tip of the catheter. The return port is in the RA the ECMO circuit without circulating to the patient
first. Or, the percentage of total ECMO blood flow
that reenters the drainage cannula, thus not contrib-
uting to patient oxygenation, is defined as the recir-
culation fraction (Rf). Major factors that influence
Rf include pump flow, cannula position and design,
cardiac output, and intravascular volume.14-18
Recirculation on ECMO is hard to assess in
clinical practice.18 Reported Rf varies from 2-60%.18-
24
Surrogate methods normally used in intensive
care to assess oxygenation (central venous oxygen
saturation ScvO2, mixed venous saturation SvO2)
are unreliable in ECMO: Rf interferes with the
Figure 13-1. A sample of different cannulae designs, measured value. Trends may be indicative though.
configurations, and cannulae placements for V-V
ECMO. Arrows indicate flow direction and poten- Rf can be calculated using the conventional
tial schematic ways for recirculation to occur related formula (Equation 1)14-15,25-26:
to cannulae matters described above. A) Cavoatrial
dual lumen cannula (DLC), B) Bicaval DLC, C)
Two single lumen cannulae (SLC) in Femoro- Rf = SPREO2 – SvO2 (Eq. 1)
jugular configuration, D) Bicaval Femoro-jugular SPOSTO2 – SvO2
SLC configuration (similar in flow pattern to the
Bicaval DLC), and E) Atrio-femoral SLC.

125
Chapter 13

SPREO2 is the saturation before the ML, SPOSTO2 Weaning

the saturation after the ML, and SvO2 the saturation
of the mixed venous blood. The difficulty in calcu- Weaning from V-V ECMO starts after successful
lating Rf arises from the complexity in measuring treatment of the underlying disease as soon as the
the true SvO2.15,27 Three methods are available to tidal volumes start to increase. It is the recovery of
estimate SvO2. In the “SvO2 method,” the sweep is one organ, the lungs, that is being assessed. Ventila-
stopped, and SvO2 is assessed when SPREO2 equals tor settings are often modest, (“lung protective”). In
SPOSTO2, but this is not recommended in critically parallel with the pulmonary improvement, ECMO
ill ECMO patients. In the “CVL method”, a cen- support has been reduced over the preceding days.
tral venous line placed in the lower SVC is used Pulmonary progression is marked by increased CO2
to approximate SvO2 from a ScvO2 sample. Both elimination compared to oxygenation. There are
methods use the conventional formula (Eq. 1) to several approaches to weaning V-V ECMO.36 As
calculate Rf. Ultrasound dilution technology (UDT) an example, the Adult ELSO Weaning Guideline8
has been validated for measurements of Rf in V-V would be to decrease QEC in steps to 1 L/min at
ECMO,15,28-29 and may improve clinical practice.18 sweep 100%, or decrease QEC to 2 L/min, reduce
Subtracting the recirculated volume flow (or sweep FiO2 to maintain SaO2 > 95%. When SaO2 is
recirculation fraction) from the total ECMO flow stable on these settings, trial off by clamping sweep,
(QEC), Equation 2 gives the effective flow (QEFF). while maintaining ventilator rest settings. If PaO2
QEFF is the volume flow of oxygenated blood sup- >95 mmHg (12.5 kPa), and PaCO2 <50 mmHg (<6.7
porting the patient. kPa) for at least 1 hour, without the development
dyspnea or other signs of stress, decannulation can
QEFF = QEC (1-Rf) (Eq. 2) be considered.

A SaO2 >75-80 % is considered adequate during

V-V ECMO.8-9 An increase in recirculation reduces
QEFF but may be compensated for by increasing
ECMO flow (QEC). However, the higher pump
flow often results in greater recirculation. When
QEC exceeds native cardiac output, recirculation
may rise such that total oxygen delivery decreases
(Figure 13-2).18,29
Hemolysis: Negative pressures pockets may
arise and the red blood cells (RBC) may also be
exposed to shear stress around the draining cannulas,
when blood passes narrower draining ports, when
chattering of the cannula occur (suction to vascular
Figure 13-2. The impact of recirculation on effec-
wall), and along the edges of the centrifugal pump tive blood flow V-V ECMO. The hashed blue line
rotor blades; cavitation/bubble formation promote shows an idealized curve for recirculation fraction
hemolysis.12,30-31 The free plasma hemoglobin re- (%, Rf, right y-axis), and the corresponding blue
leased from RBC destruction scavenges nitric oxide curve () the corresponding effective ECMO
flow (QEFF) on the left y-axis. Note that different
(NO) produced by the endothelium since NO binds Rf-curves may prevail in the same patient at dif-
1000-fold faster to free heme than to hemoglobin in ferent times, and the actual curve will be affected
the RBC.32 Decreased bioavailability of NO in the by other factors of recirculation. The dotted brown
curve is an example of another Rf-curve with its
microcirculation increases inflammation, coagula- corresponding () curve for QEFF. Note that as
tion and promotes vasoconstriction.32-35 Minimizing pump flow is increased recirculation increases and
hemolysis and coagulation activation are the reasons effective pump flow initially increases, flattens, and
finally decreases.
to keep the recirculation low.

126
 Venovenous Extracorporeal Membrane Oxygenation

Another weaning process is based on respiratory

physiology.36 Blood gas assessments before and after
the ML focus on pCO2. The sweep flow is at a mini-
mum for the respective size of ML. If the difference
between PPRECO2 and PPOSTCO2 is in equilibrium
(<2.5-3 mmHg or <0.3-0.4 kPa) almost all CO2 pro-
duced is eliminated via the patient’s own lungs. If
the delta-pCO2 is higher gaseous CO2 will be mixed
to the sweep in increments up to 5% (decrease the
diffusion gradient over the respiratory membrane,
CO2-dialyzer). The patient is observed accordingly.
When equilibrated, the patient not in stress, and the
FiO2 reasonable, a trial off is performed.

Summary

Venovenous ECMO is the primary mode for

support of refractory severe respiratory failure. In
most cases the hypoxic component predominates but
the pathophysiology is combined, i.e. both oxygen-
ation and ventilation are abnormal. In some children
and adults with severe respiratory failure, the heart
may also be affected, either by increased secondary
workload for the right ventricle, or from cytotoxic
impairment affecting global cardiac function. In
these cases, V-A ECMO should be considered from
the start.
Venovenous ECMO entails recirculation, de-
creasing the efficiency of oxygenation support to the
patient. This may, to some extent, be compensated
for by increased QEC, but the higher the flow the
higher the possibility for hemolysis and secondary
effects both in forms of increased morbidity but also
mortality. The aim should always be to keep recir-
culation at a minimum, and free plasma hemoglobin
should be monitored.

127
Chapter 13

References 11. Conrad SA, Grier LR, Scott LK, Green R, Jordan
M. Percutaneous cannulation for extracorporeal
1. Marasco SF, Lukas G, McDonald M, McMillan membrane oxygenation by intensivist: a retro-
J, Ihle B. Review of ECMO (extracorporeal spective single-institution case series. Crit Care
membrane oxygenation) support in critically ill Med. 2015;43:1010-1015.
adult patients. Heart Lung Circ. 2008;17 Suppl 12. Annich GM, Lynch WR, MacLaren G, Wilson
4:S41-S47. JM, Bartlett RH. ECMO Extracorporeal Car-
2. Rehder KJ, Turner DA, Cheifetz IM. Use of ex- diopulmonary Support in Critical Care. 4th ed.
tracorporeal life support in adults with severe United States of America: Extracorporeal Life
acute respiratory failure. Expert Rev Respir Support Organization; 2012.
Med. 2011;5:627-633. 13. Pettignano R, Labuz M, Gauthier TW, Huckaby
3. Park PK, Napolitano LM, Bartlett RH. Extracor- J, Clark RH. The use of cephalad cannulae to
poreal membrane oxygenation in adult acute monitor jugular venous oxygen content during
respiratory distress syndrome. Crit Care Clin. extracorporeal membrane oxygenation. Crit
2011;27(3):627-646. Care. 1997;1(3):95-99.
4. Peek GJ, Mugford M, Tiruvoipati R, et al. Efficacy 14. Heard M, Davis J, Fortenberry J. Principles and
and economic assessment of conventional venti- practice of venovenous and venoarterial ECMO.
latory support versus extracorporeal membrane In: ECMO Specialist Training Manual 3rd ed.
oxygenation for severe adult respiratory failure Extracorporeal Life Support Organization, Ann
(CESAR): a multicentre randomised controlled Arbor, MI, USA; 2010:59-75.
trial. Lancet. 2009;374(9698):1351-1363. 15. van Heijst AF, van der Staak FH, de Haan AF, et
5. UK Collaborative ECMO Trial Group. UK col- al. Recirculation in double lumen catheter veno-
laborative randomised trial of neonatal extracor- venous extracorporeal membrane oxygenation
poreal membrane oxygenation (ECMO Trial). measured by an ultrasound dilution technique.
Lancet 1996; 348(9020):75-82. ASAIO J. 2001;47(4):372-376.
6. ELSO Guidelines. Extracorporeal Life Support 16. Körver EP, Ganushchak YM, Simons AP, Donker
Organization, Ann Arbor, MI, USA. https:// DW, Maessen JG, Weerwind PW. Quantification
www.elso.org/Resources/Guidelines.aspx. Ac- of recirculation as an adjuvant to transthoracic
cessed April 17, 2018. echocardiography for optimization of dual-
7. Conrad SA, Broman LM, Taccone FS, et al. The lumen extracorporeal life support. Intensive
Extracorporeal Life Support Organization Care Med. 2012;38(5):906-909.
Maastricht Treaty for Nomenclature in Extra- 17. Broman LM, Hultman J. Double lumen catheter
corporeal Life Support: A Position Paper of the placement during VV ECMO in an infant with
Extracorporeal Life Support Organization. Am persistent left superior vena cava-important
J Respir Crit Care Med. 2018. considerations. ASAIO J. 2014;60(5):603-605.
8. ELSO Guidelines for Cardiopulmonary Extra- 18. Palmér O, Palmér K, Hultman J, Broman LM.
corporeal Life Support, Extracorporeal Life The influence of cannula design on recirculation
Support Organization, Version 1.4 August 2017 during venovenous extracorporeal membrane
Ann Arbor, MI, USA. oxygenation. ASAIO J. 2016;62(6):737-742.
9. Rich PB, Awad SS, Crotti S, Hirschl RB, Bartlett 19. Bonacchi M, Harmelin G, Peris A, Sani G. A
RH, Schreiner RJ. A prospective comparison of novel strategy to improve systemic oxygen-
atrio-femoral and femoro-atrial flow in adult ve- ation in venovenous extracorporeal membrane
novenous extracorporeal life support. J Thorac oxygenation: the “chi-configuration”. J Thorac
Cardiovasc Surg. 1998;116(4):628-632. Cardiovasc Surg. 2011;142(5):1197-1204.
10. Seldinger SI. Catheter replacement of the needle 20. Lindstrom SJ, Mennen MT, Rosenfeldt FL,
in percutaneous arteriography: a new technique. Salamonsen RF. Quantifying recirculation
Acta Radiologica. 1953;39(5):368-376. in extracorporeal membrane oxygenation: a

128
 Venovenous Extracorporeal Membrane Oxygenation

new technique validated. Int J Artif Organs. 30. Yasuda T, Funakubo A, Fukui Y. An investiga-
2009;32:857-863. tion of blood damage induced by static pres-
21. Wang D, Zhou X, Liu X, Sidor B, Lynch J, sure during shear rate conditions. Artif Organs.
Zwischenberger JB. Wang-Zwische double 2002;26(1):27-31.
lumen cannula-toward a percutaneous and am- 31. Toomasian JM, Bartlett RH. Hemolysis and
bulatory paracorporeal artificial lung. ASAIO ECMO pumps in the 21st Century. Perfusion.
J. 2008;54:606-611. 2011;26(1):5–6.
22. Lin TY, Horng FM, Chiu KM, Chu SH, Shieh JS. 32. Liu X, Miller MJ, Joshi MS, Sadowska-Kro-
A simple modification of inflow cannula to re- wicka H, Clark DA, Lancaster JR Jr. Diffusion-
duce recirculation of venovenous extracorporeal limited reaction of free nitric oxide with erythro-
membrane oxygenation. J Thorac Cardiovasc cytes. J Biol Chem. 1998;273(30):18709-18713.
Surg. 2009;138:503-506. 33. Ulatowski JA, Nishikawa T, Matheson-Urbaitis
23. Clements D, Primmer J, Ryman P, Marr B, Sear- B, Bucci E, Traystman RJ, Koehler RC. Region-
les B, Darling E. Measurements of recirculation al blood flow alterations after bovine fumaryl
during neonatal venovenous extracorporeal beta beta-crosslinked hemoglobin transfusion
membrane oxygenation: clinical application and nitric oxide synthase inhibition. Crit Care
of the ultrasound dilution technique. J Extra Med. 1996;24(4):558-565.
Corpor Technol. 2008;40:184-187. 34. Vogel WM, Dennis RC, Cassidy G, Apstein
24. Broman M, Frenckner B, Bjällmark A, Broomé CS, Valeri CR. Coronary constrictor effect of
M. Recirculation during venovenous extra- stroma free hemoglobin solutions. Am J Physiol.
corporeal-membrane-oxygenation – a simula- 1986;251(2:2):H413-H420.
tory study. Int J Artif Organs. 2015;31(1):23-30. 35. Vaughn MW, Kuo L, Liao JC. Effective diffusion
25. Linton R, Turtle M, Band D, O’Brien T, Jonas distance of nitric oxide in the microcirculation.
M. In vitro evaluation of a new lithium dilution Am J Physiol. 1998;274(5:2):H1705-H1714.
method of measuring cardiac output and shunt 36. Broman LM, Malfertheiner MV, Montisci A,
fraction in patients undergoing venovenous ex- Pappalardo F. (Review) Weaning from veno-
tracorporeal membrane oxygenation. Crit Care venous extracorporeal membrane oxygenation:
Med. 1998;26(1):174-177. how I do it. J Thorac Dis. 2017.
26. Sreenan C, Osiovich H, Cheung PY, Lemke
RP. Quantification of recirculation by ther-
modilution during venovenous extracorpo-
real membrane oxygenation. J Pediatr Surg.
2000;35(10):1411-1414.
27. Walker JL, Gelfond J, Zarzabal LA, Darling E.
Calculating mixed venous saturation during
veno venous extracorporeal membrane oxygen-
ation. Perfusion. 2009;24(5):333-339.
28. Walker J, Primmer J, Searles BE, Darling EM.
The potential of accurate SvO2 monitoring
during venovenous extracorporeal membrane
oxygenation: an in vitro model using ultrasound
dilution. Perfusion. 2007;22(4):239-244.
29. Darling EM, Crowell T, Searles BE. Use of dilu-
tional ultrasound monitoring to detect changes
in recirculation during venovenous extracorpo-
real membrane oxygenation in swine. ASAIO J.
2006;52(5):522-524.

129
14

Special Configurations in Adult ECLS

Vincent Pellegrino, MBBS, FCICM, Jan Bělohlávek, MD, PhD

Introduction the circulation where blood is drained from and then

returned to, determine the physiology of the support
Many forms of severe cardiac and respiratory and therefore define the mode of ECLS. These (four)
failure requiring ECLS support are inadequately regions are: the systemic venous system (including
supported with conventional forms of venovenous the right atrium), the systemic arterial system, the
(V-V ECMO) or venoarterial ECMO (V-A ECMO) left heart, and the pulmonary artery. For example,
necessitating modification of the circuit and how it V-V ECMO and V-PA ECMO are two MODES of
interacts with the native circulation. This chapter support for respiratory failure because they return
describes the physiological basis and possible oxygenated blood to different regions of the circu-
ECLS arrangements for a number of common clini- lation and provide different physiological support.
cal scenarios where nonstandard ECLS support is There are numerous drainage and return options
used and a clearly defined terminology. A lack of for any given mode of ECLS support. Thus each
consistent definitions and nomenclature in the past MODE has a number of possible configurations
has hampered the classification of ECLS support. defined by the insertion site, the type of cannula,
and the tip position of the cannulae. Table 14-1 lists
Definitions and Terminology many possible configurations for standard ECLS
and ECMO modes.
ELSO has recently proposed the following Hybrid modes of ECLS involve the simultane-
definitions for ECLS and ECMO: ous application of multiple ECLS. For example,
ECLS: A collective term for extracorporeal V-VA ECMO (veno-venoarterial ECMO) is a hybrid
therapies used for the support of various presenta- configuration of V-V and V-A ECLS in which the
tions of cardiac and/or pulmonary failure through circuit drains blood from the venous system and
the use of a pumped extracorporeal circuit. returns blood into both the venous and systemic
ECMO: The provision of oxygen and car- arterial systems.1 V-VA ECMO provides both pul-
bon dioxide exchange through the use of an monary (VV component) and cardiac support (VA
extracorporeal circuit consisting minimally of a component) in patients with combined cardiopul-
blood pump, artificial lung, and vascular access monary failure where some native cardiac output is
cannulae, using blood flows sufficient to support maintained. Patients requiring hybrid modes of sup-
port generally have higher morbidity and mortality.2
oxygenation and concomitantly enhance carbon
dioxide removal. Common Clinical Scenarios
ECLS is therefore a broad term which includes
ECMO and other forms of extracorporeal support A number of common clinical settings occur
(with or without an artificial lung). ECMO always where standard modes of ECLS cannot support
includes an artificial lung. patients adequately. Successful ECLS support re-
This chapter provides physiologically-based quires timely commencement and reconfiguration
description of ECLS support types. The region of of support when required.3 For each of the follow-

131
Chapter 14

ing clinical scenarios, the commonly associated terial septicemia, viral infections such as hantavirus
pathological conditions, important physiological cardiopulmonary syndrome, postcardiac arrest cases
considerations and available approaches to ECLS with massive pulmonary aspiration, acute mitral
support are described. valve failure and post-lung transplant cardiac and
respiratory failure.
Combined Acute Respiratory and Cardiac Failure Physiological considerations when selecting the
form of ECLS support in these settings:
Pathological conditions where support with
V-V ECMO (with inotropic support) fails to pro- • Native cardiac output (from the LV) is hypoxic
vide adequate cardiac support and (peripheral) as a result of extensive pulmonary shunt/im-
configurations of V-A ECMO are associated with paired oxygenation. It preferentially perfuses
severe differential hypoxemia due to associated lung the coronary and cerebral circulations with V-A
shunt are challenging to support adequately and ECMO support. Assessment of the severity of
may require nonstandard ECLS support.2 Common differential hypoxia is made by measuring arte-
pathological conditions associated with this state rial oxygen tension in the arterial tree away from
include pneumonia syndromes with secondary left the site of V-A ECMO return (e.g. right hand).
(LV) or right ventricular (RV) failure, advanced bac-

Cardiac Support ECLS Modes Respiratory Support ECLS Modes

V-A ECMO L-A ECLS V-PA ECLS V-V ECMO V-PA ECMO
L-A ECMO
Peripheral Peripheral Configurations Peripheral Peripheral Peripheral
Configurations Drainage Options Configuration Configurations Configuration
Drainage Options 1. LA-(Femoral-) Drainage Options 1. V-RA 1. RA-PA
1. RA-(Femero-) Return Options 1. RA- (Jugular-) (Femero- (Jug-
2. RA- (Jugular-) 1. –DA (-Femoral) Return Options Femoral) Jugular)
Return Options 2. –DA+DPC (- 2. PA- (-Jugular) 2. V-RA
1. -DA (-Femoral) Femoral +DPC*) (Femero-
2. -DA+DPC (- 3. –SA (- Subclavian) Jugular)
Femoral 3. V+V-RA
+DPC*) (Femero
3. -SA (- +Jugular-
Subclavian) Femoral)
4. V+V-RA
(Dual lumen
cannula
from
Jugular)
Central Configurations Central Configurations Central
Drainage Options Drainage Options (Sternal- Configurations
1. RA-(Sternal-) or Tunnelled-) Drainage Options
2. RA- (Tunnelled- 1. LA- 1. RA-
) 2. RSPV- Return Options
Return Options 3. LV- (Sternal or 2. PA- (Grafted-)
1. –AA (-Grafted) Thoracotomy)
2. -BCA (-Grafted) Return options (-Sternal or
3. -SA (-Grafted) -Tunnelled)
1. –AA
2. –DA
3. –SA
4. -BCA
Table 14-1. Standard ECLS MODES and configurations.

132
 Special Configurations in Adult ECLS

• Following commencement of V-A ECMO, re- provides circulatory support such as V-A ECMO
duced pulsatility/LV ejection is common with or V-VA ECMO.
changes to LV preload and afterload. Decreased • Central V-A ECMO configurations which
LV ejection reduces the extent of resulting preferentially provide cephalad perfusion may
hypoxemia. be adequate to mitigate differential hypoxia.
• Worsening differential hypoxemia occurs with Surgical arterial grafting options include the
recovering cardiac function and may signal subclavian artery, innominate artery, and proxi-
the possibility of weaning from V-A ECMO or mal aorta (sternotomy).
conversion to V-V ECMO. • Peripheral V-A ECMO support with tolerance
• V-VA ECMO provides respiratory and cardiac of- and medical management of differential
support only if some native cardiac output is hypoxia.
maintained. Where native cardiac function • (Hybrid) V-VA ECMO commenced initially or
is lost, circuit blood returning to the venous instituted after a trial of conventional (V-A or
circulation merely recirculates. V-VA ECMO V-V ECMO)
should not be instituted where cardiac ejection
is minimal or rapidly diminishing. (Figure 14-1). Hypoxemia during V-V ECMO
The quantity of circuit blood returned to the
venous system compared to the arterial system Hypoxemia during V-V ECMO support for se-
is controlled by a variable (Hoffman) tubing vere lung processes can occur despite a functioning
clamp applied to the circuit. In general, the ECMO circuit.
greater the pulsatility, the greater the proportion
of circuit blood that should be returned to the • High cardiac output with high oxygen consump-
venous system. tion due to the effects of illness, patient size,
and fever. Current adult membrane lungs can
Support options in this setting include: oxygenate up to 7 liters of deoxygenated blood
per minute.
• Progressive circulatory failure despite V-V • Poorly configured ECMO with high recircu-
ECMO, safe lung ventilation, inotropic and lation, in which large quantities of returning
volume state support is a strong indication for ECMO blood reenter the circuit with high
immediate conversion to an ECLS mode that venous saturation (>70%) due to poorly sited
return ports. This is seen if the return cannula
tip is too close to the drainage cannula or in the
case of dual lumen bicaval cannula, the return
port is oriented away from the tricuspid valve.
• Poorly configured ECMO with low circuit blood
flow due to poorly placed drainage cannula. In
this case, venous collapse (“chatter”) obstructs
the inflow into the circuit and overall circuit
flow and therefore patient oxygen delivery is
reduced. This may be seen with a low femoral
drainage cannula sitting in the low inferior vena
cava or the distal part of a dual lumen bicaval
cannula being sited in the hepatic vein.
• Adequately configured V-V ECMO with poor
circuit flow due to patient factors such as high
Figure 14-1. V-VA ECMO graphic. Note: In the work of breathing (negative inspiratory pres-
absence of native cardiac flow, post oxygenator sure), agitation, coughing, inadequate volume
blood returned to the venous circulation with simply
recirculate and provide no additional benefit.

133
Chapter 14

state, low intravascular volume, or poor native ensuring adequate separation between drainage
cardiac output. and return cannulae/ports.
• In the case of poor circuit flow due to a distal
In the presence of a large pulmonary shunt, the venous drainage cannula, improving circuit
patient arterial oxygen saturation depend predomi- drainage by placing an additional drainage can-
nantly on the proportion of deoxygenated venous re- nula to create bicaval drainage (V+V-V ECMO).
turn that can be captured by the V-V ECMO circuit.4
Draining blood from the vena cava and returning to Left Ventricular Distention
the right atrium (cavo-atrial flow) improves access
to deoxygenated blood and oxygen transfer via In patients with severe heart failure peripheral
the circuit.5 Options for improving arterial oxygen V-A ECMO is a rapidly deployable short-term sup-
saturation depend on cause and include6: port that can prevent end organ damage from under
• In the case of high oxygen consumption with perfusion, provide time for reversible forms of car-
good V-V ECMO blood flow: reducing oxygen diac failure to recover, or allow time for thorough
consumption with heavy sedation, paralysis, assessment of therapeutic options for irreversible
barbiturates, or cooling and reducing native forms of heart failure. However, despite the fact,
cardiac output with beta-blockade have been that V-A ECMO unloads the RV and thereby also
advocated. LV, in some settings V-A ECMO may contribute to
• Consideration to changing the mode of sup- gross loading of the LV and pulmonary circulation
port from V-V ECMO to V-A ECMO or V-VA which will 7require intervention to change the mode
ECMO has also been suggested but is seldom of support.
required. Conditions where V-A ECMO support can lead
• If the circuit flow is high and recirculation is to severe LV loading (Table 14-2):
not elevated, tolerating a degree of hypoxemia • Severe LV dysfunction with relative sparing of
is also commonly practiced. RV function: extensive left coronary circulation
• Reconfiguring the V-V ECMO circuit to reduce infarction, acute mitral or aortic regurgitation.
recirculation by repositioning the return cannula In this setting, loading of the pulmonary circula-
tion at the expense of the systemic circulation is

LVF > RVF Nil/Minimal LV Ejection with AR

Early/Pre-emptive Established
Pulmonary changes Gradual Early/partial Florid
Monitor Pulsatility Nil/minimal Nil
Pulmonary Hemorrhage No No Yes
Pump Function Good flows Deteriorating Access
Insufficiency
Outcome Good Variable Poor/Fatal
Associated Clinical LV process/RV sparing Late presenters
features Responds well to load Worse/advanced processes
reduction across the LV Frequent with prolonged CA
↑PEEP and ↓MAP presentations
(↓ LVSWI) Multi-organ failure
Coagulopathy
LV=left ventricle; LVF=left ventricular failure; RV=right ventricle; RVF=right ventricular
failure; AR=aortic regurgitation; PEEP=positive end expiratory pressure; MAP=mean
arterial pressure; CA=cardiac arrest; LVSWI=left ventricular stroke work index
Table 14-2. LV distention syndromes.

134
 Special Configurations in Adult ECLS

gradual and medical management of pulmonary consideration for temporary RV mechanical support
edema (high PEEP and low MAP targets) may options are given below:
prove adequate and circuit function is main-
tained. If not detected and managed, however, • V-PA ECLS: This mode of support operates
gradual loading can progress rapidly. in series with the LVAD, allows biventricular
• Minimally or nonpulsatile heart with any aortic unloading and does not bypass the newly im-
regurgitation (Figure 14-2). In this setting, fill- planted LVAD. This is desirable as there is an
ing of the LV is directly driven by circuit flow elevated risk of LVAD thrombosis with V-A
and cardiac and pulmonary circulation loading ECMO support which bypasses the new in-
is rapid and circuit function is impaired. Urgent serted LVAD. Flow settings of the (two) VADs
change of support mode is required. needs to be monitored to prevent pulmonary
edema (indicating excessive RVAD function
Table 14-3 lists the numerous possible mechani- compared to LVAD) or suction alarms on the
cal options that have been advocated for LV decom- LVAD (indicating RVAD function is inadequate
pression and cardiac support. Detailed explanation for LVAD settings).
of these forms of support is beyond the scope of this • V-PA ECMO: This mode of support may be
chapter. In general, the least invasive, cost effective required if native lung function is also impaired
form of support that allows safe patient care and in addition to RV failure. The membrane lung
can be REALISTICALLY commenced in a timely can be inserted or removed from the circuit as
manner is the approach selected. lung function changes.

ECLS Support Following Implantation of a Venopulmonary artery ECLS can be config-

Durable LVAD ured by peripheral venous cannulation (jugular or
femoral) with pulmonary artery return via a grafted
Postoperative RV failure following insertion of conduit8 or via a single dual lumen catheter (Protek
a durable implanted left ventricular assist device Duo®, Liva Nova, London, England).9 The double
(LVAD) is common and particular groups of patients lumen catheter can be inserted percutaneously and
are at higher risk.8 In some cases RV dysfunction removed at the bedside. The two cannula V-PA ap-
becomes too severe for just inotropic support but proach may allow for higher blood flow settings and
may eventually improve and not require the implan- the removal of the return catheter does not require
tation of a second long-term RVAD. Physiological sternotomy.

Figure 14-2. V-A ECMO driven LV distention. As

pulmonary circulation become filled at the expense
of the systemic circulation, ECMO circuit flow
becomes reduced.

135
Chapter 14

Mode of Support (references) Advantages Disadvantages

Thoracotomy with LV Venting on Longer term support Invasive
VA ECMO10 possible Bleeding
Thoracotomy with LV Access for Permits nursing care, Thoracotomy/Operation
temporary LVAD +/- BiVAD11 extubation, ambulation required/Costs
Central cannulation of right High flow No mechanical RV
superior pulmonary vein with VA support (with LVAD
ECMO12 mode alone)
Percutaneous transeptal LA Minimally invasive Short term only
venting with VA ECMO13,14 Easily reversible Limited flow
Maintenance of catheter
position
Specialised procedure
Percutaneous transaortic venting Minimally invasive Short term only Limited
with VA ECMO15 Easily reversible flow
Simple procedure Maintenance of catheter
position
Balloon Atrial Septostomy with Minimally invasive Repair required
VA ECMO13 Permits nursing care, Complex and specialised
extubation, ambulation procedure
Percutaneous PA venting with Minimally invasive Limited flow
VA ECMO16 Easily reversible
Bedside procedure
Impella with VA ECMO17 Simple procedure Leg ischemia risk
High cost
Immobility
LA=left atrium; LV=left ventricle; LVAD=left ventricular assist device;
BiVAD=biventricular assist device
Table 14-3. ECLS support modes options for patients with LV distention.

136
 Special Configurations in Adult ECLS

References 10. Guirgis M, Kumar K, Menkis AH, Freed DH.

Minimally invasive left-heart decompression
1. Werner NL, Coughlin M, Cooley E, et al. during venoarterial extracorporeal membrane
The University of Michigan Experience with oxygenation: an alternative to a percutaneous
Veno-Venoarterial Hybrid Mode of Extra- approach. Interact Cardiovasc Thorac Surg
corporeal Membrane Oxygenation. ASAIO J 2010;10:672-4.
2016;62(5):578-583. 11. Shuhaiber JH, Jenkins D, Berman M, et al.
2. Biscotti M, Lee A, Basner RC, et al. Hybrid con- The Papworth experience with the Levitronix
figurations via percutaneous access for extracor- CentriMag ventricular assist device. J Heart
poreal membrane oxygenation: a single-center Lung Transplant 2008;27:158-64.
experience. ASAIO J 2014;60(6):635-642. 12. Keenan JE, Schechter MA, Bonadonna DK, et
3. Shekar K, Mullany DV, Thomson B, Ziegen- al. Early Experience with a Novel Cannulation
fuss M, Platts DG, Fraser JF. Extracorporeal Strategy for Left Ventricular Decompression
life support devices and strategies for manage- during Nonpostcardiotomy Venoarterial ECMO.
ment of acute cardiorespiratory failure in adult ASAIO J 2016;62:e30-4.
patients: a comprehensive review. Crit Care 13. Eastaugh LJ, Thiagarajan RR, Darst JR, McEl-
2014;18(3):219. hinney DB, Lock JE, Marshall AC. Percu-
4. Schmidt M, Tachon G, Devilliers C, et al. Blood taneous left atrial decompression in patients
oxygenation and decarboxylation determinants supported with extracorporeal membrane oxy-
during venovenous ECMO for respiratory fail- genation for cardiac disease. Pediatr Crit Care
ure in adults. Intensive Care Med 2013;39:838- Med 2015;16:59-65.
846. 14. Swartz MF, Smith F, Byrum CJ, Alfieris GM.
5. Rich PB, Awad SS, Crotti S, Hirschl RB, Bartlett Transseptal catheter decompression of the left
RH, Schreiner RJ. A prospective comparison of ventricle during extracorporeal membrane oxy-
atrio-femoral and femoro-atrial flow in adult ve- genation. Pediatr Cardiol 2012;33:185-7.
novenous extracorporeal life support. J Thorac 15. Hong TH, Byun JH, Lee HM, et al. Initial
Cardiovasc Surg 1998;116:628-32. Experience of Transaortic Catheter Venting
6. Montisci A, Maj G, Zangrillo A, Winterton in Patients with Venoarterial Extracorporeal
D, Pappalardo F. Management of refractory Membrane Oxygenation for Cardiogenic Shock.
hypoxemia during venovenous extracorporeal ASAIO J 2016;62:117-22.
membrane oxygenation for ARDS. ASAIO J 16. Avalli L, Maggioni E, Sangalli F, Favini G,
2015;61:227-36. Formica F, Fumagalli R. Percutaneous left-heart
7. Rupprecht L, Florchinger B, Schopka S, et al. decompression during extracorporeal mem-
Cardiac decompression on extracorporeal life brane oxygenation: an alternative to surgical
support: a review and discussion of the litera- and transeptal venting in adult patients. ASAIO
ture. ASAIO J 2013;59:547-53. J 2011;57:38-40.
8. Haneya A, Philipp A, Puehler T, et al. Temporary 17. Cheng A, Swartz MF, Massey HT. Impella to
percutaneous right ventricular support using a unload the left ventricle during peripheral ex-
centrifugal pump in patients with postoperative tracorporeal membrane oxygenation. ASAIO J
acute refractory right ventricular failure after 2013;59:533-6.
left ventricular assist device implantation. Eur
J Cardiothorac Surg 2012;41:219-23.
9. Schmack B, Weymann A, Popov AF, et al. Con-
current Left Ventricular Assist Device (LVAD)
Implantation and Percutaneous Temporary
RVAD Support via CardiacAssist Protek-Duo
TandemHeart to Preempt Right Heart Failure.
Med Sci Monit Basic Res 2016;22:53-7.

137
15

Extracorporeal Carbon Dioxide Removal (ECCO2R)

Francesco Alessandri, MD, Edoardo Piervincenzi, MD, Francesco Pugliese, MD, Marco V. Ranieri, MD

Introduction ECCO2R Technology

Mechanical ventilation (MV) provides poten- The first clinical use of ECCO2R was performed
tially lifesaving gas exchange in patients with severe with a pumpless arteriovenous bypass that used
acute respiratory failure. However, MV can cause the natural gradient pressure between the femoral
adverse effects, including lung overinflation and artery and femoral vein to drive blood across the
ventilator induced lung injury (VILI), right ven- membrane oxygenator. Another configuration
tricular failure and ventilator associated pneumonia included a bypass circuit specifically designed to
(VAP). These complications have a deep impact on remove CO2 by the pump of a renal replacement
morbidity and mortality. therapy device or of an extracorporeal membrane.9-12
Extracorporeal carbon dioxide removal These “minimally invasive” devices use double
(ECCO2R) is a low flow extracorporeal technique lumen cannulas (usually 14-18 Fr), placed in the
that selectively removes CO2 that has been pro- femoral or internal jugular vein, centrifugal or roller
posed as a treatment option for patients with acute pump (200-1000 ml/min), and lung membranes
respiratory distress syndrome (ARDS).1,2 Gattinoni (Figure 15-1).
and coworkers in the early 80s first reported three
patients with ARDS refractory to the conventional Vascular Access
treatments who were successfully treated with a
CO2 removal circuit and low frequency ventilation The use of double lumen vascular cannulas
(LFPPV-ECCO2R).3 However, hemorrhagic compli- reduces the incidence of vascular complications,
cations due to high heparin doses, rudimentary and however, due to the high risk of clotting, higher
poorly biocompatible circuits were also reported doses of anticoagulation may be needed.13
with ECCO2R.3 The H1N1 pandemic in 2009 led to
a new era for the extracorporeal support techniques,
and the development of systems for extracorporeal
support aroused a growing interest. In particular,
several authors raised the hypothesis that ECCO2R
may allow ultra-protective ventilation strategies
to minimize ventilator induced lung injury (VILI)
in ARDS patients with extremely low pulmonary
compliance.4-8 Moreover, the development of tech-
nology for extracorporeal assist (circuits, pumps,
membranes) reduced the complexity of ECCO2R,
permitting their use in conditions other than ARDS,
such as acute exacerbation of chronic obstructive
pulmonary disease (COPD) and bridge to lung Figure 15-1. Schematic representation of a
transplant. venovenous extracorporeal CO2 removal system.

139
Chapter 15

Pumps Potential Clinical Indications

Pump technology continues to improve: the ECCO2R and ARDS

Hemolung Respiratory Assist System (RAS, Alung
Technologies, Inc., Pittsburgh, PA) is the only EC- Protective MV, with tidal volume (VT) of 6 ml/
CO2R system with a single component in the circuit. kg of predicted body weight (PBW) and inspira-
Both pumping and gas exchange are integrated into tory plateau pressure (Pplat) ≤ 30 cm H2O, reduces
this component and it is the safest, simplest, and mechanical stress to the lungs, decreasing lung
most efficient ECCO2R therapy. Novalung iLA inflammation, and improving survival in patients
Active (Inspiration Healthcare, Craawley, UK) with ARDS.16 However, these ventilator settings
features a centrifugal pump with diagonal flow and may not avoid ventilator induced lung injury (VILI).
a magnetic drive unit for a wide and efficient flow Respiratory acidosis from CO2 retention may limit
range (0-8 L/min); The CARDIOHELP System further reductions of VT. ECCO2R permits reduc-
(Maquet, Rastatt, Germany) is the smallest all-in- tion of VT beyond the threshold of 6 ml/kg of PBW,
one portable heart lung support system, making it keeping PaCO2 in an acceptable range. Neverthe-
ideal for patient transport. less, this “ultra-protective strategy” may decrease
alveolar ventilation, favoring alveolar collapse
Membrane Lungs with further deterioration of oxygenation. Ongoing
trials will clarify the safety and feasibility of MV
The first oxygenators utilized silicone rubber at 4 ml/kg PBW (facilitated by ECCO2R) and will
membranes for gas exchange, followed by poly- compare ECCO2R to enable lower tidal volume
propylene microporous hollow fiber membranes, ventilation vs. standard care (SUPERNOVA, Clini-
and most recently polymethyl pentene diffusion cal Trials gov NCT02282657; REST, Clinical trial
membranes utilizing hollow fiber technology with gov NCT02654327).
a nonmicroporous membrane.14
The membrane lung is divided into two cham- ECCO2R and COPD
bers, separated by a semipermeable membrane. CO2
is exchanged by diffusion across the membrane. In acute hypercapnic respiratory failure due
Blood flows along one side of the membrane, while to COPD exacerbation, noninvasive ventilation
sweep gas flows along the other side. Since CO2 (NIV) reduces the rate of endotracheal intubation
diffuses across the membrane faster than O2, CO2 and mortality and is considered the standard of
clearance is primarily determined by the gradient care.17 Despite continuous improvement in NIV
across the membrane. support, 15-26 % of patients with acute exacerba-
The baseline arterial CO2 content depends on tions of COPD may fail NIV requiring transition
tissue production and lung ventilation efficacy; the to invasive mechanical ventilation (IMV).18-20 The
higher the blood flow (usually from 750 to 1000 transition from NIV to IMV is associated with an
ml/min), the greater the CO2 removal. Membrane increased risk of death.17 ECCO2R can be used in
lung size and transit time increase gas exchange. three different clinical settings of COPD patients
Membrane surface ≥0.8 m2 more effectively corrects depending on timing:
severe hypercarbic acidosis (HCA).15 Sweep gas
flow removes CO2 from the gas chamber, thereby CO2 Removal Determinants
maintaining the CO2 gradient, allowing for contin- Baseline arterial CO2 content
ued diffusion. The determinants of extracorporeal Blood flow
CO2 removal are listed in Table 15-1. Area of membrane lung
Time of transit
Sweep gas flow

Table 15-1. Determinants of carbon dioxide

removal.

140
 Extracorporeal Carbon Dioxide Removal (ECCO2R)

• Early initiation to prevent NIV failure in pa- representing a valid bridge to lung transplantation,
tients at risk although for a limited period of time.29,30
• Initiation of ECCO2R to avoid IMV when NIV
fails Complications
• Postintubation initiation of ECCO2R in weaning
COPD patients from IMV. ECCO2R is associated with common complica-
tions of other extracorporeal devices. These can be
The paucity of evidence examining ECCO2R de- divided in device or patient related complications
creasing the rate of NIV failure and ability to wean (Table 15-2). Hemorrhage is the most frequent
hypercapnic patients from invasive mechanical complication during ECCO2R. In the ECLAIR study,
ventilation (IMV) precludes valid conclusions about 36% of ECCO2R patients had major hemorrhage
the application of ECCO2R in COPD patients.21-25 episodes, while previous studies had lower rates of
Ongoing randomized clinical studies of the effec- life threatening bleeding.21,22,31 The ECLAIR data
tiveness of ECCO2R in acute hypercapnic respira- were probably influenced by the high blood flow
tory failure patients are needed (ClinicalTrials.gov (1.3 L/min) which can cause platelet dysfunction.
NCT02107222; NCT02259335; NCT02086084; Thus, patient complications seem to be related to
NCT02586948; NCT02260583). high blood flow; whereas device side effects occur
more frequently during low flow CO2 removal.24
ECCO2R as a Bridge to Lung Transplantation Data on circuit clotting are available only in three
studies on ECCO2R, with rates from 19-30%.21,22,32
ECCO2R has successfully bridged patients Other coagulation disorders associated with
with chronic pulmonary disease waiting for lung ECCO2R include a deficiency in factor XIII-activity
transplantation (LTX), although this use remains value or the acquired von Willebrand syndrome.33,34
poorly studied.26 In 2006, Fischer et al. published a
series of 12 LTX candidate patients who developed Device Related
refractory hypercapnic respiratory failure treated Complications
with arteriovenous bypass. Ten patients received Difficulties with cannulation
transplants, and 8 survived one year later.27 Schel- Membrane clotting
longowski et al. reported the largest study with Pump failure
Heat exchanger malfunction
20 patients, including 10 treated with a VA-ILA, Cannula
a pumpless circuit, and 10 with a VV-ILA active displacement/thrombosis
circuit: 15 patients were intubated initially while Tubing rupture
5 had NIV.28 Blood flow ranged between 1-2 L/ Air in the circuit
min and the sweep gas was titrated to normalize
Patient Related
CO2. Hypercapnia and acidosis were corrected in Complications
all patients within the first 12 hrs of CO2 removal Hemorrhage due to systemic
therapy. Three patients were extubated and remained anticoagulation
on ILA until LTX, one patient was weaned from ILA Cannulation site bleeding
Hemolysis (Hct reduction
and remained intubated until LTX, and one patient
not related to hemorrhage or
was extubated and weaned from ILA before LTX. other causes of blood loss
Four patients were switched to extracorporeal mem- Hemodynamic instability
brane oxygenation (VV-ECMO=1, VA-ECMO=3) Catheter infection
after 2 days of ILA support (range 1-5 days) due Heparin induced
thrombocytopenia (HIT)
to progressive hypoxia and/or heart failure. Of the
Embolism
20 early patients, 19 received LTX and 15 were Venous severe stasis
discharged from the hospital. In several case series, Ischemia (AV circuits)
CO2 removal devices have been simple, efficient
methods to support rapidly deteriorating patients, Table 15-2. Complications

141
Chapter 15

Despite using small cannulas and low flow rates, blood promotes bicarbonate exchange with chloride,
ECCO2R has approximately the same hemorrhage increases PCO2, and enhances membrane lung CO2
incidence as ECMO. Improving CO2 removal ef- removal. The circuit is composed of a hemofilter,
ficacy has led to minimizing device complications. a membrane lung, and an electrodialysis unit.39-42
The use of sodium citrate seems to be an attractive Carbonic anhydrase (CA) catalyze the hydration
option to avoid heparin-induced bleeding due to its of CO2. A hollow fiber membrane grafted with CA
regional effect.35 facilitates the conversion of bicarbonate to CO2
accelerating CO2 removal efficacy.43 More studies
Future Perspective and Advances are needed to improve CA immobilization, thermal
stability, and hemocompatibility.44
Miniaturized and Intravenous Circuits

New miniaturized and implantable systems have

been designed to minimize the invasiveness and to
optimize safety of these procedures. Recently 12
patients with acute hypercapnia were treated with a
compact extracorporeal gas exchange system driven
by a small centrifugal pump, resembling a miniatur-
ized VV-ECMO system, achieving blood flow rates
from 0.5 to 8 L/min, depending on cannula size and
the membrane used. Blood flow rates ranged from
1.2-1.4 L/min, and venous suction pressure was kept
more positive than -80 mmHg to avoid hemolysis.
All patients had significant CO2 reduction allow-
ing less aggressive MV.36 Miniaturized centrifugal
pumps have several advantages: easy cannulation,
limited priming volumes, low extracorporeal blood
volume, and sustainable patient mobility. These
advantages suggest rationale for more widespread
use of these systems.
In the last few years, implantable respiratory
care systems that provide partial respiratory sup-
port with CO2 removal of 36-41 ml/min through a
25 Fr catheter have been developed. The Impeller
Percutaneous Respiratory Assist Catheter (IPRAC)
efficiently removes CO2 using a percutaneous ac-
cess in the jugular or femoral vein. IPRAC utilizes
an array of impellers rotating within a hollow fiber
membrane bundle to enhance gas exchange directly
in the venous system.37,38

Improving CO2 Removal

Respiratory electrodialysis (R-ED) is a promis-

ing experimental technique that enhances the CO2
extraction capacity by artificial lung membrane.
R-ED selectively modulates pH and electrolyte con-
centration. The application of an electric current to

142
 Extracorporeal Carbon Dioxide Removal (ECCO2R)

References Safety and efficacy of combined extracorporeal

CO2 removal and renal replacement therapy
1. Kolobow T, Gattinoni L, Tomlinson TA, Pierce in patients with acute respiratory distress syn-
JE. Control of breathing using an extracorpo- drome and acute kidney injury: the pulmonary
real membrane lung. Anesthesiology. 1977; and renal support in acute respiratory distress
46(2):138-141. syndrome study. Crit Care Med 2015; 43:2570-
2. Gattinoni L, Kolobow T, Tomlinson T, et al. 2581.
Low-frequency positive pressure ventilation 11. Hermann A, Riss K, Schellongowski P, et al. A
with extracorporeal carbon dioxide removal novel pump driven veno-venous gas exchange
(LFPPV-ECCO2R): an experimental study. system during extracorporeal CO2-removal.
Anesth Analg. 1978; 57(4):470-477. Intensive Care Med 2015; 41:1773-1780.
3. Gattinoni L, Agostoni A, Pesenti A, et al. Treat- 12. Schmidt M, Tachon G, Devilliers C, et al. Blood
ment of acute respiratory failure with low- oxygenation and decarboxylation determinants
frequency positive-pressure ventilation and during venovenous ECMO for respiratory
extracorporeal removal of CO2. Lancet 1980; failure in adults. Intensive Care Med 2013;
2 (8189): 292-294. 39:838-846.
4. Terragni PP, Del Sorbo L, Mascia L, et al. Tidal 13. MacLaren G, Combes A, Bartlett RH. Contem-
volume lower than 6 ml/kg enhances lung pro- porary extracorporeal membrane oxygenation
tection: role of extracorporeal carbon dioxide for adult respiratory failure: life support in the
removal. Anesthesiology. 2009;111(4):826-835. new era. Intensive Care Med 2012;38:210-220
5. Terragni PP, Rosboch G, Tealdi A, et al. Tidal 14. Robak O, Lakatos PK, Bojic A, et al. Influence
hyperinflation during low tidal volume ventila- of different oxygenator types on changing fre-
tion in acute respiratory distress syndrome Am quency, infection incidence, and mortality in
J Respir Crit Care Med. 2007;175(2):160-166. ARDS patients on veno-venous ECMO. Int J
6. Bein T, Weber-Carstens S, Goldmann A, et Artif Organs. 2014;37(11):839-846.
al. Lower tidal volume strategy (≈3 ml/kg) 15. Karagiannidis C, Strassmann S, Brodie D, et
combined with extracorporeal CO2 removal al. Impact of membrane lung surface area and
versus ‘conventional’ protective ventilation (6 blood flow on extracorporeal CO2 removal dur-
ml/kg) in severe ARDS: the prospective ran- ing severe respiratory acidosis. Intensive Care
domized Xtravent-study.Intensive Care Med. Med Exp. 2017;5(1):34.
2013;39(5):847-856. 16. Acute Respiratory Distress Syndrome Network,
7. Moss CE, Galtrey EJ, Camporota L, et al. A Brower RG, Matthay MA, Morris A, Schoen-
Retrospective Observational Case Series of feld D, Thompson BT, Wheeler A. Ventilation
Low-Flow Venovenous Extracorporeal Carbon with lower tidal volumes as compared with
Dioxide Removal Use in Patients with Respi- traditional tidal volumes for acute lung injury
ratory Failure. ASAIO J. 2016;62(4):458-462 and the acute respiratory distress syndrome. N
8. Fitzgerald M, Millar J, Blackwood B, et al. Extra- Engl J Med 2000;342(18):1301-1308.
corporeal carbon dioxide removal for patients 17. Chandra D, Stamm JA, Taylor B, et al. Outcomes
with acute respiratory failure secondary to the of noninvasive ventilation for acute exacerba-
acute respiratory distress syndrome: a system- tions of chronic obstructive pulmonary disease
atic review. Crit Care. 2014 May 15;18(3):222 in the United States, 1998-2008.Am J Respir
9. Forster C, Schriewer J, John S, Eckardt KU, Wil- Crit Care Med. 2012;185(2):152-159.
lam C. Low-flow CO(2) removal integrated into 18. Phua J, Kong K, Lee KH, Shen L, Lim TK..
a renal replacement circuit can reduce acidosis Noninvasive ventilation in hyper- capnic acute
and decrease vasopressor requirements. Crit respiratory failure due to chronic obstructive
Care 2013; 17:R154 pulmonary disease vs. other conditions: ef-
10. Allardet-Servent J, Castanier M, Signouret T, fectiveness and predictors of failure. Intensive
Soundaravelou R, Lepidi A, Seghboyan JM. Care Med. 2005; 31:533-539.

143
Chapter 15

19. Confalonieri M, Garuti G, Cattaruzza MS, et al. 29. Ricci D, Boffini M, Del Sorbo L, et al. The use of
A chart of failure risk for noninvasive ventila- CO2 removal devices in patients awaiting lung
tion in patients with COPD exacerbation. Eur transplantation: an initial experience. Transplant
Respir J. 2005; 25:348-355. Proc. 2010;42(4):1255-1258.
20. Quinnell TG, Pilsworth S, Shneerson JM, Smith 30. Ruberto F, Bergantino B, Testa MC, et al. Low-
IE. Prolonged invasive ventilation following flow veno-venous extracorporeal CO2 removal:
acute ventilatory failure in COPD: weaning first clinical experience in lung transplant recipi-
results, survival, and the role of noninvasive ents. Int J Artif Organs. 2014;37(12):911-917.
ventilation. Chest. 2006; 129:133-139. 31. Kluge S, Braune SA, Engel M, et al. Avoiding
21. Del Sorbo L, Fan E, Nava S, Ranieri VM. EC- invasive mechanical ventilation by extracorpo-
CO2R in COPD exacerbation only for the right real carbon dioxide removal in patients failing
patients and with the right strategy. Intensive noninvasive ventilation. Intensive Care Med.
Care Med 2016;42(11):1830–1831. 2012; 38:1632-1639.
22. Braune S, Sieweke A, Brettner F, et al. The 32. Morris AH, Wallace CJ, Menlove RL, et al.
feasibility and safety of extracorporeal carbon Randomized clinical trial of pressure-controlled
dioxide removal to avoid intubation in patients inverse ratio ventilation and extracorporeal CO2
with COPD unresponsive to noninvasive venti- removal for adult respiratory distress syndrome.
lation for acute hypercapnic respiratory failure Am J Respir Crit Care Med 1994;149:295-305.
(ECLAIR study): multicentre case-control study. 33. Kalbhenn J, Neuffer N, Zieger B, Schmutz A. Is
Intensive Care Med 2016;42:1437-1444. Extracorporeal CO2 Removal Really “Safe” and
23. Gattinoni L, Agostoni A, Pesenti A, et al. “Less” Invasive? Observation of Blood Injury
Treatment of acute respiratory failure with and Coagulation Impairment during ECCO2R.
low-frequency positive-pressure ventilation ASAIO J. 2017; 63(5):666-671.
and extracorporeal removal of CO2. Lancet 34. Vaquer S, de Haro C, Peruga P, Oliva JC, Arti-
1980;2:292-294. gas A. Systematic review and meta-analysis of
24. Pisani L, Fasano L, Corcione N, et al. Effects complications and mortality of veno-venous
of extracorporeal CO2 removal on inspiratory extracorporeal membrane oxygenation for
effort and respiratory pattern in patients that refractory acute respiratory distress syndrome.
fail weaning from mechanical ventilation. Am J Ann Intensive Care. 2017; 7(1):51.
Respir Crit Care Med 2015; 192(11): 1392-1394. 35. Sharma AS, Weerwind PW, Bekers O, Wouters
25. Sklar MC, Beloncle F, Katsios CM, Brochard EM, Maessen JG. Carbon dioxide dialysis in
L, Friedrich JO. Extracorporeal carbon dioxide a swine model utilizing systemic and regional
removal in patients with chronic obstructive anticoagulation. Intensive Care Med Exp. 2016;
pulmonary disease: a systematic review. Inten- 4(1):2.
sive Care Med. 2015;41(10):1752-1762. 36. Hermann A, Staudinger T, Bojic A, et al. First ex-
26. Collaud S, Benden C, Ganter C, et al. Extracor- perience with a new miniaturized pump-driven
poreal Life Support as Bridge to Lung Retrans- venovenous extracorporeal CO2 removal sys-
plantation: A Multicenter Pooled Data Analysis. tem (iLA Activve): a retrospective data analysis.
Ann Thorac Surg. 2016; 102(5):1680-1686. ASAIO J. 2014; 60(3):342-347.
27. Fischer S, Simon AR, Welte T, et al. Bridge to 37. Jeffries RG, Frankowski BJ, Burgreen GW,
lung transplantation with the novel pumpless Federspiel WJ. Effect of impeller design and
interventional lung assist device NovaLung. J spacing on gas exchange in a percutaneous
Thorac Cardiovasc Surg. 2006; 131(3):719-723. respiratory assist catheter. Artif Organs. 2014;
28. Schellongowski P, Riss K, Staudinger T, et al. 38(12):1007-1017.
Extracorporeal CO2 removal as bridge to lung 38. Mihelc KM, Frankowski BJ, Lieber SC, Moore
transplantation in life-threatening hypercapnia. ND, Hattler BG, Federspiel WJ. Evaluation of a
Transpl Int. 2015; 28(3):297-304. respiratory assist catheter that uses an impeller

144
 Extracorporeal Carbon Dioxide Removal (ECCO2R)

within a hollow fiber membrane bundle. ASAIO

J. 2009;55(6):569-574.
39. Zanella A, Castagna L, Salerno D, et al. Respi-
ratory electrodialysis. A novel, highly efficient
extracorporeal CO2 removal technique. Am J
Respir Crit Care Med. 2015;192:719-726.
40. Zanella A, Mangili P, Redaelli S, et al. Regional
blood acidification enhances extracorporeal car-
bon dioxide removal: a 48-hour animal study. J
Am Soc Anesthesiol. 2014;120:416-424.
41. Zanella A, Mangili P, Giani M, et al. Extracorpo-
real carbon dioxide removal through ventilation
of acidified dialysate: an experimental study. J
Heart Lung Transplant. 2014;33(5):536-541.
42. Zanella A, Castagna L, Abd El Aziz El Sayed
Deab S, et al. Extracorporeal CO2 removal by
respiratory electrodialysis: an in vitro study.
ASAIO J. 2016;62:143-149.
43. Manap HH, Wahab AKA. Extracorporeal car-
bon dioxide removal (ECCO2R) in respiratory
deficiency and current investigations on its
improvement: a review J Artif Organs 2017;
20:8-17.
44. Yong JKJ, Stevens GW, Caruso F, Kentish SE.
The use of carbonic anhydrase to accelerate car-
bon dioxide capture processes. J Chem Technol
Biotechnol. 2015;90:3-10.

145
16

ECMO Equipment and Circuit Design

W. Cory Ellis, CCP, Katie Butler, MSN, RN

Introduction ECMO Program Structure and Goals

Extracorporeal membrane oxygenation (ECMO) The specific goals and structure of an ECMO
provides a unique biomechanical interaction be- program should be established prior to purchasing
tween a critically ill patient and mechanical equip- the console, especially for new programs. A hub
ment. Careful consideration should be given to the and spoke program structure where patients are
type of ECMO console, disposable equipment, and cannulated at tertiary centers and then transported
monitoring devices used based on the goals and via ground or air to a designated ECMO center may
structure of the program. The purpose of this chapter consider a console that uses identical components
is to identify important considerations when choos- to facilitate standardization and simplify supply
ing an ECMO console and monitoring equipment chain management. The location where ECMO will
and when designing a standardized ECMO circuit be monitored and the personnel who will provide
for your center. oversight of the circuit may impact the choice of
an ECMO console. For example, a staffing model
ECMO Pump Console where a single ECMO Specialist is responsible for
monitoring multiple ECMO circuits may benefit
ECMO consoles can vary greatly in function- from an ECMO console with expanded safety fea-
ality, capability, size, and shape. Some consoles tures regardless of console size; whereas, a program
simply control the speed of the roller or centrifugal that seeks to transport or rehabilitate patients may
pump head while others have monitoring and inter- choose a lighter console with a smaller footprint.
ventional capabilities. Consoles come in variable
configurations with different footprints, weights, Pump Style
and sizes. Most consoles have a backup power
supply, but the amount of time available on backup Decisions about the style of pump, roller or
power varies greatly. While all ECMO consoles can centrifugal, may dictate available choices for an
mechanically support a patient, it is important for an ECMO console, especially if the program desires
ECMO program to evaluate the goals and specific integrated monitoring features. Both centrifugal and
patient population of their program prior to purchas- roller pumps can provide ECMO support; however,
ing a console. Some important considerations when each pump style has unique considerations regard-
purchasing an ECMO console include the program ing management and safety.
structure and goals, patient population, pump type
(centrifugal or roller), console size and footprint, Console Size
intra and interhospital transportability, integrated
monitoring, servo regulation, backup power require- The console size may be an important factor
ments and, of course, cost. depending on the goals of your ECMO program.
Programs that seek to develop robust ECMO trans-
port or rehabilitation programs may need smaller,

147
Chapter 16

lighter consoles. Hospitals are commonly moving The quantity and type of transport a center
to private patient rooms making the footprint of the forecasts should also influence the choice of ECMO
ECMO console a consideration, especially in the console. The console should be measured as it will
ICU setting where patient rooms often have mul- be fully equipped during ECMO support, including
tiple mechanical devices. Console size contributes blenders, ancillary monitors, IV and syringe pumps,
to transportability for both intra and inter hospital heat exchange devices, and accessory gas cylinders.
transport. Simulation of intra and interhospital transport are
an important part of determining which ECMO
Integrated Features console to use.

Integrated monitoring, servo regulation and Cost

backup power supply are available on most new
ECMO consoles; however, they vary widely in the Baseline cost of a console should be evaluated
type and capabilities of the integrated features be- critically, but a thorough evaluation of value should
tween manufacturers. Some consoles provide only include the cost of disposables, maintenance, and
integrated pressure monitoring with no interven- ancillary safety and monitoring equipment required
tional option. Other consoles can monitor multiple to support an ECMO patient at your center.
circuit pressures, in line blood parameters, have in-
tegrated safety alarms and automated interventions. ECMO Circuit Design

Transportability The ECMO circuit has evolved significantly

over with the introduction of polymethyl pentane
ECMO transport, both intra and interhospital, (PMP) oxygenators and the widespread use of
is becoming more common. ECMO patients are centrifugal pumps. Some manufacturers have de-
commonly transported within a center for both diag- veloped standardized ECLS circuits that offer off-
nostics and interventional procedures. The amount the-shelf convenience, while most provide custom
of intrahospital transport, the layout of common design options for centers to meet their needs. When
destinations including the operating room, cardiac choosing or designing an ECMO circuit, centers
catheterization lab, CT scanner or, in some centers, should consider the following principles of circuit
the rehabilitation center as well as room dimensions design: safety, suitability for long term support and
and footprint, elevator size, and door size impact adaptability (Table 16-1)
considerations of transportability.
Interhospital transport is increasingly common Safety
in the international ECMO community. In countries
where care is regionalized patients are transported ECMO circuits should be designed for patient
to a center that specializes in managing ECMO safety. Generally, safely designed circuits are simple,
patients. In the United States, many healthcare sys-
tems employ a “hub and spoke” system requiring Principles of ECMO Circuit Design
• Safe (simple)
ECMO transport capabilities. Some programs even
• Suitable for long term support
have established interhospital transport programs - Biocompatible
which provide transport of ECMO patients as a taxi - Minimal blood trauma
service between two centers for escalation of care. - Small prime volume (minimize transfusion)
Transportation occurs in ambulances, airplanes, and - Small surface area (minimize blood surface
helicopters. Each of these transport modalities has interactions)
unique regulations, space, layout, device secure- - Portable (Intra/inter hospital transport)
- High O2 and CO2 exchange
ment, power, and weight considerations that must • Easily adaptable to a wide variety of patients
be considered prior to purchasing an ECMO console.
Table 16-1. Principles of ECMO circuit design.

148
 ECMO Equipment and Circuit Design

have a limited number of connectors and access form imaging or interventions on ECMO patients
points, and easily incorporate the safety devices on may prefer a circuit designed with extra length in
your pump console. the arterial and venous lines. Access points should
The primary purpose for ECMO support is to be considered for renal replacement therapy or
provide tissue oxygen delivery. A simple circuit apheresis. Additionally, ambulatory ECMO patients
with a venous return catheter, a blood pump, an have unique circuit requirements that should be
oxygenator, and an arterial return catheter (or return considered during the design process.
to the arterial side of a dual lumen VV-ECMO cath- In addition to foreseeable problems, ECMO
eter) should be adequate to provide tissue oxygen patients can present with unique and challenging
delivery. All other components, while potentially situations that require critical thinking, novel ap-
beneficial to the way a center provides ECMO sup- proaches, and adaptability. In these situations, it
port, are ancillary to the main purpose of ECMO is important to have both adaptable personnel and
support. Ancillary components add complexity to circuits.
the circuit, so the ECMO team should determine
whether the risk of added complexity is balanced Integrated Circuits
with the benefit of added functionality or monitoring.
Integrated circuits, such as the Maquet Car-
Suitable for Long-term Support diohelp, are becoming more popular. Standardized
circuits with integrated monitoring and safety de-
ECMO is used to support patients for longer vices simplify the circuit design process as well as
periods of time than in the past, for up to months. the supply chain management. For many centers in
Even brief courses (e.g. 1 day) of ECMO are still which patient demographics are similar, integrated
a long time for support by a mechanical pump and circuits make sense, despite their lack of adaptability.
oxygenator. Factors that influence the suitability Such circuits are increasingly common in areas that
for long-term support are described in Table 16-2. use a regionalized care model, or health systems that
employ a “hub and spoke” ECMO referral model.
Adaptability
Circuit Shunts
ECMO circuits should be adaptable based on the
needs of patient management at your center. Pedi- Circuit shunts divert blood flow across a pres-
atric centers manage patients who may range from sure gradient. Many ECMO circuits are designed
neonatal and adult sized patients; hence, ECMO with one or more shunts. A common shunt in pedi-
circuits should be designed to adapt easily between atrics and for centers using roller pump technology
those patients. ECMO centers that frequently per- is an arterial to venous bridge. The bridge can cre-

Design Factor Considerations

Biomechanical interactions, sheer stresses, high peak negative
pressures and heat can produce hemolysis and thrombus formation.
Minimal blood trauma
Hemolysis and thrombosis can both diminish the longevity of the
circuit.
Maximum
Surface coatings may increase the biocompatibility of the circuit.
biocompatibility
Oxygenators should transfer gas efficiently and have a buffer to
Maximum oxygen and
allow for adequate gas exchange with deteriorating oxygenator
carbon dioxide transfer
efficiency. Oxygenators should exceed patient oxygen demands.
Circuit prime volumes should be minimized based on programmatic
Small prime volume needs to limit blood and blood component transfusions, and
maximize their efficiency.
Table 16-2. Features of the ECMO circuit that support long-term application.

149
Chapter 16

ate increased flow through the oxygenator or allow components varies by manufacturer, but all surface
for circulation through the oxygenator during an coatings are designed to provide a more biocompat-
emergency or weaning trial, but can also diminish ible interface for the blood within the circuit. The
ECMO blood flow to the patient if not adequately evidence for the use of coated surfaces remains
monitored and compensated. Another common controversial. There is not solid evidence to support
shunt moves blood from the arterial to the venous the superiority of coated ECMO surfaces; however,
side through a manifold of stopcocks. A manifold they have proven to be safe and may provide in-
shunt can provide access for arterialized blood creased biocompatibility. Surface coatings should be
sampling, volume infusion, hemofiltration, and considered where they are available when designing
CRRT. When designing a circuit, attention should be an ECMO circuit.
given to the size and placement of circuit shunts. If
a significantly sized arterial to venous shunt is used,
an ancillary flow probe should be used to ensure
adequate patient blood flow from the circuit. When
separating from ECMO for any reason, the arterial
and venous lines should be clamped between the
patient and any shunt connection to ensure that no
retrograde flow occurs through the ECMO circuit.

Access Points

Access within the ECMO circuit is often nec-

essary for pressure monitoring, volume infusion,
blood sampling, and the use of ancillary components
including a hemoconcentrator or CRRT. These ac-
cess points, while necessary, come with the risk of
infection and either air entrainment or significant
volume loss depending on the placement of the ac-
cess points in the circuit. When designing a circuit,
access points should be limited by necessity with
careful consideration of placement for safety.

Compliance Chamber

Compliance chambers provide a buffer against

the deleterious effects of high peak and prolonged
negative pressures. Compliance chambers are es-
sential when using roller pump technology. When
using centrifugal pump technology, a compliance
chamber dampens the peak negative pressure, but
may contribute to increased thrombus formation in
the circuit.

Surface Coatings

Circuits designed with surface coatings are

commonly employed for ECMO. The mechanism
of interaction between the blood and the ECMO

150
17

ECMO Circuit Components

Leen Vercaemst, RN, ECCP, Paul Kratz, CCP, CPC

Objectives ECMO circuit requires solid background knowledge

of circuit components and flow physiology. ECMO
Following the completion of this chapter, the circuits may be designed by the manufacturer;
participant should be able to: however, most extracorporeal life support (ECLS)
Compose an extracorporeal life support (ECLS) programs customize their circuits to meet the needs
circuit, while understanding and considering: of their patient population and specific institutional
requirements. Hence, this chapter reviews ECMO
• Safety & simplicity
• Blood flow physiology and biocompatibility
• The size of the patient and required ECLS
support
• The available hardware and circuit components
• Different cannula designs and how to choose
the best cannula for each type of ECMO support
• Differences in design, capacities and manage-
ment of different blood pumps
• Working principles of the oxygenator and how
to monitor functionality
• The role of inlet pressure control devices such
as automatic servo regulation or the Better-
Bladder™ Circulatory Technology Inc., Oyster
Bay, NY

Introduction

An extracorporeal membrane oxygenation

(ECMO) circuit provides mechanical cardiac and/
or respiratory support by draining blood from the Figure 17-1. The rudimentary ECMO circuit de-
sign employs outflow (1) and inflow (7) cannulae,
patient and pumping artificially ventilated and an ECMO pump (3) (roller or centrifugal), and
oxygenated blood back into the patient. The main oxygenator (5) with integrated heat exchanger, and
components of the ECMO circuit include drainage polyvinylchloride tubing (2, 4, 6) of a specific size
to connect all of the components together. In ad-
cannula, a blood pump, an artificial lung, and a dition, gas exchange is accomplished by adjusting
return cannula. These components are connected to a precision gas blending device and flowmeter to
each other, most commonly, with polyvinyl chloride provide sufficient gas flow and FiO2 to adequately
meet targets for PaO2 and PaCO2 in the patient’s
(PVC) tubing. All components can differ in size arterial blood. P1 represents the negative pressure
depending on the degree of support and patient size zone with a centrifugal pump circuit. P2 and P3 rep-
(Figure 17-1). Constructing a seemingly simple resent the monitoring of inlet and outlet pressures.

151
Chapter 17

componentry needed for optimal circuit composi- need larger diameters relative to single lumen, dual
tion and management. site cannulation.

Considerations for Cannula Selection Double Lumen Cannula

The ECMO cannulae provide the intravascular • Classical dual lumen cannula (Origen Biomedi-
connections to the patient and are the basis for cal, Maquet Getinge Group); drainage at the site
sufficient ECMO support. Appropriate cannulae of superior vena cava (SVC) and right atrium
selection must consider ECMO mode, the level (RA), and return at the level of the RA.
of support (full or partial), patient size, placement • Bicaval design (Avalon Elite®) (Figure 17-2);
site (central or peripheral), vessel size (if available), drainage from the inferior vena cava (IVC)
and underlying anatomy and pathologies which and SVC and return to the RA; with cannula
might influence the available cannulation options. positioned in IVC which makes this cannula
Blood flow through the cannulae is governed by more difficult to insert to avoid vascular rupture
Poiseuille’s Law (Equation 1). (guidewire) or malpositioning.
• Bicaval dual lumen cannulae are normally
Equation 1. ∆P = 8µLQ/πR4 inserted via the right internal jugular vein
(RIJ), with fluoroscopy or echocardiographic
Where ∆P is the pressure drop across a cannula, guidance.
µ is the dynamic viscosity, L is the length, Q is the • Most of these cannulae are available for pedi-
flow rate and R is the radius of the cannula. Thus, atric and adult support.
the length and particularly the radius (r4) have a • The TandemLife Protek Duo cannula is a dual
tremendous influence on the resistance and achiev- lumen cannula designed for either a RA-pul-
able blood flow through a cannula. Each ECMO
program should have clearly developed guidelines
for appropriate cannula selection. Cannula sizes gen-
erally refer to the outer diameter (O.D.) and can be
expressed in French size or mm, where 1mm = 3 Fr.

Flow Pressure Charts

Consider pressure gradients when selecting a

cannula for a desired flow rate. A general guideline
to avoid hemolysis is selecting a pressure drop
<100 mmHg for the arterial return cannula, and <50
mmHg for the venous drainage cannula. The flow
pressure charts available for all ECMO cannulas
are developed from in-vitro testing with water, and
since viscosity directly increases pressure changes Figure 17-2. The Avalon Elite® double lumen
(see Equation 1), the interpretation of these charts cannula for VV ECMO support is inserted to the
level of the IVC to provide blood drainage of both
must consider the greater viscosity of blood. the SVC and IVC and blood return directed across
the tricuspid valve. The appropriate positioning of
Design the Avalon Elite® cannula is critical for ensuring
effective VV ECMO support. Illustration Courtesy
of Maquet Getinge Group, Rastatt, Germany.
Manufacturers offer different cannula designs
depending on the site of insertion and the level of
required support. Dual lumen cannulae, used for
VV-ECMO, require only a single insertion site yet

152
 ECMO Circuit Components

monary artery (PA) positioning or a RA-IVC native cardiac output is relatively low in relation to
positioning. The RA-PA position essentially ECMO flow.
eliminates recirculation with VV ECMO. Harlequin Syndrome occurs with femoral VA-
ECMO, when the lower body is perfused with oxy-
Single Lumen Cannula genated pump blood while the upper body receives
deoxygenated blood from the native lung after left
• Length: The length of the cannula depends upon ventricle (LV) function has recovered. Relocation
the insertion site, and the location of blood of the arterial cannula to the subclavian artery or
drainage or return. conversion to venoarterial--venous (VAV) cannula-
• Side Holes: Multiple-stage drainage can be tion are remedies for this complication.4
beneficial for net drainage capacity but might
cause preferential flow via the holes more distal Circuit Tubing and Connectors
to the central inlet, limiting central drainage and
risking clot formation of the central drainage The ECMO circuit tubing connects the circuit
hole. The return cannulae normally have holes components together in series (drainage cannula,
only in a short portion near the tip. The mul- pump, oxygenator, return cannula). The tubing is
tistage drainage and reinfusion cannulae may typically composed of PVC that incorporates a plas-
be selected in the setting of the femoro-jugular ticizer, most commonly di-2-ethylhexyl phthalate
approach for VV-ECMO, but a different type of (DEHP) which provides resilience and flexibility
venous cannula must be selected with femoro- in the tubing.5,6 ECMO circuitry also commonly
femoral cannulation, due to the generation of a incorporates a biocompatible surface coating, that
high recirculation fraction through side holes. may be applied to the entirety of the ECMO circuit
components. This biocompatible surface coating
Potential complications associated with inser- reduces the systemic inflammatory response and
tion of cannulae are many. Femoral arterial cannu- platelet activation that would otherwise occur upon
lation carries the risk of distal limb ischemia. The contact of blood with the foreign surfaces of the
insertion of a distal perfusion cannula and sidebar circuit.7,8 Polycarbonate connectors provide con-
grafting techniques, in combination with the use of nections between lengths of circuit tubing or circuit
near-infrared spectroscopy (NIRS) probes on the components.5,6 In addition, “wye” connectors may
cannulated limb, are recommended for the preven- also be utilized in the ECMO circuit design and
tion of distal limb ischemia.1,2 Currently, a femoral should provide angled flow paths which diverge
bidirectional flow cannula is in development, by
LivaNova PLC.It provides blood flow to the body
and the ipsilateral limb, eliminating the need for a
distal perfusion cannula (see Figure-17-3).
Compartment syndrome has been reported with
femoral venous cannulation using large cannulae.
The cannulae compromise venous return, leading to
compartment syndrome, and ultimately limb isch-
emia.3 A distal drainage cannula may be employed
to provide venous drainage to the compromised limb.
Recirculation with VV-ECMO (dual site or sin-
gle site), occurs when the drainage and return can- Figure 17-3. A bidirectional cannula under develop-
nula or ports are in close proximity (see Figure 17-2). ment by LivaNova. This femoral cannula employs
a bidirectional port (upper image) which provides
The returned oxygenated blood is drawn back into blood flow to the vasculature of the distal limb
the ECMO circuit without circulating to the body. (lower image), and eliminates the need for a distal
Recirculation may increase when cannulae are in- limb perfusion cannula. Illustration courtesy of
correctly positioned (dual lumen cannula) or when LivaNova PLC, London, UK.

153
Chapter 17

gently.5,6 All polycarbonate connectors should at its center of rotation, and an area of high pres-
provide low profile connections and avoid abrupt sure (positive) at its periphery through centrifugal
changes in the direction of the flow path in order to forces, which in turn imparts the energy required
reduce turbulence to the greatest extent possible.5,6 to generate blood flow with the centrifugal pump.
A key distinction between the roller pump and the
Blood Pumps-Rationale for Choosing a Blood centrifugal pump is that the former is an occlusive
Pump for ECLS pump, where generated flow is insensitive to pre-
load and afterload, while the latter is nonocclusive
Type of Pump: Roller or Centrifugal Pump? and flow is preload and afterload sensitive. The
drawbacks of roller pumps include potential blood
The choice of a blood pump type depends on a cell damage by the sequential roller occlusion of
variety of factors including local expertise, desired raceway tubing, tubing rupture and tubing spallation
pump and hardware features, local economics and or embolization via leached tubing fragments. In
availability. The roller pump is a positive displace- addition, roller pump configurations are much less
ment pump that generates blood flow by sequentially transportable than centrifugal pumps.
compressing a length of tubing within the roller
raceway. In contrast, the centrifugal pump gener- Design
ates blood flow by the formation of a constrained
vortex, through the rotational motion of an impeller Centrifugal pumps have had three distinctive
or cone within the pump housing. This constrained design generations (Figure 17-4). The first cen-
vortex creates an area of low pressure (negative) trifugal pump, the Biomedicus® CP Medtronic,
Minneapolis, MN, had a fixed central shaft, and was
considered a technical revolution, but did not pair
well with long-term silicone membrane oxygenators.
The high resistance of the silicone oxygenator ne-
cessitated the use of high revolutions per minute
(RPM) to generate a given blood flow, resulting
in significant heat generation, thrombosis, and
hemolysis. The introduction of second generation
centrifugal pumps (heat resistant bearings), third
generation centrifugal pumps (magnetically levi-
tated and bearingless) (Figure17-5), and low resis-
tance hollow fiber oxygenators, have alleviated the
issue of heat generation and allow for lower RPM
settings to produce equivalent pump flows. With
these developments there has been a progressive
transition from the use of roller pump to centrifugal
pumps for ECMO support (Figure 17-6).
The centrifugal pump design influences the flow
dynamics inside the pump head. Computational
flow dynamics are used to design the pump in an
effort to optimize flows. Reducing the size of the
pump can be advantageous in reducing prime and
Figure 17-4. Centrifugal blood pumps classi- foreign surfaces; however, smaller designs require
fied by generation. Centrifugal pump design greater RPM to create forward blood flow, which in
was engineered to reduce heat generation at the
pivot point of the pump and improving overall turn can produce greater levels of heat and injury to
blood handling with successive pump generations. blood elements. Smaller pump designs are also more
(CP=Centrifugal Pump). vulnerable to failure than larger pumps as a result

154
 ECMO Circuit Components

of blood clotting, which may become trapped and Desired Blood Flow Rate
block the small impeller. Blood pumps with impel-
lers exhibit superior performance at a given RPM For roller pumps the size of the circuit tubing,
level than those without, but more easily propel and especially the raceway tubing, is dependent on
clots and air forward within the circuit. Magnetically the desired blood flow rate and has a direct bearing
levitated pumps can tilt partially or block entirely on the RPM required to achieve the desired blood
because of clot, resulting in severe hemolysis or flow rate. For lower flow rates ¼” size tubing is
sudden pump failure. A thorough understanding of selected, for higher flows 3/8” size tubing is rec-
the dynamics of each pump (Table 17-1) is neces- ommended.
sary in order to enable the diagnosis of pump related Centrifugal blood pumps may be either high
technical failures, and to develop specific protocols flow range pumps or low flow range pumps. High
to prevent or troubleshoot these complications. flow range pumps provide blood flow from 1-10
LPM, making them suitable for use with the adult

Figure 17-5. Summary of centrifugal pump physiologic principles. (CP=Centrifugal Pump).

Figure 17-6. Worldwide utilization of roller and centrifugal pumps for adult and pediatric ECLS application.

155
Chapter 17

population. All centrifugal pumps with 3/8” outlets the blood pump increases substantially and the red
are considered high flow pumps, meaning that they blood cells are susceptible to rupture. In low flow/
have a low hemolysis index for a high range of flows high pressure situations, such as pediatric ECLS or
with a physiological pressure head window. These extracorporeal carbon dioxide removal (ECCO2R)
pumps can safely be used for low pump flows with indications, a suitable low flow pump should be
adults, for example in a weaning scenario, where considered (Table 17-1).
the pressure lowers in relation to the decrease in
flow. However, these pumps lose their optimal Hemocompatibility
hemocompatibility when their optimal flow/pres-
sure head window is deserted, which may occur Roller pumps are considered more or less he-
if used in the pediatric population. In the pediatric mocompatible depending on the occlusive settings
setting, elevated RPM are required for low flows, of the rollers in the housing. Apart from potential
as higher resistances in smaller circuit tubing and blood damage via inappropriate roller occlusion,
cannulae result in similar pressures as adults with roller pumps can also damage the circuit tubing,
high flows. Hence blood cell contact time within resulting in leaching of toxic substances or by re-
leasing artificial particulate emboli into the patient.
Regular ‘walking’ of the tubing, or changing the site
of tubing occlusion minimizes this complication
Centrifugal Max Priming
Pump flow Volume/ Features but it remains a concern. Roller pump circuits com-
Rate Outlet monly have larger foreign surface contact areas due
Port Size
Medtronic 8 LPM 86ml/ Carmeda coating to the longer tubing required for gravity drainage
Biomedicus BP80 3/8” Fixed shaft pump (heat generation) and intermittent “walking” of the tubing.
Back-up manual hand-crank
Medtronic 1,5 LPM 48ml/ Carmeda coating Centrifugal pumps are extensively tested using
Biomedicus BP50 3/8” Fixed shaft pump (heat generation) computational fluid dynamics to predict red blood
(Low Flow) Back-up manual hand-crank
cell trauma. Centrifugal pumps are recommended by
Medtronic 10 LPM 40ml/ Low heat generation mono pivot bearings
Affinity CP 3/8” impeller pump their manufacturers for certain flow ranges but not
Back-up manual hand-crank for physiological pressure environments. When used
LivaNova 8 LPM 57ml/ Low heat generation mono pivot bearings
Revolution CP 3/8” impeller pump outside of safe flow-high pressure head windows,
Back-up manual hand-crank they become less hemocompatible. The use of high
Xenios Medos 8 LPM 16ml/ Low heat generation mono pivot bearings
Deltastream DP3 3/8” Small pump design requiring high rpms flow centrifugal pumps in pediatrics requires down-
Pump only to be turned on when primed sizing of the tubing to fit the ¼” circuit, which may
Requires back-up moter
Xenios Medos 2,4 LPM 16ml/ Low heat generation mono pivot bearings expose blood cells to turbulence and shear stress,
Deltastream DP3 ¼” Small pump design requiring high rpms causing damage and local thrombus formation.
pediatrics Pump only to be turned on when primed
(Low Flow) Requires back-up motor Numerous reports and data from the ELSO Registry
Maquet-Getinge 10 LPM 32ml/ Low heat generation mono pivot bearing confirm these findings, indicating that centrifugal
Rotaflow 3/8” Impeller pomp
Back-up manual hand-crank pumps in neonates, with the exception of the Abbott
Maquet-Getinge 2.8 LPM 18ml/ Low heat generation mono pivot bearings Pedivas™ St. Jude Medical Inc., Minneapolis MN,
Rotassist ¼” Impeller pump
pediatrics 2.8 Integrated pressure sensors cause more blood cell damage than roller pumps
(Low Flow)
5/7 LPM 240-
Back-up manual hand-crank
Pump integrated in oxygenator
with adult patients.9-11 In addition, all high flow
Maquet-Getinge
Cardiohelp 5.0- 273ml/ Low heat generation bearings range, magnetically coupled or levitated pumps are
7.0 3/8” Impeller pump with 4 outlet ports comparable in terms of hemocompatibility when
Integrated pressure sensors
Manual hand-crank used with adults.12-13 In any case, regular measure-
Abbott Centrimag 8 LPM 31ml/ Magnetic levitated impeller
3/8” Requires back-up motor
ment of plasma free hemoglobin (PFH) is required
Abbott Pedivas 1,7 LPM 14ml/ Magnetic levitated impeller to monitor circuit derived hemolysis. Hemolysis,
(Low Flow) ¼” Requires back-up motor
as defined by ELSO, is a concentration of PFH in
the blood exceeding 50 mg/dl. Free hemoglobin is
Table 17-1. Commercially available low and nephrotoxic and binds rapidly and irreversibly to
high flow centrifugal blood pumps.
endogenous nitric oxide (NO), inducing a range

156
 ECMO Circuit Components

of complications unfavorable in ECMO patients, ministration, decreasing RPM, reducing ultrafiltra-

such as increased systemic and pulmonary vascular tion rates, etc.) can mitigate or prevent damaging
resistance, increased thrombin generation, platelet peak negative pressures. Incorporating a bladder in
dysfunction, and clotting disorders.14 the drainage tubing not only dampens fluctuations
of venous line pressures but also a means to avoid
Cost direct air-blood interaction at the site of pressure
measurement. Dampened pressure fluctuations
Cost is complicated as blood pump prices vary result in substantially reduced gaseous microem-
in different countries. Generally, the prices of most boli formation in the venous line.16 In contrast,
1st and 2nd generation centrifugal pumps fall into the bladder only collapses at negative pressures
the same range. Cardiohelp is more expensive due exceeding -250 mmHg and that turbulences inside
to the integrated pump, oxygenator, and sensor the device may activate clotting or accelerate blood
technology. The Centrimag™/Pedivas™, the only cell damage.
completely levitated pump, is substantially more
expensive. Cost calculation can also substantially Oxygenators
vary, depending on whether centers use the pump
according to label or manufacturers recommenda- Description and Characteristics
tions, or whether change-out can be done upon
discretion of the user. Blood oxygenators are in most cases membrane
In the United States, all blood pumps for ECLS oxygenators, based on a microporous hollow fiber
are FDA approved for 6 hours; however, change-out membrane design. The fibers are wound or bundled
of the pump every 6 hours is impractical. Centrimag in a hard shell, with a large surface area-to-volume
is FDA approved for 30 days as a VAD but not in
the setting of ECMO support, when an oxygenator
is included in the circuit. In contrast, European CE
markings offer longer periods of validated use for
ECMO support based mostly upon the manufacturer
recommendations. For example, recommendations
are currently 7 days for the DP3, 30 days for the
Cardiohelp, 28 days for the Lifebox (ECLS set with
the Sorin Revolution Centrifugal Pump).

Bladders

The composition (length and size) of the circuit

and the site of pressure measurement (inflow/nega-
tive pressure) affect the forces needed to produce
forward flow. While most centrifugal circuit users
avoid inflow/negative pressures more negative than
-50 to -70 mmHg, the most important aspect is to
avoid sudden or peak inflow/negative pressures
at all times. Pressure servo regulation or volume Figure 17-7. The Better Bladder™ volume buffer
buffer capacitance devices such as the Better Blad- capacitance device, and charts of circuit pressure
der™ Circulatory Technology Inc., Oyster Bay, conditions when used with a centrifugal pump, in
the setting of a significant peak inflow/negative
NY can help in avoiding sudden peaks of inflow/ pressure, with and without the use of the Better
negative pressure (see Figure 17-7).15 In the event Bladder™ device. Illustrations, pictures, and charts
these devices are not incorporated, close pressure courtesy of Circulatory Technologies Inc. Oyster
Bay, New York, USA.
observation and immediate intervention (fluid ad-

157
Chapter 17

ratio (1-2.5 m2), with a priming volume of 100-350 and the plasma infiltration and leakage through the
ml. The gas usually flows inside the lumen of the pores; consequently, the membrane permeability
fibers, while blood flows externally, around the declines markedly over time. PMP fiber oxygen-
fiber bundle (extraluminal flow). The opposite flow ators are a kind of hybrid oxygenator in which the
pattern where blood flows inside the fibers (intra- microporous fibers are layered by a true membrane,
luminal flow), is far less common, as extraluminal preventing plasma from leaking through the pores.
flow results in a lower resistance to gas transfer Most PMP oxygenators use the Membrana® coated
through the blood film. Furthermore, the resistance fiber (e.g. Medos Hilite®, Maquet Getinge Quadrox
to blood flow and pressure drop are reduced with ID, LivaNova PLC Eos) to make their oxygenators
extraluminal flow, making these oxygenators less more plasma leak resistant. The Membrana® fiber
prone to blood cell damage and hemolysis. Passive has been developed not to avoid but to reduce
secondary flow may be achieved by creating undula- plasma leakage, as this is a function of the thickness
tion or texturing of the membrane surface or using of the nonporous fiber skin which must be limited
a specific flow geometry in order to enhance blood to preserve gas transfer capacity.17,18
contact with the gas exchange surface. Turbulence When choosing an oxygenator, the physiologi-
and secondary flow increase the pressure drop and cal needs of the patient and the level of support must
the shear rate experienced by red blood cells, again be considered (Table 17-2, Table 17-3). Oxygen
causing shear-induced hemolytic damage. transfer capacity depends on the intrinsic properties
The use of highly hydrophobic materials pre- of the oxygenator, circuit blood flow, hemoglobin
vents plasma leakage through the membrane pores. content, the oxygen pressure gradient across the
In the early 1980s, the first commercial hollow fiber hollow fiber membrane (fraction of oxygen set
oxygenators used silicone-coated microporous poly- on blending device), and inlet O2 saturation. The
propylene membranes; more recently, poly-methyl-
pentene (PMP) membranes have become available Selection of an Oxygenator Referenced to Required VO2

and exhibit very good performance. Although Oxygen Consumption

at Rest
VO2/BW
mL/kg/min
Total VO2
mL/min
ECMO Oxygenators
VO2max mL/min

microporous membranes exhibit high gas transfer Neonates < 6 Kg 5-8 < 30-48 Medos 800 LT:
Avecor 0800:
48
70

rates, in long-term use they undergo progressive

Eurosets Newborn A.I. One: 90
Pediatric < 35 Kg 4-6 < 140-210 Medos 2800 LT: 170

alteration of surface properties due to protein and

Quadrox ID Ped: 180
Avecor 2500: 113
Eurosets Pediatrics: 250
lipid adsorption, the wetting of membrane pores, Adults 70 ± 25 Kg 3-5 < 350 ± 100 Medos 7000 LT: 520
Quadrox ID Adult: 425
Avecor 4500: 400
EOS ECMO LivaNova: 320
Selection of an ECMO Oxygenator Eurosets Adult A.I. One: 350
Selecting Appropriate An oxygenator must transfer Normal Adult Lungs = 3500 mL/min with 21% Oxygen
Oxygenator Size the amount of O2 consumed Rated Blood Flow of Current ECMO Oxygenators
by the patient. Adult Device 7000 mL/min
Rated Flow The maximum bloodflow Pediatric Device 2800 mL/min
through the oxygenator Neonatal Device 800 mL/min

enabling the inlet blood Required Patient Blood Flow

saturation to increase from Adults 60 mL/Kg/min (up to 7000/60 = 115 Kg)

Pediatrics 80 mL/Kg/min (up to 2800/80 = 35 Kg)
75% to 95% for blood with a Neonates 120 mL/Kg/min (up to 800/120 = 6.2 Kg)
hemoglobin of 12 g/dL. Patient Oxygen Consumption at Rest
Ideal Flow The flow necessary to Adults 3-5 mL/Kg/min
increase the saturation to ≥ Pediatrics 4-6 mL/Kg/min
99%, which can only be Neonates 5-8 mL/Kg/min

achieved when the desired Rated Blood Flow Must Be Above Required Patient Support

flow is below the rated flow.

The oxygenator will be able to transfer the amount of Table 17-3. Selection of an ECMO oxygenator in
oxygen consumed by the patient as follows:
relation to required oxygen consumption (VO2) and
Patient O2 consumption (VO2) < Oxygenator maximum
expected blood flow requirements for adult, pediat-
gas transfer capacity (VO2max)
ric, and neonatal patient ranges. Manufacturer speci-
fied maximum oxygen consumption (VO2max)
Table 17-2. Selection of and ECMO oxygenator capacities for the Medos, Maquet-Getinge, Avecor,
with reference to oxygen consumption and rated Liva Nova, and Eurosets oxygenators, available in
flow capabilities. the neonatal, pediatric, and adult size ranges.

158
 ECMO Circuit Components

amount of CO2 transfer is relatively independent The incidence of oxygenator failure, as reported
of blood flow while the sweep gas flow rate is the in the ELSO Registry, is related to the duration of
major determinant of CO2 clearance. The different support (Table 17-4).19
oxygenators available on the market, are character- A wet membrane may occur from the accumu-
ized mainly by: lation of condensation in the gas phase of the oxy-
genator due to the temperature differences between
• Maximum flow rate or rated blood flow the ventilating gas and blood phase, causing the
• Internal resistance to flow membrane to become less permeable to CO2. Drying
the oxygenator with high gas flow (‘burping’) for
The internal resistance to flow is inherent to the several seconds can solve this problem, warming
oxygenator design and is expressed by the pressure the sweep gas to body temperature has also been
drop (Pin–Pout) across the oxygenator. Oxygenators suggested to prevent this phenomenon.
with high pressure drop result in more shear stress or
friction to blood cells, although shear stress within Heat Exchangers
an oxygenator is also a function of the length of
the pathway.19 Description of a Heat Exchanger

Oxygenator Related Complications Heat exchangers may be either separate in an

ECMO circuit, positioned before the oxygenator, or
Despite the optimization of anticoagulation and more commonly, integrated in the oxygenator within
the use of heparin bonded surfaces, there remains a separate compartment. Nonsterile water from the
a significant incidence of oxygenator clotting as heater device is separated from the blood by stain-
indicated by the 2017 ELSO Registry (Table 17-4).19 less steel, aluminum, or polyurethane. These areas
Oxygenator clotting is indicated by gradually in-
creasing transmembrane pressures, rising D-dimers,
decreased gas transfer, and hemolysis.20 Routine
monitoring of these parameters permits prompt
intervention and elective oxygenator change-out,
rather than emergent change-out. As acute clotting
can occur, there should be a protocol and regular
water drill for urgent oxygenator change-out.20
Oxygenator failure results in its inability to meet
the manufacturer specified capacity for gas transfer.
Oxygen transfer can be calculated by measuring the
O2 content of the blood at the outlet of the oxygen-
ator minus the oxygen content at the inlet of the
oxygenator and should be equivalent to published
data for the given blood flow and hemoglobin level.
Circuit Related International International Circuit Related International International
Complications Summary Summary Complications Summary Summary
Respiratory Neonates Adults Cardiac Neonates Adults
ECMO (%) (%) ECMO (%) (%)
Oxygenator 4 6 Oxygenator 3 3
Failure Failure
Oxygenator 14 13 Oxygenator 12 7
Clots Clots
Pump Failure 1 1 Pump Failure 1 1
Air in Circuit 3 1 Air in Circuit 3 1
Severe 14 5 Severe 14 5 Figure 17-8. The Maquet-Getinge HU-35 heater
Hemolysis Hemolysis
unit, employs a centrifugal pump, which circulates
temperature regulated water to an integrated or ex-
Table 17-4. Most prevalent ECMO circuit related ternal heat exchanger device. Photograph courtesy
complications as reported to ELSO to July 2017. of Maquet Getinge Group, Rastatt, Germany.

159
Chapter 17

are glued or welded to separate the two compart- precautions with respect to usage, disinfection,
ments and maintain sterility of the blood. Heat ex- and maintenance of heaters/heater-coolers.24-26
changers in ECMO keep the patient normothermic,
except in some special situations where cooling is
desired. ECCO2R, oxygenators do not have a heat
exchanger compartment as this is not required in
small, low flow devices.

Heat Exchanger Related Complications

Heat exchanger leakage is a rare but potentially

fatal complication of ECMO and is associated with
massive blood hemolysis. Heat exchanger integrity
must be assessed during the initial assembly of
equipment and monitored throughout the ECMO
course so that any suspicions are met with timely
intervention. Water should preferably be circulated
through the heat exchanger before priming to detect
any leakage, but this is not always possible (prep-
rimed circuits) and does not guarantee an intact
heat exchanger. Alternatively, pressurized air can
be delivered to the water chamber while a pressure
manometer is monitored for losses in air pressure,
which would indicate a chamber leak.21
Depending on the type of heater-cooler device,
water is either pumped through the heat exchanger or
drawn into it; the latter provides a greater degree of
safety as leakage is indicated by blood stained water
circulating through the semitransparent heater cool-
er tubing (Figure 17-8). Reports describe the buildup
of electrostatic charge from roller pumps creating
a spontaneous discharge, which compromises the
integrity of the heat exchanger fibers.22 In 2011, in
response to customer complaints, a manufacturer
issued a warning noting oxygenators primed with
sodium chloride solution for a prolonged period of
time may develop microholes in the heat exchanger
interface.23
Recently, infectious complications via con-
taminated heat exchangers/heater-coolers have
been reported in patients who underwent cardiac
surgery with cardiopulmonary bypass. More spe-
cifically, the reports express concerns about My-
cobacterium chimaera growing in heater-coolers
and the association with some Mycobacterium
chimaera infections in surgical patients. Health-
care institutions worldwide have since been issued
recommendations for prevention and the necessary

160
 ECMO Circuit Components

References 11. Botrell S, Bennett M, Augustin S, et al. A com-

parison study of haemolysis production in 3
1. Lamb KM, Hirose H. Vascular complications contemporary centrifugal pumps. Perfusion.
in extracoporeal membrane oxygenation. Crit 2014.
Care Clin. 2017;33(4):813-824. 12. Lehle K, Philipp A, Müller T, et al. Flow
2. Cakici M, Ozcinar E, Baran C, et al. A retro- dynamics of different adult ECMO systems:
spective cohort analysis of percutaneous versus A clinical evaluation. Artificial Organs. 2014;
side-graft perfusion techniques for veno-arterial 38(5):391–398.
extracorporeal membrane oxygenation in pa- 13. Palanzo DA, El-Banayosy A, Stephenson
tients with refractory cardiogenic shock. Perfu- E, Brehm C, Kunselman A, Pae WE. Com-
sion. 2017;32(5):363-371. parison of hemolysis between CentriMag and
3. Gates JD, Bichell DP, Rizzo RJ, Couper GS, RotaFlow rotary blood pumps during extracor-
Donaldson MC. Thigh ischemia complicating poreal membrane oxygenation. Artif Organs.
femoral vessel cannulation for cardiopulmonary 2013;37(9):162-166.
bypass. Ann Thorac Surg. 1996;61(2):730-733. 14. Rother RP, Bell L, Hillmen P, Gladwin MT.
4. Rupprecht L, Lunz D, Lubnow M, Schmid C. The clinical sequelae of intravascular hemolysis
Pitfalls in percutaneous ECMO cannulation. and extracellular plasma hemoglobin: a novel
Heart Lung Vessel. 2015; 7(4): 320–326. mechanism of human disease. JAMA 2005;
5. Lequier L, Horton SB, McMullan DM, Bartlett 293(13):1653-1662.
RH. Extracorporeal membrane oxygenation cir- 15. Tamari Y, Lee-Sensiba K, King S, Hall, MH.
cuitry. Pediatr Crit Care Med. 2013; 14:S7–S12. An improved bladder for pump control during
6. Thiagarajan, RR. Extracorporeal membrane ECMO procedures. Journal of Extra-Corporeal
oxygenation in infants and children. In: Car- Technology. 1999; 31(2)
diopulmonary Bypass and Mechanical Support, 16. Ganushchak YM, Ševerdija EE, Simons AP,
Principles & Practice. 4th ed. New York, NY : van Garsse L, Weerwind PW. Can minimized
Wolters Kluwer; 2016: 709-728. cardiopulmonary bypass systems be safer?
7. Rais-Bahrami K, Nunez S, Revenis ME, Luban Perfusion. 2012;27(3):176-182.
NLC, Short BL. Follow-up study of adolescents 17. Eash HJ, Jones HM, Hattler BG, Federspiel,
exposed to Di(2-Ethylhexyl) Phthalate (DEHP) WJ. Evaluation of plasma resistant hollow
as neonates on extracorporeal membrane fiber membranes for artificial lungs. ASAIO J.
oxygenation (ECMO) support. EnvironHealth 2004:491-497
Perspect. 2004; 112:1339–1340. 18. Peinemann KV, Pereira Nunes S. Membrane
8. Kopp R, Mottaghy K, Kirschfink M. Mechanism technology: Membranes for life science. J Wiley
of complement activation during extracorpo- & Sons; 2011:57-59.
real blood-biomaterial interaction: effects of 19. Extracorporeal Life Support Organization.
heparin coated and uncoated surfaces. ASAIO ELSO registry report, July 2017. https://www.
J. 2002;48(6):598-605. elso.org/Registry/Statistics/Reports.aspx. Ac-
9. Barrett BC, Jaggers JJ, Cook EF, et al. Pediatric cessed August 7, 2017.
ECMO outcomes: Comparison of centrifugal 20. Dornia C, Philipp A, Bauer S, et al. D-dimers
versus roller blood pumps using propensity are a predictor of clot volume inside membrane
score matching. ASAIO J. 2013; 59(2):145-151. oxygenators during extracorporeal membrane
10. O'Kelly P. Neonatal extracorporeal membrane oxygenation. Artificial Organs. 2015.
oxygenation outcomes by pumps type – ELSO 21. Hamilton C, Stein J, Seidler R, et al. Testing
Data Registry analysis. 31st Annual Children's of heat exchangers in membrane oxygenators
National Symposium: ECMO and Advanced using air pressure. Perfusion. 2006;21:105-107.
Therapies for Respiratory Failure. February 22. Snijders J, De Btuijn P, Bergmans M, Bastianen,
2015 G. Study on causes and prevention of electro-

161
Chapter 17

static charge build-up during extracorporeal

circulation. Perfusion.1999;14:363-70.
23. LivaNova PLC. Notice safety warning from
the Sorin group. https://www.scps.org.uk/pdfs/
sorin-12:12:11.pdf. Accessed
24. U.S. Food and Drug Administration (FDA).
Medical device safety, alerts and notices. www.
fda.gov/MedicalDevices/Safety/AlertsandNo-
tices/ucm466963.htm. Accessed October 15,
2016.
25. Centers for Disease Control and Prevention
(CDC). Contaminated heater cooler devices.
https://www.cdc.gov/hai/outbreaks/heater-
cooler.html. Accessed Oct 27, 2015.
26. European Centre for Disease Prevention and
Control (ECDC). Invasive cardiovascular infec-
tion by Mycobacterium chimaera potentially
associated with heater-cooler units used during
cardiac surgery. https://ecdc.europa.eu/sites/
portal/files/media/en/publications/Publications/
mycobacterium-chimaera-infection-associated-
with-heater-cooler-units-rapid-risk-assessment-
30-April-2015.pdf. Accessed May 4, 2015.

162
18

ECLS Safety and Other Monitoring Devices

Larissa Yalon, BSN, RN, CCRN, Kenneth A. Schenkman, MD, PhD

Introduction nula and the arterial pressure of the patient. An

obstructed or kinked arterial line will lead to
Monitoring patients on ECLS requires constant an increased pressure. If this pressure exceeds
attention. Although many monitoring devices exist 400 mmHg, the risk of circuit interruption and
to optimize safety of the ECLS patient, none of hemolysis increases.
these can replace the vigilance of the bedside clini- • Pre-oxygenator/internal pressure is measured at
cian. This chapter divides safety and monitoring the inlet to the oxygenator and is used in cor-
into monitoring the state of the ECLS circuit, and relation with the post-oxygenator pressure to
monitoring the clinical state of the patient. diagnose oxygenator issues. If an outflow prob-
lem exists, both the pre- and post-oxygenator
Circuit Related Monitoring pressures rise together. An independent increase
in pre-oxygenator pressure is an indicator of
oxygenator clot formation or flow disruption
Pressure Monitoring within the oxygenator housing.
• Delta Pressure (ΔP)/transmembrane pressure
Most systems measure pressures at three points is the calculated difference between the pre-
in the circuit (Figure 18-1). Traditionally, pressures oxygenator and post-oxygenator pressures. A
have been monitored using transducers added to cir- rise in the delta (rising pre-oxygenator pressure
cuits. Newer ECLS systems have integrated pressure with falling post-oxygenator) without a corre-
monitoring, decreasing circuit access points. Pumps
can be set to servo regulate and maintain pressures
within given parameters. As centers look to sim-
plify systems and decrease access points, pressure
monitoring may be limited to venous line pressures.

• Venous line/ inlet pressure is measured prepump

and monitored for excessive negative pressure,
which can lead to vessel or right atrial damage,
as well as cavitation and resulting hemolysis.
The venous pressure also reflects circuit volume
status. Pressure depends on patient volume sta-
tus, circuit length, cannula position, and FR size.
If a bladder is present, the venous pressure may
be obtained from an access site on the bladder.
• Post oxygenator/arterial line pressure is mea-
sured after the oxygenator and is a function
of the pump speed, the tubing resistance, can- Figure 18-1. Basic centrifugal pump ECLS circuit.

163
Chapter 18

sponding augmentation in flow reflect increas- Oxyhemoglobin Monitoring

ing resistance within the oxygenator, usually the
result of thrombosis. Pressure trends should be At a minimum, circuits have a noninvasive
followed rather than the absolute number. monitor to measure oxygen saturation of the blood
on the venous limb. In VA-ECLS this value fairly
Blood Flow Meters accurately reflects mixed venous saturation (SVO2)
provided the probe is placed before any shunts or
Extracorporeal blood flow is monitored via non- vents enter the circuit. In VV-ECLS this value is
invasive ultrasonic flow probes. A probe integrated falsely elevated by recirculation, but can be trended
into most systems monitors blood flow to the patient. and used to monitor recirculation fraction. A probe
If additional blood flow measurements are needed, placed on the arterial limb can be used to ensure
standalone transonic probes/boxes can be utilized. adequately oxygenated blood is being returned to
Low flow and high flow alarms can be set to alert the patient as well as monitor oxygenator function.
staff to changes in baseline flow. Retrograde flow is Many oxygen saturation sensors also have the abil-
a risk with nonocclusive centrifugal pumps, if the ity to monitor hemoglobin and hematocrit. More
distal pressure exceeds the pump generated pres- complex systems have the ability to monitor ECLS
sure. This risk can be mitigated by appropriate low gas parameters and calculate patient PaO2, PCO2 and
flow alarms. Many flow sensors have the ability to physiologic parameters, including oxygen consump-
detect zero or negative flow, resulting in an alarm, tion and CO2 production. These systems also permit
and if available, servo regulation of the circuit. The remote monitoring of the ECLS circuit. Inline blood
mode of servo regulation depends upon the system gas monitoring systems require a sensor in the blood
being used. The Maquet CARDIOHELP can detect path but can provide continuous monitoring of acid
and react to backflow of blood by automatically base status, oxygenation, and ventilation.
activating zero flow mode. In zero flow mode, the
CARDIOHELP® aims at a flow of 0 L/min by Emergency Power and Backup Systems
controlling the pump accordingly. The Livonova/
Sorin electrical remote controlled tubing clamp Modern ECLS circuits have battery back up
(ERC) instantaneously occludes the arterial line in systems which allow the pump to run for some
the event of a bubble, or level or retrograde flow. To time during a power failure. A backup system must
ensure accurate flow measurement the flow probe be available in case of complete pump or power
must be calibrated and corrected for tubing size and failure. The majority of systems have a hand crank
outer diameter. mechanism, but if hand cranking is not supported,
a backup console must be available and properly
Bubble/Air Detectors programmed.

Air/bubble detectors can identify microscopic Temperature Monitoring

air bubbles in the blood path. The ultrasonic sensors
sense a change from the normal blood path, trigger The ECLS circuit exposes the blood to ambient
an alarm, and if enabled, automatically discontinue temperature. Neonates and children can quickly
support to the patient via altering/stopping flow or become hypothermic should the heat exchanger
clamping a line. The flow meter may have a dual systems fail. The ability to monitor the tempera-
function also serving as a bubble detector. Bubble ture of the blood coming from and returning to the
detectors can be placed on the venous or arterial patient is available in integrated systems, or via
limbs or both and interventions set independently. oxygenator outlet sensors. In adult patients a heat
exchanger may not be necessary as the percentage
of extracorporeal blood volume is less, allowing
for normal thermoregulation. Close patient tem-

164
 ECLS–Safety and Other Monitoring Devices

perature monitoring should always be included in oxygen than blood in the descending aorta that is
the assessment. supplemented with flow from the ECLS circuit.

Patient Related Monitoring Cerebral and Somatic Oxygenation

Patients on ECLS may also be monitored with

Patient Hemodynamics cerebral and/or somatic oxygenation devices. The
current devices for these measurements are based
Blood pressure is typically monitored continu- on near-infrared spectroscopy (NIRS). In contrast
ously by invasive arterial catheters. Pulsatility of to pulse oximetry, these devices generally result
the arterial line may be diminished by poor native in trends only and do not report absolute values of
cardiac function in patients on VA-ECLS or by high tissue oxygenation. However, since they do not rely
ECLS blood flow. As the native cardiac function on pulsatility, they will provide oxygenation values
improves, pulsatility usually increases, due to the even with poor or absent cardiac function.
force generated by the improving left ventricle. The
mean arterial pressure is affected by the patient’s Interpreting SVO2
vascular tone, by native cardiac output, and by ECLS
flow. Thus, the mean arterial pressure may be a more In patients on VA-ECLS, the circuit venous
useful monitored measurement for ECLS patients. blood usually represents the true patient mixed
In patients with femoral arterial cannulation, venous oxygen status, if the patient does not have
oxygenated blood from the ECLS circuit may not intracardiac or pulmonary shunting. As oxygen
effectively reach the upper body arterial circulation delivery increases, SVO2 also rises, but as patient
due to the flow of poorly oxygenated blood pumped oxygen consumption, SVO2 decreases. SVO2 inter-
from the heart. In patients with severe lung disease, pretation thus requires consideration of oxygen
perfusion of the upper body may be from blood delivery and consumption. Sedation, anesthesia,
that is not fully oxygenated. Thus arterial blood or full mechanical ventilator support decrease
gas sampling from a right radial arterial catheter metabolic demands and oxygen consumption. Fever,
may be helpful to assure adequate oxygen is being seizures, and agitation increase metabolic activity
delivered to the brain. and oxygen consumption, thus decreasing SVO2 if
all else stays the same.
Pulse Oximetry In VV-ECLS, interpreting SVO2 is more com-
plex. While the above considerations remain, recir-
Pulse oximetry is routinely used for critically culation of delivered oxygenated blood back to the
ill patients on ECLS. In infants with a patent duc- circuit results in an increase in the measured SVO2.
tus arteriosus, differences in pre- and post-ductal When other conditions remain the same, increases
oxygenation can be monitored by pulse oximetry in SVO2 may be interpreted as increased recircula-
probes on the right hand and on any of the other tion in the ECLS circuit. Taking the other potential
limbs where perfusion includes flow from the ductus contributions into account, trends in SVO2 may be
to the aorta. Depending on cannula placement in VA used to assess relative changes in recirculation.
ECLS, streaming of arterial flow from the aorta, as
well as partial obstruction of flow from the cannula, Harlequin (North-South) Syndrome
may also result in differential saturation readings
in the right hand from the rest of the body. If there In patients on VA-ECLS with femoral arterial
is sufficient blood flow out from the heart due to cannulation, blood flow from the ECLS circuit head-
native cardiac function and the lungs are severely ing up the descending aorta competes with blood
impaired, blood coming up the ascending aorta and flow coming down the aorta from the heart, as de-
going to the right subclavian artery may have lower scribed above. This can result in the patient having
lower saturations in the top half of the body and

165
Chapter 18

higher saturation in the lower half of the body, the

so-called "Harlequin" or "North-South Syndrome."

Limb Perfusion

With femoral cannulation, perfusion to the

limb distal to the cannula placement may be com-
promised. Use of a limb reperfusion catheter helps
to mitigate this issue, but distal ischemia may still
occur. Currently, no consensus on optimal moni-
toring for distal limb ischemia exists, but clinical
exam, temperature in the affected limb, and the
use of tissue oximeters (NIRS) on the limb distal to
the catheter may all help assess adequacy of distal
perfusion.

Neuromonitoring

Adequate oxygenation and perfusion of the brain

is an essential goal of ECLS. However, monitoring
brain function in these patients can be challenging.
In infants, routine cranial ultrasound examination
is typically used to monitor for intracranial hemor-
rhage. In older children and adults, clinical monitor-
ing may be most useful. However, in heavily sedated
patients or in patients receiving muscle relaxation,
clinical examination may be insufficient. Intermit-
tent or continuous EEG monitoring can provide a
general sense of brain activity and identify seizures.
The availability of portable CT scanners has made
neuroimaging less challenging for patients on ECLS,
although traditional CT imaging can be performed
in the radiology suite, if the patient on ECLS can
be safely transported for study.

166
19

Management of the Neonate on ECMO

Barbara Haney, RN, MSN, RNC-NIC, CPNP-AC, FELSO, Robert DiGeronimo, MD

Introduction inadequate venous drainage, need for additional

oxygenation support and technically challenging
Extracorporeal membrane oxygenation (ECMO) cannulation.8 However, some centers primarily
has been routinely used for over three decades in utilize VV with similar outcomes and low rates of
neonates with severe respiratory failure. In 1975, Dr.conversion to VA-ECMO.9 Courses of ECMO for
Robert Bartlett reported the first successful use of CDH average 320 hours, significantly longer than
ECMO in a term neonate with meconium aspiration other diagnoses.5
and went on to publish landmark articles proving Meconium aspiration syndrome (MAS) remains
the efficacy of ECMO in neonates.1,2 Subsequent the second most common indication for neonatal
randomized controlled trials performed in the 1980s ECMO (22%), typically requiring relatively short
and 90s confirmed the benefit of ECMO in neonates courses, averaging 145 hours, with excellent sur-
with severe respiratory failure in preventing death vival results of 92%.5 These neonates are usually
and long-term disability.3,4 To date, over 35,000 excellent candidates for VV support5 while new-
neonates have been treated with ECMO in the borns with septic shock often require VA support
Extracorporeal Life Support Organization (ELSO) to provide increased oxygen delivery and hemody-
Registry with an overall survival to discharge or namic support. Reported ELSO survival for isolated
transfer of 71%.5 pneumonia is 60%, and when associated with sepsis,
The use of ECMO for neonatal respiratory fail- survival drops to 45%.5
ure (see Chapter 7) has significantly declined since
its peak in 1992 of 1516 cases, likely reflective of Initiation and Circuit Considerations
other beneficial therapies such as inhaled nitric
oxide (iNO), surfactant and improved mechanical Pre-ECMO evaluation of neonates should in-
and high frequency ventilation (HFV) technology.6,7 clude a thorough medical history and physical exam,
ECMO, however, remains a valuable therapy for head ultrasound, echocardiogram, measurement
selected neonates 34 weeks or greater with severe of coagulation parameters, and renal ultrasound if
respiratory disease. Although the best ECMO mode indicated. For VA-ECMO, standard single lumen
for any neonate must be determined individually, venous cannulas range from 10 to 14 French (Fr)
neonates may receive venoarterial (VA) or veno- and arterial catheters from 8 to 10 French size. VV
venous (VV) ECMO. double lumen (VVDL) cannulas designed for use
Congenital diaphragmatic hernia (CDH) has be- in the right internal jugular (IJ) vein are available as
come the most common diagnosis reported to ELSO 13 Fr so smaller neonates may not be candidates for
in this population, representing approximately one this mode. Typical catheter placement in a neonate
third of all cases. Survival to discharge or transfer involves insertion of the arterial catheter 2 to 3cm
is 50%, the lowest for any of the major diseases for into the right common carotid artery (target tip
neonatal ECMO.5 VA-ECMO historically has been junction of innominate and ascending aorta) and 6
the preferred mode of support for CDH patients to 7cm for the venous catheter into the right internal
based on concerns of cardiovascular instability, jugular vein (target tip mid to lower right atrium).

167
Chapter 19

Catheter placement is confirmed with CXR and/or Daily Patient Management

echocardiography per center protocol (Figure 19-1).
An important complication of VV-ECMO is Fluid, Electrolytes, Nutrition
recirculation which occurs when blood flows back
through the ECMO pump without it passing out the Neonates on ECMO have high rates of protein
heart and into the systemic circulation. Poor cath- catabolism, thus providing appropriate nutrition
eter position, low cardiac output, low intravascular is essential. The American Society for Parenteral
volume, and high pump flow may contribute to and Enteral Nutrition (ASPEN) ECMO guidelines
recirculation. Echocardiogram may help with can- recommend nutritional support be initiated expedi-
nula positioning, and evaluating cardiac function or tiously.10 Recommended parenteral nutrition goals
intravascular volume status are 100-120 Kcal/kg/day and up to 4 grams/kg/day
Circuits for neonates up to 10 kg typically use ¼ protein. Enteral nutrition has been increasingly
inch tubing, have a volume of 200 to 300 ml and are employed in neonates despite historical concerns
primed with saline, albumin and type-specific blood, regarding altered gut perfusion and possible nec-
respectively. If cannulation is emergent, O-negative rotizing enterocolitis.11 Serum electrolytes, renal
emergency release, blood can be used. For VV- function, and other appropriate labs to monitor nutri-
ECMO, blood should be irradiated and fresh (less tion and toxicity associated with TPN use should be
than 5 days old) or washed/hemofiltered to reduce measured routinely (Table 19-1). Ionized calcium
potassium concentration prior to priming. ECMO levels should be maintained >1.1 mmol/L to facili-
centers continue to use both roller and centrifugal tate cardiac function. Minerals and trace elements
pumps for neonates as no definitive published litera- should also be added to the TPN as appropriate.12,13
ture currently exists demonstrating the superiority Transient renal dysfunction occurs commonly, but
of either technology. However, some centers have spontaneously resolves over the first 48-72 hours
reported more hemolysis and adverse outcomes in many patients. Restricting daily fluid intake to
using centrifugal pumps in neonates. 60-100 ml/kg/day and diuretics facilitates diuresis
that resumes as cardiac output improves, capillary
leak resolves, and fluid mobilization occurs.14

Figure 19-1. Neonatal VA and VV catheter positioning on CXR.

168
 Management of the Neonate on ECMO

Cardiovascular System who have suffered severe hypoxia or in whom the

tip of the arterial catheter is placed too close to
ECMO blood flow in neonates should be gradu- the coronary arteries.17 Increased afterload with
ally increased after initiation to 120 ml/kg/min to higher flows of VA-ECMO may also contribute.14
provide adequate tissue perfusion and oxygenation Cardiac stun is characterized by a pulse pressure
(range from 80-150 ml/kg/min).14,15 Critically ill <10 mmHg.13 Treatment involves adjusting the ar-
neonates frequently have compromised cardiac terial cannula position as necessary and providing
function and hemodynamic instability, requiring adequate ECMO blood flow support while awaiting
resuscitation prior to cannulation. After cannulation, myocardial recovery.
vasoactive agents can typically be weaned but this Systemic hypertension can also complicate
depends on the underlying pathology. With VV- ECMO, mostly commonly with VA support. It
ECMO vasopressor weaning must be performed should be aggressively treated due to its association
with close attention to native cardiac output. Typi- with an increased risk of intracranial hemorrhage.18
cally cardiac function improves with VV-ECMO, Treatment includes lowering ECMO pump flow as
probably related to improved myocardial oxygen tolerated, weaning vasopressors, and the selective
delivery and decreases in right ventricular (RV) use of antihypertensive medications as indicated.
afterload.16
A unique complication of VA-ECMO is cardiac
stun. While the etiology of cardiac stun remains
undetermined, it occurs more often in neonates

TEST FREQUENCY
ACT Every one hour, more frequently if either out of range or if
heparin is turned off. Many centers are utilizing other
parameters in addition to or instead of dependence on ACT.
anti Xa and Antithrombin III Every 6-12 hours
Bilirubin, LFTs Daily or per center protocol
Blood cultures Per center protocol
Some centers daily or every other day; others only when
clinically indicated
Blood Gases - Circuit Every 12 hours
(Pre- and Post-oxygenator) Especially when using the CDI blood gas monitoring system,
necessary to calibrate the CDI
Blood Gases - Patient Every 4-6 hours
(Venous and Arterial) Some centers monitor more frequently to track lactic acid
levels, CO2 levels and pH status
Coagulation studies Some centers monitor PT, PTT, fibrinogen, anti-thrombin III
levels, and FSP, while others only monitor fibrinogen daily or
more often
CBC with differential and Every 6 to 12 hours depending on the coagulation status
platelet count
Na, K, Cl, HCO3, iCa and Every 12 hours or as needed
glucose
Plasma free hemoglobin Daily
Some centers only when significant hemolysis is suspected
Total protein, albumin, Ca, Daily or per center protocol
Phos, Mg
TEG/Rotem Some centers use thromoelastography as needed to direct
blood product administration

Table 19-1. Common laboratory tests and frequency.

169
Chapter 19

Oxygen Delivery When conventional ventilation is used, various

positive end expiratory pressure (PEEP) strategies
The underlying goal of ECMO support is to are employed. In a recent review of the ELSO Reg-
provide adequate oxygen delivery (see Chapter 2). istry evaluating ventilation strategies, low (4-6 cm
Oxygen delivery is the amount of oxygen content H2O), moderate (7-9 cm H2O), and high (10-12 cm
in the blood times blood flow (Oxygen Delivery H2O) PEEP were all used, with high PEEP being
= Arterial Oxygen Content x Flow). Flow equals the most common (43%).19 Additional rest settings
the combination of ECMO pump flow and native included minimal respiratory rates of 10-20 bpm and
cardiac output of the patient (in VA-ECMO). When peak inspiratory pressures (15-20 cm H2O) or tidal
oxygen delivery to the brain and other vital organs volume (4 to 6 ml/kg). FiO2 typically was set be-
does not meet demands, oxygen debt occurs result- tween 0.21-0.40.12,19,25 Bronchoscopy with or with-
ing in lactic acidosis, and end organ injury. Oxygen out mucolytics or surfactant is used during ECMO
consumption is defined as the difference in arterial to assist with clearance of secretions, evaluation
and venous oxygen content times flow and increases of infection, and to improve lung recruitment.14,24
with sepsis, catecholamine release, hyperthermia, Ventilation on ECMO is managed by titrating the
and patient activity. Oxygen sufficiency is best mea- sweep gas flow rate to maintain the desired pCO2,
sured in neonates on ECMO by using a combination typically 40-50 mmHg. There is variation between
of monitoring mixed venous oxygen saturations as centers regarding sweep gas methods, with some
well as other measures of perfusion including lactate using a sweep gas of 100% oxygen, while others
levels and near-infrared spectroscopy (NIRS). use a mixture of oxygen and compressed room air,
via an oxygen-air blender. To manage the pCO2,
Pulmonary System carbogen gas (5% CO2, 95% O2) or 100% CO2 may
be titrated into the sweep gas. It is critical to adjust
ECMO minimizes the requirement for pul- sweep gas flow slowly to avoid rapid changes in
monary gas exchange, so typically the ventilator cerebral pCO2 that have been associated with brain
can be weaned to “rest” settings to avoid further injury in the neonate.14 Patients are most at risk for
ventilator induced lung injury. On VA support, the large changes in pCO2 at cannulation when ventila-
ECMO circuit bypasses the pulmonary vasculature tion becomes dominated by the circuit sweep gas
so ventilator settings can be lowered fairly quickly. flow therefore setting a modest sweep flow is en-
In contrast, with VV-ECMO, the ventilator should couraged. 25 Daily chest radiographs help to assess
be weaned more slowly and target rest settings lung recovery, evaluate for air leak, and monitor
may be higher than with VA. When the ventilator cannula and other line positions. Blood gases are
is weaned quickly on VV, lung collapse can ensue, drawn every 6-12 hours, to include patient arterial
resulting in acute increases in pulmonary vascular blood gas and circuit pre- and post-oxygenator blood
resistance with acute increases in RV afterload as gases. (Table 19-2).
well as inadequate oxygenation and ventilation.12
The optimal ventilation strategy for neonates while Infection
on ECMO is not standardized and varies by institu-
tion, the underlying disease process of the patient, as Instrumentation, prolonged use of invasive sup-
well as the mode of ECMO support. With regards port devices such as central lines, urinary catheters
19

to mode of ventilation, both conventional and HFV and endotracheal tubes, as well as frequent access
(oscillator, jet or other) can be successfully used.19- into the circuit, places neonates on ECMO, with
22
Increasingly, patients are being extubated while their immature immune system, at increased risk of
on ECMO in an effort to lessen sedation needs and infection. Nosocomial infections can have signifi-
avoid ventilator related complications; however, cant consequences to the ECMO patient; however,
the benefit of this strategy on outcomes in neonates there is no evidence that routine prophylactic use
remains unclear.23,24 of antibiotics decreases the risk of such infections.14,
26
Furthermore, no evidence currently supports ob-

170
 Management of the Neonate on ECMO

taining routine surface or blood cultures on ECMO, every 12 to 24 hours or 0.05 to 0.4 mg kg/hr continu-
but maintaining a low threshold for evaluating for ous infusion or other diuretics may facilitate urine
sepsis is critical to initiate treatment promptly. Some output.28 In cases in which the urine output does not
centers, however, do routinely monitor daily or improve after 48 to 72 hours, hemofiltration or con-
every other day circuit blood cultures in neonates tinuous renal replacement therapy (CRRT) through
on ECMO. An initial drop in white blood cells the circuit can additionally be utilized to support
and platelets often follows placement on ECMO fluid removal and kidney function (see Chapter 28).
or circuit changes due to adhesion of these cells
to the oxygenator and tubing. For antibiotic dos- Neurologic System
ing, ECMO may alter medication pharmokinetics,
necessitating doses or dosing interval adjustments Severe intracranial hemorrhage (ICH) is the
(see Chapter 6).27 most dreaded morbidity for neonates on ECMO.29-
32
Neuromonitoring of the neonatal ECMO patient
Renal System should include frequent neurologic exams and rou-
tine neuroimaging (e.g. cranial ultrasound [CUS]).
Transient renal dysfunction commonly occurs One study found that 93% of ICHs occurred during
in neonates after placement on ECMO. Hypoxia, the first five days of ECMO, so some centers stop
hypotension, and the inflammatory response to the routine CUS screening after this period while others
ECMO circuit probably contribute to this phenom- continue with daily or every other day screening.32
enon. In the majority of cases, renal dysfunction Near-infrared spectroscopy (NIRS) monitoring
spontaneously resolves within the first 48 to 72 may be helpful for early identification of patients
hours of ECMO. Acute kidney injury (AKI) is char- at risk for poor cerebral and systemic oxygen de-
acterized by poor urine output with elevated BUN livery.12,33 Continuous video or amplitude integrated
and creatinine levels. Furosemide, 1 to 2 mg/kg EEG monitoring has also been used increasingly to

PARAMETERS GOAL
Blood flow – initial 120 ml/kg/min (range 80-150 ml/kg/min)
Blood flow – trial off idle 50-100 ml/min
Blood flow – no trial off idle 20 ml/kg/min
Circuit pCO2 40-45 mmHg
Circuit pO2 150-400 mmHg
Calories 100-120 Kcal/kg/day
Fibrinogen > 150 mg/dL
Fluid intake - initial 60-100 ml/kg/day
Hematocrit > 35%, increasing to 40% when weaning off ECMO
iCa > 1.1 mmol/L
Plasma free hemoglobin < 50 mg/dL
Platelets 75,000-100,000
Potassium >3 mEq/L - may need to be increased
Protein 4 gm/kg/day
Conventional rest ventilator Rate: 10-20 bpm
settings (High frequency PIP: 15-20 cm H2O
settings as per center protocol)
PEEP: 10-12 cm H20
FiO2: 21%-40%
Sodium >135 mEq/L - may need to be decreased
SpO2 VV ECMO: > 85%
VA ECMO: 92-99%
Sweep gas FiO2 60-100%
SvO2 >= 65%
Table 19-2. Neonatal respiratory ECMO goals.

171
Chapter 19

monitor for seizures, which occur in 9% of neonatal and antifactor Xa levels to titrate anticoagulation
ECMO cases.5 (Table 19-2). For patients at high risk or with active
Therapeutic hypothermia (core temperature at bleeding, lower anticoagulation levels are targeted.
33-34°C) has been shown to improve outcome for For planned surgeries while on ECMO (e.g. CDH
term and near-term infants with hypoxic ischemic repair), pretreatment with antifibrinolysis agents
encephalopathy and treatment should be continued such as aminocaproic acid may help reduce surgical
in neonates requiring ECMO.14 This can easily be site bleeding.8
done by using the ECMO heater/cooling device or Platelet levels are routinely maintained above
a cooling blanket with an esophageal/rectal servo- 80,000 to 100,000 cells/mm3 in neonates due to their
regulated temperature probe. higher risk of bleeding. Most centers target a mini-
Sedatives and antianxiety drugs, such as mum hematocrit between 35% and 40% depending
morphine, dexmedetomidine and midazolam are on the hemodynamic status and oxygen delivery
commonly used for neonates on ECMO although needs of the patient (Table 19-2). Blood product
clinical guidelines for sedation and analgesia have transfusion has been shown to be associated with
yet to be published.14,34 Prolonged high cumulative increased morbidty so limiting transfusion and do-
doses of opiods and benzodiazepines are associated nor exposure to all patients on ECMO should be the
with adverse outcomes so strategies to minimize goal (see Chapter 5).31,39-41 Additonally, minimizing
their use should be incorporated into a care plan. laboratory sampling as much as possible should also
Although there are theoretic neurologic benefits be attempted.42 Hemolysis occurs more commonly
to routine placement of cephalad jugular cannula, in neonates, likely secondary in part to the use of
including augemented venous return, reduced VV- smaller cannulas and increased circuit resistance.
ECMO recirculation, cerebral venous decompres- In addition to overall patient examination and
sion, and cerebral mixed venous oxygen saturation circuit condition, plasma free hemoglobin levels
monitoring, their use is not universal and benefits can monitor for hemolysis, however threshold for
on outcome remain unproven.14,35 intervention varies by center.
Due to the high risk nature of ECMO and de-
gree of critical illness necessary to meet criteria for ECMO Weaning
treatment, all surviving neonates should receive
long-term followup (see Chapter 26). Neurodevel- The timing of ECMO weaning consitutes a part
opment outcomes in survivors treated with ECMO of daily rounds and is specific for each patient. The
have shown to be equivalent or better compared ongoing risks of continuing ECMO must be weighed
to conventionally supported neonates with similar versus any additional gain achieved with more time
disease severity. Significant neurodevelopmental on ECMO. In general, once adequate lung rest has
impairment has been reported to occur in 15 to been achieved to include significant improvement
25% of ECMO treated neonates; however, severe of the underlying lung disease and pulmonary hy-
or profound impairment is uncommon and seen in pertension, as well as resolution of airleak, ECMO
<5% of survivors.15,32,37,38 Many centers routinely support can be gradually reduced. For VA patients,
perform post-ECMO imaging, as 10 to 15% will adequate cardiac function needs to be present when
have moderate to severe injury noted on head CT ECMO flows are weaned. Additional considerations
or MRI.38 include overall fluid status as decannulation would
coincide with good renal function and at or near
Hematology ideal body weight with minimal edema; however,
some patients may not be able to achieve this goal
Anticoagulation remains a particular and con- and may need ongoing CRRT for residual AKI
stant challenge for the neonate on ECMO and is de- following decannulation. Details on techniques for
tailed in Chapter 4. Most centers use a combination weaning from VA and VV ECMO can be found in
of activated clotting time (ACT), activated partial Chapter 25.
thromboplastin time (APTT), antithrombin activity,

172
 Management of the Neonate on ECMO

Conclusion

ECMO remains a vitally important modality

for improving survival and outcomes in neonates
with severe reversible respiratory disease who fail
conventional medical management. Timely referral
of candidates for ECMO is critical to help prevent
morbidty and mortality in neonates that would po-
tentially benefit from ECMO. Continued research
efforts and collective knowledge sharing between
centers to improve safety and lessen morbity associ-
ated with ECMO, as well as better defining which
target population of neonates will most benefit from
ECMO support, will be important considerations
moving forward.

173
Chapter 19

References 1990-1999. J Pediatr Surg. 2001;36(8):1199-

204.
1. Bartlett RH, Gazzaniga AB, Jefferies MR, 10. Jaksic T, Hull MA, Modi PB, Ching YA, George
Huxtable RF, Haiduc NJ, Fong SW. Extra- D, Compher C. A.S.P.E.N. clinical guidelines:
corporeal membrane oxygenation (ECMO) nutrition support of neonates supported with
cardiopulmonary support in infancy. Trans Am extracorporeal membrane oxygenation. J Par-
Soc Artif lntern Organ. 1976;22:80–88. enter Nutr. 2010;34(3):247-253.
2. Bartlett RH, Andrews AF, Toomaisian JM 11. Hanekamp MN, Spoel M, Sharman-Koend-
Haiduc NJ, Gazzaniga AB. Extracorporeal jibiharie I, Peters JWB, Albers MJIJ, Tibboel
membrane oxygenation (ECMO) for newborn D. Routine enteral nutrition in neonates on
respiratory failure: forty-five cases. Surgery. extracorporeal membrane oxygenation. Pediatr
1982;92(2):425–433. Crit Care Med 2005;6(3):275-279.
3. Bartlett RH, Roloff DW, Cornell RG, Andrews 12. Rais-Bahrami K, Van Meurs KP. Venoarte-
AF, Dillon PW, Zwischenberger JB. Extracor- rial versus venovenous ECMO for neonatal
poreal circulation in neonatal respiratory failure: respiratory failure. Semin Perinatol. 2014
a prospective randomized study. Pediatrics. Mar;38(2):71-77.
1985;76(4):479–487. 13. Short BL. Extracorporeal Membrane Oxygen-
4. UK collaborative randomized trial of neonatal ation. In: Avery G, Mhairi G, Macdonald M, Se-
extracorporeal membrane oxygenation. UK shias M, Mullett M, eds. Avery’s Neonatology
Collaborative ECMO Trial Group. Lancet. 6th ed. Philadelphia, PA: Lippincott Williams
1996;348(9020):75–82. & Wilkins; 2005:622-33.
5. ELSO International Summary, ELSO Interna- 14. Bhombal S, Sheehan AM, Van Meurs KP. Medi-
tional Registry, July 2017. Extracorporeal Life cal management of the neonate with respiratory
Support Organization, Ann Arbor, MI. failure on ECLS. In Brogan TV, Lequier L,
6. Hintz SR, Suttner DM, Sheehan AM, Rhine Lorusso R, MacLaren G, Peek G, eds. Extracor-
WD, Van Meurs KP. Decreased use of neo- poreal Life Support: The ELSO Red Book. 5th
natal extracorporeal membrane oxygenation ed. Ann Arbor, MI: Extracorporeal Life Support
(ECMO): how new treatment modalities Organization; 2017:183-199.
have affected ECMO utilization. Pediatrics. 15. Mugford M, Elbourne D, Field D. Extracor-
2000;106(6):1339-1343. poreal membrane oxygenation for severe re-
7. Roy BJ, Rycus P, Conrad SA, Clark RH. The spiratory failure in newborn infants. Cochrane
changing demographics of neonatal extracorpo- Database Syst Rev. 2008 Jul 16;(3).
real membrane oxygenation patients reported 16. Roberts N, Westrope C, Pooboni SK, et al. Veno-
to the extracorporeal life support organization venous extracorporeal membrane oxygenation
(ELSO) registry. Pediatrics. 2000;106(6):1334- for respiratory failure in inotrope dependent
1338. neonates. ASAIO J. 2003;49(5):568-571.
8. Harting T, Davis CF, Lally KP. Congenital dia- 17. Schiller P, Vikholm P, Hellgren L. Experimental
phragmatic hernia and ECMO. In Brogan TV, venoarterial extracorporeal membrane oxy-
Lequier L, Lorusso R, MacLaren G, Peek G, eds. genation induces left ventricular dysfunction.
Extracorporeal Life Support: The ELSO Red ASAIO J. 2016;62(5):518-24.
Book. 5th ed. Ann Arbor, MI: Extracorporeal 18. Sell LL, Cullen ML, Lerner GR, Whittlesey
Life Support Organization; 2017:133-143. GG, Shanley CJ, Klein MD. Hypertension
9. Dimmitt RA, Moss RL, Rhine WD, Benitz WE, during extracorporeal membrane oxygen-
Henry MC, Van Meurs KP. Venoarterial versus ation: cause effect, and management. Surgery.
venovenous extracorporeal membrane oxygen- 1987;102(4):724-730.
ation in congenital diaphragmatic hernia: the 19. Alapati D, Aghai ZH, Hossain MJ, et al. Lung
extracorporeal life support organization registry, rest during extracorporeal membrane oxygen-
ation for neonatal respiratory failure - practice

174
 Management of the Neonate on ECMO

variations and outcomes. Pediatr Crit Care Med. extracorporeal membrane oxygenation. Crit
2017;18(7):667-674. Care. 2007;11(5):R111.
20. Reiss I, Schaibel T, van den Hout L, et al: Stan- 29. Bulas D, Glass P. Neonatal ECMO: neuroimag-
dardized protocol management of infants with ing and neurodevelopmental outcome. Semin
congenital diaphragmatic hernia in Europe: The Perinatol. 2005;29(1):58-65.
CDH EURO Consortium Consensus. Neonatol- 30. Mok YH, Lee JH, Cheifetz IM. Neonatal ex-
ogy. 2010;98(4):354-64. tracorporeal membrane oxygenation: update
21. Kuluz MA,Smith PBMears SP, et al. Prelimi- on management strategies and long - term out-
nary observations of the use of high frequency comes. Adv Neonatal Care. 2016;16(1):26-36.
jet ventilation as rescue therapy in infants with 31. van Heijst AFJ, de Mol AC, Ijsselstijn J. ECMO
congenital diaphragmatic hernia. J Pediatr Surg. in neonates: neuroimaging findings and out-
2010;45(4):698-702. come. Semin Perinatol. 2014;38(2):104-113.
22. Snoek KG, Capolupo I,van Rosmalen J, et al. 32. de Mol AC, Liem KD, van Heijst AF. Cere-
Conventional mechanical ventilation versus bral aspects of neonatal extracorporeal mem-
high-frequency oscillatory ventilation for brane oxygenation: a review. Neonatology.
congenital diaphragmatic hernia: a random- 2013;104(2):95-103.
ized clinical trial(the VICI-trial). Ann Surg. 33. Clair MP, Rambaud J, Flahault A, et al. Prog-
2016;263(5):867-74. nostic value of cerebral tissue oxygen satura-
23. Anton-Martin P, Thompson MT, Sheeran PD, tion during neonatal extracorporeal membrane
Fischer AC, Taylor D, Thomas JA. Extubation oxygenation. PLoS One. 2017; 12(3): e0172991.
during pediatric extracorporeal membrane oxy- 34. Wagner D, Pasko D, Phillips K, Waldvogel J,
genation: a single-center experience. Pediatr Annich G. In vitro clearance of dexmedetomi-
Crit Care Med. 2014;15(9):861-869. dine in extracorporeal membrane oxygenation.
24. Jenks CL, Tweed J, Gigli KH, Venkataraman R, Perfusion. 2013;28(1):40-6.
Raman L. An international survey on ventilator 35. Roberts J, Keene S, Heard M, McCracken
practices among extracorporeal membrane oxy- C, Gauthier TW. Successful primary use
genation centers. ASAIO J. 2017;63(6)787-792. of VVDL+V ECMO with cephalic drain
25. Bembea MM, Lee R, Masten D, et al. Magni- in neonatal respiratory failure. J Perinatol.
tude of arterial carbon dioxide change at initia- 2016;36(2):126-131.
tion of extracorporeal membrane oxygenation 36. Khambekar K, Nichani S, Luyt DK, et al. De-
support is associated with survival. J Extra velopmental outcome in newborn infants treated
Corpor Technol 2013;45(1):26-32. for acute respiratory failure with extracorporeal
26. Bizzarro MJ, Conrad SA, Kaufman DA, Rycus membrane oxygenation: present experience.
P. Extracorporeal Life Support Organization Arch Dis Child Fetal Neonatal Ed. 2006;91(1):
Task Force on Infections. Infections acquired F21-F25.
during extracorporeal membrane oxygenation 37. McNally H, Bennett CC, Elbourne D, Field DJ;
in neonates, children, and adults. Pediatr Crit UK Collaborative ECMO Trial Group. United
Care Med. 2011;12(3):277-281. Kingdom collaborative randomized trial of
27. Kirk C, Abel EE, Muir J, Dzierba AL. Strate- neonatal extracorporeal membrane oxygen-
gies for medication management. In Brogan TV, ation: follow-up to age 7 years. Pediatrics.
Lequier L, Lorusso R, MacLaren G, Peek G, eds. 2006;117(5):e845-54.
Extracorporeal Life Support: The ELSO Red 38. Short BL, Soghier L. Neonatal Respiratory
Book. 5th ed. Ann Arbor, MI: Extracorporeal Diseases. In Brogan TV, Lequier L, Lorusso
Life Support Organization; 2017:795-808. R, MacLaren G, Peek G, eds. Extracorporeal
28. van der Vorst MM, den Hartigh J, Wildschut Life Support: The ELSO Red Book. 5th ed.
E, Tibboel D, Burggraaf J. An exploratory Ann Arbor, MI: Extracorporeal Life Support
study with an adaptive continuous intravenous Organization; 2017:123-129.
furosemide regimen in neonates treated with

175
Chapter 19

39. Smith A, Hardison D, Bridges B, Pietsch J. Red

blood cell transfusion volume and
mortality among patients receiving extracor-
poreal membrane oxygenation. Perfusion.
2013;28(1):54-60.
40. Winkler AM. Transfusion management during
extracorporeal support. In Brogan TV, Lequier
L, Lorusso R, MacLaren G, Peek G, eds. Ex-
tracorporeal Life Support: The ELSO Red
Book. 5th ed. Ann Arbor, MI: Extracorporeal
Life Support Organization; 2017:105-122.
41. Sawyer AA, Wise L, Ghosh S, Bhatia J, Stans-
field BK. Comparison of transfusion thresholds
during neonatal extracorporeal membrane oxy-
genation. Transfusion. 2017 Sept;57(9):2115-
2120.
42. Dalton H, Reeder R, Garcia-Filion P, et al. Fac-
tors associated with bleeding and thrombosis in
children receiving extracorporeal membrane
oxygenation (ECMO).Am J Respir Crit Care
Med. 2017;196(6):762-782.

176
20

Pediatric Respiratory ECLS: Patient Management

Jana ASSY, MD, Micheal L. Heard, RN, FELSO

Introduction Patient Management

Despite the advent of newer therapies for pediat- While most aspects of patient management may
ric respiratory failure, ECMO remains an important have been already covered in other chapters, a few
therapy when conventional management fails or deserve special consideration in this unique and
becomes dangerous. In selected cases, it can serve challenging population.
as a treatment to prevent ventilator associated lung
injury, or a bridge to lung transplantation. In the lat- Mode of Support
est report from ELSO, 8287 pediatric patients with
respiratory indications for ECMO were supported Venovenous (VV) ECMO is the most common
with ECMO with an overall survival rate of 67%. configuration for respiratory failure in pediatric
Pediatric respiratory diseases that are com- patients. Double lumen cannulae have been estab-
monly considered for ECMO therapy are listed in lished as an effective and safe device for VV-ECMO
Table 20-1, and more detail regarding these diseases in neonates and children. Newer cannulae have
is described in Chapter 7. improved flow profiles, decreasing recirculation
While there are no absolute contraindications with the use of a single jugular vein cannulation
for respiratory ECMO, consideration must be given site.1 The use of these cannulas can be associated
for the duration of mechanical ventilation with high with less sedation and early patient mobilization,
levels of support (after 14 days there appears to be which enhances lung recovery. Echocardiography
a decline in survival) but also should be individual- or fluoroscopy is necessary for the initial placement
ized.1 of the cannula, for repositioning, and to manage
complications caused by cannula malposition. 2
Percutaneous cannulation of the RIJV and femoral

TOTAL AVG RUN LONGEST SURVIVED % SURVIVED

RUNS TIME RUN TIME
Viral pneumonia 500 298 1,680 365 73%
Bacterial pneumonia 204 302 4,286 133 65%
Pneumocystis 4 483 671 3 75%
pneumonia
Aspiration pneumonia 65 210 1,932 43 66%
ARDS, postop/trauma 29 256 859 20 68%
ARDS, not 149 357 3,086 92 61%
postop/trauma
Acute resp failure, 495 303 7,503 308 62%
non-ARDS
Other 1,370 261 2,699 781 57%
Table 20-1. Pediatric Respiratory ECMO runs by diagnosis from 2012
(from the International ELSO Registry, July 2017, Ann Arbor, MI)

177
Chapter 20

vein, under ultrasound guidance, is preferred over In the absence of acknowledged ideal PEEP lev-
surgical method, even in small children (<10 kg). els in children, echocardiographic assessment can
The double lumen cannula is available in mul- aid in setting PEEP levels, avoiding RV dysfunction
tiple sizes and can support patients from neonatal and decreased venous return. Close hemodynamic
to adult sizes. In larger children, it remains the monitoring is essential while choosing the adequate
recommended option, but VV-ECMO may also be PEEP level, in order to assess the impact on the RV
performed via two separate cannulas inserted in two and to anticipate hemodynamic impairments trans-
separate sites, usually the RIJV and the femoral vein. lating into lower oxygen deliver (i.e. NIRS values,
In these cases, however, the recommended configu- higher lactates and low SvO2).
ration in children is drainage from the femoral vein Lung improvement must be carefully and re-
and return via the superior vena cava/right atrium, peatedly assessed using:
dependent on patient and vessel size. Use of the
jugular-femoral configuration optimizes ECMO • Chest X-ray; and chest CT if necessary
pump flow when the femoral vessels are small (i.e. • Lung compliance
younger children) but at the expense of higher risk • Changes in tidal volumes
of recirculation. Regardless of configuration, in • Changes in lung volumes on ultrasound
cases of suboptimal venous drainage, an additional Tracking the improvement in lung function
venous cannula may increase drainage. helps to decrease sedation and encourage sponta-
Conversion to VA-ECMO is recommended in neous breathing. With decreased sedation, assisted
cases of associated cardiovascular compromise or ventilation should replace controlled modes as early
when hypoxia causes acidosis and associated tissue as possible, to promote muscle and lung recovery,
damage. When considering VA-ECMO, pump flow and shorten ECMO duration. The combination of
should be carefully managed in order to 1) allow ECMO and neuronally adjusted ventilator assist
some blood flow into the native pulmonary circula- (NAVA) has been reported to improve patient-venti-
tion, avoiding ischemia due to lung hypoperfusion, lator synchrony and reduce the work of breathing.5,6
and 2) minimize increased left ventricular afterload Extubating patients on ECMO may avoid the
that can impair left heart ejection. complications of prolonged sedation, and posi-
tive pressure ventilation.6 However, this practice
Ventilation requires a dedicated and experienced team with
established protocols and adequate expertise in
Once VV-ECMO settings are optimized, pro- noninvasive ventilation methods (please see chapter
tective lung ventilation is implemented. Many 27) . Again, if extubation leads to complete lung
uncertainties remain regarding pediatric ARDS ven- collapse the patient may experience increased PVR
tilation strategies, with identified guidelines based and RV dysfunction.
primarily on adult data.3 During the acute phase of
the disease, many centers use pressure-controlled Hypoxia
ventilation with low peak pressure (<28 cmH20),
high PEEP levels (10-15 cm H2O), and low FiO2 Low oxygen saturations (as low as 75 to 80%)
(<50%). Despite these ELSO recommendations, can be tolerated when adequate tissue oxygenation
clinical practice is moving toward lower settings.4 is maintained. Monitoring of lactate levels, urine
It seems more pertinent to target the mean airway output, peripheral perfusion, and NIRS should be
pressure and driving pressure rather than peak inspi- carefully observed. Measurement of SvO2 is not a
ratory pressure. Additionally, once ECMO support reliable indicator of oxygenation on VV-ECMO
has been achieved, ventilator settings should not due to recirculation, but is useful in VA-ECMO
be lowered to rest settings too rapidly as this may and should be maintained at 65-75%. In VV-ECMO
result in rapid lung volume loss, sudden rises in it is very important to manage recirculation (see
pulmonary vascular resistance, and possible acute Chapter 13). Cannula(e) placement should be as-
right ventricular (RV) dysfunction.

178
 Pediatric Respiratory ECLS: Patient Management

sessed carefully to assure appropriate arterial flow receive diuretics on ECMO but may benefit from
toward the tricuspid valve. Optimizing oxygenation renal replacement therapy (RRT, hemofiltration,
by red blood cell (RBC) transfusion if the patient hemodialysis, or peritoneal dialysis). No agree-
has low hemoglobin level, increasing pump flow ment concerning the routine use of RRT during
(or decreasing in the case of recirculation), or in- ECMO currently exists. However, this technique
creasing drainage, either through a larger cannula or can improve and accelerate fluid removal, support
adding another venous catheter to allow additional dysfunctional kidneys, and allow adequate caloric
pump flow, may all help. Additionally, the oxygen intake in children on VV-ECMO.7 Whenever pos-
extraction ratio, (DO2/VO2), should be maintained sible, RRT should be performed through the ECMO
at 3:1 or greater and this may be added by targeted circuit to avoid the use of an additional large bore
temperature management, sedation, and occasion- catheter. More information detailing the use of renal
ally neuromuscular blockade (NMB). VA-ECMO replacement therapies can be found in Chapter 28.
may be considered when tissue oxygenation is Children require appropriate nutrition (enteral
insufficient and additional drainage is not feasible. or parenteral) for adequate healing and recovery.
Enteral feeding via a feeding tube has been shown
Cardiac Output to be safe and effective in pediatrics. Oral nutrition
may be easily delivered in extubated and awake
VV-ECMO primarily supports respiratory func- patients. Caloric goals should be set daily and must
tion. Consequently, maintaining a normal cardiac include critical illness needs. Achieving nutritional
output is required to provide adequate oxygenation. goals may be facilitated by placing a patient on
ECMO initiation usually leads to a significant RRT to help manage the fluid required to deliver
improvement in cardiac function due to improved the nutrition.
oxygenation of blood reaching the left ventricle,
increased pulmonary blood flow due RV afterload Sedation
(decreased PVR) in response to improved pulmo-
nary arterial oxygenation, and decreased mechanical Sedation with muscle paralysis can be used
ventilator settings. If necessary, inotropic support during the initial critical phase of the disease, but
(milrinone, levosimendan, catecholamines) may NMB should be discontinued as soon as possible in
help optimize cardiac output. Also RV function order to avoid muscle deconditioning and enhance
should be closely monitored since dysfunction may spontaneous breathing. Spontaneous breathing also
persist for several days. improves regional ventilation.
Most patients with severe respiratory failure Benzodiazepines and opioids are most com-
require inotropic support before ECMO initiation. monly used and can be associated with tachyphy-
When high doses of inotropes are used, conversion laxis, as well as delirium and withdrawal syndrome.
to VA-ECMO may be indicated, in order to avoid Dexmedetomidine may reduce the need for these
adverse effects of catecholamine doses and to op- medications and minimize their adverse effects,
timize myocardial oxygen consumption. Nursing while assuring patient comfortable.
assessment of cardiac function should include aus- While it is preferable to awaken patients and
cultation and telemetry. VA-ECMO patients may encourage movement for myriad benefits, certain
have muffled heart sounds due to high pump flow pediatric patients may not tolerate the activity or
and little to no pulse pressure. the cannula position may be compromised. The
RESTORE team has developed a nurse-led, goal-
Fluid Management and Nutrition directed patient sedation management protocol,7 that
helps reduce sedative exposure without significant
Children receiving ECMO often have signifi- clinical discomfort.
cant fluid overload due to high fluid resuscitation Awake patients require age-appropriate stimula-
prior to and during ECMO. Fluid overload may ex- tion and activities. Infants enjoy music or mobiles,
acerbate the underlying lung injury. Patients usually while older children appreciate games, books, or

179
Chapter 20

electronics. Child Life specialists can provide the Anticoagulation

resources to determine the best activities for patients.
Additionally, physical therapy and occupational Unfractionated heparin (UFH), which is widely
therapists assure proper body alignment and provide available and easily reversed, remains the most com-
conditioning exercises to promote efficient recovery. monly used anticoagulant on ECMO. Laboratory
Certain ECMO patients may sit up, stand or even monitoring of anticoagulation varies widely be-
walk (see Chapter 27). The right conditions may tween centers (activated clotting time (ACT), anti-
even allow a child to ride a tricycle! The nurse must Xa, PTT, thromboelastography). ACT is performed
promote a safe environment where each pediatric routinely in many ECMO centers; however, it does
patient can participate in activities that are age ap- not exclusively represent heparin effect. APTT is
propriate and fun. poorly reliable in critically ill children and subject to
Maintaining normal sleep-wake cycles for pa- inter and intrapatient variability due to the influence
tients is of paramount importance. Intensive care of coagulation factors and antithrombin levels and
units are noisy and lights may be on 24 hours a day.8 the presence of a lupus inhibitor. The anti-Xa assay
Patients who are sedated also benefit from quiet reflects best UFH effect.14 Multimodal anticoagula-
time. Although the ECMO bedside may be one of tion monitoring associating ACT with anti Xa activ-
the busiest in the unit, caregivers should maintain ity is one of the most recommended protocols for
an unobtrusive demeanor in the patient room. En- anticoagulation monitoring in children on ECMO.
suring that ‘quiet time’ is observed in the afternoon Thromboelastography and thromboelastometry
and an appropriate ‘lights out for nighttime’ ritual evaluate clot strength and fibrinolysis and may be of
contribute to decreased delirium, sleep deprivation, a particular help in challenging cases of bleeding or
and hallucinations while encouraging healing.9 excessive clotting. Antithrombin (AT) levels should
Deeply sedated patients require frequent as- be followed and deficiency should be suspected
sessment of the eyes. Critically ill patients who are in case of heparin resistance. In case of decreased
mechanically ventilated, sedated, or unconscious, levels, AT supplementation may optimize heparin
who have decreased tear production, and reduced activity but may place the patient at increased risk
or absent blink reflexes are at high risk for exposure of bleeding.15,16
keratopathy or ‘dry eye’, and other ophthalmic com- The goal of anticoagulation is to balance ECMO
plications.10 Nurses should perform a focused eye circuit fluidity and patient hemostasis. Bleeding
assessment every shift to evaluate for eyelid swell- and thrombosis causing need for intervention are
ing, conjunctiva redness, corneal hazing, ability to the most common complications listed it the ELSO
maintain eye closure, and discharge or crusting on Registry. The testing regimens need to be tailored
the eyelid. These represent early signs of dry eye specifically and repeatedly during the ECMO run.
that can lead to corneal ulceration or infection.11,12 The use of direct thrombin inhibitors such as bivali-
The nurse should use normal saline soaked gauze rudin or argatroban, in ECLS patients (see Chapter
to clean from the inner to the outer canthus every 8 4) has gained some attention.
hours. Any sign of abnormality should be reported
to the physician team and an ophthalmologist con- Associated Measures
sulted.
The provision of eye care is part of the routine Prone Positioning and Bronchoscopy
care provided to all ICU patients. Standard practice
includes instillation of a lubricant every two hours. Early prone positioning improves oxygen-
Patients who have an inability to maintain eyelid ation, decreases ventilation-perfusion mismatch
closure should use properly installed polyethylene and pulmonary artery pressure, and minimizes
covers which are more effective at providing a cardiopulmonary interactions. Prone positioning
barrier against tear evaporation and exposure to air during ECMO is feasible, has not been associated
currents, thus decreasing complications.13 with significant adverse events, and may decrease
mortality.17 However, cannulae may become mal-

180
 Pediatric Respiratory ECLS: Patient Management

positioned with proning. Some awake children do important role in the management of ECMO patients,
not tolerate this position well. Turning a patient especially with concomitant dual cerebral and so-
requires multiple staff and careful positioning of matic sensors.19,20 The sensors immediately capture
the child with positioners and comfort measures to decreased oxygenation or perfusion from the brain
assure cannulas do not become dislodged. Prone and the kidney. The isolated decrease in cerebral
positioning may improve pulmonary compliance, oxygenation reliably indicates decreased cardiac
minimizing cardiopulmonary interactions, and im- output/ECMO flow, decreased arterial saturation,
proving cardiac function. Chest physiotherapy and or a decrease in PaCO2 with subsequent cerebral
percussive ventilation in the prone position may be vasoconstriction.
used to enhance recovery and recruitment.18 Ultrasound is routinely used for cardiac, lung,
Flexible bronchoscopy is safe in anticoagulated and brain monitoring and diagnosis, and to assess
patients and helps in both diagnosis and treatment cannula position, drainage, and intracardiac vol-
of lung disease. Additional support from the ECMO ume. Chest computed tomography (CT), electrical
pump may be required, especially on VV-ECMO. impedance tomography, transpulmonary pressure
Again, careful positioning of the head and neck dur- measurement, and diaphragmatic electrical activ-
ing the procedure is required to assure the cannulas ity measured with NAVA may be useful to monitor
remain in proper position. pulmonary compliance and assess lung recovery.

Blood Products Nursing Care

Transfusion of blood products often occurs sec- Skin Care

ondary to bleeding, hemolysis, and frequent blood
sampling (see Chapter 5). RBC transfusion is used Skin care is considered a tenet of nursing care.
to increase oxygen content, and thus improve oxy- ECMO patients are at risk for pressure ulcers due to
gen delivery. The ideal hemoglobin level on ECMO the severity of their illness, the use of anticoagulant
remains a matter of debate. Most centers maintain therapy, and immobility. The head and neck are
hemoglobin of 8-12 g/dl. Fresh frozen plasma (FFP)frequently edematous due to the lack of mobility
may be administered to maintain a fibrinogen level,
(i.e. NMBs). Large cannulas sutured in place, en-
or to correct coagulopathy after large transfusions
dotracheal and oronasal tubes and other intensive
of RBCs. Platelets are given to replace losses due to
care devices increase the risk for pressure ulcers.21
clot formation within the ECMO circuit and patient.
In 2007, the National Pressure Ulcer Advisory Panel
Additional blood products such as albumin may be made the prevention and treatment of pressure ulcers
given for volume needs. Antithrombin and factor in infants and children a key priority.22 In 2015, the
replacements may be given as indicated by labora- Joint Commission listed prevention of health care
tory results. Many ECMO programs use standard- associated pressure ulcers a National Patient Safety
ized protocols for treatment of lab results. In most
Goal.23 The use of the Pediatric Braden Q Scale for
pediatric ECMO programs, the circuit has ample pressure ulcer staging can assist the nurse in the
infusion sites. Packed RBCs, FFP, and albumin can early assessment of pressure ulcers, providing ap-
be administered directly into the ECMO circuit by propriate and timely interventions. Basic nursing
experienced staff. Platelet transfusion may be given
interventions that may lower the risk of develop-
directly into the circuit but it is not recommended
ment of pressure ulcers include avoiding injury
pre-oxygenator because of the risk of clotting. due to shear forces, turning the patient every two
hours, use of positioning aids, and changing dirty
Monitoring or wet linen accordingly. Pediatric patients should
have soiled diapers changed as early as possible.
Multimodal monitoring allows rapid recogni- Avoiding the use of multiple layers and plastic lined
tion and management of complications (see Chapter protective barriers is recommended to assist in the
18). Near-infrared spectroscopy (NIRS) plays an flow of oxygen to prevent skin ulcers. Additionally,

181
Chapter 20

difficulty in cannula positioning may render regular be coated with dental wax. Finally, lip care should
patient turning challenging. In these cases, the nurse not include lubricants such as petroleum jelly as
should use fluidized positioners for extremities, they can increase the dryness of tissues. Preferably
head, and shoulders. The careful manipulation of use water based or aloe based lip balms.26
these positioners every 2 hours helps reposition the
patient so that pressure areas are rotated. Finally, the Wound Care
ECMO patient may be placed on a specialty mat-
tress or overlay such as an egg crate, air-filled bed The pediatric patient on ECMO experiences
or action bed that reduces pressure.24 particular risk for wounds. Most wounds are associ-
ated with indwelling devices, often placed prior to
Oral Care ECMO cannulation. As such, they may have had
time to heal. Yet, all breaks in the skin must be con-
Intensive care patients are at risk for poor oral sidered wounds that carry the risk of hemorrhage or
health and the ECMO patient has the additional risk infection. Additionally, the insertion of the ECMO
of anticoagulant therapy. Pediatric patients present cannula(e) is undertaken just before or immediately
several different oral health dilemmas including after anticoagulation therapy has begun. These sites
the presence of braces or other orthodontic devices are at particular risk for bleeding and for nosocomial
and natural tooth loss. The nurse must perform a infection. Most ECMO centers have an institutional
comprehensive oral assessment and provide mouth protocol for dressing of intravascular devices as part
care as prescribed. Frequency of mouth care varies, of a Central Line-Associated Bloodstream Infection
but is recommended hourly for high risk patients, policy (CLABSI). The protocol may include types
(e.g. heavily sedated or those receiving oxygen). of catheters, assessment tools, frequency and types
Absence of care for 2 to 6 hours can significantly of dressing changes, and prevention techniques
reduce the benefits of any oral intervention previ- that may be used. Maintaining the same protocol is
ously carried out.25 Mouth care may make use of oral important as this decreases the likelihood of compro-
rinses. Many mouthwashes contain alcohol which mising an established CLABSI program; however,
can cause a burning sensation and may increase careful review of the protocol by the ECMO team
existing inflammation, and hence are not recom- is important, so that it may be appropriately adapted
mended. Chlorhexidine gluconate (CHG) has been to the needs of the ECMO patient. Rigid adherence
shown to be a very effective antibacterial mouth- to the protocol may not allow for atypical situations
wash, effectively removing dental plaque where such as this with excessive hemorrhage.
mechanical tooth brushing cannot be performed. Nurses should routinely assess all cannulation
Saline is a nonirritant mouthwash and can be use- sites, including type and integrity of dressing, drain-
ful in situations where other rinses are unavailable, age or bleeding, catheter position, stability, and
especially since the frequency of mouth care is security. Additionally, the skin around the cannula
paramount. Toothbrushes can effectively remove is assessed for redness, swelling, and breakdown.
plaque when used correctly. When appropriate a Cleaning the cannula and incision site is best done
patient should brush their own teeth since nurses utilizing CHG swabs or pads. Other cleaning agents
may have difficulty in assessing how hard to brush. include betadine and sterile saline wipes. The types
ECMO patients receiving anticoagulation are at risk of dressing used on ECMO cannulas vary greatly,
for oral bleeding so firm bristled brushes should be but are typically thought of as dry dressings. They
avoided. Foam swabs are useful for cleaning the may be simple gauze or gauze pads. Transparent
oral cavity but only when soaked in CHG. Other film dressings have a thin layer of plastic that covers
tools, such as metal instruments or cotton swabbed a wound area, creating a barrier. They allow some
fingers should be avoided on ECMO patients. If the oxygen exchange to reduce bacteria growth. These
patient has orthodontic devices in place, determine dressings are best for dry, nonexudative sites. Re-
if the device is removable as that is preferable to moval of transparent dressings can tear the underly-
leaving it in place. Otherwise, the inner lips should ing skin, so caution must be exercised. Once a site

182
 Pediatric Respiratory ECLS: Patient Management

begins to bleed or drain, the use of these dressings

is contraindicated due to the frequent removal that
may result in skin tears. Stable bloody drainage is
preferentially left alone until the dressing is required
to be changed. This allows the clot to strengthen and
prevent further blood loss. Foam dressings have an
adsorptive and protective effect for at risk sites or
pressure ulcers present. They are self-adherent and
easily cut to fit for specific difficult to dress sites.
Tape may be used to secure dressings in place. Avail-
able in paper, cloth, or plastic, selecting a type that
best fits the need of the dressing, without putting
the skin under unnecessary risk, is important. Many
patients also have sensitivities to tape which must
be taken into consideration. Cannula sites, which
are typically in the neck or groin, can be difficult
to dress. Pediatric patients vary in size and move-
ment, which must be taken into consideration when
choosing a dressing.27

183
Chapter 20

References 10. Xie H, Kang J, Mills GH. Clinical review: The

impact of noise on patients’ sleep and the ef-
1. Maslach-Hubbard A, Bratton SL. Extracor- fectiveness of noise reduction strategies in in-
poreal membrane oxygenation for pediatric tensive care units. Critical Care 2009;13(2):208.
respiratory failure: History, development and 11. Marshall AP, Elliott R, Rolls K, Schacht
current status. World J Crit Care Med 2013; S, Boyle M. Eyecare in the critically ill:
2(4):29-39. clinical practice guideline. Australian Critical
2. Javidfar J, Wang D, Zwischenberger JB, et al. Care;21(2):97-109.
Insertion of bicaval dual lumen extracorporeal 12. Hernandez EV, Mannis MJ. Superficial kera-
membrane oxygenation catheter with image topathy in intensive care patients. American
guidance. ASAIO J 2011; 57(5):203-205. Journal of Ophthalmology 1997;124(2):212-
3. Gattinoni L., Tonetti T., and Quintel M. How 216.
best to set the ventilator on extracorporeal 13. Joyce N. Eye care for the intensive care patients.
membrane lung oxygenation. Curr Opin Crit A systematic review. The Joanna Briggs Insti-
Care 2017; 23(1):66-72. tute for Evidence Based Nursing and Midwifery,
4. MacLaren G, Coombes A, Bartlett RH. Contem- Systematic Review. 2002; 21.
porary extracorporeal membrane oxygenation 14. Shan H, Min D. Prevention of exposure kera-
for adult respiratory failure: life support in the topathy in intensive care unit. International
new era. Intensive Care Med 2012; 38(2):210- Journal of Ophthalmology 2010: 3(4):346-348.
220. 15. Wong TE, Nguyen T, Shah SS, Brogan TV,
5. Mauri T, Grasselli G, Suriano G et al. Control Witmer CM. Antithrombin concentrate use in
of respiratory drive and effort in extracorporeal pediatric extracorporeal membrane oxygen-
membrane oxygenation patients recovering ation: A multicenter Cohort stud. Pediatr Crit
from severe acute respiratory distress syndrome. Care Med 2016;17(12):1170-1178.
Anesthesiology 2016;125(1):159-167. 16. Zabrocki LA, Brogan TV, Statler KD, Poss
6. Karagiannids C, Lubnow M, Philipp A et al. WD, Rollins MD, Bratton SL. Extracorporeal
Autoregulation of ventilation with neutrally membrane oxygenation for pediatric respiratory
adjusted ventilator assist on extracorporeal lung failure: Predictors of mortality. Crit Care Med.
support. Intensive Care Med 2010;36(12):2038- 2011; 39(2): 364-370.
44. 17. Kimmoun A, Roche S, Bridey C, et al. Pro-
7. Schneider JB, Sweberg T, Asaro LA,et al ; longed prone positioning under VV ECMO is
Randomized Evaluation of Sedation Titration safe and improves oxygenation and respiratory
for Respiratory Failure (RESTORE) Study compliance. Ann Intensive Care 2015; 5(1):35.
Investigators. Sedation management in chil- 18. Yeha N, Dominick CL, Connely JT, et al. High
dren supported on extracorporeal membrane frequency percussive ventilation and bron-
oxygenation for acute respiratory failure. Crit choscopy during extracorporeal life support in
Care Med 2017;45(10):e1001-1010. Children. ASAIO J 2014; 60(4):424-428.
8. Tian F, Jenks C, Potter D, Miles D, Raman L. 19. Ghanayem NS, Wernovsky G, Hoffman GM.
Regional cerebral abnormalities measured by Near-infrared spectroscopy as a hemodynamic
frequency-domain Infrared spectroscopy during monitor in critical illness. Pediatr Crit Care Med.
extracorporeal membrane oxygenation. ASAIO 2011;12(4):S27-32.
J 2017; 63(5):252-e59. 20. Reyrson L, Lequier L. Anticoagulation manage-
9. AL-Samsan RH, Cullen P. Clinical investiga- ment and monitoring during pediatric extracor-
tions: Sleep and adverse environmental factors poreal life support: A review of current issues.
in sedated mechanically ventilated pediatric Front Pediatr 2016;4:67.
intensive care patients. Pediatric Critical Care 21. Schindler CA, Mikhailov TA, Kuhn EM, et al.
2005;6(5):562–567. Protecting fragile skin: Nursing interventions
to decrease development of pressure ulcers in

184
 Pediatric Respiratory ECLS: Patient Management

pediatric intensive care. American Journal of

Critical Care 2011;20(1):26–34.
22. Baharestani MN, Ratliff CR. Pressure ulcers
in neonates and children: An NPUAP white
paper. Advances in Skin and Wound Care
2007;20(4):208–220.
23. National Patient Safety Goals Effective
January 1, 2015. The Joint Commission. Avail-
able at: http://www.jointcommission.org/as-
sets/1/6/2015_npsg_lt2.pdf. Accessed January
18, 2016.
24. Butler CT. Pediatric skin care: Guidelines for
assessment, prevention, and treatment. Pediatric
Nursing 2006;32(5):443–450.
25. Murphy A. Clinical review - Oral hygiene a
priority of care. Available at: https://www.inmo.
ie/article/printarticle/2764. Accessed January
18, 2016.
26. DeKeyser-Ganz F, Fink NF, Asher O, Bruttin M,
Nun MB. ICU nurses’ oral-care practices and
the current best evidence. Journal of Nursing
Scholarship 2009;41(2):132–138.
27. Hess CT. Clinical Guide: Skin and Wound Care.
6th ed. Ambler, PA: Lippincott, Williams &
Wilkins; 2008.

185
21

Patient Management: Neonatal and Pediatric Cardiac ECLS

Linda Edwards, MB ChB, MRCPCH (UK), Mark Todd, HBSc, RRT

Introduction Indictions, Contraindications, and Cannulation

Extracorporeal Life Support (ECLS) is used The indications and contraindications for car-
increasingly for patients with cardiac failure unre- diac ECLS are reviewed in Tables 21-1 and 21-2,
sponsive to maximal conventional medical therapy. respectively. Approaches to cannulation ultimately
In 2016 the Extracorporeal Life Support Organiza- rest on the capability of the team and the experi-
tion (ELSO) reported survival to discharge in neo- ence of the available cannulating surgeon. Cannula
natal and pediatric patients as 51% for all cardiac selection is also based on the flow requirements
diagnoses combined. Survival varies according and size of the patient. Use the largest diameter
to underlying diagnosis as well as the availabil- and shortest length cannula for each patient. ECLS
ity and provision of other longer term mechanical centers should have an available size/flow chart and
devices (ventricular assist/Berlin Heart) or organ variety of available cannulas for use. (Figure 21-1)
transplantation. These resources vary with ECLS
centers and country of practice. A working diagno- Central (Transthoracic) Approach
sis is imperative in order to guide decision making
related to the suitability for ECLS. Venoarterial The advantages of this approach include:
ECLS (VA-ECLS) is the support mode of choice 1) quick and direct access in the immediate postop-
for cardiac patients and is covered in this chapter. erative period after sternotomy, 2) right atrial (RA)
Indications for Cardiac ECLS
and aorta sites allow optimal venous drainage and
Preoperative
arterial return, providing high flows when necessary
Surgically correctable lesions and cannot be medically (single ventricle physiology). The disadvantages
stabilized prior to surgery/cardiac catheterization are: 1) increased risk of infection and bleeding,1
Bridge to cardiac transplantation if endstage cardiac failure and
transplant candidate
Postoperative
Failure to separate from cardiopulmonary bypass following
corrective or palliative cardiac surgery
Contraindications for Cardiac ECLS
Refractory low cardiac output following corrective or Cardiac condition is irreversible and not
palliative cardiac surgery transplant candidate
Low cardiac output following resuscitation from Uncontrolled bleeding and/or
cardiopulmonary arrest contraindications to using anticoagulation
Nonsurgical Patient too small or premature for adequate
Refractory low cardiac output secondary to reversible cause vessel cannulation
(myocarditis/dysrhythmia) Brain death or significant irreversible
Refractory pulmonary hypertension
endstage neurological impairment
Refractory hypoxia/respiratory failure exacerbated by
congenital heart disease
Futility (endstage malignancy/fatal
Cardiac Arrest diagnosis)
Cardiac arrest without return of spontaneous circulation Family directives limit further intensive
(ECPR) therapy
Table 21-1. Indications for Cardiac ECLS. Table 21-2. Contraindications for Cardiac ECLS.

187
Chapter 21

2) increased tissue fatigue in longer ECLS runs (>1 4) increased risk of brain emboli and/or impaired
week) so may require conversion to peripheral can- cerebral blood flow with neck vessel cannulation.1
nulation, 3) more interruptions in cardiac massage
during cardiac arrest, which may impact outcomes,2 Daily Management of the Cardiac ECLS Patient
4) inadequate SVC decompression in single ven-
tricle physiology.3 A clear understanding of the goals of ECLS
for each patient can guide the practitioner in daily
Peripheral Approach management and the expected duration of mechani-
cal support. Management strategies mostly depend
Peripheral cannulas can be placed by either sur- on the CICU or PICU where the patient receives
gical cutdown or percutaneously accessing neck and/ care but general ECLS management considerations
or groin vessels in older children. The advantages of exist. A thorough understanding of the underlying
this approach include: 1) often less bleeding around cardiac anatomy must also guide care. Most deci-
the cannulation site, 2) side grafts can be used in sions in cardiac ECLS patients involve a balance of
the elective setting to reduce the incidence of distal risk and benefit. Use of a daily order set/parameters
ischemia,4 3) often reduced sedation requirements, for the patient is a useful bedside tool for the ECLS
and 4) reduced infection risk. The disadvantages Specialist. (Figure 21-2) Many pediatric intensive
are: 1) higher resistance to blood flow in periph- care units use checklists to standardize care plans
eral cannulas due to the smaller radius and greater for their patients. (Figure 21-3)
length, 2) SVC drainage may provide only partial
support in bidirectional cavopulmonary shunt, 3)
limb ischemia can complicate femoral cannulation, *G120*
The Labatt Family
Heart Centre
Critical Care Unit
Daily ECMO Goals & Guidelines
ECLS CANNULATION
Weight _________ kg Height _________ cm BSA _________ m2
NECK/FEMORAL
Maintain parameters and notify physician if unable to achieve goals.
Arterial Use target guidelines when none specified. c VA c VV c Other________
Cannula Size Weight Max. Blood Flow For all additional details please refer to ECMO Policy & Procedures.
Cannulation Cart 1:
COMPLETE ON INITIATION & DAILY DURING ECMO ROUNDS
Biomedicus - Pediatric 8 Fr up to 3 kg 600 ml/min Date / Time
Biomedicus - Pediatric 10 Fr 3 - 5 kg 800 ml/min ECMO Day (0, 1,…)
Biomedicus - Pediatric 12 Fr 5 - 8 kg 1000 ml/min Systolic BP mmHg
Biomedicus - Pediatric 14 Fr 8 - 15 kg 1500 ml/min MAP mmHg

Cannulation Cart 2: CVP mmHg Default Target Guidelines*

Biomedicus - Femoral 15 Fr 15 - 25 kg 2500 ml/min Temp ºC 35.5 – 37.0 °C

Patient pH 7.35 – 7.45
Biomedicus - Femoral 17 Fr 25 - 35 kg 3500 ml/min
Patient paCO2 mmHg 35 – 45 mmHg
Biomedicus - Femoral 19 Fr 35 - 50 kg 4500 ml/min Patient paO2 mmHg > 100 mmHg
Biomedicus - Femoral 21 Fr 50 - 60 kg 5500 ml/min SvO2 > % > 65%
Biomedicus - Femoral 23 Fr > 60 kg 6000 ml/min Hematocrit % 32 – 37% or *42 – 45%
Venous Platelet count mm3 > 80,000 – 120,000 mm3

Cannulation Cart 1: Fibrinogen g/L > 1.5 – 4.0 g/L

Biomedicus - Pediatric 8 Fr up to 3 kg 600 ml/min Antithrombin units/ml 0.6 – 1.2 units/ml

Biomedicus - Pediatric 10 Fr 3 - 5 kg 800 ml/min Standard Heparin units/ml 0.4 – 0.6 units/ml
ACT sec 180 – 220 sec
Biomedicus - Pediatric 12 Fr 5 - 7 kg 1000 ml/min
Pump Flow ml/kg/min
Biomedicus - Pediatric 14 Fr 7 -10 kg 1500 ml/min *Single ventricle physiology requires specialized goals (higher flows, Hct, etc) depending on the stage of palliation.
Cannulation Cart 2: Heparin Infusion: NEVER less than 20 IU/kg/h unless specific physician order prescribed on separate order sheet.
Biomedicus-Fem-Arterial * 15 Fr 10 - 15 kg 2000 ml/min
Biomedicus-Fem-Arterial * 17 Fr 15 - 25 kg 2500 ml/min FLUID BALANCE GOALS & PHYSICIAN NOTIFICATION PARAMETERS – Notify physician if:

Biomedicus-Fem-Arterial * 19 Fr 25 - 30 kg 3500 ml/min

Blood loss > ml/kg/h
Urine output < ml/kg/h
Biomedicus-Fem-Arterial * 21 Fr 30 - 35 kg 4500 ml/min Albumin < 20 or > 40 g/L
Biomedicus-Fem-Arterial * 23 Fr > 35 kg 5500 ml/min Pressure < –30 mmHg if ¼" circuit or –60 mmHg if 3/8" circuit

* These arterial cannulae should be used in the venous position Fluid Balance: even or
neg.
Venous - Femoral SCUF net rate ml/h

Biomedicus-Fem-Venous 15 Fr 20 - 30 kg 2000 ml/min CVVH net rate ml/h

Biomedicus-Fem-Venous 17 Fr 30 - 35 kg 2800 ml/min Physician Initials

Biomedicus-Fem-Venous 19 Fr 35 - 40 kg 3500 ml/min Consultations:

Biomedicus-Fem-Venous 21 Fr 40 - 45 kg 4800 ml/min c Thrombosis c Physiotherapy c Head ultrasound daily if fontanel open

Biomedicus-Fem-Venous 23 Fr 45 - 50 kg 5500 ml/min c Nephrology Mandatory for Dialysis c Dietician

Biomedicus-Fem-Venous 25 Fr 50 - 55 kg 6000 ml/min

Biomedicus-Fem-Venous 27 Fr 55 - 60 kg 6000 ml/min Initial day physician print name Physician signature

Biomedicus-Fem-Venous 29 Fr 55 - 60 kg 6000 ml/min

Form 52539/G120 (Rev 2009/12) Chart Copy GC#31125 Page 1 of 1

Version: June 2017

Figure 21-2. Example of a daily order set to guide
Figure 21-1. Example of an ECLS peripheral can- ECLS Specialists in maintaining patients on ECMO.
nula selection chart. Used with permission from Used with permission from the Hospital for Sick
the Hospital for Sick Children, Toronto, Canada. Children, Toronto, Canada.

188
 Patient Management: Neonatal and Pediatric Cardiac ECLS

Cardiovascular Although minimal blood returns to the left heart

on full VA-ECLS, this volume must be ejected ad-
General equately to prevent ventricular distension and clot
formation. Some degree of ventricular contraction
Cardiac recovery and maximizing myocardial is preferred as this is the main source of coronary
oxygen delivery underlies ECLS support. Nonpulsa- blood flow. In severe cardiac failure, or in ECPR
tile extracorporeal pump flow results in a flat arterial with minimal to no myocardial activity, patients
tracing while the patient receives full venoarterial risk atrial and ventricular over distension which
support. Pulse pressure widens with cardiac recov- inhibit cardiac recovery. Timely echocardiography
ery. The common targets for ECLS blood flow are assessing left-sided overdistention permits appropri-
based on age and disease states. Goals for neonates ate intervention.
are 100-150 mls/kg/minute; for single ventricle Interventions to address atrial or ventricular
physiology flows of 150-200 mls/kg/minute help over distension include atrial venting which con-
compensate for pulmonary runoff and in children sists of direct placement of a cannula into the left
flows are usually 80-120 mls/kg/minute although atrium (requires an open sternum) or enlarging or
some centers target 2.4-3.2 L/min/m2. creating an interatrial connection via a transcatheter
Central venous oxygen saturation goals are usu- approach. The atrial shunt ensures that blood return-
ally a Sv02 >70-75% but all targets are influenced ing to the LA flows into the RA then onto the ECLS
by patient examination and markers of end organ circuit, rather than into the LV.
perfusion (urine output/lactate/Sv02).
Hypertension
Patient Sticker
Name
Hospital Number
DOB Increased peripheral vascular resistance should
be avoided due to its negative influence on heart
PATIENT SURVEILLANCE WHILST ON ECLS
ONLY tick if task completed, please initial all tasks
Day Day Day Day Day Day Day recovery. Hypertension soon after cannulation is
Haematology
FBC and coagulation 05:00 05:00 05:00 05:00 05:00 05:00 05:00 common but usually transient and may not require
and and and and and and and
treatment. The patient must be well sedated and sub-
screen
17:00 17:00 17:00 17:00 17:00 17:00 17:00
Free Haemoglobin Day 0 Day 1 Minimum

clinical seizures excluded as an etiology. A variety

(Send 1 ml in FBC bottle. Inform For For Alternate
labs you are sending sample) Baseline comparison days
(Send sample between 9 am and

of agents treat hypertension but each ECLS provider

11 am)
Date when new X match
sent. Required every 72 hrs
(MUST arrive in Blood bank by 07.00
for new blood to be available by
12.00.)
should use drugs that they understand. Common
Clinical Chemistry
U & E’s and LFT’s
05:00 05:00 05:00 05:00 05:00 05:00 05:00
agents include sodium nitroprusside, nicardipine,
Imaging
X-ray (Daily)
Echocardiogram
Morning Morning Morning Morning Morning Morning Morning and hydralazine.
Day 0 within 4 hours of Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No
cannulation
Head Ultrasound. Infants
6/12 and less. Try to get one
done pre ECLS. (Day 0
Essential then alternate days)
Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No
Dysrhythmia
CFAM
(On all ECPR patients)

The goal is AV synchrony. Dysrhythmias can oc-

Any Other Tests
Microbiology
Circuit blood cultures.
(Please indicate if culture taken from
ECPR circuit before initiation of
ECLS.)
Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No
cur in patients after cannulation as the myocardium
is vulnerable following periods of low cardiac out-
(Day 0 essential then alternate days)

*Daily blood cultures from Day 10*

Wound swabs
(If possible when dressing down)
Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No Yes/No put, acidosis, and the use of vasoactive medications.
Urine / ETT Secretions
(Day 0 and then as indicated)
Yes/No Yes/No Yes/No

Circuit Maintenance
Yes/No Yes/No Yes/No Yes/No
Patients may require pacing at a slow rate to prevent
Sample 3 way tap 12 hourly over distension and potential clot formation within
the heart, but this must be balanced against the
Pressure transducers
(Every 3 Days)
Other 3 way taps
(Every 3 days)
need to rest the heart. In patients with dysrhythmia
requiring ECLS for low cardiac output state, antiar-
rhythmic agents (amiodarone, esmolol) should be
Figure 21-3. Example of a daily checklist to stan-
Approved by ECLS Working Group 05/18 Version 2.1.0 Review 05/2020

dardize patient care plans on ECMO. Used with optimized. It is important to correct and optimize
permission from Birmingham Children’s Hospital, metabolic disturbance (K+/Ca2+/acidosis/Mg2+) and
Birmingham, UK. maintain temperature control (avoid fever).

189
Chapter 21

Renal deposition within the circuit leading to blood clot

formation, many centers use fat emulsion without
Fluid overload should be avoided or managed difficulty. Appropriate caloric intake may necessi-
aggressively as this too can impede the patient tate fluid removal/diuresis to accommodate nutrition.
recovery. Fluid overload is common and due to
systemic inflammatory response and capillary leak Neurological
associated with CPB and initial exposure to the
ECLS circuit, fluid resuscitation and administra- Sedation/Pain Management
tion of blood products on ECLS, and acute kidney
injury/renal failure. Adequate pain relief and sedation ensures safety
Start intravenous (IV) diuretics immediately of the ECLS patient. Drug doses often escalate
with close attention to fluid balance. Fluid removal quickly due to increased volume of distribution
may be challenging so early institution of renal from the ECLS circuit so dose medications to effect
replacement therapy (RRT) for fluid removal may (see Chapter 6). Intravenous agents are preferred
improve outcomes. Mode of dialysis, configurations but oral agents can be added as adjuncts in patients
into the circuit, and RRT prescriptions will vary receiving enteral feeding. Sedation goals vary with
between centers (see Chapter 28). individual patients and diagnoses. Older infants and
Important aspects of patient management in- children cannulated peripherally often require less
clude documenting accurate fluid input/output with sedatives when compared to the neonatal cardiac
ordered daily target balance parameters, restricting patient cannulated via the central approach.
fluid intake in the cardiac population (60-100 mls/ Continuous IV neuromuscular relaxants are
kg/day) while assuring appropriate caloric intake, only required if there is a specific need to prevent
repositioning the patient when possible to aid mo- all patient movement (rare) or there are limitations
bilization of fluid, and ensuring that cannulas are to ECLS flows when the patient moves. Allowing
well secured as their position can change as fluid some spontaneous patient movement can aide fluid
overload resolves. mobilization and permit basic neurological assess-
ment and evaluation of seizures.
Gastrointestinal/Nutrition
Patient Evaluation
Adequate nutrition promotes recovery from
critical illness. Patients experience a high protein ECLS can cause acute neurological injury
catabolic state and neonates can lose 15% lean body (hemorrhage, seizures, infarction and brain death)
mass in the first 7 days on ECLS. Historically, par- associated with an increased risk of mortality and
enteral nutrition has been preferred due to concerns long-term neurological sequelae.5
of pre-ECLS hypoxia and high dose vasoactive Routine cerebral ultrasound is recommended in
medications as well as nonpulsatile ECLS blood all patients with an open fontanelle. The risk of hem-
flow to the splanchnic circulation. Previous reports orrhage is highest in the initial 3-5 days of support.
suggest that secondary intestinal ischemia can lead ELSO Registry reports 11% intracranial hemorrhage
to bacterial gut translocation and ultimately necro- in neonatal cardiac and 6% in pediatric cardiac pa-
tizing enterocolitis, resulting in a wide variation in tients. Intracerebral injury should be monitored and
feeding practices in ECLS patients. Most centers alter anticoagulation or increase coagulation factors/
establish nasogastric enteral feeding as soon as platelets targets when appropriate. Risk factors for
feasible and safe but many centers remain cautious intracranial hemorrhage include 1) acidosis - pH
in the acute phase. Patients must be well nourished <7.10 at time of cannulation, 2) cardiopulmonary
during this catabolic phase. Parenteral nutrition resuscitation (CPR) before cannulation or during
should be started immediately if feeding difficulties ECLS, 3) difficult anticoagulation, and 4) ongoing
arise and should ideally be administered directly coagulopathy.
to the patient. Although there are reports of lipid

190
 Patient Management: Neonatal and Pediatric Cardiac ECLS

Electroencephalography (EEG) of ECLS management is covered in Chapter 4. How-

ever, several important points should be emphasized.
ELSO Registry reports clinical electrical sei- Any baseline pre-ECLS coagulopathy should be cor-
zures between 3-4%. EEG may be warranted if there rected. Bolus anticoagulation (usually 50-100 units/
are suspected seizures or the patient is on continuous kg UFH) pre-cannulation with dose adjustments
neuromuscular relaxation. based on individual patient circumstances (bleed-
ing, post CPB, DIC). After cannulation and once
Near Infrared Spectroscopy (NIRS) monitoring tests (e.g. ACT) have fallen to protocol
levels, commence continuous anticoagulation. (e.g.
Used routinely in many centers, NIRS provides UFH infusion at 10-20 units/kg/hour).
noninvasive monitoring of cerebral oxygenation Hourly circuit and patient checks are manda-
when stickers are placed on the forehead. This pro- tory to monitor clots and fibrin strands as well as
vides an adjunct in neuromonitoring. bleeding. Clear guidelines and targets for the ECLS
specialist are paramount to adjust and maintain
Computed Tomography of the Brain anticoagulation. Anticoagulation monitoring is
center-specific but many monitoring tests are af-
This may be warranted on ECLS to evaluate fected by other factors including thrombocytopenia,
for neurological insult which can support the team hypofibrinogenemia and anemia.
with reorientation of care and early separation from Treatment/investigation algorithms for out of
ECLS. The risks of moving any patient on ECLS range coagulation studies should guide bedside man-
cannot be underestimated and clear guidelines for agement. All decisions surrounding anticoagulation
the safe transfer of patients should be developed. in ECLS patients are guided by the clinical status
of the patient as well as the integrity of the circuit.
Respiratory
Infections/Surveillance
The goal is to avoid/reduce raised pulmonary
vascular resistance (PVR) and ensure good coronary Early recognition and prompt treatment of
perfusion. Assess lung compliance, chest radiograph, infections is essential as the ECLS circuit can be-
and blood gases daily. Most ECLS patients are man- come contaminated. No standardized international
aged on ventilator “rest” settings during their ECLS guidelines exist for infection surveillance or antibi-
run which depend on local practice but follows the otic prophylaxis in ECLS but surveys suggest most
rules of low rate, low pressure, enough PEEP to (but not all centers) use prophylactic antibiotics and
prevent atelectasis, and minimize ventilator associ- almost half of centers use a standardized protocol.6
ated injury. Gas exchange is performed by the circuit. Local treatment guidance should be created with
Patients who develop respiratory complications infectious diseases/microbiology departments. Care-
should be investigated and managed appropriately. givers should have a low threshold to treat infec-
The ECLS team should establish guidelines with tions as they can disseminate rapidly and consider
respiratory therapy/physiotherapist so that patients antifungal therapy in longer ECLS courses or in
receive appropriate respiratory toilet and physio- patients with chronic illness or immunodeficiencies.
therapy while mitigating the risks of bleeding and
cannula displacement with appropriate team educa- ECLS Emergencies
tion. Finally, invasive investigations (bronchoscopy)
must be performed by a skilled operator. Emergencies on ECLS occur and protocols must
be in place to aid the bedside providers to respond
Anticoagulation quickly. ECLS emergencies should form the founda-
tion of scenarios and simulation practice for teams
Anticoagulation balances the risk of circuit (see Chapter 35). Emergencies often result in the
thrombosis with bleeding. This paramount feature need to separate the patient rapidly from ECLS so

191
Chapter 21

the specialist can quickly diagnose and correct the be created and practiced with the bedside ECLS
problem. Often this means that the patient suffers providers.
a low cardiac output or arrest state. The priority is All centers should have a system to rapidly
to promptly restore ECLS flows and to support and mobilize the ECLS team during any emergency.
resuscitate the patient prior to that. Clear delineation
of roles in this setting should be well established and Bleeding
rehearsed regularly. Some centers have an emer-
gency ECLS resuscitation box or emergency drugs A recent analysis found that nearly half of chil-
available at all times at the bedside. Other ECLS dren with cardiac disease on ECLS had hemorrhagic
programs run epinephrine 0.01 mcg/kg/minute complications, associated with a significant risk of
intravenously at all times during an ECLS run so mortality.7 As the number of hemostatic complica-
that the drug is always available. tions increase, the outcome worsens. Severe bleed-
ing in the early postoperative period occurs due to
Air Embolism a combination of causes including long CPB times,
trauma at surgical sites, dilution of clotting factors
Air can be entrained into the ECLS circuit or air during CPB, and with volume replacement, hypo-
may also come out of solution, through cavitation. thermia, hypoxia, and acidosis.
The risk of air entrainment is increased when there Severe bleeding requires timely correction with
are excessive negative venous line pressures. Air can blood products and coagulation factors to maintain
enter via any port or pigtail when the circuit is ac- sufficient preload to support ECLS flows (see Chap-
cessed, during the administration of medications via ter 5). Appropriate blood products for replacement
the central venous line, or at cannula insertion sites. include packed red blood cells (PRBCs), fresh
Methods to reduce high negative venous pres- frozen plasma (FFP), platelets, and cryoprecipi-
sure include: 1) providing sufficient preload to the tate. Transfusion targets should be determined by
patient with volume infusion, 2) adding a secondary the ECLS team. Common transfusion thresholds
venous drainage cannula, 3) upsizing inadequately include
sized cannula, 4) decreasing the pump speed, and
5) verifying correct cannula position. • PRBCs to maintain hematocrit 32-37% (higher
for single ventricle patients).
Unplanned Accidental Decannulation • Platelets counts >80-150,000.
• Cryoprecipitate to keep fibrinogen >1.5 (150
This potentially fatal complication must be mg/dL).
managed with immediate clamping of the circuit
to prevent further blood loss and the risk of air en- Discussion among the ECLS team is necessary
trainment and embolism. The ECLS surgeon must to determine when it may be appropriate to explore
immediately be alerted as intervention is required. the chest or cannulation site to avert further surgical
As with all acute separations, the patient may be bleeding losses. It may be necessary to reduce or
left in a low cardiac output or cardiac arrest state so stop anticoagulation to allow bleeding to slow but
the appropriate team must support and resuscitate the risk of circuit thrombosis must be foreseen and
the patient. a replacement circuit should be prepared.
The ECLS team should have a system to quickly
Machine Failure replace the ECLS circuit and for this to be simulated
regularly in ECLS training.
Pump failure due to technical malfunction or
loss of power (both electrical and battery backup) Tamponade
can also occur. Some ECLS devices offer the ability
to support the patient manually during pump failure. Recognition of the warning signs of tampon-
An institutional policy to manage this event must ade before complete loss of ECLS blood flow is

192
 Patient Management: Neonatal and Pediatric Cardiac ECLS

tantamount. These signs include increased central

venous or filling pressures in the heart, worsening
negative venous pressure and reduction in ECLS
blood flow, and sudden reduction in chest drain
output in bleeding patients.
Blood collecting in the pericardial space or
mediastinum increases the intrapericardial pressure,
limiting venous return to the heart.8 Decreased ve-
nous return impacts the ability to provide adequate
ECLS flows as the preload to the pump is diminished
and can even occur in patients centrally cannulated
via an open sternum. Tamponade requires early
intervention by drainage of any blood collecting
in the pericardial space and/or exploration of the
mediastinum to locate and correct surgical sites of
bleeding. However, any surgical exploration can
further exacerbate bleeding.

Weaning from ECLS

The goal of management is myocardial recov-

ery with safe weaning and separation from ECLS
followed by decannulation. The underlying diag-
nosis usually provides a general timeframe for an
attempted wean and trial off ECLS. There should
be scheduled daily ECLS team meetings to discuss
the feasibility of a potential trial off. Failure to wean
cardiac ECLS after 1 week indicates a poor prog-
nostic sign and the parents must be appropriately
counselled and supported.
Important issues prior to trials off ECLS include
confirming that all lines, tubes, and pacing wires are
correctly sited and functional. Fluid overload has
been managed. All medications should be delivered
to the patient except UFH, although some centers
deliver 50% UFH to the patient and 50% to the cir-
cuit. This may include inotropic support during the
wean in ECLS flows. Ventilator settings should be
adjusted before weaning ECLS flows. Many centers
wean ECLS flows gradually over 6 to 12 hours to
an idling flow rate of 100 ml/minute although some
target 50 mls/kg/minute. Tight anticoagulation
control during weaning decreases the risk of circuit
thrombosis. For further details on weaning please
see Chapter 25.

193
Chapter 21

References

1. Butt W, Heard M, Peek G. Clinical management

of the extracorporeal membrane oxygenation
circuit. Pediatr Crit Care Med. 2013;14(5 Suppl
1):S13-19.
2. Chan T, Thiagarajan RR, Frank D, et al. Survival
after extracorporeal cardiopulmonary resuscita-
tion in infants and children with heart disease. J
Thorac Cardiovasc Surg. 2008;136(4):984-992.
3. Kirsch R, Schwartz S. Medical Management
of Neonates and Children with Cardiovascular
Disease. In: Brogan TV, ed. Extracorporeal Life
Support: The ELSO Red Book. 5th ed. Ann
Arbor, MI: ELSO; 2017:347-355.
4. Peek GJ, Hammond I. Neonatal/Pediatric Car-
diac Cannulation. In: Brogan TV, ed. Extracor-
poreal Life Support: The ELSO Red Book. 5th
ed. Ann Arbor, MI: ELSO; 2017:347-355.
5. Mehta A, Ibsen L. Neurologic complications
and neurodevelopmental outcome with extra-
corporeal life support. World J Crit Care Med.
2013;2(4):40-47.
6. Lao LS, Fleming GM et al. Antimicrobial
prophylaxis and infection surveillance in ex-
tracorporeal membrane oxygenation patients:
a multi-institutional survey of practice patterns.
ASAIO J. 2011;57(3):231-238.
7. Werho DK, Pasquali SK, Yu S, et al. Hemor-
rhagic complications in pediatric cardiac pa-
tients on extracorporeal membrane oxygenation:
An analysis of the extracorporeal life support
organization registry. Pediatr Crit Care Med.
2015;16(3):276-288.
8. Zwischenberger JB, Cilley RE, Hirschl RB, et
al. Life-threatening intrathoracic complications
during treatment with extracorporeal membrane
oxygenation. J Pediatr Surg. 1988;23(7):599-
604.

194
22

Patient Management for the Adult Patient with Respiratory Failure on ECLS

Daniel L. Herr, MD, John C. Greenwood, MD

General Principles of Care PaCO2, patients will often remain tachypneic, and
appear extremely uncomfortable. Standard assess-
When considering the medical management of ment scales such as the Richmond Agitation and
the patient on venovenous extracorporeal membrane Sedation Scale (RASS) or a pain assessment scale,
oxygenation (VV-ECMO), the most important guid- such as the Behavioral Pain Scale (BPS), are used
ance to remember is that the same general principles for the titrating sedation and analgesic medications
of managing a critically ill patient with severe lung in the critically ill patient.1 Unfortunately, these
disease on ECMO should be no different than the scales are often not helpful because the VV-ECMO
care of a patient with severe lung disease not on patient’s respiratory drive is no longer compensat-
ECMO. The ECMO machine is in place to support ing for an iatrogenic respiratory acidosis, but rather
the body’s organs until the primary disease process a pulmonary-neurological driven reflex that will
resolves. The patient who requires VV-ECMO for occur even if heavily sedated. Large amounts of
pulmonary failure is most commonly being treated intravenous narcotics will often be administered in
for acute respiratory distress syndrome (ARDS). an attempt to reduce the work of breathing, but this
ARDS is caused by diffuse lung injury, characterized is usually an ineffective strategy. High dose narcot-
by acute inflammatory changes, associated with in- ics may actually be contraindicated in this type of
creased pulmonary vascular permeability, increased patient because pain is rarely the cause of exces-
extravascular lung water, and loss of aerated lung sive work of breathing, and may only contribute to
tissue. The etiology of ARDS is most commonly disorientation and dependence.
related to primary lung infections or injuries that As a result of this rapid breathing pattern, there
result in shock leading to a total body inflammatory can be dramatic fluctuations in intra-thoracic pres-
response. The management of the patient on VV- sure, which often results in decreased ECMO flows
ECMO should be set around goals of care set out and acute desaturation. When this patient-ventilator
by a multidisciplinary team, and these goals should dyssynchrony occurs, temporary paralysis may be
be assessed on a daily basis. The role of the ECMO required. The choices of neuromuscular blocking
Specialist is to be an integral part of the team and agents include cisatracurium, rocuronium, or ve-
to understand how ECMO can impact the team’s curonium. Short-term paralysis in the acute stage
ability to achieve those goals. of severe ARDS has been found to decrease mortal-
ity.2 Cisatracurium is often the paralytic of choice
Systems Management and Goals for patients on VV-ECMO because the dose can
be titrated, so that the minimal necessary dose can
Sedation, Analgesia, and Neurologic Assessment be administered to control respiration and avoid
interference with ECMO flows, yet keep the patient
Probably one of the more difficult treatment comfortable and able to move.
goals when managing a patient with respiratory The most common sedative agent used dur-
failure is to control the work of breathing. Despite ing paralysis is propofol, others can be found in
ECMO’s ability to normalize both the PaO2 and Table 22-1. It is important to note that propofol is

195
Chapter 22

highly lipophilic and protein bound, and can have a in the femoral vein the reading of central venous
high propensity to be adsorbed by the extracorporeal pressure will not be accurate.
circuit.3 An alternative medication used for sedation, The hemodynamic goals for a patient on VV-
that is very useful when using low dose continuous ECMO for respiratory failure are not much different
paralysis, is dexmedetomidine. Dexmedetomidine than the patient on a ventilator for respiratory failure
induces a non-REM sleep state, with minimal without ECMO. As previously mentioned, the pa-
amnestic effects, and mild analgesia.4 Dexmedeto- tient’s neurological state can influence both ECMO
midine has no effect on respiratory drive, except flow and hemodynamics. Agitation and anxiety
to control anxiety induced hyperventilation.5 For can cause large fluctuations in heart rate (HR) and
intermittent paralysis, rocuronium and vecuronium blood pressure (BP). Generally, these fluctuations
can be used in conjunction with intermittent ben- have much less of a clinical impact, compared
zodiazepines, and may be useful to improve both to the patient requiring VA-ECMO. Patients on
patient-ventilator-circuit synchrony and the need of VV-ECMO often have significant pulmonary hy-
excessive sedative use.6 The use of oral or intermit- pertension which places a large afterload pressure
tent atypical antipsychotic agents such as risperi- on the right ventricle (RV). As a consequence of
done or haloperidol can be considered to minimize this pulmonary hypertension, the RV will provide
the need for continuous infusion therapy. inadequate forward flow into the lungs and left
ventricle (LV), which ultimately can lead to hemo-
Hemodynamic Monitoring and Management dynamic compromise. ECMO flows can influence
both CVP and arterial pressure, thus the specialist
The ECMO Specialist should be aware of the should be aware of the patient’s blood pressure and
type of devices being used for monitoring hemody- CVP goals, understanding that those goals may be
namic parameters. At the minimum, most patients individually set based on RV function, along with
will have an arterial pressure catheter and central other considerations. Pulse pressure can also be
venous pressure catheter used to monitor arterial monitored to assess the patient’s pulse estimate LV
and central venous pressures (ABP and CVP). The function, similar to the patient on VA-ECMO. De-
arterial monitor is usually placed in the radial or creasing pulse pressure with an increasing CVP in
femoral artery and the central venous catheter in the the patient on VV-ECMO should raise concern for
internal jugular or subclavian vein. It is important to worsening RV dysfunction. Strategies to improve
remember that if the central venous access is placed RV function include conservative fluid resuscitation,

AGENT ONSET EMINIMATION LOADING MAINENANCE ADVERSE ECMO

HALF LIFE DOSE DOSE EFFECTS EFFECT
Midazolam 2-5 min 3-11 hrs 0.01-0.05 0.02–0.1 Respiratory ↑ dose
mg/kg mg/kg/hr depression + ↑ Vd
hypotension
Lorazepam 15-20 8-15 hrs 0.02–0.04 0.01–0.1 Respiratory ↑ dose
min mg/kg mg/kg/hr depression, ↑ Vd
hypotension,
propylene glycol
related acidosis
Propofol 1-2 min Short term 5 5-50 μg/kg/min Hypotension, ↑↑ dose
3-12 hrs μg/kg/min respiratory over time
Long term depression,
30-70 hrs hypertriglyceridemia,
pancreatitis, propofol
infusion syndrome
Dexmedetomidine 5-10 1-3 hrs None 0.2–0.7 μg/kg/hr Bradycardia, Unknown
min hypotension
Abbreviations: Volume of distribution (Vd), micrograms (μg), kilograms (kg)

Table 22-1. Common sedatives medications used in patients requiring ECMO.

196
 Patient Management for the Adult Patient with Respiratory Failure on ECLS

the addition of inotropic support, and pulmonary oxygenator CO2. Often, the patient’s PaCO2 will
vasodilator therapy. In addition, a temporary reduc- be higher than normal in the setting of low tidal
tion of ECMO flow may reduce PA pressure and volume ventilation and permissive hypercapnia.
reduce RV strain. However, if the purpose of VV-ECMO is to be used
Dysrhythmias are also common in the critically as support therapy until the patient’s lungs heal and
ill patient on VV-ECMO, the most common being can ventilate on their own, the use of permissive
atrial fibrillation (AF). This can be the result of over hypercapnia may not be appropriate. Permissive
diuresis, atrial distention, and more often the use hypercapnia also increases the need for the patient
of inhaled beta agonists. Any time the patient has to maintain a higher respiratory rate on the ventilator,
changes in ECMO flow and pressure, the specialist which can result in increased respiratory agitation
should review the hemodynamic monitor and make and reduced ECMO flow. An appropriate goal of
sure the patient is not in an irregular rhythm. This patient and post-oxygenator PCO2 is 35-40 mmHg.
change in rhythm should be immediately brought If the post-oxygenator PCO2 is high, then appropri-
to the attention of the clinical provider. Patient’s ate ECMO sweep titration should be performed to
developing atrial arrhythmias on VV-ECMO will optimize the patient’s PaCO2.
often require cardioversion to restore a normal When considering ventilator management goals,
sinus rhythm. Bradycardia (heart rate <60) is also the most important concepts are: 1) VV-ECMO is
a common arrhythmia, that may be the result of the intervention that will provide support to allow
medications (beta-, calcium channel blockade, dex- the lungs to rest and heal, 2) the ventilator should not
medetomidine, etc.) but also can be a late sign of be a dependent factor in patient care, 3) mechanical
progressive hypoxia. In the case of progressive bra- ventilation should be minimized so that it does not
dycardia, the specialist should immediately check cause additional ventilator associated lung injury
for new patient-ECMO blood mixing (recirculation) (VILI).8,9 The exact ventilator settings for a patient
and circuit integrity. on VV-ECMO are controversial, but most physi-
cians would agree that mild positive end-expiratory
Respiratory/Ventilator Management (Lung Protec- pressure (PEEP) is helpful, very low tidal volumes
tive Ventilation) are accepted, and lowest possible FiO2 setting are
beneficial.
Respiratory mechanical ventilation goals in the Prone positioning is a therapy that the ECMO
VV-ECMO patient will often significantly influence Specialist should understand. Usually this proce-
ECMO management. In general, respiratory goals dure is done early in the course of severe ARDS
for most patients on VV-ECMO for severe lung (PaO2:FiO2 <150) in order to better improve alveolar
disease should be the same as a patient who is not ventilation.10,11 The proning process requires a large
receiving ECMO support. amount of teamwork and coordination. The role
The respiratory and physiologic oxygenation of the specialist during this process is to maintain
and ventilation goals in a patient on VV-ECMO cannula patency and appropriate positioning. Often,
consist of three blood gas measurements: the patient while in the process of turning of the patient, there
arterial, pre-oxygenator, and post-oxygenator blood can be decreased oxygenation events that can be
gases. The goals for oxygenation monitoring are associated with or without flow changes. It is good
patient arterial saturation around 88%7 (although practice to anticipate these events, and increase flow
some suggest that even lower saturation can be and sweep prior to proning while on VV-ECMO.
tolerated), and a post-oxygenator PaO2 greater >300
mmHg. The pre-oxygenator saturation value, unlike Fluid and Renal Management
in VA-ECMO, is highly influence by recirculation
and therefore does truly reflect the patient’s mixed Fluid and renal management of the VV-ECMO
venous saturation. patient can be challenging. There is a general
The goals for ventilation monitoring should consensus that fluid overload is associated with
be based upon the patient PaCO2 and the post- increased morbidity in ARDS12,13 and yet this is a

197
Chapter 22

very common occurrence in patients with severe are for the patient and to have some instruction on
respiratory failure on VV-ECMO. The controversy when to give or remove volume.
occurs as to what defines fluid overload, and how As far as the management of fluid, the specialist
does one truly assess or monitor fluid status. Most can contribute significantly to this process. Many
institutions at the minimum monitor two param- VV-ECMO patients develop acute kidney injury
eters, the actual measured daily fluid balance and (AKI) which makes them even more susceptible
CVP (intravascular fluid compartment). The use of to fluid overload. Progressive AKI can often lead
patient weight is another parameter that includes to ineffective medical treatment for diuresis, and
extravascular fluid gains and losses, but even with an eventual need for continuous renal replacement
internal bed scales, this measurement can be highly therapy (CRRT) for volume removal. Alternatively,
variable on a day to day basis. if no electrolyte or solute clearance is needed, a
Many clinicians will target a CVP to be less than hemoconcentrator can be added to the circuit for
10 mmHg. Unfortunately, concurrent physiologic more rapid fluid removal. Knowledge of how to
changes in patients (pulmonary hypertension, RV incorporate these devices into the VV-ECMO circuit
failure, high intrathoracic pressure, etc.) can signifi- is vital to the specialist’s training.
cantly distort the CVP so that it does not accurately
reflect intravascular fluid status.14 Continuous cardi- Infection Control Management
ac output monitoring devices, including the NICOM
(Cheetah Medical Ltd, Newton Center, MA, USA), The care of the ECMO cannula and circuit is a
Flo-trac (Edwards Lifesciences Corp., Irvine, CA, shared responsibility of the clinical nurse and spe-
USA), and others have not been validated in patients cialist. A new infection in an ECMO patient can be
on ECMO. Point-of-care echocardiography can also devastating. The use of prophylactic antibiotics for
be used to determine intravascular fluid status, but ECMO is controversial. Institutions often will use
can be challenging. Visualization of both the right prophylactic antibiotics for emergent cannulations
and left heart during a fluid challenge, along with but usually the antibiotics are discontinued in less
IVC, may be beneficial. than 48 hours. The cannula insertions sites should
Fluid assessment for the specialist consists of be checked daily with the patient’s nurse and ob-
assessing the patient, both by physical exam and served for loose stiches, cannula holders, and a clean
how the patient responds to fluid. A common event insertion site. Any ECMO cannula holder that fits
referred to as “chugging” or “chatter” of the ECMO tight to the skin or made of hard material can poten-
circuit occurs when the pump is unable to spin at the tially cause skin breakdown over time. Devices that
desired RPM due to inadequate intravascular vol- stick to the skin and have a tube anchoring device
ume, and as a result ECMO flows decrease. During attached, are better for skin integrity and are very
this clinical scenario, often times the first interven- secure. Wound dressings should be clear and contain
tion is to increase intravascular volume. Continued some form of chlorhexidine or antimicrobial barrier.
or excessive administration of fluid therapy will The ECMO Specialist should limit the number of
often result in progressive RV distention and failure. circuit blood access points, and all free stopcock
Other causes of “chatter” include changes in patient endings should have dead end caps, preferably al-
positioning, cannula migration, cannula malposition, cohol impregnated. Anytime the specialist accesses
kinked tubing, or even thrombosis. the circuit, those access points should be cleaned
If the drainage cannula is too low in the inferior vigorously with alcohol.
vena cava and/or the patient has a distended or large Heater-cooler devices (HCD) attached to
abdomen on exam, the cannula may not be able ECMO circuits can be a source in severe infections.
to function well, then a request to manipulate the In 2015 there were several case reports of infections
cannula is appropriate. The specialist must keep the with the bacteria, mycobacterium chimera, that
providers aware of how often chugging is occurring were associated with heater-cooler devices.15 It is
and the amount of volume given. In the end, the the responsibility of the ECMO Specialist team to
role of the specialist is to know what the fluid goals maintain HCD devices and follow the proper main-

198
 Patient Management for the Adult Patient with Respiratory Failure on ECLS

tenance and positioning of these units when used in

combination with ECMO circuits.16

Hematological Management

The involvement of the specialist with the hema-

tological management of the VV-ECMO patient is
centered on the detection of clots and/or hemolysis.
Avoidance of these events is related to the close
observation of ECMO circuit changes that indicate
that the process of clotting and or hemolysis may
be occurring. Clinically significant oxygenator clot
burden can be detected with transmembrane pres-
sures >60 mmHg or decreases in the pre- and post-
oxygenator PaO2 differential. Direct visualization
of the oxygenator and circuit tubing using a high
intensity flashlight is important to detecting fibrin
deposition and thrombosis.
The most common agents used for therapeutic
anticoagulation include unfractionated heparin or di-
rect thrombin inhibitors (i.e. argatroban, bivalirudin).
Each of these anticoagulants has their own distinct
pharmacology, pharmacokinetics, and therapeutic
endpoints. Therapeutic anticoagulation targets
may include activated partial thromboplastin time
(APTT), anti-factor Xa (anti-Xa) levels, activated
clotting time (ACT), or can even utilize throm-
boelastography (TEG) to ensure adequate levels
of anticoagulation. It is important that the ECMO
Specialist is aware of the current anticoagulation
strategy and targets, as this can have a significant
impact on the treatment of bleeding and thrombotic
complications. Given that most patients requiring
VV-ECMO support require higher flow rates, the
risk of thrombosis is much less when compared
to VA-ECMO. As a result, more conservative an-
ticoagulation strategies are often prescribed during
VV-ECMO use.

199
Chapter 22

References distress syndrome: Updated study-level Meta-

Analysis of 11 randomized controlled trials.
1. Barr J, Fraser GL, Puntillo K, et al. Clinical critical care medicine. 2014;42(5): 1252-1262.
practice guidelines for the management of 12. National Heart, Lung, and Blood Institute Acute
pain, agitation, and delirium in adult patients Respiratory Distress Syndrome (ARDS) Clini-
in the intensive care unit. Crit Care Med. cal Trials Network, Wiedemann HP, Wheeler
2013;41(1):263-306. AP, et al. Comparison of two fluid-management
2. Papazian L, Forel J-M, Gacouin A, et al. strategies in acute lung injury. N Engl J Med.
Neuromuscular blockers in early acute re- 2006;354(24):2564-2575.
spiratory distress syndrome. N Engl J Med. 13. Rosenberg A, D. R. Association of cumulative
2010;363(12):1107-1116. fluid balance on outcome in acute lung injury;
3. Shekar K, Roberts JA, Mcdonald CI, et al. a retrospective review of the ARDSnet Tidal
Sequestration of drugs in the circuit may lead volume study cohort. Journal of Intensive Care
to therapeutic failure during extracorporeal Medicine. 2009;24(1):35-46.
membrane oxygenation. Crit Care Lond Engl. 14. Marik PE, Cavallazzi R. Does the central venous
2012;16(5):R194. pressure predict fluid responsiveness? An updat-
4. Triltsch AE, Welte M, von Homeyer P, et al. ed meta-analysis and a plea for some common
Bispectral index-guided sedation with dex- sense. Crit Care Med. 2013;41(7):1774-1781.
medetomidine in intensive care: a prospective, 15. Götting T, Klassen S, Jonas D, et al. Heater–
randomized, double blind, placebo-controlled cooler units: contamination of crucial de-
phase II study. Crit Care Med. 2002;30(5):1007- vices in cardiothoracic surgery. J Hosp Infect.
1014. 2016;93(3):223-228.
5. Mahmoud M, Mason K. Dexemedetomidine: 16. Stammers AH, Riley JB. The Heater Cooler
reveiw, update, and future considerations of as a Source of Infection from Nontubercu-
pediatric perioperative and periprocedural ap- lous Mycobacteria. J Extra Corpor Technol.
plications and limitations. BJA.2015; 171-182 2016;48(2):55-59.
6. Timofte I,Terrin M, Barr E. adaptive periodic
paralysis allows weaning deep sedation, over-
comingthe drwoning syndrom in EMCO patietn
bridge for lung tranaplantation: A CAse series
Journal of Critical Care. 2017;42: 157-191.
7. Brodie D, Bacchetta M. Extracorporeal mem-
brane oxygenation for ARDS in adults. N Engl
J Med. 2011;365(20):1905-1914. doi:10.1056/
NEJMct1103720.
8. Schmidt M, Pellegrino V, Combes A, Scheink-
estel C, Cooper DJ, Hodgson C. Mechanical
ventilation during extracorporeal membrane ox-
ygenation. Crit Care Lond Engl. 2014;18(1):203.
doi:10.1186/cc13702.
9. Slutsky A, R. V. Ventilator-Induced Lung Injury.
NEJM 2014;(369): 2126-2136.
10. Guérin C, Reignier J, Richard J-C, et al. Prone
positioning in severe acute respiratory distress
syndrome. N Engl J Med. 2013;368(23):2159-
2168.
11, Lee JM, B. W. The efficacy and safety of prone
positional ventilation in acute respiratory

200
 Patient Management for the Adult Patient with Respiratory Failure on ECLS

Addendum

On May 24, 2008, the study “Extracorporeal

Membrane Oxygenation for Severe Acute Respira-
tory Distress Syndrome (EOLIA),” a multicenter,
international, randomized trial, was published in to
New England Journal of Medicine.1 This trial was
designed to determine the effect of early initiation
of ECMO in patients with the most severe forms
of ARDS. The trial was stopped early because it
would not reach its primary endpoint of ECMO
decreasing mortality by 20% at 60 days. At 60 days
35% of the ECMO group and 46% of the control
group had died (relative risk [RR] 0.76; 95% con-
fidence interval [CI]: 0.55-1.04; P=0.09). The key
secondary endpoint was treatment failure defined
as crossover to ECMO for persistent hypoxia (Sa02,
Sat <80%> 6 hours and investigator’s discretion) or
death in the control group and as death in patients
in the ECMO group. The trial did show significant
differences for treatment with the RR being 0.62
(95% CI, 0.47 – 0.82; P<0.001). A total of 35%
in the control group received crossover to ECMO
at a mean of 6.5 +/- 9.7 days after randomization
and their mortality reached 57% vs 41% among the
other patients in the control group (RR 1.39; 95%
CI 0.95-2.03). The second analysis used a rank-
preserving structural failure time model approach
to recover the causal effect of ECMO and produced
and estimated hazard ratio for death within 60 days
of 0.51 (95% CI, 0.24-1.02). There is controversy
over whether this trial can be used as a positive or
negative trial for the use of ECMO for severe ARDS.
The editorial which accompanies the study suggest
that the primary and secondary analyses direct us
to a common conclusion that ECMO probably has
some benefit in this population despite it not being
positive in the traditional sense.2

1. Combes A, Hajage D, Capellier G et al. EOLIA

Trial Group, REVA and ECMONet. Extracorpo-
real Membrane Oxygenation for Severe Acute
Respiratory Distress Syndrome. N Engl J Med
2018;378(21): 1965-1975.
2. Harrington D, Drazen JM: Learning from a trial
stopped by a Data and Safety Monitoring Board.
N Engl J Med 2018;378(21):2021-2032.

201
23

Adult Cardiac ECLS: Patient Management

Chirine Mossadegh, RN

The Extracorporeal Life Support Organization Cannulation Strategies

(ELSO) Registry reports a 57% survival rate for
adults supported with ECMO for cardiac indica- There are various cannulation strategies pos-
tions.1 The indications for cardiac ECMO include sible depending on indications and logistics.
acute myocardial infarction, postcardiotomy, myo- Two techniques are available, each with benefits
carditis, acute pulmonary embolism, preoperative and side effects:
support, acutely decompensated chronic cardiomy-
opathy, graft dysfunction after heart transplantation, • Surgical insertion via a cut down, allowing a
and other rare causes of cardiogenic shock such as direct visualization of the vessels
refractory dysrhythmias or Takotsubo cardiomy- • Percutaneous insertion using the Seldinger
opathy.2 technique is ECHO guided, similar to a central
The monitoring of any ECMO patient resembles line insertion. This technique is preferred in an
that of any ICU patient starting with a head-to-toe emergency setting or when cannulation occurs
assessment: outside of the OR.

• Vital signs: heart rate, mean arterial blood pres- Cannulation sites are numerous:
sure (MAP), temperature, arterial saturation,
central venous pressure (CVP) • Femoral-femoral insertion provides easier ac-
• Physical assessment: hypoperfusion (e.g. cool cess, especially in an emergency setting
extremities), diaphoresis • Femoral-axillary insertion: this choice can be
• Neurological status: consciousness, pupillary made if the femoral cannulation is not possible
reaction • Central cannulation: commonly done in the OR
• Check of all the devices: IV lines, dressings, following cardiopulmonary bypass.
ventilator settings, infusion pumps.
Pain and Sedation
In addition to this regular assessment, the
ECMO pump and circuit must also be monitored ECMO patients are now more commonly awake
for the potential risks linked to ECMO described and even extubated.3 Patients on VA-ECMO can be
in Chapters 32 and 33. awakened shortly after cannulation, some teams
The complete assessment of the ECMO system even cannulate on nonsedated and extubated pa-
includes plugs, fluid connectors, alarms, color of tients with local anesthetic. For patients on ECMO
the tubings, the integrity of the circuit, charting of after heart surgery, analgesia and sedation during the
the different parameters (RPM, Flow, Sweep, FiO2, first 24 hours may keep the patient safe, preventing
circuit pressures), and checking lines and oxygen- pain and potential blood pressure spikes that could
ator for clots and/or fibrin. damage sutures.
The ECMO circuit traps medications, altering
pharmacokinetics and pharmacodynamics of anal-

203
Chapter 23

gesics and sedatives such as propofol, midazolam, or As we will see later, a balloon pump can be
opioids (see Chapter 6).4 Higher doses of sedatives inserted to prevent pulmonary edema. In that case,
and analgesics are required to obtain appropriate the balloon generates arterial pulsatility. To assess
patient sedation and comfort. Hence, protocols for whether the pulsatility is due to the balloon pump or
the management of pain and sedation should be the heart of the patient, the balloon can be paused
altered for ECMO patients. briefly and the arterial tracing monitor watched. A
flat tracing suggests that the native heart has not
Preventing Complications yet recovered.

ECMO is a simplified version of the extracorpo- Limb and Hand Ischemia

real circuits used in the operating room for thoracic
surgery that places the patient at risk for a variety The femoral arterial catheter can partially or
of complications. It is crucial then to prevent such totally occlude the vessel. Blood flow to the leg is
complications or to detect them early. All caregiv- then compromised. To prevent limb ischemia, inser-
ers (doctors, nurses, perfusionists, physiotherapists, tion of a reperfusion line in the superficial femoral
respiratory therapist) must be trained to recognize artery or a retrograde arterial catheter in the foot
these complications including bleeding, thrombo- (posterior tibial or dorsalis pedis artery) connected
embolism, hemolysis, decannulation, etc. to the reinjection cannula to allow leg perfusion is
recommended (Figure 23-2).5-7
Hemodynamic Monitoring The nurse monitors the leg by:

ECMO is a nonpulsatile device and generates • Comparing the temperature of both legs by
laminar flow. Initially after cannulation many VA- touch or using oximetry or NIRS
ECMO patients have poor or no heart pulsatility. • Checking the aspect of the leg including stiff-
Hence, the pump generates the blood pressure. The ness, color change which may evolve from
arterial waveform can appear dampened or flat with white to blisters, and in the most extreme cases
identical systolic, mean and diastolic arterial pres- foot necrosis (Figure 23-3).
sure numbers (i.e. minimal or absent pulse pressure).
Inexperienced staff can think that the arterial catheter The reperfusion line should always be visible
has failed. (Figure 23-1). through a clear dressing, so the ECMO specialist
When monitoring these patients, the aim is to can check for kinks, clots, or fibrin deposition.
maintain the mean arterial pressure above 65mmHg. Figure 23-4 demonstrates clots and fibrin on an
Recovery of pulsatility is one of the signs of left occluded reperfusion line. The same ischemia can
ventricular function improvement.

Figure 23-1. Dampened arterial waveform (red Figure 23-2. Reperfusion cannula with femoral
tracing) caused by continuous flow of ECMO pump. cannulation. The reperfusion cannula assures flow
Systolic and diastolic pressures may be approximate to the lower leg which can be compromised by the
or identical. arterial catheter.

204
 Adult Cardiac ECLS: Patient Management

occur with axillary cannulation. A reperfusion line oxygenated blood to the right atrium which then
can be inserted to allow perfusion in the arm. perfuses the brain.

Differential Hypoxia Fluid Management

Differential hypoxia can occur in patients can- The right balance between hypovolemia and
nulated for VA-ECMO via the femoral vessels. Also fluid overload is complex for patient undergoing
called the Harlequin syndrome, two circulations VA-ECMO especially after cardiac surgery. Mas-
syndrome, or north/south syndrome, occurs when sive blood losses are frequent after heart surgery
the heart recovers and fully saturated blood from and ECMO itself can worsen this hemorrhage.
the circuit mixes with the blood ejected from the Hypovolemia is suspected with chattering of the
left ventricle (LV) that is poorly oxygenated, creat- lines and occurs with sudden variations of ECMO
ing a mixing cloud. When cardiac function is very flows resulting in hypotension. The nurse must then
poor and the aortic valve does not open, ECMO administer enough volume to maintain efficient
provides 100% of the blood flow and oxygenation. ECMO flow (MAP >65 mmHg). Hypovolemia
The ECMO blood flow infused into the femoral ar- induces hypotension and unstable or low ECMO
tery is in a direction that is retrograde to the native flow. Variations of flow increase the risk of fibrin
blood flow. If the heart recovers before the lung, it and clot formation.
ejects poorly oxygenated blood. The location of Large volume administration in association with
the mixing cloud depends on the amount of ECMO muscle relaxants and venodilators can contribute
support provided and the degree of left ventricular
ejection. The mixing cloud occurs proximally in
the aorta with poor LV function and high ECMO
flow. The mixing cloud moves distally in the aorta
when LV function improves and/or the ECMO flow
is decreased. This can result in desaturated blood
perfusing the aortic arch, the brain, and coronary
arteries while fully saturated blood perfuses the
lower body. The head appears blue, whereas the
lower extremities appear pink.
To detect differential hypoxia, the pulse oxim-
eter is placed on a finger of the right hand and blood
gases should be measured in the right radial artery,
which reflects the native cardiac output (i.e. from
the LV). This result is then compared to saturations
on a lower extremity. To correct and treat this hy- Figure 23-4. Clots and fibrin in tubing leading to re-
perfusion catheter. The serum and cellular elements
poxia, most teams add a jugular cannula to return are separating due to the absence of flow.

Figure 23-3. Ischemia caused by cannulation of femoral artery. Initial appearance of leg and foot due to com-
promised blood flow (A). Developing blisters and necrosis of foot and toes (B). Frank necrosis of foot (C).

205
Chapter 23

to tissue edema. Diuretics can treat fluid overload.

For patients who do not respond to diuretics (urine
output less than 0.5 mL/kg/hr, positive fluid balance
>500 mL in the past 24 hours), renal replacement
therapy should be started.
Also, fluid overload can result in pulmonary
edema. VA-ECMO can increase LV afterload, po-
tentially worsening pulmonary edema.
Placing an aortic balloon pump at ECMO can-
nulation may diminish fluid overload and pulmonary
edema by partially unloading the LV.8,9
When LV function is very poor, pulmonary
edema can result from LV distention increasing
left-sided pressures. Treatment requires unloading
the LV. If conventional treatment (e.g. diuretics) are
inadequate, several approaches are then possible:

• Performing an atrial septostomy10

• Placing an Impella, a pump inserted percutane-
ously to take blood from the LV and reinject it
into the aorta11,12
• Unloading both ventricles by cannulating for
central double ECLS.

Ventilator Management

Ventilator management is normally similar to

an ICU patient not on ECMO. However, after car-
diopulmonary bypass, some lung damage can occur
and impair efficient oxygen delivery. After trying
to optimize ventilator setting with rest settings and
decrease pulmonary vascular resistance, conversion
to VAV-ECMO should be considered.

206
 Adult Cardiac ECLS: Patient Management

References vere ventricular dysfunction on extracorporeal

membrane oxygenation. Catheter Cardiovasc
1. ELSO. (2017). ECLS Registry Report: Interna- Interv. 1999 ;46(2):179-186.
tional Summary. 10. Ward KE, Tuggle DW, Gessouroun MR, Ove-
2. Lorusso R, Belliato M, Weerwind P, Gelsomino rholt ED, Mantor PC. Transseptal decompres-
S, Maessen J. Adult cardiovascular diseases and sion of the left heart during ECMO for severe
procedures that predispose to ECLS. In: Brogan myocarditis. Ann Thorac Surg 1995;59(3) :749-
TV, Lequier L, Lorusso R, MacLaren G, Peek, 751.
ed. Extracoporeal Life Support : The ELSO 11. Cheng A, Swartz MF, Massey HT Impella to
Red Book, 5th Edition. Ann Arbor, MI, USA: unload the left ventricle during peripheral ex-
Extracorporeal Life Support Organization; 2017. tracorporeal membrane oxygenation.ASAIO J.
479-500. 2013 ;59(5);533-536.
3. Linden V, Palmer K, Reinhard J, et al. High 12. Koeckert MS, Jorde UP, Naka Y, Moses JW,
survival in adult patients with acute respiratory Takayama H. Impella LP 2.5 for left ventricular
distress syndrome treated by extracorporeal unloading during venoarterial extracorporeal
membrane oxygenation, minimal sedation, and membrane oxygenation support.J Card Surg.
pressure supported ventilation. Intensive Care 2011;26(6):666-668.
Med 2000;26(11):1630-1637.
4. Shekar K, Roberts JA, Smith MT, Fing YL, Fra-
ser JF. The ECMO PK Project: an incremental
research approach to advance understanding of
the pharmacokinetic alterations and improve pa-
tient outcomes during extracorporeal membrane
oxygenation. MC Anesthesiol 2013;13(7).
5. Russo CF, Cannata A, Vitali E, Lanfranconi
M. Prevention of limb ischemia and edema
during peripheral venoarterial extracorporeal
membrane oxygenation in adults. J Card Surg.
2009; 24(2): 185-187.
6. Kasirajan V, Simmons I, King J, Shumaker MD,
DeAnda A, Higgins RS. Technique to prevent
limb ischemia during peripheral cannulation for
extracorporeal membrane oxygenation. Perfu-
sion 2002;17(6):427-428.
7. Greason KL, Hemp JR, Maxwell JM, Fetter
JE, Moreno-Cabral RJ. Prevention of distal
limb ischemia during cardiopulmonary support
via femoral cannulation. Annals Thorac Surg.
1995;60(1):209-210.
8. Petroni T, Harrois A, Amour J, et al. Intra-aor-
tic balloon pump effects on macrocirculation
and microcirculation in cardiogenic shock
patients supported by venoarterial extracor-
poreal membrane oxygenation. Crit Care Med
2014;42(9):2075-2082.
9. Seib PM, Faulkner SC, Erickson CC, et al.
(1999)Blade and balloon atrial septostomy for
left heart decompression in patients with se-

207
24

Procedures on ECLS

Carl Davis, MD, FRCS, Gillian Wylie, RCN, RSCN, BSc

Objectives hemorrhagic, and thoracic complications.1,3 It did,

however, find increased blood product administra-
• Describe the incidence of noncardiac surgical tion was associated with mortality.
procedures on ECLS and their impact on patient Although a wide range of procedures can now
mortality and morbidity. be undertaken on ECLS, they still carry a risk; there-
• Understand the general principles for prepar- fore, an assessment of the need for the procedure
ing and supporting the ECLS patient during a and the proposed benefits must be balanced against
procedure. the significant risk of morbidity caused by bleeding,
• Describe the most common noncardiac surgical longer length of stay, or infection.1,2 If feasible, sur-
procedures and specific management consid- gery should be delayed until the patient is close to
erations. decannulation (or after) as best results are achieved
if rapid postprocedural weaning and decannulation
Introduction are possible.4 If surgery is warranted, then it should
be undertaken by an experienced, skilled operator
Developments in ECLS technology over the and, where possible, semielectively allowing the
past 10 years, including the use of simplified team to prepare themselves, the environment, and
centrifugal circuits, have changed patient selec- the patient for surgery.
tion criteria and allowed for the management of
more complex patients, often for extended periods Preparation and Support
on ECLS. As a result, the incidence and scope of
procedures carried out on ECLS support have also Communication
increased. Trauma patients, including those with
head injuries receiving ECLS support and patients Once the need for a procedure has been agreed
receiving tracheostomy, cesarean section, and crani- upon within the multidisciplinary team, effective
otomy while on ECLS have become more common. communication must review the scope of the
Early reports describe an incidence of 14% ECLS planned procedure, assessing any specific risks
patients undergoing a surgical procedure mainly involved, such as air entrainment, problems with
in response to bleeding complications.1 In 2015, a ECLS support, bleeding, and to detail what equip-
study by Taghavi et al. of 563 adult ECLS patients ment is needed and the responsibility for retrieval.
found an incidence of 47.8% with an average of A firm but achievable timeframe should be agreed
1.41 procedures per patient. Abdominal procedures upon, though this will be dictated by the condition
were most common (18%), followed by control of of the patient and how emergent the procedure.
hemorrhage (13%), and exploratory thoracotomy/ Discussion should include identifying key personnel
thoracoscopy (8%).2 This study did not find a result- required, anticoagulant management, antibiotic pro-
ing increase in mortality following surgery, which phylaxis, ventilation management, optimal ECLS
contrasts with earlier studies describing augmented settings as well as management in case of compli-
mortality related to multiple surgical procedures, cation. Personnel who may be needed during the

209
Chapter 24

procedure (i.e. nursing, medical, ECLS, perfusion, Care should be taken that diathermy consoles or
surgery, cardiology, and anesthesiology) should be cables are not placed on the ECLS console which
assembled for the preoperative brief and their roles might be vulnerable to malfunction from electrical
and responsibilities should be clearly defined. If ex- interference.
tra blood products are needed, then the blood bank
should be contacted and recent blood results should Coagulation Management
be available. The use of checklists can streamline
this process and prevent omissions, which can be Bleeding poses a significant problem during
critical in the emergent setting and a postprocedural any procedure and may not relate to the invasive-
debrief can highlight key learning opportunities.5 ness of the procedure but rather to the underlying
patient condition, making emergency surgery more
The Environment challenging. Therefore, the ECLS specialist must
minimize this risk by close attention to anticoagu-
Most surgical procedures on ECLS can be safely lation management before, during, and following
managed within the intensive care environment with surgery. Platelets are optimized, >100,000/mm3
the benefit of familiarity to the critical care team, for adults and >150,000/mm3 in pediatrics (again
thus removing the risks associated with transfer this may be center or surgeon specific), fibrinogen
and providing immediate access to a wide range of should be >1.5-2.0 g/L (150-200 mg/dL) and any
drugs, equipment, and supplies that might be needed underlying coagulopathy corrected. In our pediatric
during the surgery. In our center patients are only center, heparin is titrated to maintain ACT’s of 160-
transported if specific imaging, such as fluoroscopy, 180 (Hemachron Elite Accriva Diagnostics, San
is required or if transitioning to cardiopulmonary Diego, CA) and an anti-factor Xa level of 0.3–0.5
bypass for a cardiac procedure is needed, but this IU/mL (and sometimes possibly lower) for several
may be center dependent. The room should be hours prior to and during the procedure. In high
cleared of unnecessary equipment and people giv- risk patients, heparin may be stopped during the
ing space for the surgical team and its equipment procedure while maintaining high flows >120mLs/
and to minimize disruptions and ‘chatter’ during min, although this mandates close attention to the
the procedure. Strict adherence to aseptic condi- circuit observing for clot/fibrin generation. Our
tions must be maintained by the surgical team and center routinely has a standby circuit available.
the immediate bedside team should wear hats and In adult patients with significantly higher flows,
masks and avoid contaminating the field. common practice includes stopping heparin for 1-2
The type of procedure and surgeon preference hrs prior to the procedure and restarting only when
will dictate patient positioning, but adequacy of it becomes clear that bleeding is not an active issue;
flows and cannula position should be assessed fol- this may occur over several days.6 Some evidence
lowing any movement. Prior to draping the patient, exists that antifibrinolytics (Amicar, Aprotinin,
an extension line should be attached for central IV Tranexamic acid) reduce bleeding and our institution
access so that drugs and fluids can be administered has used both Aprotinin and Tranexamic acid with
as necessary (although in an emergency the circuit good effect. Dosing varies between centers, but the
may be utilized) and the ET tube, ECLS cannula, loading dose and infusion should be started prior
and circuit tubing secured. Clear drapes provide safe to surgery and continued until bleeding has ceased.
visualization of the patient during surgery although Although it has been reported in the literature to
they may distract the surgeon. Adequate pain relief increase the need for ECLS circuit replacement, this
should be administered prior to the procedure and has not been our center experience, particularly with
the patient usually requires muscle relaxation. Easy Tranexamic acid.7 During the procedure, the team
access to equipment for suctioning and hand ven- monitors blood loss, ensures replacement keeps
tilation should be within reach, ECLS and patient pace to prevent ECLS flow issues, and these data
alarm settings checked, QRS monitor tones enabled, are communicated to the entire medical team. If
and the diathermy pad positioned on the patient. significant volumes of clotting factors are adminis-

210
 Procedures on ECLS

tered, the ECLS circuit and transmembrane pressure ment is removed, emergency drugs and volume are
(TMP) should be assessed frequently, particularly available, a portable oxygen source is attached to
if heparin has been paused. the oxygenator, cannulae and tubing are secured,
Postprocedure, blood loss should be assessed and that they lead the team, maintaining a ‘physi-
initially every 15 mins, then hourly. If paused, cal bridge’ between the patient and ECLS circuit
heparin should be restarted postprocedure if ongoing to prevent tension or inadvertent decannulation.9
blood loss remains low with an ACT goal of 160-180 The pathway to CT should be cleared and best po-
(or other center targets) to maintain the circuit and sitioning of the patient ‘head first’ or ‘feet first’ will
minimize the risk of bleeding. During the postop- depend on cannulation site, organ to be imaged, and
erative period, regular monitoring of patient vital the room configuration. On arrival the water heater,
signs, ECLS parameters, oxygenation, blood loss, power mains, and gases should be reinstated. The
complete blood count, and coagulation is necessary ECLS specialist should stay with the circuit during
to direct aggressive blood product replacement and the procedure and the extremes of patient/table
recognize ongoing bleeding. The specialist should movement mapped slowly prior to beginning the
note failure to augment hematocrit despite transfu- study. During cardiac catheterization ‘air entrain-
sion, increased negative access pressures, falling ment’ can occur as sheaths or lines are inserted into
SvO2 or problems with ECLS flows, all of which the vessels, especially if circuit access pressures are
suggest occult bleeding. high. Reducing flows or small aliquots of volume
will reduce the risk as will ensuring that the opera-
Common Noncardiac Surgical Procedures tor is cognisant of the possibility. The puncture site
and limb should be checked regularly for bleeding
Bronchoscopy and perfusion. 10

Patients on ECLS often benefit from flexible or Abdominal Surgery

rigid bronchoscopy to facilitate diagnosis or help
clear secretions or areas of persistent atelectasis. A Patients may require abdominal procedures
review of 44 ECLS patients in our center who re- including laparotomy and bowel resection. During
ceived 80 flexible bronchoscopies elicited positive surgery ECLS flows may be affected when handling
findings in 43% of patients, including compression the bowel or manipulating the liver, posing a risk
of airway, suboptimal positioning of ET tubes, and of cavitation. Achieving hemostasis can also be
airway obstructive from clot/secretions. In skilled difficult despite optimization of clotting parameters
hands the risk of complications is low.8 The ECLS and anticoagulation. Liberal use of diathermy and
specialist, however, should ensure that the platelet packing the cavity can help and often an abdominal
and ACT values are within parameters, sedation is patch is needed to avoid abdominal compartment
appropriate for comfort and if desaturation occurs, syndrome and facilitate reexploration. Bowel re-
the blender is at 100% and flows increased as re- section carries significant morbidity and mortality.
quired. Minor blood staining of secretions occurs Fluid shifts, sepsis, and inferior vena cava (IVC)
in 20-30% of examinations and normally resolves compression can all affect access pressures and
spontaneously, but more persistent bleeding can flows with significant volume resuscitation often
be treated with a diluted epinephrine solution of required.11 Our own experience has been that de-
1:10,000 given via the ET tube. spite surviving the actual procedure and ECLS, few
patients requiring bowel resection have survived
Radiological Procedures to discharge. CDH repair is covered in Chapters
7 and 19.
Cardiac catheterization and computed tomog-
raphy (CT) can be safely carried out on ECLS but
requires internal transfer of the patient. The ECLS
specialist should ensure that all unnecessary equip-

211
Chapter 24

Thoracic Procedures

Thoracic procedures including thoracotomy,

lung biopsy, sternotomy, left atrial (LA) vent and
chest drain placement may be required in respiratory
or cardiac ECLS patients to aid circuit flow or wean-
ing if pleural collections are significant. Although
these procedures carry a significant risk of bleeding,
procedures such as inserting a chest tube do not
appear to be associated with decreased survival if
the risk is minimized with judicious management of
anticoagulation.12,13 In patients with central cannula-
tion, bleeding into the pericardial or pleural space
can impair venous return or pump flow. Warning
signs include tachycardia, decreased pulse pressure
and increasing access pressures, early recognition
and evacuation of clot will prevent serious interrup-
tions to circuit flows.
During thoracic exploration or insertion of
an LA vent, there is a risk of air entrainment and
significant cannula site bleeding as the heart is
manipulated and the ECLS specialist should be
prepared for either. During the preprocedural brief,
the correct diameter of connectors and tubing should
be selected to minimize the period off ECLS sup-
port and a bolus of heparin given to ensure adequate
anticoagulation during the procedure when blood
may be static in the cannulae.

212
 Procedures on ECLS

References 11. Lam MC, Yang PT, Skippen PW, Kissoon N,

Skarsgard ED. Abdominal compartment syn-
1. Atkinson JB, Kitagawa H, Humphries B. drome complication paediatric extracorporeal
Major surgical intervention during extra-cor- life support: diagnostic and therapeutic chal-
poreal membrane oxygenation. J Pediatr Surg. lenges. Anaesth Intensive Care. 2008;36(5):726-
1992;27(9):1197-1198. 31.
2. Taghavi S, Jayarajan SN, Mangi AA, et al. Ex- 12. Jackson HT, Longshore S, Feldman J, Zirschky
amining nonsurgical procedures in patients on K, Gingalewski CA, Collin G. Chest tube
extracorporeal membrane oxygenation. ASAIO placement in children during extracorporeal
J. 2015;61(5):520-525. membrane oxygenation (ECMO). J Ped Surg.
3. Nagaraji HS, Mitchell KA, Fallat ME, Groff DB, 2014;49(1):51-53;discussion 53-54.
Cook LN. Surgical complications and proce- 13. Tashiro J, Perez EA, Lasko DS, Sola JE. Post-
dures in neonates on extracorporeal membrane ECMO chest tube placement: A propensity
oxygenation. J Pediatr Surg.1992;27(8):1106-9; score-matched survival analysis. J Pediatr Surg.
discussion 1109-1110. 2015;50(5):793-797.
4. Chestovich PJ, Kwon MH, Cryer GH, Tillou
A, Hiatt JR. Surgical procedures for patients
receiving mechanical cardiac support. AM
Surgeon. 2011;77(10):1314-1317.
5. Roeleveld PP, Peek GJ. Procedures on ECLS.
In: Extracorporeal Life Support: The ELSO Red
Book. 5th ed. Ann Arbor, MI: ELSO; 2017.
6. Sidebotham D, Allen SJ, McGeorge G, Ibbott
N, Willcox T. Venovenous extracorporeal mem-
brane oxygenation in adults: Practical aspects
of circuits, cannulae, and procedures. J Cardio-
thorac Vasc Anesth. 2012;26(5):893-909.
7. Van der Staak FH, De Haan AF, Geven WB,
Festen C. Surgical repair of congenital diaphrag-
matic hernia during extracorporeal membrane
oxygenation: hemorrhagic complications and
the effect of Tranexamic acid. J Pediatr Surg.
1997;32(4):594-599.
8. Kamat PP, Popler J, Davis J, et al. Use of flexible
bronchoscopy in pediatric patients receiving
extracorporeal membrane oxygenation (ECMO)
support. Pediatr Pulmonol. 2011;46(11):1108-
1113.
9. Goodwin SJ, Randle E, Iguchi A, Brown K,
Hoskote A, Calder AD. Chest computed tomog-
raphy in children undergoing extra-corporeal
membrane oxygenation: a 9-year single-centre
experience. Pediatr Radiol. 2014;44(6):750-760.
10. Callahan R, Trucco SM, Wearden PD, Beerman
LB, Arora G, Kreutzer J. Outcomes of pediatric
patients undergoing cardiac catherization while
on extracorporeal membrane oxygenation. Pe-
diatr Cardiol. 2015;36(3):625-632.

213
25

Weaning and Trial Off

Chris Harvey, MB, ChB, MRCS

Introduction and there should be clearing of the chest radiograph

(Figure 25-1). Often radiological clearance of the
While ECMO has been shown to improve sur- lungs lags behind clinical improvement by 24-48
vival in neonatal1 and adult patients,2 it is important hrs. In neonates, the pulmonary artery pressures as
that attempts are made to utilize ECMO support only estimated with a Doppler signal on echocardiogra-
when indicated. Very long ECMO runs have been re- phy should have reduced, preferably to less than 2/3
ported3 but prolongation of ECMO support beyond the systemic blood pressure.
what is necessary may lead to development of circuit In patients receiving cardiac support, evidence
or patient related complications. The ECMO team of recovery in myocardial function should precede
must regularly assess the need for continued ECMO weaning and may be assessed at the bedside with
support. This concept holds true across all patient echocardiography. However, cardiac function on
ages and ECMO modalities. There are seldom any ECHO can be influenced by the amount of ECMO
adverse consequences from a failed attempt at a trial flow the patient is receiving. Higher pump flows lead
off ECMO support. to a reduced preload and increased afterload with a
resultant reduction in myocardial contractility. It is
When to Wean therefore advantageous to assess cardiac function
with the circuit clamped briefly.
Understanding the natural history of the under-
lying condition necessitating ECMO support can Window of Opportunity
help in deciding on the timing of initial weaning.
For example, runs of 3 days are not uncommon For some patients there exists a window of
in neonates with meconium aspiration syndrome opportunity where they can be weaned from sup-
whereas adults with H1N1 influenza usually re- port. As the aphorism states we need to be mindful
quire a median of 10 days ECMO support.4 Most that perfection is not the enemy of good enough.
improvement in cardiac function occurs within 72 For example, a respiratory patient appeared to be
hrs of cannulation for ECMO postcardiac function progressing and the flows were weaned accordingly.
with diminishing improvement as time progresses.
Mortality has been shown to increase by as much
as 15% per week in myocarditis patients supported
on ECMO.5 However, the natural history of the
underlying condition can serve only as a guide
with significant interpatient variability and regular
assessment mandatory.
As the ECMO duration progresses, improve-
ment in organ function hopefully becomes appar-
ent. For patients on respiratory support blood gases
should improve, lung compliance should increase, Figure 25-1. Improvements seen on chest x-ray.

215
Chapter 25

One day she is considered to be suitable for trial off; An alternate approach can be utilized in which
however, she remains on ECMO, with the assump- only minimal weaning of circuit flow is performed.
tion that she would continue to improve and post- The flow within the circuit is kept high to reduce
ECMO care would be simpler with further support. the risk of circuit thrombosis and allow for the use
Unfortunately, that night her condition deteriorates of less systemic anticoagulation. Typically for an
due to pulmonary hemorrhage and ECMO flow adult patient the flow is kept above 3000 ml/min
again increases. She remains on ECMO and cannot and 300 ml/min for an average neonate. Weaning of
be weaned. On day 21 her pupils become dilated. the sweep gas FiO2 can demonstrate improved lung
A CT scan confirms massive intracerebral hemor- function. Patient improvement is assessed clinically,
rhage and care is withdrawn. While the bleeding in radiologically, and with echocardiography and trial
the lung may have occurred off ECMO with sig- off attempted when improvement is evident.
nificant consequences, it may be that it could have
been avoided in a decannulated, non-anticoagulated Oxygen Responsiveness
patient and the outcome may have proven different.
A useful indicator of patient improvement is the
Push Off response to increased concentration of oxygen in
the ventilator inspired gas. The FiO2 is temporally
It may be necessary to “push” a patient off increased to 1.0 and peripheral saturations observed.
ECMO support. This usually occurs due to an A rapid rise of peripheral saturations to >95% sug-
ECMO related complication, most commonly bleed- gests that a trial off has a good likelihood of success.
ing (either newly diagnosed intracerebral bleeding
or massive uncontrollable hemorrhage). The risk Weaning of Sweep Gas
of continued ECMO support may significantly
outweigh the benefit of continuing ECMO. Under As well as weaning the ECMO blood flow
these circumstances it may be necessary to tolerate sweep gas, flow rate is weaned in response to im-
higher than usual levels of conventional support to provements in patient carbon dioxide levels. It may
allow the patient to be cannulated but should only be useful to run the patient with a mild respiratory
considered as a last resort. acidosis (pH 7.25-7.35) to encourage renal compen-
sation that may be useful in minimizing the length
How to Wean of the ECMO run.
The composition of the sweep gas can also be
The traditional approach to weaning ECMO adjusted to alter patient oxygenation. This is use-
support is to reduce pump flow slowly in a step- ful in cardiac patients who, due to the underlying
wise fashion in response to improvements in the cardiac anatomy, are not expected to reach normal
patient condition. This works well and applies to O2 saturations post-ECMO. These patients benefit
both respiratory and cardiac support. The method from acclimatization to lower O2 saturations while
aims to reach “minimum flow” at which point the on ECMO
contribution of the circuit to oxygenation and or car-
diac output is negligible. Exactly what constitutes The “One Way Wean”
minimal flow varies according to circuit design, the
components used, and the amount of anticoagula- In some patients, it may be that further continua-
tion utilized in the ECMO center. In general flows tion of ECMO is unlikely to increase the probability
of 100-150 ml/min are acceptable for a 1/4 inch of achieving a successful outcome, but the patient
neonatal circuit and around 750-1000 ml/min in a condition is not sufficiently without hope to prompt
3/8 inch adult circuit. If the patient remains stable full withdrawal of care. In these rare patients the
on minimal flow, then consideration should be given weaning process may be kept to a minimum and all
to formal trial off. that is required is to establish sufficient stability to

216
 Weaning and Trial Off

permit decannulation. The ECMO team and family arteriovenous bridge is required to maintain blood
need to be prepared in the event of sudden demise. flow through the oxygenator reducing the risk of
circuit thrombosis.
Trial Off ECMO
1. Any drugs on the circuit need attaching to the
The exact technique of ECMO trial is more a patient
function of the mode of ECMO utilized rather than 2. Ensure ACT is adequate (a bolus of heparin may
the underlying diagnosis be it cardiac or respiratory. be appropriate pretrial off)
The degree of conventional support throughout 3. Commence lung protective ventilation
the trial off does need to be tailored to the specific 4. Increase inotropic support
clinical scenario. For example, more emphasis on 5. Wean pump flow and clamp cannula (avoiding
ventilator settings is applicable to respiratory ECMO connectors and wire wound portions of the
and more attention to inotropic and fluid manage- cannula)
ment in the cardiac patient. 6. Unclamp or open the bridge
7. Adjust ventilation, inotropic support and fluid
VV Trial Off according to arterial blood gases, invasive pres-
sures and echocardiogram appearance
1. Commence lung protective ventilation 8. Flush the cannula at least every 10 min by un-
2. Disconnect the sweep gas from the oxygenator clamping the cannula and clamping the bridge.
3. Assess the patient with arterial blood gasses The blood remains stagnant in the cannula,
every 20 min creating a significant risk of circuit thrombosis;
4. Adjust ventilation accordingly therefore, the length of the trial off needs to be
5. If arterial blood gases are not satisfactory with short. With patients undergoing cardiac support it
appropriate patient work of breathing then the is often possible to decide whether decannulation
sweep gas is simply reconnected is appropriate relatively quickly and a prolonged
As the blood flow through the circuit is main- trial off is often unnecessary. In the respiratory
tained throughout the trial off, it can be extended patient cannulated for VA-ECMO a longer period
without complications for long periods. Usually a off ECMO may be required to assess recovery of
period of 2 hrs suffices to assess organ function native lung function and thrombosis is more likely.
and to decide whether decannulation is appropriate. An alternative approach to trialling off VA-
However, longer trials are possible, especially if the ECMO has been proposed that negates the need to
patient is borderline or complicated by premorbid clamp the circuit. This approach will be described
condition. for centrifugal ECMO systems but by reversing the
If the arterial blood gas is satisfactory direction of the raceway it would also be applicable
(>60 mm Hg) on lung protective ventilation to a roller pump circuit. This method of trial off
(TV <6 ml/kg, peak inspiratory pressure <25 cm relies on permitting retrograde flow in the circuit.6
H2O and FiO2 <0.6) then decannulation should be 1. Commence lung protective ventilation
considered. 2. Increase inotropic support if required
3. Reduce pump rpm until circuit flows backwards
VA Trial Off 4. Assess respiratory and cardiac function as with
standard trial off
Trialling off VA-ECMO is more complex, as
the patient must be separated from the circuit and By reducing the rpm the pump pressure falls
it is not possible to simply disconnect the sweep below the systemic arterial pressure and blood
as this creates a right to left shunting and patient flows backwards through the circuit, from the arte-
desaturation. Separation is achieved by intermit- rial cannula, through the oxygenator and returns to
tently clamping and unclamping the cannula. An the patient through the venous cannula. The pump

217
Chapter 25

is used to reduce the extent of arteriovenous shunt Special Considerations

acting as a brake rather than a pump so increasing
the rpm will reduce the amount of flow and decreas- In patients with a shunt dependent circulation
ing the rpm will permit increased retrograde flow. the practice within some centers is to clip the shunt
Some pumps and flow probes will measure ret- to prevent excess run off into the pulmonary circula-
rograde flow without the need to alter their position; tion if ECMO is required. If this has been performed,
however, other systems require the ECMO specialist then the shunt will need unclipping prior to trialling
to reverse the probe to confuse the pump console the patient off ECMO. It is often best to open the
into sensing that flow is antegrade. In neonates, chest, remove the clip and then commence the trial
flows of 100 ml/min are sufficient to prevent circuit off immediately. This will allow direct visualisa-
thrombosis without significant hemodynamic com- tion of the heart under both the on and off ECMO
promise. Occasionally a small dose of fluid (around loading conditions, permit assessment of filling
10 cc/kg) or low dose inotropic support (0.1µg/kg/ status and contractility and aid progression towards
min epinephrine) is required to maintain adequate decannulation.
retrograde flow and systemic blood pressure. In adults cannulated for peripheral VA-ECMO
As blood flows continuously through the cir- with an additional cannula placed to provide blood
cuit, the risk of circuit thrombosis is reduced and distally (“legmo” cannula) may need some ad-
prolonged VA trial off is possible. Utilizing this ditional measures to ensure adequate perfusion
method, patients have been trialled off for up to 8 is maintained throughout the trial off. It may be
hrs without complication. possible to simply clamp the circuit in the standard
Retrograde trial off is possible in most patients fashion above the bifurcation in the arterial limb of
but it may be prudent to avoid this technique in pa- the ECMO circuit. Blood will then shunt from the
tients with abnormal coronary circulation in whom main arterial cannula and perfuse the leg through
even a modest reduction in diastolic pressure may the “legmo” cannula. In some patients the systemic
lead to myocardial ischaemia (Figure 25-2). This blood pressure driving this shunt may be insufficient
technique is often not possible in patients with to provide adequate perfusion. In this situation, by
small arterial cannula (8 Fr or less) as systemic clamping only the main arterial cannula, perfusion
pressure is insufficient to overcome the resistance to the distal limb may be augmented by the circuit.
of the cannula.

Figure 25-2. Flow probes in retrograde trial off.

218
 Weaning and Trial Off

References

1. UK Collaborative ECMO Trial Group. UK

collaborative randomised trial of neonatal ex-
tracorporeal membrane oxygenation. Lancet.
1996;348(9020):75-82.
2. Peek GJ, Mugford M, Tiruvoipati R et al. Ef-
ficacy and economic assessment of conventional
ventilator support versus extracorporeal mem-
brane oxygenation for severe adult respiratory
failure (CESAR): a multicentre randomised con-
trolled trial. Lancet. 2009;17;374(9698):1351-
1363.
3. Gupta P, McDonald R, Chipman CW et al. 20-
year experience of prolonged extracorporeal
membrane oxygenation in critically ill children
with cardiac or pulmonary failure. Ann Thorac
Surg. 2012;93(5):1584-90.
4. Australia and New Zealand Extracorporeal
Membrane Oxygenation (ANZ ECMO) Influ-
enza Investigators. Extracorporeal Membrane
Oxygenation for 2009 Influenza A(H1N1)
Acute Respiratory Distress Syndrome. JAMA.
2009 ;302(17):1888-1895.
5. Rajagopal SK, Almond CS, Laussen PC, Rycus
PT, Wypij D, ThiagarajanRR. Extracorpo-
real membrane oxygenation for the support
of infants, children and young adults with
acute myocarditis: A review of the Extracor-
poreal Life Support Registry. Crit Care Med
2010;38(2):382-387.
6. Westrope C, Harvey C, Robinson S, Speggio-
rin S, Faulkner G, Peek GJ. Pump controlled
retrograde trial off from VA ECMO. ASAIO J
2013;59(5):517-519.

219
26

Outcomes and Followup of Patients Receiving ECLS

Hanneke IJsselstijn, MD, PhD, Raisa M. Schiller, PhD, Aparna U. Hoskote, MBBS, MD, MRCP

Introduction usually have overt morbidity at discharge and their

followup should be directed by neurologists along
The use of extracorporeal life support (ECLS) with community based child development centers
for life-threatening, reversible cardiorespiratory and neurorehabilitation centers.6 However, as the
failure has revolutionized intensive care manage- majority of ECLS survivors do not have neurologic
ment resulting in excellent survival outcomes over complications, their outcome looks favorable at
the years. To date, almost 90,000 patients have been discharge. This chapter primarily covers long-term
treated with ECLS with neonates being the largest neurodevelopmental outcomes in children without
group. The overall survival to discharge or transfer overt neurologic comorbidity who have survived
is 55%, with the highest survival rates in patients critical illness and includes those who receive ECLS
who received ECLS for respiratory failure.1 in the first weeks or months of life.
Improved survival rates in a population of criti-
cally ill children, particularly neonates, result in a Long-term Outcomes
growing population of childhood survivors who
would otherwise have died. However, the long-term Medical Outcomes
medical and neurodevelopmental outcome remains
of some concern.2 If not appropriately identified and The long-term medical outcomes are primar-
managed, these outcomes may evolve into signifi- ily determined by the underlying disease, and fol-
cant long-term neuropsychological sequelae with lowup care should be arranged after discharge. For
wide ranging implications for the health, education, survivors of neonatal ECLS, specific issues include
and integration of these children into society.2 To regular screening for chronic kidney disease, not
date, most data on long-term outcomes are from only in those who developed acute kidney injury
children who needed ECLS in the neonatal period during ECLS but also in children without signs of
and to a lesser extent in children who have needed acute kidney injury.2 Pulmonary morbidity including
ECMO for cardiovascular disease.2,3 Relevant data airflow obstruction and reduced exercise tolerance
on long-term outcomes in children who required after ECLS is mainly determined by the underly-
ECLS at older age or in adult life are currently ing disease (e.g. congenital diaphragmatic hernia
lacking.4,5 or respiratory distress syndrome) and duration of
Long-term morbidities are primarily determined ECLS support or artificial ventilation, in neonatal
by condition of the underlying disease and associ- ECLS survivors and adults.2,7
ated comorbidities and by neurologic complications.
Neurologic complications during ECLS have been Neurodevelopmental Outcomes
reported to range from 0 to 15%; the most common
being central nervous system hemorrhages or infarc- Motor function has been studied in children
tion in neonates. Mortality is high in children who who have survived respiratory or cardiac ECLS
have developed neurologic complications.1 Those in the neonatal period. Gross motor performance
children who survive with neurologic complications was found to be impaired in approximately 50%

221
Chapter 26

neonatal ECLS survivors studied at school age. are applicable to children and adults who survive
Longitudinal assessments have shown that deficits ECLS. Anxiety, depression, PTSD, and persistent
become apparent as children age, potentially due emotional related difficulties have been reported in
to early brain injuries only becoming evident later adults following ECLS.7
in life when more complex skills are required.2 In
children with cardiovascular disease who required Long-term Followup
ECLS, only two studies describe motor function
outcome. In children seen between 1 and 4.5 years, For large sample studies, several instruments
72% had normal motor outcome. Similar results have been developed to evaluate functional status
were reported in children who were followed up to 2 after pediatric critical care, e.g. Functional Status
years after ECLS, of whom 87.5% had normal motor Scale (FSS) and Pediatric Overall Performance
performance. Studies on the subsequent long-term Category/Pediatric Cerebral Performance Category
outcome of motor function until school age and (POPC/PCPC) scales. The scores of these scales are
beyond in these survivors are currently lacking.3 closely associated and are suitable for global assess-
Below average intelligence has been reported in ments of functional status in pediatric patient groups
4-year-old children who received ECLS for cardio- but not for patient-specific assessment.13
vascular disease.3 In neonatal ECLS survivors, intel- In general, recommendations for followup de-
ligence is usually average and stable over time.2,6 pend on the presence of neurological comorbidity,
Nevertheless, many children have problems with the nature and extent of the underlying disease, and
school performance later in life.1 Rather than general the indication for ECLS.14 Patients with neurological
intellectual functioning, these academic problems comorbidity should be referred for followup by a
seem to be associated with specific impairments ob- neurologist and/or a community based child devel-
served in the domains of sustained attention, verbal opment center at discharge. The need for regular
memory, and visuospatial memory at school-age.8,9 assessments and intervention depends on the extent
These deficits have been specifically associated with of impairment. In case of suspected neurological
small hippocampal volume (memory) and impaired comorbidity that is not obvious at discharge, pa-
global white matter microstructure (attention) in tients should be considered as having neurological
school age children, irrespective of underlying comorbidity and referred for followup to a pediatric
disease or type of cannulation.6,10 Recently, a shared neurologist. Moreover, if patients suffer from under-
vulnerability of the hippocampus in critically ill lying disease (e.g. cardiovascular disease or lung
neonates (including ECLS survivors and those with function impairment as a result of severe ARDS),
congenital heart disease) to hypoxia, neuroinflam- additional disease-specific followup should be ar-
mation, stress, and anesthetics has been postulated. ranged. Patients without neurological comorbidity
Exposure to these factors associated with critical at discharge should be referred for disease-specific
illness may lead to early hippocampal alterations followup if applicable. For children, irrespective
and long-term memory impairments.11 of the presence of an underlying disease, a long-
A systematic review evaluating health-related term followup program with regular assessments
quality of life (HRQL) after survival of pediatric covering various medical and neurodevelopmental
critical illness has identified the following determi- domains is recommended (see Guidelines). Such
nants for impaired HRQL: 1) underlying diagnosis a followup program would preferably be offered
(sepsis, meningoencephalitis, trauma), 2) chronic within the ECMO center but could also be provided
comorbid conditions, 3) treatment related factors by community pediatricians and other health care
such as cardiopulmonary resuscitation, long stay, providers closer to their home. A single center 1 year
invasive technology, 4) psychological outcome ECMO followup program offered an opportunity for
(posttraumatic stress disorder [PTSD], parent families to return to the ECMO center, so that any
anxiety/depression), and 5) social/environmental neurodevelopmental concerns could be identified,
characteristics (low socioeconomic status, parental and children could be sign-posted to appropriate
functioning, genetics).12 Many of these determinants services.15 For adults, specific attention should be

222
 Outcomes and Followup of Patients Receiving ECLS

paid to followup of psychiatric disorders, emotional

functioning and participation.

223
Chapter 26

References critical illness. Crit Care Med. 2018;46(3):401-

410.
1. Extracorporeal Life Support Organisation. Inter- 9. Schiller RM, Madderom MJ, Reuser JJ, et al.
national summary. July 2017. https://www.elso. Neuropsychological follow-up after neonatal
org/Registry/Statistics/InternationalSummary. ECMO. Pediatrics. 2016; 138(5). doi:10.1542/
aspx. Accessed October 20, 2017. peds.2016-1313.
2. IJsselstijn H, Madderom MJ, Hoskote AU. Out- 10. Schiller RM, IJsselstijn H, Madderom MJ, et
comes, complications, and followup of neonates al. Neurobiological correlates of attention and
with respiratory failure. In: Brogan TV, Lequier memory deficits following critical illness in
L, Lorusso R, MacLaren G, Peek G, eds. Extra- early life. Crit Care Med. 2017;45(10):1742-
corporeal Life Support: The ELSO Red Book. 1750.
5th Ed. Ann Arbor, MI: Extracorporeal Life 11. Schiller RM, IJsselstijn H, Hoskote A, et al.
Support Organization; 2017: 217-230. Memory deficits following neonatal critical
3. Brown KL, Hoskote AU. Outcomes, complica- illness: a common neurodevelopmental path-
tions, and followup of neonates and children way. The Lancet Child & Adolescent Health.
with cardiovascular disease. In: Brogan TV, Published online January 10, 2018. doi:10.1016/
Lequier L, Lorusso R, MacLaren G, Peek G, eds. S2352-4642(17)30180-2.
Extracorporeal Life Support: The ELSO Red 12. Aspesberro F, Mangione-Smith R, Zimmerman
Book. 5th Ed. Ann Arbor, MI: Extracorporeal JJ. Health-related quality of life following pedi-
Life Support Organization; 2017: 395-403. atric critical illness. Intensive Care Med. 2015;
4. Prodhan P. ECLS outcomes, complications, and 41(7):1235-1246.
followup of children with respiratory failure. In: 13. Pollack MM, Holubkov R, Funai T, et al. Pedi-
Brogan TV, Lequier L, Lorusso R, MacLaren G, atric intensive care outcomes: development of
Peek G, eds. Extracorporeal Life Support: The new morbidities during pediatric critical care.
ELSO Red Book. 5th Ed. Ann Arbor, MI: Ex- JAMA Pediatr. 2014;168(7):671-676.
tracorporeal Life Support Organization; 2017: 14. IJsselstijn H, Hunfeld M, Schiller RM, et al. Im-
297-305. proving long-term outcomes after extracorpo-
5. Camboni D, Schmid C. Neurologic and pulmo- real membrane oxygenation: from observational
nary complications in adult ECLS. In: Brogan follow-up programs towards risk stratification.
TV, Lequier L, Lorusso R, MacLaren G, Peek Frontiers Pediatr 2018: In press.
G, eds. Extracorporeal Life Support: The ELSO 15. Kakat S, O’Callaghan M, Smith L, et al. The
Red Book. 5th Ed. Ann Arbor, MI: Extracorpo- 1-Year Follow-Up Clinic for Neonates and
real Life Support Organization; 2017: 575-581. Children After Respiratory Extracorporeal
6. IJsselstijn H, Hoskote A, Schiller RM. Guide- Membrane Oxygenation Support: A 10-Year
lines for Long-term follow up after neonatal and Single Institution Experience. Pediatr Crit Care
pediatric ECMO. Extracorporeal Life Support Med. 2017;18(11):1047-1054.
Organization ECLS Guidelines 2018.
7. Schmidt M. Outcomes and complications of
adult respiratory ECLS. Neurologic and pulmo-
nary complications in adult ECLS. In: Brogan
TV, Lequier L, Lorusso R, MacLaren G, Peek
G, eds. Extracorporeal Life Support: The ELSO
Red Book. 5th Ed. Ann Arbor, MI: Extracorpo-
real Life Support Organization; 2017: 471-478.
8. Leeuwen L, Schiller RM, Rietman AB, et al.
Risk factors of impaired neuropsychological
outcome in school-aged survivors of neonatal

224
27

Ambulation and ECLS

Ira M. Cheifetz, MD, FCCM, FAARC, Ali McMichael, MD, Jan Hau Lee, MBBS, MRCPCH, MCI

Introduction lodgement as well as dyspnea in venovenous (VV)

patients. The largest patient population in which
During ECMO support, patients can become there is experience with awake and ambulatory
physically debilitated due to prolonged periods of ECMO are those awaiting lung transplantation.12,16,20
pharmacologic sedation and immobility.1 Ambula- While the potential risk of cannula dislodgement
tory ECMO presents a promising approach to allow exists when mobilizing ECMO patients, an increas-
active physical therapy, promote physical rehabili- ing number of studies demonstrate that with careful
tation, and optimize neuromuscular recovery after planning, patients, including children, can be awake
critical illness. Initial small studies and case series and ambulate with overall low risk.3,11-14
have shown benefit in patients who were awake and A case series of three children (10-14 years of
rehabilitated on ECMO, including decreased days age) supported on VV-ECMO (12-109 days) while
of mechanical ventilation posttransplant, intensive awaiting lung transplantation demonstrated the
care unit stay, and total hospital length of stay.2-4 A feasibility of awake ECMO in this patient popula-
carefully coordinated, multidisciplinary approach tion.14 These patients had chronic lung conditions
is essential to successful physical rehabilitation prior to the institution of ECMO support. As such,
and ambulation for patients receiving ECMO. In their tolerance for pauses of mechanical ventila-
this chapter, we review the available literature, tion support may be higher than patients who were
describe a process to ambulate patients on ECMO, previously well prior to severe respiratory failure.15
and consider potential challenges and successes of A recent case report demonstrated that ambulatory
this type of novel program. ECMO for a previously well patient with ARDS
can be successful.16
Rationale for Ambulatory ECMO
A Programatic Approach
Pharmacologic sedation, neuromuscular block-
ade, and immobility are well accepted risk factors Each ambulatory ECMO program should
for increased morbidity and mortality in the criti- define their specific goals and patient populations
cally ill population.6-7 This knowledge has led to a (i.e. adult and/or pediatrics; VA and/or VV-ECMO;
shift toward decreasing sedation and encouraging bridge to transplantation and/or bridge to recovery)
earlier mobilization of critically ill patients.8-10 to be included. More specific components of each
This trend in clinical management has also been program that must be defined at a high level include
applied in ECMO patients.3,11-16 The reports of defining ambulation goals unique to each patient
good outcomes following active rehabilitation and population, approach for airway management (e.g.
ambulation on ECMO, should result in decreased intubation, extubation, or tracheostomy placement),
use of the traditional approach of heavy sedation and disciplines to be included in the overall program
and even intermittent paralysis in some centers.17-19 as well as with each patient who ambulates. An
The traditional sedation approach to ECMO ultimate goal of such a program may be to have
management stems from the concern for cannula dis-

225
Chapter 27

all age appropriate patients participate in physical in internal jugular or axillary vessels. Cannulation
rehabilitation while on ECMO. in the femoral vein risks potential dislodgment
and limits lower extremity mobility, potentially
Initial Assessment and Defining Rehabilitation impacting rehabilitation efforts. Still, some centers
Goals actively mobilize ECMO patients with femoral-
jugular VV cannulation.22,23 It should be noted that
Physical rehabilitation for ECMO patients central cannulation does not necessarily preclude
begins just as any other inpatient rehabilitation pro- ambulation as novel cannulation techniques have
gram with a general assessment, including cognition, been described.24-25
range of motion, strength, and functional mobility.
To ambulate on ECMO, the patient must be awake, Team Roles
able to follow simple commands, and answer ‘yes’
or ‘no’ questions either verbally or with hand com- Ambulating ECMO patients must be a carefully
munication. The patient must also demonstrate ad- coordinated, multidisciplinary effort. Depending on
equate body strength, generally as determined by a the specific equipment used, need for mechanical
physical therapist, and be able to maintain adequate ventilation, patient stability, and physical endur-
oxygen saturation and hemodynamics in sitting and ance, 4 to 6 team members, including bedside care
standing positions. nurses, ECMO specialists (RN, RT, and/or perfu-
The interdisciplinary team, including at least sion), and physical therapists are needed. The need
physicians, nurses, respiratory therapists, and for a medical provider to accompany ambulatory
physical therapists, should define the specific reha- ECMO patients should be at the discretion of each
bilitation goals for each patient. Goals may include individual institution and potentially based on each
number of days walked and/or number of steps individual situation.
taken each day. The patient and/or parent/guardian Roles must be clearly delineated for each team
must clearly be informed of the risks and benefits member. For example, during ambulation the nurse
of ambulation. is responsible for monitoring and safeguarding pa-
tient vitals, lines, and IV poles. The successful uti-
Airway Management and Location of Cannulas lization of a nurse-led ambulation protocol has been
reported.26 The respiratory therapist is responsible
Patient-specific decisions are important and for the ventilator and endotracheal/tracheostomy
may depend on the type of ECMO (i.e. VA vs. VV) tube, as applicable. An ECMO specialist manages
as well as the cannulation details (i.e. location of the ECMO machine and the cannulae, which may
cannulation, type of cannulas used). However, each require more than one person. A physical therapist
program should decide on their general approach guards the patient during transfers and throughout
to airway management for all patients. Benefits of the entire ambulation process. If family members
tracheostomy over endotracheal intubation include are present, they can stand in front of the patient and
improved patient comfort leading to decreased provide encouragement and/or travel behind and
sedation and decreased risk of accidental dislodge- help push supporting equipment, such as a wheel-
ment during exercise. An approach to bedside chair. To decrease patient anxiety and eliminate
tracheostomy in ECMO patients has recently been confusion, one person, usually a physical therapist
reported.21 Extubation may seem most optimal, but or nurse, is designated as the primary communicator
clinicians must consider the scenario of a pump with the patient.
catastrophe during an ambulation in a patient with-
out direct access to the airway. Ventilator support, Equipment and Preparation for Ambulation.
if any, should be individualized. As a general rule,
if a patient requires maximal inspired oxygen and/ Prior to ambulation, the team should iden-
or sweep gas flow, ambulation should be deferred. tify and collect the equipment needed for the walk
ECMO cannulas should generally be located either (Table 27-1). This may include an ambulatory assist

226
 Ambulation and ECLS

device, ‘transport’ (portable) ventilator, monitor(s), during the ambulation process. Figure 27-1 shows
and a wheelchair. Assistive devices should be de- a typical ECMO ambulation.
termined by a physical therapist based on patient
strength and balance. Emergency supplies, including Prior to Ambulation
bag-valve-mask and suction, may be necessary de-
pending on the length of travel. The ECMO console Each team member involved, the patient, and
should be fully charged and accompanied by a hand his/her family must agree to ambulate at the time
crank and clamps, in case of emergency. chosen. If the patient, a family member, or a team
If the patient requires significant ventilator sup- member has any concerns, those issues must be ad-
port, end-tidal carbon dioxide measurements can dressed prior to ambulation. All participants must
be used during the walk and should be considered agree to proceed for each ambulation activity. The
prior to the start of the ambulation process. The team and patient should predetermine the daily goal
ECMO Specialist titrates the sweep gas as needed for the walking distance. Prior to starting, prepare a
based on the end-tidal carbon dioxide and oxygen large walking path for the equipment and staff. Look
saturation values. ahead and clear items from the hallway and have a
The ICU charge nurse and other team members plan to open doors, as needed. If the team anticipates
should know the planned location of the ambula- walking through a narrow area, such as a doorway,
tion, if beyond the walls of the ICU. An emergency plan ahead as to how to position staff, patient, and
response plan should be in place prior to each am- equipment. For example, place the ventilator in front
bulation. Mobile phones should always be readily of the patient and the ECMO machine behind the
accessible. patient. If the patient is unable to speak or vocalize,
Intravenous lines should be well secured. We determine a clear hand signal that the patient can
have found that wrapping 3MTM CobanTM Self-Ad- use to communicate a need to stop, rest, or sit.
herent tape around the patient’s head to stabilize the Assess the patient for electrolyte imbalances,
cannulas for those cannulated via the neck helps. A bleeding/thrombosis, excessive intravascular vol-
key responsibility of an ECMO specialist is to moni- ume requirements, and other signs of a change in
tor the ECMO cannulas for any signs of movement clinical status prior to starting. Any bleeding around

SUGGESTED EQUIPMENT
• Adequately charged ECMO console
with clamps, hand-crank, and
oxygen tank.
• Ambulation assist device(s).
• Wheelchair to follow patient, if
indicated.
• Equipment to hold/support
cardiorespiratory monitor(s), oxygen
for the patient, chest tubes, etc. May
be able to use ECMO cart depending
on configuration.
• Bag-valve-mask bag.
• Transport (mobile) ventilator, if
needed.
• Consider portable suction.
• Mobile phones to call for assistance.
Table 27-1. Suggested equipment for ambulation. Figure 27-1. Typical ECMO ambulation.

227
Chapter 27

a tracheostomy or cannula site should be assessed function, eligibility for transplantation (if appli-
and cleared by the appropriate medical provider cable), and neurodevelopmental effects, especially
prior to ambulating. in the younger patients.
ARDS survivors are well known to have
Challenges/Lessons Learned increased long-term morbidities and mortality.27
Numerous studies have demonstrated that qual-
Designating a time that is acceptable for the ity of life scores at six months and beyond were
entire mobility team is essential for successful and lower in ARDS survivors as compared to a general
consistent ambulation. This approach helps the population.28-29 One year later, mortality of ARDS
nursing staff prepare for the day and may prevent survivors was higher compared to initial in-hospital
an incident such as the patient receiving sedation mortality.30 Unfortunately, there are somewhat lim-
immediately prior to ambulation. ited studies that examine the long-term outcomes of
It is possible that an individual patient may patients supported with ECMO. Persistent physical
need an increase in inspired oxygen, sweep gas, or limitations and emotional impact have been seen
ECMO flow during ambulation. To help predict if a six-months post ECMO.31 A small study compar-
patient will require more respiratory support when ing survivors of ARDS supported with and without
walking, the ECMO Specialist should observe for ECMO did not demonstrate any difference between
desaturation and increased work of breathing as the quality of life and psychological impairment at
patient moves from standing to sitting or ‘marches one year.27,32 Future larger comparative studies
in place.’ While ambulating, the ECMO Specialist are needed to examine the long-term outcomes in
should observe the ECMO monitor for decreased ECMO survivors.
flow alarms as well as the patient monitor for de-
saturation and should make adjustments as needed. Conclusion
Patient anxiety can present a significant chal-
lenge to ambulation. Waking up after being cannu- In summary, ambulation on ECMO is a co-
lated onto ECMO and having a tracheostomy that ordinated, multidisciplinary event that can help
prevents speaking can be emotionally distressing. rehabilitate patients whether the scenario is bridge
Adding aggressive physical therapy can further ex- to recovery or transplantation. Ambulation can be
acerbate patient stress. To help alleviate anxiety and performed safely with a thoughtful programmatic
motivate the patient, goals should be discussed with approach and the coordinated efforts of an ECMO
the patient (and his/her family), including long-term ambulation team along with a motivated patient and
goals required to list the patient for lung transplanta- family. A general goal for all ECMO patients may
tion, if applicable. When possible (i.e. in nonurgent eventually be to participate in ambulation and/or
cannulation situations), it may be advantageous to other rehabilitation exercises to their optimal ability.
discuss these goals prior to ECMO cannulation to
help mentally prepare the patient and family for the
challenges once cannulated. Patient family or sup-
port members can be essential in relieving anxiety.
For some patients, it may be helpful for family or
support members to hold signs of encouragement
along the walk or play music.

Long-term Outcomes

An important consideration in deciding to initi-

ate ECMO support for patients with severe cardiac
and/or respiratory failure pertains to long-term
outcomes, including exercise tolerance, lung/cardiac

228
 Ambulation and ECLS

References awaiting lung transplantation: a practical ap-

proach. Crit Care Med. 2011;39(12):2593-2598.
1. Hermans G, De Jonghe B, Bruyninckx F, Van 12. Fuehner T, Kuehn C, Hadem J, et al. Extracor-
den Berghe G. Clinical review: Critical ill- poreal membrane oxygenation in awake patients
ness polyneuropathy and myopathy. Crit Care. as bridge to lung transplantation. Am J Respir
2008;12(6):238. Crit Care Med. 2012;185(7):763-768.
2. Inci I, Klinzing S, Schneiter D, et al. Outcome 13. Garcia JP, Kon ZN, Evans C, et al. Ambula-
of Extracorporeal Membrane Oxygenation tory veno-venous extracorporeal membrane
as a Bridge To Lung Transplantation: An In- oxygenation: innovation and pitfalls. J Thorac
stitutional Experience and Literature Review. Cardiovasc Surg, 2011;142(4):755-761.
Transplantation, 2015;99(8): 1667-1671. 14. Schmidt F, Sasse M, Boehne M, et al. Concept
3. Rehder KJ, Turner DA, Hartwig MG, et al. of ‘awake venovenous extracorporeal mem-
Active rehabilitation during extracorporeal brane oxygenation’ in pediatric patients await-
membrane oxygenation as a bridge to lung ing lung transplantation. Pediatr Transplant.
transplantation. Respir Care. 2013;58(8):1291- 2013;17(3):224-230.
1298. 15. Cheifetz IM. Extracorporeal membrane oxygen-
4. Nosotti M, Rosso L, Tosi D, et al. Extracorpo- ation of the future: smaller, simpler, and mobile.
real membrane oxygenation with spontaneous Pediatr Transplant. 2013;17(3):202-204.
breathing as a bridge to lung transplantation. In- 16. Turner DA, Rehder KJ, Bonadonna D, et al.
teract Cardiovasc Thorac Surg. 2013;16(1):55- Ambulatory ECMO as a Bridge to Lung Trans-
59. plant in a Previously Well Pediatric Patient with
6. Shehabi Y, Bellomo R, Reade MC, et al. Early ARDS. Pediatrics. 2014;134(2):e583-585.
intensive care sedation predicts long-term mor- 17. Schmidt M, Stewart C, Bailey M, et al. Me-
tality in ventilated critically ill patients. Am J chanical ventilation management during ex-
Respir Crit Care Med. 2012;186(8):724-731. tracorporeal membrane oxygenation for acute
7. Shehabi Y, Chan L, Kadiman S, et al. Sedation respiratory distress syndrome: a retrospective
depth and long-term mortality in mechanically international multicenter study. Crit Care Med.
ventilated critically ill adults: a prospective 2015;43(3):654-664.
longitudinal multicentre cohort study. Intensive 18. Marhong JD, Telesnicki T, Munshi L,. Mechani-
Care Med. 2013;39(5):910-918. cal ventilation during extracorporeal membrane
8. Shehabi Y, Bellomo R, Reade MC, et al. Early oxygenation. An international survey. Ann Am
goal-directed sedation versus standard sedation Thorac Soc. 2014;11(6):956-961.
in mechanically ventilated critically ill patients: 19. Polastri M, Loforte A, Dell’Amore A, Nava S.
a pilot study. Crit Care Med. 2013;41(8):1983- Physiotherapy for Patients on Awake Extracor-
1991. poreal Membrane Oxygenation: A Systematic
9. Morris PE, Goad A, Thompson C, et al. Early Review. Physiother Res Int. 2016;21:203–209.
intensive care unit mobility therapy in the treat- 20. Lehr CJ, Zaas DW, Cheifetz IM, Turner DA.
ment of acute respiratory failure. Crit Care Med. Ambulatory extracorporeal membrane oxygen-
2008;36(8):2238-2243. ation as a bridge to lung transplantation: walk-
10. Schweickert WD, Pohlman MC, Pohlman AS, ing while waiting. Chest. 2015;147(5):1213-
et al. Early physical and occupational therapy 1218.
in mechanically ventilated, critically ill pa- 21. Schwartz SP, Bonadonna D, Hartwig MG,
tients: a randomised controlled trial. Lancet. Cheifetz IM. Bedside Tracheostomy on Pe-
2009;373(9678):1874-1882. diatric ICU Subjects Supported by Extracor-
11. Turner DA, Cheifetz IM, Rehder KJ, et al. Ac- poreal Membrane Oxygenation. Respir Care.
tive rehabilitation and physical therapy during 2017;62(11):1447-1455.
extracorporeal membrane oxygenation while

229
Chapter 27

22. Reeb J, Olland A, Renaud S, et al. Vascular 32. Luyt CE, Combes A, Becquemin MH, et al.
access for extracorporeal life support: tips and Long-term outcomes of pandemic 2009 influ-
tricks. J Thorac Dis. 2016;8(Suppl 4):S353-363. enza A(H1N1)-associated severe ARDS. Chest.
23. Wells CL, Forrester J, Vogel J, et al. Safety 2012;142(3):583-592.
and Feasibility of Early Physical Therapy for
Patients on Extracorporeal Membrane Oxygen-
ator: University of Maryland Medical Center
Experience. Crit Care Med. 2018;46(1):53–59.
24. Ranney DN, Benrashid E, Meza JM, et al.
Central Cannulation as a Viable Alternative to
Peripheral Cannulation in Extracorporeal Mem-
brane Oxygenation. Semin Thorac Cardiovasc
Surg. 2017;29(2):188-195.
25. Keenan JE, Schechter MA, Bonadonna DK, et
al. Early Experience with a Novel Cannulation
Strategy for Left Ventricular Decompression
during Nonpostcardiotomy Venoarterial ECMO.
ASAIO J. 2016;62(3):e30-34.
26. Boling B, Dennis DR, Tribble TA, Rajagopalan
N, Hoopes CW. Safety of Nurse-Led Ambula-
tion for Patients on Venovenous Extracorporeal
Membrane Oxygenation. Prog Transplant. 2016
Jun;26(2):112-116.
27. Herridge MS, Cheung AM, Tansey CM, Matte-
Martyn A, Diaz-Granados N, et al; Canadian
Critical Care Trials Group. One-year outcomes
in survivors of the acute respiratory distress
syndrome. N Engl J Med. 2003;348(8):683-693.
28. Cheung AM, Tansey CM, Tomlinson G, et al.
Two-year outcomes, health care use, and costs of
survivors of acute respiratory distress syndrome.
Am J Respir Crit Care Med. 2006;174(5):538-
544.
29. Dowdy DW, Eid MP, Dennison CR, et al.
Quality of life after acute respiratory distress
syndrome: a meta-analysis. Intensive Care Med.
2006;32(8):1115-1124.
30. Wang CY, Calfee CS, Paul DW, et al. One-year
mortality and predictors of death among hospital
survivors of acute respiratory distress syndrome.
Intensive Care Med. 2014;40(3):388-396.
31. Schmidt M, Zogheib E, Rozé H, et al. The
PRESERVE mortality risk score and analysis
of long-term outcomes after extracorporeal
membrane oxygenation for severe acute respi-
ratory distress syndrome. Intensive Care Med.
2013;39(10):1704-1713.

230
28

Acute Kidney Injury and Renal Support Therapy during ECLS

Rachel Sirignano, MD, Matthew L. Paden, MD

Acute Kidney Injury in ECLS Patients present, and hopefully preventing them, as a goal
to improve overall ECLS outcomes.
Background
Evidence for Treatment of AKI and FO during
Acute kidney injury (AKI) is commonly seen ECLS
and has been demonstrated as a risk factor for death
in critically ill patients of all ages in the ICU.1-3 How- In general, we know very little about the optimal
ever, it has been difficult to say how common and treatment of either AKI or FO during ECLS. There
how much of a risk factor because the definitions are essentially no data in the medical literature
of AKI have varied widely and often based solely regarding optimal dose, timing, method, frequency,
on changes in creatinine. Progress has been made or technique of AKI or FO therapy during ECLS.
over the last decade, and now multiple standard- While focusing mostly on continuous renal replace-
ized scoring systems exist which utilize both urine ment therapy (CRRT) use, Chen et al. have provided
output and serum creatinine to provide a standard the most comprehensive and contemporary evidence
language allowing classification of AKI in the ICU based review of AKI management during ECLS.21
population.4-6 The Acute Dialysis Quality Initiative (ADQI) re-
Previous reports of ECLS patients that evalu- cently published expert consensus guidelines on
ated AKI showed highly variable prevalences which AKI, as well as pharmacological and mechanical
varied depending on age, type of ECLS, and AKI fluid removal and these principles are, in general,
classification system. Reports have demonstrated applicable for ECLS patients.22,23 In addition, the
AKI in 19-71% of neonates, 20-72% of pediatrics, Extracorporeal Life Support Organization (ELSO)
and up to >70% of adult ECLS patients.7-11 A con- publishes a series of guidelines (found at https://
sistent theme in the outcomes of these studies is an www.elso.org/ Resources/Guidelines.aspx) to as-
association between AKI and mortality for ECLS sist ECLS providers. These guidelines provide
patients. Renal injury can also be demonstrated information concerning treatment of AKI, and the
beyond traditional definitions of urine output and importance of fluid management in order to reach
serum creatinine utilizing the concept of fluid over- a homeostasis similar to the patient’s normal (dry
load (FO), or the inability of the kidneys to maintain weight) extracellular fluid volume. Understanding
euvolemia over time. Similar to what was seen in the that the basic clinical science remains sparse, and
traditional definitions of AKI, the previous studies of that most guidance is based on expert opinion, in
FO in ECLS patients consistently demonstrate cor- this chapter, we will cover the therapies and methods
relations with FO and increased mortality, impaired used in ECLS patients (unless otherwise specified),
oxygenation, and increased duration of extracor- with the caveat that additional research is necessary
poreal support.12-20 As both AKI and FO have been to determine best practices and optimal therapy.
associated with negative clinical outcomes, focus
has turned to treating these comorbidities, when

231
Chapter 28

Indications for Treatment of AKI and FO during monly used therapy, CRRT, as an example of how
ECLS to implement and integrate these systems together.

For ECLS patients, indications for renal re- CRRT during ECLS
placement therapy (RRT) are broadly divided into
removal of drugs/toxins, FO, AKI, and electrolyte Hemofilters are common to all CRRT imple-
imbalances that are refractory to medical therapy, mentations. The predominant design consists of
with usage for each of these indications varying multiple hollow fibers through which blood flows,
among institutions.12,24 Some centers use >10% FO surrounded by an open space where ultrafiltrate can
to start RRT, based upon association with increased be created or dialysate can be infused. The hollow
mortality of this population in pediatric intensive fibers can be made from various materials with
care patients not supported with ECLS.25 However, polysulfone and polyacrylonitrile being common
as stated above, both timing and dose for adding worldwide. The hollow fibers have very small pores
RRT to treat AKI and FO remain poorly established in them, which allow the passage of small solutes
in ECLS patients and vary between centers. Man- and fluid, but retain larger blood components such
agement of AKI and FO have included filtration and as blood cells, platelets, and albumin. The size of
dialysis based RRTs as intermittent or continuous these pores also vary by manufacturer, material,
therapies. The majority of RRTs are blood based and model and influence the substances (includ-
therapies; however, peritoneal dialysis is used in ing drugs) that will pass through the filter into the
an important minority of patients, especially in the ultrafiltrate/dialysate. Choice of hemofilters for a
neonatal cardiac population. Based on data from program in general, and for an individual patient,
the ELSO Registry, the most common method for should be made based upon these and other design
treating AKI and FO is by adding CRRT to the features such as fiber internal diameter, fiber thick-
ECLS circuit. ness, and sieving coefficients when targeting specific
molecule removal.
Technical Aspects of RRT during ECLS We now review the 3 main methods for provid-
ing CRRT during ECLS.26,27 The first and simplest
Performing two different extracorporeal thera- method utilizes a second vascular access point for
pies, such as RRT and ECLS, together is difficult and a commercially available CRRT device without
requires a combination of biomedical engineering any connection to the ECLS circuit. This approach
expertise, local knowledge of available device op- is limited to patients with appropriate vascular ac-
tions, and clinical skill. These independent devices cess prior to cannulation. Placement of a dialysis
were not engineered or designed to work together, catheter is possible during ECLS, but an elevated
and so simply achieving forward flow of blood bleeding risk exists once systemic anticoagulation
in them can be a significant hurdle by itself. Also, has begun. Using this approach, CRRT is provided
when looking at this problem worldwide, due to as would be for other intensive care patients, with
country specific regulations and choices by device the notable exception that the systemic anticoagula-
manufacturers, a wide variety of devices may or may tion for ECLS precludes the need for the standard
not be available to any individual center. This vastly local CRRT circuit anticoagulation. This method
increases the number of permutations of choices that has the additional drawback of increasing risk of
can be made when designing these platforms, and catheter associated line infections.
makes a discrete review of all possibilities difficult. The second method for CRRT during ECLS
We thus focus on the engineering, physics, and involves creating a shunt of blood postpump from
physiology commonly seen in currently available the ECLS circuit which flows through an isolated
devices. The reader should determine how best to hemofilter. The positive pressure of the ECLS
balance those factors based on the local choices of pump drives blood through the hemofilter, creat-
devices. Additionally, we focus on the most com- ing an ultrafiltrate as fluid and small solutes exit
the pores in the hemofilter. In many setups, the

232
 Acute Kidney Injury and Renal Support Therapy during ECLS

amount of ultrafiltrate produced is limited by using the hemofilter circuit (usually ~100-200 mL/min).
a standard intravenous (IV) pump programmed for Additionally, the return of blood exiting the hemo-
the number of mL/hour of ultrafiltrate desired and filter prepump, makes CRRT less efficient because
then this volume is measured with a bedside urim- filtered blood can recirculate through the hemofilter
eter as an additional method to improve accuracy. circuit. While this is a theoretical concern, clinically,
The ultrafiltrate can then simply be discarded to the ECLS flow rates greatly exceed the hemofilter
provide slow continuous ultrafiltrate (SCUF), or a circuit flow rates, making clinically significant re-
replacement fluid with an appropriate electrolyte circulation negligible. Also, using SCUF for smaller
composition for the patient can be delivered into children is not recommend as it can create multiple
the circuit via an additional standard IV pump to electrolyte disturbances. The “in-line” circuit often
provide continuous venovenous hemofiltration has no pressure monitoring, which makes identify-
(CVVH). When providing CVVH, a dialysate can ing hemofilter malfunction, thrombosis, or rupture
also be added (in a counter current fashion with challenging. The most important disadvantage to
respect to blood flow) to the outside of the hemo- “in-line” hemofiltration of CRRT relates to inaccu-
filter fibers which provides continuous venovenous racy of fluid balance. The standard IV pumps used
hemodiafiltration (CVVHDF). The blood exiting for control of ultrafiltrate and replacement fluid are
the hemofilter, is now returned to the ECLS circuit not engineered for this therapy and actually func-
pre-pump (Figure 28-1). tion as flow restrictors (and not pumps), having an
This method of “in-line” hemofiltration was inherent error rate of approximately ~12.5% when
the earliest form of tandem extracorporeal support used in an ECLS setting.28 An in vitro setup of “in-
therapies and has the advantage of ease of use, line” CRRT with ECLS, demonstrated a measured
simplicity for all ECMO specialists (did not require vs. prescribed flow error of up to 34 mL/hour (>800
ECMO specialists specifically trained in commercial mL/day), which in a small patient (5 kg) could equal
CRRT devices), and low cost of supplies. However, their daily fluid goals (~150 mL/kg/day).29 Careful
many disadvantages have been noted with this monitoring of the replacement fluid volume as well
method. The creation of a shunt returning blood as ultrafiltrate volumes is essential to the ECMO
prepump creates a discrepancy between the blood specialist duties when providing these therapies to-
flow listed on the ECLS pump and what is actually gether. This can be done either via volume measure
being delivered to the patient. Patient flow rates are (mL) or based on highly accurate scales (+/- 1 gram).
now more accurately measured in the distal “arterial” The need for hourly monitoring increases ECMO
limb of the ECLS circuit just prior to return to the specialist workload and contributes to a reduction
patient. The difference between ECLS pump flow, of use of this “in-line” technique.
and patient delivered flow must be monitored and The third and preferred method for CRRT on
that number represents the amount of flow through ECLS patients is via introduction of a commercially
available CRRT device into the ECLS circuit. Mul-
tiple factors contribute to determining the optimal
method to connect these two, separate, extracorpore-
al support devices including, but not limited to, type
and placement of pressure monitors, CRRT device
software limitations for variables such as pressure
and flow, type of ECLS pump, and ECLS circuit
design. In general, the roller head pump driven
ECLS circuits have positive pressures in the venous
limb, allowing for prepump venous limb placement
and return of the CRRT device (Figure 28-2). In this
configuration, the driving pressure for the CRRT
circuit relies on the CRRT device pump generating
Figure 28-1. “In-line” CRRT during ECMO.
adequate negative pressure to pull blood from the

233
Chapter 28

ECLS circuit and then creating a positive displace- device allows for more accurate fluid balance, a
ment and pressure to drive it through the hemofilter longer hemofilter life, standard pressure and flow
and return it to the ECLS circuit. The pressures in monitoring of the CRRT circuit, and the ability to
the venous limb of a roller head circuit are often use a CRRT device that has been engineered for the
not very negative and are similar to what would be purpose of providing CRRT.29-31 All commercially
expected in a well functioning dialysis catheter, thus available CRRT devices are not specifically ap-
allowing the CRRT device software to function in proved or designed for use during ECLS. Using a
its designed fashion. commercially available CRRT device is this manner
Alternately, a centrifugal pump driven ECLS with ECLS has the disadvantages of higher costs and
circuit has a much larger negative venous limb pres- additional training for the ECMO specialist.
sure that the CRRT software pressure limits typically
inhibit functioning and so an alternative configura- Outcomes of CRRT during ECLS
tion is usually required. Additionally, when making
the physical connections of the CRRT device to Although there are advantages and disadvan-
the ECLS circuit in this highly negative pressure tages of differing methods for providing simulta-
environment the risk of air entrainment into the neous CRRT and ECLS, few outcome data exist
ECLS circuit is high. For a centrifugal ECLS system, to compare methods. Prior studies comparing “in-
recommendations include placing the CRRT device line,” commercial CRRT devices, and stand alone
postpump, but pre-oxygenator, to move the CRRT CRRT via a second vascular site, found all methods
device to a region of positive pressure where from a “adequate” for fluid and electrolyte control. As men-
software alarm based standpoint it is more likely to tioned above, there appears to be disadvantages to
function consistently, as well as decrease the risk of the “in-line” system with decreased fluid accuracy
air entrainment (Figure 28-3). In this configuration, compared to commercially available devices.30-32
the blood is returned from the CRRT device to the These studies also demonstrated a longer filter life
ECLS circuit postpump venous limb, just prior to the when a CRRT device was incorporated into the
membrane oxygenator. Having the CRRT both drain ECLS circuit (138.4 hour) compared to 36.8 hours
and return postpump but pre-oxygenator is optimal seen in non-ECLS children with standalone CRRT,
because this configuration allows the oxygenator to and 27.2 hours seen in CRRT during ECLS with
act as a clot and air trap, and decreases the amount citrate anticoagulation added. However, no differ-
of recirculation through the CRRT circuit. ence was noted in ICU length of stay (LOS), hospital
When compared to the “in-line” method, the LOS, or mortality.
incorporation of a commercially available CRRT

Figure 28-2. CRRT using a commercially available Figure 28-3. CRRT using a commercially available
device and a roller head ECLS circuit. device and a centrifugal ECLS circuit.

234
 Acute Kidney Injury and Renal Support Therapy during ECLS

The largest set of outcome data for concomitant initiation (p=0.05), and the peak FO on ECLS (p
CRRT on ECLS comes from the ELSO Registry.33 <0.0001) predicted duration of ECLS in survivors.
The ELSO Registry defines “renal failure” as a com-
plication of ECLS, with three levels of injury coded: Summary of Concomitant CRRT and ECLS
creatinine (Cr); 1.5-3, Cr >3, and use of dialysis/he-
mofiltration/continuous arteriovenous hemodialysis. AKI and FO occur commonly in ECLS patients,
These definitions are limited compared to modern and RRT is commonly used in this population.
scoring systems and historically the Registry only However, methods of providing these extracorporeal
counts only one episode per run with no duration support therapies together are not standardized and
information. This complicates analysis of outcomes, no products are commercially designed for this use.
because there is no differentiation between a patient As in all critically ill patients, mortality of ECLS
who received CRRT for their entire 14 day ECLS patients who develop AKI exceeds those patients
course vs. someone who received it only several without AKI. In contrast to previous thinking, it has
hours before death. With those caveats, analysis of been demonstrated that the presence and degree of
the ELSO Registry demonstrates that renal failure AKI is the risk factor for death, not the use of the
in each age category of respiratory failure has as- CRRT device. Additionally, FO has also been shown
sociated worse survival (neonatal 37%, pediatric to be correlated to higher mortality and longer ECLS
34%, adult 45%) compared to the overall survivals duration. Both AKI and FO remain potential clini-
of these categories (neonatal 63%, pediatric 58%, cal targets to improve mortality in ECLS patients.
adult 61%). Similar data are noted in the cardiac However, specifics for optimal treatment of AKI and
population. FO for ECLS patients do not exist, and therefore
As coded in the ELSO Registry, both kidney in- wide variation among institutions exists. Additional
jury and the use of renal replacement therapy (RRT) multicenter data and trials with standardized proto-
are associated with increased mortality. A subset of cols are needed to better guide AKI management in
6 ELSO centers across North America came together ECLS patients.
and formed the Kidney Injury During Membrane
Oxygenation (KIDMO) research network to further
investigate relationships between RRT use, AKI,
and survival in pediatric patients (<19 years old).12,22
They evaluated 835 ECLS patients at their centers
from 2007-2011 using a modern definition of AKI
and found that AKI affects the majority of pediatric
ECLS patients (50-69%), occurs early (99% within
first 48 hours) and is associated with longer duration
of ECLS (~ 48 hours) and higher mortality (Odds
Ratio=2). There was a clear association in these
data between increasing AKI stage being associated
with increased risk of death. Fluid overload >10% is
present in almost half (46.4%) and >20% in almost
one quarter (24.1%) of ECLS patients at the time
of cannulation.34 Often FO worsens once on ECLS,
with 84.8%, 67.2%, and 29% of patients having a
peak FO greater than or equal to 10%, 20%, and
50% respectively. The magnitude of FO correlated
with survival, with median peak FO being lower in
patients who survived to hospital discharge (24.8%
vs. 43.3%; p <0.0001). The degree of FO at ECLS

235
Chapter 28

References 11. Luo XJ, Wang W, Hu SS, et al. Extracorporeal

membrane oxygenation for treatment of cardiac
1. Selewski DT, Charlton JR, Jetton JG, et al. failure in adult patients. Interact Cardiovasc
Neonatal Acute Kidney Injury. Pediatrics. Thorac Surg. 2009;9(2):296-300.
2015;136(2):e463-73. 12. Selewski DT, Askenazi DJ, Bridges BC, Cooper
2. Fortenberry JD, Paden ML, Goldstein SL. DS, Fleming GM, Paden ML, et al. The Impact
Acute kidney injury in children: an update on of Fluid Overload on Outcomes in Children
diagnosis and treatment. Pediatr Clin North Am. Treated With Extracorporeal Membrane Oxy-
2013;60(3):669-688. genation: A Multicenter Retrospective Cohort
3. Honore PM, Jacobs R, Hendrickx I, et al. Pre- Study. Pediatr Crit Care Med. 2017.
vention and treatment of sepsis-induced acute 13. Arikan AA, Zappitelli M, Goldstein SL, Nai-
kidney injury: an update. Ann Intensive Care. paul A, Jefferson LS, Loftis LL. Fluid overload
2015;5(1):51. is associated with impaired oxygenation and
4. Bellomo R, Kellum JA, Ronco C. Defining and morbidity in critically ill children. Pediatr Crit
classifying acute renal failure: from advocacy to Care Med. 2012;13(3):253-258.
consensus and validation of the RIFLE criteria. 14. Sutherland SM, Zappitelli M, Alexander SR,
Intensive Care Med. 2007;33(3):409-413. et al. Fluid overload and mortality in children
5. Bellomo R, Ronco C, Kellum JA, Mehta RL, receiving continuous renal replacement therapy:
Palevsky P. Acute Dialysis Quality Initiative the prospective pediatric continuous renal re-
w. Acute renal failure - definition, outcome placement therapy registry. Am J Kidney Dis.
measures, animal models, fluid therapy and 2010;55(2):316-325.
information technology needs: the Second In- 15. Davison D, Basu RK, Goldstein SL, Chawla
ternational Consensus Conference of the Acute LS. Fluid management in adults and children:
Dialysis Quality Initiative (ADQI) Group. Crit core curriculum 2014. Am J Kidney Dis.
Care. 2004;8(4):R204-212. 2014;63(4):700-712.
6. Mehta RL, Kellum JA, Shah SV, et al. Acute 16. Roy BJ, Cornish JD, Clark RH. Venovenous
Kidney Injury Network: report of an initiative extracorporeal membrane oxygenation affects
to improve outcomes in acute kidney injury. Crit renal function. Pediatrics. 1995;95(4):573-578.
Care. 2007;11(2):R31. 17. Swaniker F, Kolla S, Moler F, et al. Extracor-
7. Askenazi DJ, Ambalavanan N, Hamilton K, et poreal life support outcome for 128 pediatric
al. Acute kidney injury and renal replacement patients with respiratory failure. J Pediatr Surg.
therapy independently predict mortality in 2000;35(2):197-202.
neonatal and pediatric noncardiac patients on 18. Weber TR, Kountzman B. Extracorporeal mem-
extracorporeal membrane oxygenation. Pediatr brane oxygenation for nonneonatal pulmonary
Crit Care Med. 2011;12(1):e1-6. and multiple-organ failure. J Pediatr Surg.
8. Smith AH, Hardison DC, Worden CR, Fleming 1998;33(11):1605-1609.
GM, Taylor MB. Acute renal failure during 19. Kelly RE, Jr., Phillips JD, Foglia RP, et al.
extracorporeal support in the pediatric cardiac Pulmonary edema and fluid mobilization as
patient. ASAIO J. 2009;55(4):412-416. determinants of the duration of ECMO support.
9. Lou S, MacLaren G, Paul E, Best D, Delzoppo J Pediatr Surg. 1991;26(9):1016-1022.
C, Butt W. Hemofiltration is not associated 20. Selewski DT, Cornell TT, Blatt NB, et al. Fluid
with increased mortality in children receiving overload and fluid removal in pediatric patients
extracorporeal membrane oxygenation. Pediatr on extracorporeal membrane oxygenation re-
Crit Care Med. 2015;16(2):161-166. quiring continuous renal replacement therapy.
10. Gadepalli SK, Hirschl RB. Extracorporeal life Crit Care Med. 2012;40(9):2694-2699.
support: updates and controversies. Semin Pe- 21. Chen H, Yu RG, Yin NN, Zhou JX. Combina-
diatr Surg. 2015;24(1):8-11. tion of extracorporeal membrane oxygenation
and continuous renal replacement therapy in

236
 Acute Kidney Injury and Renal Support Therapy during ECLS

critically ill patients: a systematic review. Crit in series with extracorporeal membrane oxygen-
Care. 2014;18(6):675. ation. Kidney Int. 2009;76(12):1289-1292.
22. Goldstein S, Bagshaw S, Cecconi M, et al. Phar- 32. Shum HP, Kwan AM, Chan KC, Yan WW.
macological management of fluid overload. Br The use of regional citrate anticoagulation
J Anaesth. 2014;113(5):756-763. continuous venovenous hemofiltration in ex-
23. Rosner MH, Ostermann M, Murugan R, et al. tracorporeal membrane oxygenation. ASAIO
Indications and management of mechanical J. 2014;60(4):413-418.
fluid removal in critical illness. Br J Anaesth. 33. Thiagarajan RR, Barbaro RP, Rycus PT, et
2014;113(5):764-771. al. Extracorporeal Life Support Organization
24. Fleming GM, Askenazi DJ, Bridges BC, et Registry International Report 2016. ASAIO J.
al. A multicenter international survey of renal 2017;63(1):60-67.
supportive therapy during ECMO: the Kidney 34. Selewski DT, Askenazi DJ, Bridges BC, et al.
Intervention During Extracorporeal Membrane The impact of fluid overload on outcome of
Oxygenation (KIDMO) group. ASAIO J. children treated with extracorporeal membrane
2012;58(4):407-414. oxygenation: A multicenter retrospective cohort
25. Goldstein SL, Currier H, Graf JM, Cosio CC, study. Pediatr Crit Care Med 2017;18(12):1126-
Brewer ED, Sachdeva R. Outcome in children 1135.
receiving continuous renal veno-venous hemo-
filtration. Pediatrics 2001;107:1309-1312
26. Askenazi DJ, Selewski DT, Paden ML, et al. Re-
nal replacement therapy in critically ill patients
receiving extracorporeal membrane oxygen-
ation. Clin J Am Soc Nephrol. 2012;7(8):1328-
1336.
27. Laverdure F, Masson L, Tachon G, Guihaire J,
Stephan F. Connection of a Renal Replacement
Therapy or Plasmapheresis Device to the ECMO
Circuit. ASAIO J. 2017.
28. Jenkins R, Harrison H, Chen B, Arnold D, Funk
J. Accuracy of intravenous infusion pumps in
continuous renal replacement therapies. ASAIO
J. 1992;38(4):808-810.
29. Sucosky P, Dasi LP, Paden ML, Fortenberry JD,
Yoganathan AP. Assessment of current continu-
ous hemofiltration systems and development of
a novel accurate fluid management system for
use in extracorporeal membrane oxygenation.
Journal of Medical Devices, Transactions of the
ASME. 2008;2(3).
30. Symons JM, McMahon MW, Karamlou T,
Parrish AR, McMullan DM. Continuous renal
replacement therapy with an automated monitor
is superior to a free-flow system during extra-
corporeal life support. Pediatr Crit Care Med.
2013;14(9):e404-408.
31. Santiago MJ, Sanchez A, Lopez-Herce J, et al.
The use of continuous renal replacement therapy

237
29

Extracorporeal Support with Therapeutic Apheresis

Rachel Sirignano, MD, Meral Patel, MD, Matthew L. Paden, MD, James D. Fortenberry, MD, MCCM

Introduction Therapeutic cytapheresis involves the removal

of a specific cellular element. In erythrocytapheresis,
As providers have become more comfortable red blood cells (RBC) are removed from the blood
with extracorporeal support, the number of patients and replaced with banked RBC. The most common
with increased complexity and with multiple organ use for erythrocytapheresis occurs in the setting of
failure has grown. ECLS provides the physiologic complications from sickle cell disease, but has other
stability that can serve as a “platform” upon which rare uses as well. Leukapheresis, the removal of
multiple organ support therapies can be delivered. In white blood cells from the blood, is most commonly
this chapter, we discuss the use of therapeutic apher- performed to harvest stem cells for bone marrow
esis procedures during ECLS, review the limited transplantation. It also has indications in acute
clinical experience with apheresis procedures during leukemias with hyperviscosity syndrome leading
ECLS, and offer a guide for successful provision of to organ failure as well as a potential therapy for
these tandem procedures. severe pertussis infection.
TPE is the separation and replacement of the
Apheresis during ECLS plasma from blood. The aim of TPE is to remove
large molecular weight substances (such as cy-
Apheresis separates the blood into its individual tokines and antibodies), or highly protein bound
components. Commonly used therapeutic apheresis molecules while also restoring depleted coagulation
procedures include therapeutic plasma exchange factors, proteins, and enzymes to regain homeostasis
(TPE), erythrocytapheresis, and leukapheresis. necessary for clinical recovery.2 TPE is the most
These therapies have an evolving history in medi- commonly provided apheresis procedure during
cal care and comprehensive reviews of this topic ECLS, and will be the focus of this review.
are available.1 Use of these therapies represents the Commercially available apheresis devices
standard of standalone care for diseases such as separate the blood into its component parts, either
acute stroke in sickle cell disease or for thrombotic by centrifugation or by filtration. Historically, cen-
thrombocytopenic purpura and may play a valuable trifugation techniques are the most commonly used.
role in the treatment of other diseases. However, use However, the introduction of plasma filters used
of apheresis procedures is much less established in conjunction with continuous renal replacement
in critically ill patients, especially those requiring devices has increased filtration applications during
ECLS. Concomitant ECLS was previously consid- the last decade. Plasma filtration devices have the
ered a contraindication to providing apheresis due to disadvantage of performing only TPE while dedicat-
perceived patient instability. Increasingly, providers ed centrifugal apheresis devices can perform many
perform any or all of these therapies during ECLS, different apheresis procedure types. No evidence
even placing patients on ECLS to achieve the car- supports outcome differences between centrifuga-
diorespiratory stability necessary to complete the tion and filtration, and choice of method is based on
apheresis procedure. local equipment, staffing resources, and physician
experience. Regardless of method, anticoagulation

239
Chapter 29

is required to avoid excessive thrombosis due to certain patients. These groups of conditions are
activation of blood in the extracorporeal circuit. considered Category IV and a contraindication to
Citrate is most commonly used outside of the ECLS performing therapeutic apheresis (Table 29-1).3
setting. However, the heparin provided for ECLS
anticoagulation may be sufficient for this circuit as Technical Aspects of Apheresis during ECLS
well (see Chapter 4).
Similar to the CRRT methods discussed in
Indications and Contraindications Chapter 28, apheresis procedures can be provided
as standalone therapy through a separate venous ac-
Currently no ELSO guidelines exist pertaining cess or through the ECLS circuit for venous access.
to indications for the use of apheresis procedures No evidence exists to suggest superiority of either
on ECLS. Instead, the decision to proceed with method, although integration using the ECLS circuit
apheresis should be based on evidence of effective- appears to be more common in the literature, likely
ness for the underlying disease and not presence of due to ease of use, decreased risk of infection, and
concomitant ECLS support. General indications for the potential complications of attaining new venous
apheresis are established by the American Society access on an anticoagulated ECLS patient.
for Apheresis (ASFA). AFSA publishes a set of For standalone therapy, apheresis procedures
evidence-based guidelines for the use of apheresis, can potentially be done with two large bore pe-
most recently in 2016.3 These guidelines review
the medical evidence for apheresis use based on CATEGORY INDICATIONS/CONTRAINDICATIONS
underlying disease, and present a detailed summary Category I Renal and Liver transplantation*
of recommendations. Based on the AFSA clinical Diffuse alveolar hemorrhage (ANCA
associated granulomatosis with polyangiitis)
evidence review, indications for TPE are identified Diffuse alveolar hemorrhage (Goodpasture’s
as categories I-IV: Category I, apheresis is consid- syndrome)
Atypical HUS (Factor H Antibodies)
ered a first line therapy; Category II, a second line Thrombotic thrombocytopenic purpura
therapy; Category III, data are unclear on its benefit; Hyperviscosity in monoclonal gammopathies
Fulminant Wilson’s disease
and Category IV, evidence suggests apheresis is Category II Acute disseminated encephalomyelitis
harmful or ineffective. Catastrophic antiphospholipid syndrome
Category I indications which patients may ABO-incompatible HSCT
Atypical HUS (complement mutations)
experience while on ECLS are listed in Table 29-1. Mushroom poisoning/overdose
Increasingly TPE has been used during ECLS for the Severe acute chest syndrome (Sickle cell
related)
management of sepsis with multiorgan dysfunction Severe systemic lupus erythematosus
syndrome (MODS) and thrombocytopenia associ- Category III Cardiac Transplantation Antibody Mediated
Rejection
ated multiorgan failure (TAMOF).4 These uses are Sepsis with multiorgan failure
listed as Category III indications in the most recent Thrombocytopenia associated multiorgan
failure
AFSA guidelines (Table 29-1), as little evidence Category IV Amyloidosis
exists for concomitant use of apheresis and ECLS. Amyotrophic lateral sclerosis
ASFA provides guidance for the clinician, Dermatomyositis/polymyositis
Atypical HUS (membrane cofactor protein
which us a modification of McLeod’s Criteria mutation)
(Table 29-2).3 In cases where guidelines are not pres- Refractory immune thrombocytopenia
ABO incompatible deceased donor renal
ent, prior to initiating treatment, clinicians should transplant
document their reasoning for potential mechanism Lupus nephritis
Idiopathic polyarteritis nodosa vasculitis
of benefit for the specific disease state, potential ASFA=American Society for Apheresis; ANCA=anti-neutrophil
for the apheresis procedure to improve the patient’s cytoplasmic antibody; HUS=hemolytic uremic syndrome;
condition, the proposed therapeutic plan and dura- HSCT=hematopoietic stem cell transplantation
*Antibody mediated rejection and desensitization.
tion, and plan for assessment of response to therapy.
Based on 2016 evidence-based AFSA guide- Table 29-1. Apheresis category indications and
lines, apheresis is ineffective or even harmful in contraindications. From Schwartz et al.3

240
 Extracorporeal Support with Therapeutic Apheresis

ripheral catheters. Ideally, however, patients should or increasing the temperature goals for the ECLS
have a double lumen central venous line designed circuit may be necessary. Therapeutic apheresis
for dialysis that is greater than or equal to 7 French is an intermittent procedure, and when not in use,
in diameter, or a preexisting subcutaneous port venous access catheters should be locked with an
designed for apheresis use. A commercially avail- anticoagulant solution, per institutional protocol, to
able automated blood cell separator that controls prevent thrombosis.
fluid balance and maintains normal plasma volume More commonly, the apheresis circuit connects
is used to perform the apheresis procedure. Most directly to the ECLS circuit in a similar configura-
apheresis circuits have an extracorporeal volume of tion to CRRT (see Chapter 28).5 Again, careful atten-
approximately 350 mL. For patients who weigh less tion to device placement is required based on type of
than 20 kg, a blood prime approximating that used ECLS pumping system. As with CRRT, none of the
for ECLS can be used to avoid dilutional anemia. commercially available apheresis devices have been
For larger patients, priming with crystalloid is suf- engineered and validated for use in the ECLS circuit.
ficient. Blood prime should also be considered in any A common complication is a negative pressure alarm
patient with hemodynamic instability or inadequate on the venous limb of the apheresis device. This
oxygen delivery. When used as standalone therapy, can be ameliorated by improving cannula position
attention should be paid to patient temperature to improve drainage, reducing ECLS flow (if clini-
as hypothermia may be exacerbated by this addi- cally acceptable), altering the position of apheresis
tional extracorporeal volume, especially in smaller drainage from the ECLS circuit, or changing the
patients. Warming of either the apheresis circuit parameters for the alarm limits on the apheresis
device. For patients already receiving concomitant
MCLEOD’S
EVIDENCE CRITERIA EXPLANATION ECLS and CRRT (Figure 29-1), the apheresis device
Mechanism “Plausible The current can be added in series to the CRRT limb, obviat-
pathogenesis” understanding of ing the need for an additional dialysis catheter.6 At
the disease Emory University/Children’s Healthcare of Atlanta,
process supports a
we attach two three-way stopcocks to the pigtail
clear rationale for
the use of the connection coming off the venous bladder. During
therapeutic the procedure, the apheresis device pulls blood from
apheresis the connection of the first stopcock and returns to
modality. circuit connection of the second stopcock, where the
Correction “Better blood” The abnormality,
which makes
blood then continues on into the CRRT circuit and
therapeutic ultimately back to the ECLS circuit. In this method,
apheresis the CRRT and apheresis blood flow rates must be
plausible, can be the same to avoid recirculation and pressure alarms.
meaningfully
Determination of dose and duration of therapy
corrected by its
use. relies on the ASFA guidelines. However, important
Clinical “Perkier There is a strong dosing calculation differences should be considered
Effect patients” evidence that when providing these therapies during ECLS. In the
therapeutic ASFA guidelines, dosing is often based on a multi-
apheresis confers
benefit that is plier of “plasma volume,” “whole blood volume,” or
clinically “red blood cell mass”.3 These numbers are usually
worthwhile, and calculated based on weight or body surface area.
not just For example, estimated blood volume in newborns
statistically
to 3 month old infants is 80-90 mL/kg, for children
significant.
over 3 months old 70 mL/kg, and approximately 5
Table 29-2. Modified McLeod’s criteria for L for an adult. Many of these numbers are estimates
evaluation of therapeutic apheresis efficacy. From of the Nadler equation which provides a more
Schwartz et al.3 thorough calculation of total blood volume. Total

241
Chapter 29

plasma volume (PV) is often assumed to be 60% of the extracorporeal volume would lead to a greater
total blood volume, based on a normal hematocrit than 50% underdosing in this child.
of 40%. During ECLS, the assumptions built into For TPE (depending on disease), ASFA guide-
these equations are invalid and should be considered lines usually recommend removal of 1-1.5 PV,
prior to prescribing therapy. which is replaced with either albumin or fresh frozen
During ECLS, one must calculate both the total plasma (FFP).3 Strict dosing to the individual mL is
patient blood volume and the total extracorporeal not necessary for apheresis procedures, and for prac-
circuit volume. Failure to do so will provide inad- tical purposes the dose can be rounded to the nearest
equate dosing of the apheresis therapy. Total extra- unit of product (albumin or FFP). For plasma-based
corporeal circuit volume should include the sum of molecules capable of extraction by TPE, removal is
each individual extracorporeal circuit being used estimated to be 63.2% of total concentration with
(ECLS+CRRT+Apheresis+etc.) Similarly, total processing of 1 PV, 77.7% with 1.5 PV, 86.5% with
PV should be calculated prior to each procedure 2 PV, and 95% with 3 PV. If processing more plasma
using a recent hematocrit. The discrepancies can results in greater reductions, then one might ask why
be significant, especially with small children. For the recommendations usually only call for 1-1.5 PV
example, using standard AFSA calculations, a 10 kg dosing. Many of the molecules targeted for TPE
child would have a total blood and plasma volume are not solely confined to the blood compartment,
of 700 mL and 420 mL respectively, for dosing. and redistribute with time necessitating multiple
Accounting for an ECLS volume of 400 mL, CRRT therapies over time. Additionally, the benefit of
volume of 250 mL, and apheresis volume of 350 increased percentage removal must be balanced by
mL, and hematocrit of 35%, the same child has a both the amount of additional blood products needed
total blood and plasma volume for dosing of 1700 for replacement, and the additional time involved
mL and 1105 mL respectively. Not accounting for in performance of the procedure. Also, other blood

Figure 29-1. Representative schematic diagram of multi-tandem extracorporeal procedures: ECMO, TPE, and
CRRT. Note: ECMO flow rates: 250 mL-5000 mL/min; TPE flow rates: 30 mL/min-70 mL/min; CRRT flow
rates: 50 mL/min- 150 mL/min. (ECMO, extracorporeal membrane oxygenation; TPE, therapeutic plasma
exchange; CRRT, continuous renal replacement therapy; ▪ = three way stopcock valve). Reproduced with
permission from ASAIO J.

242
 Extracorporeal Support with Therapeutic Apheresis

components and medications are removed while volume ordered and either may fail to run, or default
focusing on removal of the target molecule. Drug to providing the procedure over extended periods
dosing in particular becomes exceptionally com- of time (>8 hours). In such situations, the dura-
plicated with the use of these multiple methods tion can be extended, but an alternative, preferred
of extracorporeal elimination, such as when using solution is to adjust the entered patient weight to
ECLS, CRRT, and apheresis concomitantly. While reflect a similar native (non-ECLS) blood volume.
some guidelines and pharmacokinetics data exist, For example, the 10 kg patient on ECLS described
further study needs to be performed. Careful, daily, above with a total blood volume of 1700 mL could
repeated clinical assessment of the effect of each be entered as a 24 kg patient and this would allow
prescribed drug is required to evaluate for signs of processing of a similar volume of plasma over a
both potential over- and under-dosing.7,8 much more reasonable time period (~2 hours). In
In general, additional anticoagulation for any circumstance where one is considering overrid-
apheresis is unnecessary while on ECLS, as hepa- ing or altering manufacturers recommended usage, a
rin anticoagulation (or direct thrombin inhibitors) formal protocol should be established for the center,
usually suffices for total extracorporeal circuit an- with multiple independent practitioners checking
ticoagulation. Increased monitoring of total circuit the calculations prior to initiating each procedure.
anticoagulation both during and immediately after
apheresis should be considered. Provision of citrate Use and Outcomes of Apheresis during ECLS
anticoagulation for apheresis, while possible in the
ECLS circuit, was associated with hypocalcemia A recent review of the literature identified nu-
and hypotensive complications in a large cohort merous reports (172 total patients) of patients treated
of both children and adults.5 Consideration should with TPE during ECLS. No randomized clinical
be made of the effect of citrate anticoagulation and trial on any single disease has been published that
of the additional blood products used during the includes the use of apheresis techniques during
procedure. Citrate provides anticoagulation through ECLS. A report by Dyer et al. provides the largest
depletion of free, ionized extracellular calcium experience of 293 concomitant apheresis and ECLS
needed for many steps in the coagulation cascade. in 76 adult and pediatric patients.5 In this cohort, the
Cardiac function may be negatively affected (es- most common pediatric indication was multisystem
pecially in neonates) by low ionized calcium. The organ failure and the most common adult indication
anticoagulant effect of large amounts of citrate may was humoral transplant rejection. Both ECLS and
exacerbate bleeding potential in ECLS patients. In apheresis teams provided the combined therapies.
both cases, the negative effects of citrate on calcium Anticoagulation was provided to the ECLS circuit
can be mitigated by infusion of additional calcium. with heparin and citrate for the apheresis circuit.
As none of the commercially available apheresis Citrate related complications of hypocalcemia
devices are engineered or validated for use during (47% pediatric, 27% adult) and hypotension (22%
ECLS, anticoagulation issues remain a concern. pediatric, 34% adult) were both seen and success-
For example, some of the commercially available fully managed. Over the remainder of the published
apheresis devices will not run without a bag of cases, the most common indication for apheresis
anticoagulant attached and infusing. During ECLS was sepsis with either TAMOF or MODS.4,6,9-12
a bag of normal saline can be substituted for the The next most prevalent category was for antibody
citrate allowing the apheresis device to mitigate this removal in active autoimmune diseases13-22 and
limitation. Similarly, some commercially available organ transplantation.23-27 Individual case reports
apheresis devices calculate estimated blood and exist for use in ingestions,28-30 hypoxic ischemic
plasma volumes based on Nadler’s formula and encephalopathy,31 hemophagocytic lymphohistiocy-
provider entered weight and height. As seen above, tosis,32 severe hemolysis while on cardiopulmonary
when the much larger values are provided based bypass for cardiac transplant,33 and drug reaction
on patient plus extracorporeal volume, the devices with eosinophilia and systemic symptoms (DRESS)
will potentially alarm due to the large exchange syndrome associated myocarditis (see Table 29-1).34

243
Chapter 29

Significant heterogeneity exists in these case re-

ports, due to the variety of rare diseases, differences
in techniques of ECLS (mode, equipment, etc.), dif-
ferences in techniques of apheresis (centrifugation
vs. filtration, dose, timing of initiation, etc.), and
lack of severity of illness scoring. Combined, these
factors make analyzing outcomes, including compli-
cations and survival, difficult. As has been done with
the ELSO Registry for ECLS, capturing additional
data about patient, prescription, and device related
factors in an international database would permit
the accrual of data that may allow definition of best
practices and ultimately improve outcomes.

Summary of Apheresis during ECLS

Use of apheresis procedures during ECLS is

technologically feasible and may provide benefit.
AFSA guidelines can provide assistance regarding
appropriateness of apheresis techniques during
ECLS, as well as timing and dosing information.
When evidence for use is not present in the medi-
cal literature, reliance on the modified McLeod’s
criteria and the underlying pathophysiology of the
disease is necessary. Variation exists on how apher-
esis procedures are performed during ECLS and
anticoagulation strategies and deserves further study.

244
 Extracorporeal Support with Therapeutic Apheresis

References modynamic shock. Pediatr Crit Care Med.

2011;12(2):e87-89.
1. Ward DM. Conventional apheresis therapies: a 11. Tabbutt S, Leonard M, Godinez RI, et al. Severe
review. J Clin Apher. 2011;26(5):230-238. influenza B myocarditis and myositis. Pediatr
2. Kawai Y, Cornell TT, Cooley EG, et al. Thera- Crit Care Med. 2004;5(4):403-406.
peutic plasma exchange may improve hemo- 12. Mok Q, Butt W. The outcome of children admit-
dynamics and organ failure among children ted to intensive care with meningococcal septi-
with sepsis-induced multiple organ dysfunction caemia. Intensive Care Med. 1996;22(3):259-
syndrome receiving extracorporeal life support. 263.
Pediatr Crit Care Med. 2015;16(4):366-374. 13. Hohenforst-Schmidt W, Petermann A, Visouli
3. Schwartz J, Padmanabhan A, Aqui N, et al. A, et al. Successful application of extracorpo-
Guidelines on the Use of Therapeutic Apheresis real membrane oxygenation due to pulmonary
in Clinical Practice-Evidence-Based Approach hemorrhage secondary to granulomatosis with
from the Writing Committee of the American polyangiitis. Drug Des Devel Ther. 2013;7:627-
Society for Apheresis: The Seventh Special Is- 633.
sue. J Clin Apher. 2016;31(3):149-162. 14. Barnes SL, Naughton M, Douglass J, Murphy
4. Nguyen TC, Cruz MA, Carcillo JA. Throm- D. Extracorporeal membrane oxygenation with
bocytopenia-Associated Multiple Organ Fail- plasma exchange in a patient with alveolar
ure and Acute Kidney Injury. Crit Care Clin. haemorrhage secondary to Wegener’s granulo-
2015;31(4):661-674. matosis. Intern Med J. 2012;42(3):341-342.
5. Dyer M, Neal MD, Rollins-Raval MA, Raval 15. Ahmed SH, Aziz T, Cochran J, Highland K. Use
JS. Simultaneous extracorporeal membrane of extracorporeal membrane oxygenation in a
oxygenation and therapeutic plasma exchange patient with diffuse alveolar hemorrhage. Chest.
procedures are tolerable in both pediatric and 2004;126(1):305-309.
adult patients. Transfusion. 2014;54(4):1158- 16. Yusuff H, Malagon I, Robson K, Parmar J, Ham-
1165. ilton P, Falter F. Extracorporeal membrane oxy-
6. Bridges BC, Hardison D, Pietsch J. A case series genation for Life-threatening ANCA-positive
of the successful use of ECMO, continuous re- pulmonary capillaritis. A review of UK experi-
nal replacement therapy, and plasma exchange ence. Heart Lung Vessel. 2015;7(2):159-167.
for thrombocytopenia-associated multiple organ 17. Agarwal HS, Taylor MB, Grzeszczak MJ, et al.
failure. J Pediatr Surg. 2013;48(5):1114-1117. Extra corporeal membrane oxygenation and
7. Hites M, Dell’Anna AM, Scolletta S, Taccone plasmapheresis for pulmonary hemorrhage
FS. The challenges of multiple organ dysfunc- in microscopic polyangiitis. Pediatr Nephrol.
tion syndrome and extra-corporeal circuits for 2005;20(4):526-528.
drug delivery in critically ill patients. Adv Drug 18. Kolovos NS, Schuerer DJ, Moler FW, et al.
Deliv Rev. 2014;77:12-21. Extracorporal life support for pulmonary hem-
8. Jamal JA, Economou CJ, Lipman J, Roberts JA. orrhage in children: a case series. Crit Care Med.
Improving antibiotic dosing in special situations 2002;30(3):577-580.
in the ICU: burns, renal replacement therapy 19. Di Maria MV, Hollister R, Kaufman J. Case
and extracorporeal membrane oxygenation. report: severe microscopic polyangiitis suc-
Curr Opin Crit Care. 2012;18(5):460-471. cessfully treated with extracorporeal membrane
9. Sirignano RM, Meyer EK, Fasano R, Paden oxygenation and immunosuppression in a pedi-
ML. Pediatric Tandem Therapeutic Apheresis: atric patient. Curr Opin Pediatr. 2008;20(6):740-
A Multidisciplinary Approach. ASAIO J. 2017. 742.
10. Patel P, Nandwani V, Vanchiere J, Conrad SA, 20. Dalabih A, Pietsch J, Jabs K, Hardison D, Bridg-
Scott LK. Use of therapeutic plasma exchange es BC. Extracorporeal membrane oxygenation
as a rescue therapy in 2009 pH1N1 influenza as a platform for recovery: a case report of a
A--an associated respiratory failure and he- child with pulmonary hemorrhage, refractory

245
Chapter 29

hypoxemic respiratory failure, and new onset acute necrotizing encephalopathy. Pediatr Neu-
goodpasture syndrome. J Extra Corpor Technol. rol. 2015;52(1):110-114.
2012;44(2):75-77. 32. Kitazawa Y, Saito F, Nomura S, Ishii K,
21. Gupta T, Khera S, Kolte D, et al. Back from the Kadota E. A case of hemophagocytic lympho-
brink: catastrophic antiphospholipid syndrome. histiocytosis after the primary Epstein-Barr
Am J Med. 2015;128(6):574-577. virus infection. Clin Appl Thromb Hemost.
22. Dornan RI. Acute postoperative biventricular 2007;13(3):323-328.
failure associated with antiphospholipid anti- 33. Hei F, Irou S, Ma J, Long C. Plasma exchange
body syndrome. Br J Anaesth. 2004;92(5):748- during cardiopulmonary bypass in patients with
754. severe hemolysis in cardiac surgery. ASAIO J.
23. Jhang J, Middlesworth W, Shaw R, et al. Thera- 2009;55(1):78-82.
peutic plasma exchange performed in parallel 34. Lo MH, Huang CF, Chang LS, et al. Drug reac-
with extra corporeal membrane oxygenation for tion with eosinophilia and systemic symptoms
antibody mediated rejection after heart trans- syndrome associated myocarditis: a survival
plantation. J Clin Apher. 2007;22(6):333-338. experience after extracorporeal membrane
24. Wang SS, Chou NK, Ko WJ, et al. Effect of oxygenation support. J Clin Pharm Ther.
plasmapheresis for acute humoral rejection 2013;38(2):172-174.
after heart transplantation. Transplant Proc.
2006;38(10):3692-3694.
25. Saito S, Matsumiya G, Fukushima N, et al. Suc-
cessful treatment of cardiogenic shock caused
by humoral cardiac allograft rejection. Circ J.
2009;73(5):970-973.
26. Stendahl G, Berger S, Ellis T, et al. Humoral
rejection after pediatric heart transplantation: a
case report. Prog Transplant. 2010;20(3):288-
291.
27. Dellgren G, Koirala B, Sakopoulus A, et al. Pe-
diatric heart transplantation: improving results
in high-risk patients. J Thorac Cardiovasc Surg.
2001;121(4):782-791.
28. Kolcz J, Pietrzyk J, Januszewska K, Pro-
celewska M, Mroczek T, Malec E. Extracor-
poreal life support in severe propranolol and
verapamil intoxication. J Intensive Care Med.
2007;22(6):381-385.
29. Koschny R, Lutz M, Seckinger J, Schwenger
V, Stremmel W, Eisenbach C. Extracorpo-
real life support and plasmapheresis in a case
of severe polyintoxication. J Emerg Med.
2014;47(5):527-531.
30. Maclaren G, Butt W, Cameron P, Preovolos
A, McEgan R, Marasco S. Treatment of poly-
pharmacy overdose with multimodality extra-
corporeal life support. Anaesth Intensive Care.
2005;33(1):120-123.
31. Wang KY, Singer HS, Crain B, Gujar S, Lin DD.
Hypoxic-ischemic encephalopathy mimicking

246
30

Mechanical Liver Support

Patrick D. Brophy, MD, MHCDS

Abstract Introduction

A variety of approaches to extracorporeal The presence of acute or acute on chronic liver

support for liver failure (both acute and acute on failure in patients in the intensive care unit portends
chronic) have existed for well over half a century. a high mortality rate. Acute liver failure (ALF) may
In general, the hepatic support strategies have been be defined as the development of severe acute liver
partitioned into biologic or nonbiologic approaches.1 injury in a patient without cirrhosis or existing liver
As far back as the 1950s hemodialysis was employed disease and an illness less than 26 weeks in dura-
as a strategy for approaching hepatic failure. As the tion as evidenced by encephalopathy and impaired
mechanical options changed so too did the evolution synthetic clotting functions (International Normal-
of approaches to hepatic support. These included; ized Ratio [INR] of >1.5 or higher).3 The most com-
hemadsorption, blood/plasma exchange, charcoal mon causes of ALF include drug-induced hepatitis
hemoperfusion, continuous hemofiltration, and (predominantly acetaminophen) and viral induced
more recently, dedicated hepatic replacement ap- hepatitis. Other less common causes include; sepsis,
proaches (Figure 30-1).1 These dedicated hepatic HELLP (hemolysis, elevated liver enzymes, low
replacement approaches include (but are not limited platelets), malignancies, venoocclusive disease,
to); Single Pass Albumin Dialysis (SPAD), Molecu- Wilson disease, and mushroom ingestions.3
lar Adsorbent Recirculating System (MARS), and Acute-on-chronic liver failure (ACLF) involves
Fractionated Plasma Separation and Adsorption an acute deterioration of liver function in patients
(Prometheus, Figure 30-2).2 The following chapter with preexisting cirrhosis. This may be due to
focuses on the topic of acute and acute on chronic extrahepatic precipitating factors such as sepsis or
liver failure, outlining the primary approaches for
extracorporeal replacement including SPAD, Pro-
metheus, and MARS. Appendix 1 provides a general
approach for MARS setup as it is a primary me-
chanical hepatic specific support therapy available.

Figure 30-2. Schematic drawings of A) the Molecu-

lar Adsorbent Recirculating System (MARS), B)
the Fractionated Plasma Separation and Adsorption
Figure 30-1. Historical evolution of treatments for (Prometheus) device, and C) Single Pass Albumin
hepatic dysfunction. Dialysis (SPAD) device.

247
Chapter 30

secondary to acute superimposed liver injury from excellent reviews on scoring systems for acuity and
drug ingestions, etc. Although it may be reversible, medical management of ALF and ACLF exist and
it does have a high short-term mortality (~50% the readers are directed to these for more granular
within 30 days).4 Given the high association rate of information.3-7
multiorgan failure (MOF) and mortality with these In patients with worsening disease, particularly
conditions, avoidance of the progressive develop- increasing cerebral edema (the primary cause of
ment of MOF and early supportive treatment should mortality in ALF) and evolving MOF, a decision
be the goal of therapy.3 point must be reached in terms of escalation of sup-
port and treatment approaches. Given the generation
Approach to Treatment of a significant inflammatory response including
bile acids, toxic fatty acids, vasoactive molecules,
High index of suspicion and prompt interven- among others, in such settings (Figure 30-3),5 it
tion in patients with ALF/ACLF are necessary. makes intuitive sense to consider mechanical hepat-
Hospitalization at a center with available hepatic ic support approaches in order to temper the immune
transplantation is recommended. Approaches to and pathophysiologic responses. In general, such
treatment include specific targeting of the acute approaches are implemented as a potential bridge
etiology when available (i.e. for patients with to liver transplantation or hepatic regeneration, for
acetaminophen-induced ALF, activated charcoal specific toxin removal (i.e. mushroom poisoning8)
and prompt administration of N-acetyl cysteine or attempting to improve/stabilize clinical status
[NAC] should be implemented). Use of NAC may (especially cerebral edema).
also be indicated for patients with ALF due to other
causes and can be useful in those with early grades Selected Mechanical Extracorporeal Hepatic
of encephalopathy. Prevention of further injury Support
(including cessation of any potential offending
medications), identification of possible ingestions Hemodialysis and Continuous Renal Replacement
(i.e. mushroom poisoning), supportive care, and Therapy (CRRT)
management of complications should be the initial
goals. Once stabilized, prognosis and decision The first reasonably documented attempt at
whether to pursue hepatic support therapy or liver providing hepatic support was performed via hemo-
transplantation should be entertained.3 dialysis and reported in 1956.9 While 4 of 5 patients
As in any patient with organ failure, assessment responded with improved clinical and neurologic
for cardiovascular stability, fluid and vasopressor parameters, no improvement in overall survival
management need to be meticulously monitored.
Hypoglycemia must be anticipated and preemptively
addressed. Airway management and ongoing as-
sessment for bleeding, acute kidney injury (AKI),
cerebral edema (resulting in intracranial hyperten-
sion, ischemic brain injury and herniation), and
the possible development of sepsis are standard
approaches. In fact, laboratory monitoring includ-
ing ammonia measurements, serial blood, urine,
and sputum cultures should be obtained and any
suspicion for infection should be treated with the
appropriate broad spectrum antibiotics. In general,
coagulopathy should not necessarily be treated with
blood products unless overt bleeding is present. Ev- Figure 30-3. Pathophysiology of liver failure and
ery attempt should be made to avoid the progressive therapeutic interventions to the components of
development of multiorgan dysfunction. Several liver failure.

248
 Mechanical Liver Support

was observed. The use of CRRT (or continuous demonstrated clearance of a toxic level of carba-
hemofiltration) is likely the widest approach that mazepine using a 5% albumin enhanced dialysate
clinicians have implemented for stabilizing and sup- for CVVHD. In fact, the clearance half-life of drug
porting patients with ALF or ACLF. The concept that elimination was reduced from an expected 20 hours
CRRT can buy time, prevent further deterioration, to approximately 8 hours. An additional case study
maintain hemostability, manage cerebral edema, demonstrated effective clearance of methotrexate in
and allow potential recovery of functional hepatic the setting of acute kidney injury;13 however, this
cellular mass or bridge to liver transplantation has study also employed hemoperfusion as part of the
made it a relatively standard therapy. Additionally, clearance strategy.
given its ability to be integrated inline into other The specific dosing of albumin required to
continuous therapies like extracorporeal membra- provide adequate clearance remains uncertain;
nous oxygenation (ECMO) (Figure 30-4) makes however, the lower the concentration the lower the
it a versatile option as well. While CRRT may not costs. Chawla et al.14 reported that albumin dialysate
improve overall outcomes in terms of mortality, its concentrations as low as 1.85% had similar bilirubin
value has clearly been demonstrated in the man- removal properties as 5% albumin dialysate. A more
agement of cerebral edema10 as well as controlling comprehensive, well-controlled in vitro study was
elevated ICP in a stable and continuous fashion due performed by Churchwell et al.15 which compared
to fulminant hepatic failure.11 In addition to these the clearance of phenytoin, valproic acid, and
key factors, CRRT enables the clinician to provide carbamazepine controlling for dialyzers (0.6 m 2
adequate nutritional support and maintain fluid bal- acrylonitrile 69 and1.5 m 2 polysulfone), various
ance in patients with hepatic dysfunction. blood flow (180–270 mL/mn) rates, dialysate flow
(1–4 L/h) rates, using dialysate albumin concentra-
Single Pass Albumin Dialysis (SPAD) tions ranging from 0%, 2.5%, and 5%. Interestingly,
the use of 5% albumin dialysate concentrations
SPAD utilizes a typical CRRT setup but with with the larger filter provided optimized clearance
albumin added to the standard dialysate (continuous for valproic acid and carbamazepine. Neither blood
venovenous hemodialysis--CVVHD, Figure 30-2)12

Figure 30-4. In line VA-ECMO using standard CRRT machine or SPAD (albumin dialysate only). Venous sys-
tem drains from the central catheter and the arterial system returns to the femoral catheter, arrows denote flow
pattern. CRRT/SPAD intake and outflow lines placed between the ECMO pump and oxygenator to prevent air
entrapment and bypass of the oxygenator (setup will vary depending on equipment used).

249
Chapter 30

nor dialysate flow rates had a significant effect on MELD score above 30), treated with Prometheus.
clearances in the ranges used within the study. Like other acute extracorporeal therapies, at this
While SPAD confers the convenience of being time it is unclear which subgroups of patients are
integrated into other techniques including ECMO best served by this therapy and further evaluation
(similar to CRRT) since the dialysate fluid can be needs to be considered.
manipulated as needed, the inability to regenerate
albumin and the high cost of using in center albumin Molecular Adsorbent Recirculating System
and pharmacy preparation reduce its attractiveness (MARS)
for broader and more sustained applications.
Like the other hepatic therapies, the clinical
Fractionated Plasma Separation and Adsorption rationale for utilizing MARS includes providing an
(Prometheus) environment conducive to allow regeneration of na-
tive hepatic cells or bridging to liver transplantation,
Like SPAD or MARS, the Prometheus uses a clinically supportive tool for clearing toxins and
albumin to remove protein-bound toxins, vasoactive improving hepatic encephalopathy, particularly in
factors and allows a bridge to hepatic regeneration the face of MOF. MARS is available in the United
or liver transplantation.16 The Prometheus is a Fre- States and is FDA approved for use in protein-bound
senius based device that is not currently available drug poisoning and treatment of hepatic encepha-
in the United States. Conceptually, it differs from lopathy in decompensated chronic liver disease. In
MARS in that it utilizes the patient’s own albumin this setting the MARS machine is utilized in tandem
to enter the Fractionated Plasma Separation and with the Prismaflex CRRT circuit, providing both
Adsorption via a specialized high molecular weight renal and hepatic support. In this circuit (see Figure
cutoff filter. The albumin is then returned to the 30-2) blood is dialyzed against a self-contained 20%
patient’s circulation after being reactivated in the albumin solution across an albumin impregnated
circuit and the cycle continues (see Figure 30-2). high flux polysulfone filter.19 The MARS flux filter
Many of the studies utilizing Prometheus involve extracorporeal volume is 150 ml + lines, 600 ml of
single center analyses. In a small, retrospective, 20% albumin and may not be suitable for smaller
adult focused study, Komardina et al.17 evaluated infants. Outside of the United States the MARSmini
the use of the Prometheus in 39 post cardiac surgi- (56 ml+ lines, 500 ml 20% albumin) is also available.
cal patients (predominantly acquired valvular dis- This system does not allow filtration of the patient’s
ease) that developed ALF as a complication over a albumin but rather allows albumin bound toxins to
6-year period. Both APACHE II and MELD scores be transferred across the filter via the impregnated
were evaluated in patients. Interestingly, approxi- albumin. Once the toxins have transferred to the
mately 80% of patients also received concomitant intra-circuit, albumin solution it is dialyzed against
CRRT due to AKI, which may have confounded a standard low flux dialysate filter and bath whereby
these outcomes. The patients were placed on the water soluble molecules are removed. The toxin
Prometheus treatments (6 hr duration treatments, 1 laden albumin solution is then scrubbed and regener-
treatment [median] per patient) and monitored for ated by passing through an activated charcoal filter
safety related issues and successful clearance of and an ion exchange column. The process repeats
toxic metabolites, mainly bilirubin metabolites and itself for up to 8 hours/session. (see appendix 1 for
bile salts. The survival rate at 28 days in this patient setup and initiation guide).
group was 23%. In a larger randomized control During MARS therapy a multitude of albumin
trial known as the HELIOS study,18 145 patients bound toxins are removed including aromatic amino
with ACLF received 8-10 Prometheus treatments. acids, bilirubin, bile acids, copper, middle and
There was no significant increase in survival in the short chain fatty acids, nitric oxide (S-nitrosothiol),
treatment group. The authors noted a statistically in- protoporphyrin. Water soluble substances such as
creased survival rate in the most severely ill patient ammonia, creatinine, tryptophan, tumor necrosis
subgroup (with hepatorenal syndrome type I and factor-α, urea, and interleukin 6 (IL-6) are also

250
 Mechanical Liver Support

removed. Larger molecules like clotting factors, support over standard medical therapy in the setting
immunoglobulin, hormone binding proteins, and of ALF as well as ACLF. While these metaanalyses
albumin are not removed.20 are encouraging, interpretations must be tempered
Early in the investigative phase, MARS was because the confidence intervals were wide and so
evaluated in a single center, small prospective, the conclusions must be drawn in that light.
randomized controlled trial in 13 patients with Devices such as MARS do appear to have some
Type 1 hepatorenal syndrome. All patients under- positive benefit in terms of reducing ALF/ACLF
went standard supportive medical therapy (SMT) treatment costs.24, 25 It is also clear that using devices
as well as hemodiafiltration (HDF). Eight patients that have the ability to regenerate albumin within
were assigned to the MARS group and underwent the system (MARS/Prometheus) can significantly
an average of 5.2 therapy runs lasting 6-8 hours reduce pharmacy/institutional costs compared to
over successive days. In the control (SMT+HDF) techniques like SPAD.26 Albumin based extracor-
group, the 7-day mortality rate was 100%; whereas, poreal liver support also appears to be beneficial
in the MARS (SMT+HDF+MARS) group, the 7-day in an outpatient setting in patients with intractable
mortality rate was 62.5% and the 30-day mortality pruritus in terms of quality of life27,28 and this ap-
rate was 75%.21 These promising initial results were plication must be more fully explored.
not replicated by other investigators. More recently,
the RELIEF study group randomized 179 patients Conclusions
with ACLF failure to SMT with or without MARS.
Due to high patient dropout the Per Protocol group The use of albumin based extracorporeal liver
ended up with SMT; N=85 and MARS; N=71. The support appears to be a relatively safe therapy and
average MARS treatment number was 6.5 lasting provides demonstrated benefits in removing albumin
approximately 6.8 hours per session. The primary bound toxins in patients with ALF/ACLF and toxic
endpoint of liver transplantation-free survival within overdoses. It is also associated with improvements
28 days did not differ between groups, and a non- in hepatic encephalopathy; however, no significant
significant improvement in hepatic encephalopathy survival benefit has been clearly demonstrated to
and bilirubin levels was observed in the MARS date. It may well be that there are specific subsets
group. The authors concluded that in this setting of patients that will have a survival benefit when
MARS conferred no beneficial effect on survival albumin based extracorporeal liver support devices
in patients with ACLF.22 are used. These studies are being conducted and
A metaanalysis by Tsipotis et al.2 reviewed 70 as evidence mounts, such therapies will take their
full text articles excluding those articles that were place among the tools we use in patients with MOF.
nonrandomized or case reports. Their final analysis
included 10 randomized controlled trials. The ma-
jority of these utilized the MARS therapy, one used
Prometheus and 2 trials used both devices. Also a
majority of these studies used some form of renal
replacement therapy. The metaanalysis revealed a
net change in serum mean bilirubin level favored
the albumin based modality. Additionally, albumin
based modality was favored when evaluating risk
ratio improvement in the West Haven grading of
hepatic encephalopathy. Finally, the risk ratio for
all cause mortality was favored with albumin based
modalities over standard medical therapy. In a
smaller metaanalysis done by Stutchfield et al.23 us-
ing 8 randomized control trials demonstrated a risk
ratio for mortality that favored extracorporeal liver

251
Chapter 30

Appendix 1: (Adapted from the MARS Product B. Initiating the Molecular Adsorbent Recirculat-
Users Guideline) ing System (MARS)

A. Setting Up the Molecular Adsorbent Recirculat- PURPOSE:

ing System (MARS) Initiate the Molecular Adsorbent Recirculating Sys-
tem (MARS) while assuring patient safety.

PURPOSE: EQUIPMENT:
Prepare the Molecular Adsorbent Recirculating Sys- MARS Monitor
tem (MARS) for the treatment of pediatric patients Prismaflex System
requiring hepatic supportive therapy. Gown
Gloves
EQUIPMENT: Face Shield
MARS Monitor PROCEDURE:
Prismaflex System 1. Prepare the Prismaflex-MARS System.
PrisMARS 1115/1 Treatment Kit 2. Place face shield, gown. Wash hands/hand
5 Liter bag of Prismasate hygiene. Glove.
1 Liter bags Normal Saline 3. If Albumin circulation has not stopped automat-
Prismaflex Effluent Bag ically prior to initiating MARS therapy, press
Replacement Solution ENTER on the MARS menu and select STOP.
Dialysate Solution 4. Set treatment parameters on MARS Main menu.
Albumin Solution 5. Select treatment on MARS Main menu.
PROCEDURE: 6. Start Prismaflex by pressing START softkey.
1. Assemble equipment. Wash hands/hand hygiene. 7. When Prismaflex blood circuit is stable, press
2. Turn the power switch on the MARS Monitor START/STOP button on MARS to start MARS
to the “On” position. monitor.
3. Turn the power switch on the Prismaflex CRRT 8. Press START to begin treatment.
machine to the “On” position. 9. Remove gown, gloves, and face shield. Hand
4. Choose New Patient or Same Patient. hygiene.
5. If New Patient is chosen, the control unit ad- 10. Document procedure.
vances to the Set Excess Pt. Fluid Loss or Gain
Limit screen. Set limit per Pediatric Nephrolo-
gist’s order.
6. Follow the instructions displayed on the Pris-
maflex Control Unit. The user will be guided
through the complete setup of the Prismaflex
and MARS system by the Prismaflex device.
7. Several of the MARS screen commands are
no longer required and must be avoided. The
Prismaflex will prompt the user when to ignore
the MARS screen instructions and when it is
required to follow the MARS screen prompts.
Always follow the instructions on the Prismaflex
Control Unit during setup.

252
 Mechanical Liver Support

References safely reverse methotrexate nephrotoxicity.

Pediatr Nephrol 2016;31(10):1699–1703.
1. Millis JM, Losanoff JE. Technology Insight: 14. Chawla LS, Georgescu F, Abell B, Seneff
liver support systems. Nat Clin Pract Gastro- MG, Kimmel PL. Modification of continuous
enterol Hepatol. 2005; 2(9), 398–405. venovenous hemodiafiltration with single-pass
2. Tsipotis E, Shuja A, Jaber BL. Albumin Dialysis albumin dialysate allows for removal of serum
for Liver Failure: A Systematic Review. Adv bilirubin. Am J Kidney Dis 2005;45(3):e51–e56.
Chronic Kidney Dis. 2015;22(5):382-390. 15. Churchwell MD, Pasko DA, Smoyer WE, Muel-
3. Shah NJ, John S. Liver Failure, Acute. Stat- ler BA. Enhanced clearance of highly
Pearls [Internet]. Treasure Island (FL): Stat- protein-bound drugs by albumin-supplemented
Pearls Publishing; 2018 Jan-.2018 Jan 10. dialysate during modeled continuous.
4. Mikolasevic I, Milic S,Radic M,Orlic L,Bagic h e m o d i a l y s i s . N e p h r o l D i a l Tr a n s p l a n t .
Z, Stimac D. Clinical profile, natural history, 2009;24(1):231-238.
and predictors of mortality in patients with 16. Nevens F, Laleman W. Artificial liver support
acute-on-chronic liver failure (ACLF). Wien devices as treatment option for liver fail-
Klin Wochenschrift. 2015; 127(7-8): 283–289. ure. Best Pract Res Clin Gastroenterol 2012;
5. Stadlbauer V, Jalan R. Acute liver failure: liver 26(1):17–26.
support therapies. Curr Opin Crit Care. 2007; 17. Komardina E, Yaroustovsky M, Abramyan
13(2):215–221. M, Plyushch M. Prometheus therapy for the
6. Li H, Chen HS, Nyberg SL. Extracorporeal liver treatment of acute liver failure in patients after
support and liver transplant for patients with cardiac surgery. Kardiochir Torakochirurgia.
acute-on-chronic liver failure. Sem Liver Dis. 2017;14(4):230-235.
2016; 36(2):153–160. 18. Rifai K, Kribben A, Gerken G et al. HELIOS
7. Jain V, Dhawan A. Extracorporeal Liver Support Study Group Extracorporeal liver support by
Systems in Paediatric Liver Failure. J Pediatr Fractionated Plasma Separation and Adsorption
Gastroenterol Nutr. 2017;64(6):855–863. (Prometheus) in patients with acute-on-chronic
8. Lionte C, Sorodoc L, Simionescu V. Successful liver failure (HELIOS study): a prospective ran-
treatment of an adult with Amanita phalloides- domized controlled multicenter study. J Hepatol.
induced fulminant liver failure with molecular 2010;52:S3.
adsorbent recirculating system (MARS). Rom 19. Kapoor D. Molecular adsorbent recirculating
J Gastroenterol. 2005;14(3):267-271. system: albumin dialysis-based extracorporeal
9. Kiley JE, Welch HF, Pender JC, Welch CS. liver assist device. J Gastroenterol Hepatol.
Removal of blood ammonia by hemodialysis. 2002;17 Suppl 3:S280-286.
Proc Soc Exp Biol Med. 1956;91(3):489-490. 20. Mitzner SR, Stange J, Klammt S, Koball S,
10. Matsubara S, Okabe K, Ouchi K, et al. Continu- Hickstein H, Reisinger EC. Albumin dialysis
ous removal of middle molecules by hemofiltra- MARS: knowledge from 10 years of clinical
tion in patients with acute liver failure. Crit Care investigation. ASAIO J. 2009;55(5):498-502.
Med. 1990;18(12):1331-1338. 21. Mitzner SR, Stange J, Klammt S, et al. Improve-
11. Davenport A. Haemofiltration in patients ment of hepatorenal syndrome with extracor-
with fulminant hepatic failure. Lancet. poreal albumin dialysis MARS: results of a
1991;338(8782-8783):1604. prospective, randomized, controlled clinical
12. Askenazi DJ, Goldstein SL, Chang IF, Elenberg trial. Liver Transpl. 2000;6(3):277-286.
E, Feig DI. Management of a severe carba- 22. Bañares R, Nevens F, Larsen FS, et al; RELIEF
mazepine overdose using albumin-enhanced study group. Extracorporeal albumin dialysis
continuous venovenous hemodialysis. with the molecular adsorbent recirculating
Pediatrics 2004; 113(2): 406–409. system in acute-on-chronic liver failure: the
13. Chan WK, Hui WF. Sequential use of hemo- RELIEF trial. Hepatology. 2013;57(3):1153-
perfusion and single-pass albumin dialysis can 1162.

253
Chapter 30

23. Stutchfield BM, Simpson K, Wigmore SJ. Sys-

tematic review and meta-analysis of survival
following extracorporeal liver support. Br J
Surg. 2011;98(5):623-631.
24. Kantola T, Mäklin S, Koivusalo AM, et al.
Cost-utility of MARS treatment in ALF. World
J Gastroenter 2010; 16(18); 2227-2234.
25. Hessel FP, Bramlage P, Wasem J, Mitzner SR.
Cost-effectiveness of the artificial liver support
system MARS in patients with acute-on-chronic
liver failure. Eur J Gastroenterol Hepatol 2010;
22(2): 213-220.
26. Drexler K, Baustian C, Richter G, Ludwig
J, Ramlow W, Mitzner S. Albumin dialysis
molecular adsorbents recirculating system:
impact of albumin dialysate concentration on
detoxification efficacy. Ther Apher Dial 2009;
13(5); 393-398.
27. Leckie P, Tritto G, Mookerjee R, Davies N,
Jones D, Jalan R. Outpatient albumin dialysis
for cholestatic patients with intractable pruritus.
Aliment Pharmacol Ther 2012; 35(6): 696-714.
28. Schaefer B, Schaefer F, Wittmer D, Engelmann
G, Wenning D, Schmitt CP. Molecular Adsor-
bents Recirculating System dialysis in children
with cholestatic pruritus. Pediatr Nephrol 2012;
27(5): 829-834.

254
31

Staffing for the ECMO Patient

Tracy Morrison, BSN, MSQA, FELSO, Micheal L. Heard, RN, FELSO

Introduction The nurse-oriented staffing model is a common

approach to staffing. Nursing is already present
The choice of how to provide care for the at the bedside and addition of pump observation
patient on extracorporeal life support (ECLS) is tasks is usually not overwhelming. The ICU staff-
multifactorial. The choice may be between, but not ing model utilizing bedside nurses to monitor the
limited to, a team of ECMO Specialists or Regis- ECMO patient and circuit has been shown to save
tered Nurses (RN), Certified Clinical Perfusionists money over a perfusion model, where perfusionists
(CCP), Registered Respiratory Therapists (RRT), or “round” on ECMO patients and have responsibili-
some combination. Each ECLS program must make ties managing the ECMO circuit. The ICU staffing
the decision based on patient population, acuity, model did not increase the number adverse patient
institutional guidelines, licensure and regulatory events.2
requirements, and the many intangibles that guide While the existing bedside nurse may be the
the administrator. easiest solution for staffing the ECLS patients, what
is the overall impact to care delivery for the patient?
The Staffing Models and the Decision Process A 2012 survey of nurses in English National Health
Service hospitals reported that 86% of nurses re-
The care of an ECLS patient is usually more ported one or more care activities were left undone
complex than that of other ICU patients. The ECLS due to lack of time on their shift.3 Care activities
patient is connected to a blood pump, via cannula reported missed ranged from the most likely activity
in large vessels. The pump requires frequent moni- missed: comforting or talking to their patients (66%)
toring including assessment the circuit for clots, air, to the least likely activity missed: pain management
proper functioning of the mechanical components, (7%). Of interest, “a mean of 7.8 activities per shift
and in-depth laboratory assessment. Additionally, were left undone on wards that were rated as ‘failing’
the blood pump or disposables may fail, causing on patient safety, compared with 2.4 where patient
interruption of the support to the patient, which safety was rated as ‘excellent’ (p<0.001).”3 The
can result in a life-or-death situation. The bedside distress that this causes nurses may be insurmount-
caregiver must be trained to intervene immediately able and may lead to increased job dissatisfaction.
in the event of such a situation. Respiratory therapists are the next group consid-
Many ECLS programs begin staffing patients ered for staffing ECMO patients. These profession-
with their in-house perfusionists who are expert in als have a deep knowledge of respiratory failure and
extracorporeal technology. However, the perfusion- management, as well as a propensity for technology
based model requires the addition of perfusion staff and equipment. Therapists and their departments
as an increasing number ECMO patients may cause have an ease of flexibility that is required with the
significant demands on personnel and necessitate ebb and flow of ECMO patient cases, allowing
cancelling cardiac operating room cases.1 The cost the ability to increase staff in order to care for the
of these additional personnel may be substantial. ECMO patient. However, simply adding the ECMO
patient into their ventilator checks may provide a

255
Chapter 31

new set of workload challenges that could affect failures cause 80–90% of all medical errors. Humans
patient safety. have known limitations that are knowledge-based
(relating to education and training), rule based
Choosing the Right Staffing Model for Your (relating to errors in interpreting data or applying
Program the wrong rule), or skill based (missed tasks and
memory failures). Active failures are errors commit-
When choosing a staffing model, the assessment tee by frontline staff and may be honest mistakes due
of job workload as compared to human capability to poor decisionmaking or lack of skill, or they may
and attitude is important to assure maximum staff be violations such as work arounds or knowingly not
performance. Workload consists of a portion of following policy. Latent failures are system weak-
the staff’s resource expended when performing a nesses caused by poor design, poor leadership, and
work task. Avoiding task overload can reduce work inadequate systems. Latent failures may occur due to
stressors that can affect overall job performance. lack of resources, poorly designed processes, lack of
Figure 31-1 details the job demands, or work stress- supervision, and a care environment preconditioned
ors, that interact with staff capabilities and ultimately for unsafe acts.8
may affect job performance.4 Job demand should There is increasing evidence that a relationship
only require a portion of the employee’s abilities or exists between hospital nurse staffing levels and
attention. Employee ability is positively influenced hospital-associated morbidities. Staffing levels
by initial and continuing training and education.5 refer to nurse to patient ratio, inadequate nursing
Decisions around staffing must be made judi- skill mix, excessive overtime use, and the use
ciously, without considering the impact of staffing of float nurses to staff patient units. The rates of
on the cost of care alone. Inadequate staffing can hospital-associated infections rise when increased
have a detrimental impact on a range of performance numbers of float nurses are used for staffing. The
and safety issues. Factors that should be considered relationship of registered nurse (RN) staffing level
when evaluating a new staffing model include: and mortality has been well studied. A higher level
of RN staffing is related to lower patient mortality
• Increasing job scope for some staff and increased patient safety. A European study of
• Maintaining skills for new and existing staff over 400,000 surgical patients >50 years of age
• Having enough staff with appropriate skills/ found a 7% increase in the odds of death within 30
experience 24/7 days of admission occurs when an RN’s workload
• Potential to eliminate outside contractors was increased by one additional patient per shift.9
• Collaboration with other departments (perfu-
sion, existing hospital ECMO program) as Location of Patients and Staffing Impact
expert resources
• Opportunities to benchmark with other ECMO Choosing an area to place ECMO patients is
centers to learn from experience often determined by physicians and the specialty
The safety of each patient relies on having the diagnosis of the patient. However, a dedicated
right staff, with the appropriate skills, caring for specific unit or space for ECMO patients increases
the patient 24/7.6 The number of patients an ECMO staff exposure to ECMO patients and technology,
Specialist can monitor safely has not been studied.
However, adverse events, morbidity, and mortality
Job demands
rise as the number of patients per nurse increases.
Nursing workload, as it relates to patient acuity, is Task type
Task quality
Person’s
capability and Performance
also linked to patient outcomes.7 Task schedule
Task environment
attitude

Staffing should be designed to provide the Task conditions

needed leadership, necessary resources, and culture Work “stress” Work “strain”
for employees to perform their work safely.8 Human Figure 31-1.

256
 Staffing for the ECMO Patient

augmenting the pool of staff available to respond When new technology is implemented into
to patient emergencies. patient care areas, medical errors and near misses
To understand the human factors that influence are common if the staff has not received thorough
ECMO Specialist performance, you must look at the training. Patient safety suffers with rapid implemen-
physical layout of the unit (e.g. patient cohorting, tation of new medical devices when they are inad-
convenient supply storage), what is the cognitive equately integrated into caregiver’s workflow.12 To
load of the assignment (e.g. new specialist, difficult do so appropriately relies on human factors research
case), and what are the contextual environmental and real-life situations.11 New and existing ECMO
factors (e.g. great teamwork, open communication). programs should designate resources to assist with
Factors that can reduce specialist performance staffing and/or patient emergencies due to the com-
include task unfamiliarity, time constraints to plex nature of the ECMO patient and equipment.
complete tasks, poor communication, inadequate
risk assessment, and poorly designed workflow. Current State of ECMO Programs
Psychologist James Reason found that the likeli-
hood of error reaches 75% when a person performs The Extracorporeal Life Support Organiza-
an unfamiliar procedure. The error rate reduces to tion (ELSO) recognizes ECMO programs with the
0.05% when a skilled person with knowledge of all Center of Excellence (COE) designation based on
the potential risks, performs the same procedure. a scored application consisting of seven categories,
Many of the elements that staff members endure at including education, process improvement, and
an ECMO bedside including noise, interruptions, quality improvement. Each center must complete
distractions, stress, inadequate lighting, reliance a demographic overview of their program for the
on memory to perform skills, fatigue, and fear; all application. The awarded Centers of Excellence of
promote human error. As you design your staffing 2017 and 2018 represent children’s hospitals, adult
model, you must try to mitigate the elements that only hospitals, and a combination of both as well
promote error within the context of your unit and as international programs.
healthcare delivery system.10 Figure 31-2 describes the most common staff-
ing for the coordinator position. Generally, ECMO
Linking Staffing, Education, and Quality coordinators are RNs, followed by RRTs, and then
CCP. Two centers had physicians as coordinators
Utilization of a standardized educational and one center used biomedical engineering as a
platform to train all members of the ECMO team coordinator.
(registered nurses, respiratory therapists, perfusion- Table 31-1 outlines the types of staffing models
ists, mid-level providers, and physicians) improves most commonly used in the 2017 and 2018 cohort of
patient safety by reducing variability in ECMO com-
petency that may exist between professions. The COE Applicants 2017 & 2018

educational outline may be tailored to the specific

Coordinator Profession
70

type of patients being cared for (adult or pediatric)

as well as the staff providing care (perfusion or
60

nursing). Deficits that each profession may have 50

may be focused on to fill in the gaps.

Continuing education plays a key role in error
40

prevention as well. The use of simulation scenarios 30

that reinforce Crisis Resource Management skills 20

have a specific impact on staff ability to perform un-

der emergent situations (see Chapter 35). Improving 10

communication across disciplines and teams, who 0

are placed under stressful situations reproduced in

RN RRT CCP MD/DO Other

simulation, decreases the potential for errors.11 Figure 31-2.

257
Chapter 31

COE applicants. The most common staffing models in the United States, hourly rates for additional
are the 1:1 model, with each specialist monitoring compensation begin at as low at $2/hour and can be
1 ECMO pump, and a mixed model, with some as high as $75/hour on weekends, with the average
ECMO patients being monitored as 1:1 and some as hourly pay increase of $8/hr. Base pay increases
1:2. The difference in staffing for the centers that use start at 3% and are reported up to 10% of base pay.
both 1:1 and 1:2 can be due to acuity of the patient, Bonuses reported were up to $4000/yearly prorated
the type of equipment being used such as a simpli- for hours worked.
fied circuit, or an unexpected increase in patient
numbers. Eleven centers use the single caregiver Conclusion
model with most of those centers from 2018. Only
four centers staff ECMO pumps at 1 Specialist to 3 When deciding to implement an ECMO pro-
or 4 ECMO pumps. gram or restructure an existing center, staffing may
Figure 31-3 describes the professions of staff be one of the most difficult aspects to design. Mul-
who sit at the pump for the COE applicants 2017 and tiple types of staffing models are being effectively
2018. The most common profession sitting pumps utilized by ELSO centers. Taking advantage of
is nursing at 68.7% of the total staff reported on the the Award for Excellence in Life Support centers
demographic form. Respiratory therapy follows at permits benchmarking with programs that may
20.5%, perfusion at 10.3% and physicians at less resemble yours in context or patient population.
than 1% of staff sitting pump.
Sixty percent of COE applicants compensate COE Applicants 2017 & 2018
Number of Staff by Profession who Sit Pump
their team members for managing the ECMO 2500

pump. Most centers use an increase in hourly rate

as compensation while some use a combination 2000

of strategies to reward staff, including increasing

base pay, advancing clinical ladder, and a bonus 1500

for hours sitting pump (Figure 31-4). For example,

1000

Staffing Model Total

Number
500

1:1 40
(1 Specialist per pump, patient has 0

RN RRT CCP MD/DO

a dedicated nurse 1:1)

1:1 9 Figure 31-3.
(1 Specialist per pump, patient's
nurse may have 1 or more patients)
1:2 11 Most Common Types of Compensation
Paid for Sitting Pump

(1 Specialist for 2 pumps, patients

have a dedicated nurse 1:1)
1:2 1
(1 Specialist for 2 pumps, patient's
5%

nurse may have 1 or more patients)

8%

Single Caregiver 11
1:3 2 9%
38%

(1 Specialist to 3 pumps)
1:4 2
(1 Specialist to 4 pumps)
Mixed Model with both 1:1 and 1:2 16
Hourly % Base Pay Clinical Ladder Bonus

Table 31-1. 2017 & 2018 COE Applicants Reported

Staffing Model Figure 31-4.

258
 Staffing for the ECMO Patient

Evaluate your current staffing in your ICU by job

description and availability to see which profession
would be the best fit for your program. Provide
leadership, staff support, and excellent training
and education for your staff. Consider providing an
incentive for staff sitting pump; this may positively
affect your staff retention and save money while
reducing patient mortality and morbidities. While
ELSO cannot recommend one staffing model over
another, design your model with patient safety at
forefront and continuously review the workload of
your staff, available staff resources, education and
training, and any patient safety events.

259
Chapter 31

References tion laboratory and clinical environment. Simul

Healthc 2011;6(5): 284-291.
1. Moll V, Teo E, Grenda D, et.al. Rapid develop- 12. Rice S. New medical technology poses safety
ment and implementation of an ECMO Program. problems if users not trained properly. Mod-
ASAIO J. 2016;62(3): 354-358. ern Healthcare. (2014). http://www.modern-
2. Cavarocchi NC, Wallace S, Hong EY, et al. healthcare.com/article/20140816. MAGA-
A cost-reducing extracorporeal membrane ZINE/308169986.
oxygenation (ECMO) program model: a single
institution experience. 2015. Thomas Jefferson
University Department of Surgery Faculty
Papers. Paper 120. h p://jdc.je erson.edu/sur-
geryfp/120.
3. Ball, JE, Murrells T, Rafferty AM, Morrow E,
Griffiths P. ‘Care left undone’ during nursing
shifts: associations with workload and per-
ceived quality of care. BMJ Qual Saf. 2014;23:
116-125.
4. Kroemer, K, Kroemer, H, Kroemer-Elbert, K.
(2001). Ergonomics: How to design for ease
and efficiency. 2nd ed. Prentice Hall.
5. Carayon P, Xie A, Kianfar S. Human factors and
ergonomics as a patient safety practice. BMJ
Qual Saf. 2014; 2:196-205.
6. Humanfactors101.com. (2018). Search Results
for “staffing levels and workload” – Human
Factors 101. [online] Available at: https://hu-
manfactors101.com/?s=staffing+levels+and+
workload.
7. Psnet.ahrq.gov. (2018). Nursing and Patient
Safety | AHRQ Patient Safety Network. [online]
Available at: https://psnet.ahrq.gov/primers/
primer/22/nursing-and-patient-safety?q=nursi
ng+and+patient+safety.
8. The Joint Commission. Human factors analy-
sis in patient safety systems. The Source.
2015;13(4): 1-10.
9. Griffiths P, Ball J, Murrells T, Jones S, Raf-
ferty AM. Registered nurse, healthcare support
worker, medical staffing levels and mortality in
English hospital trusts: a cross-sectional study.
BMJ. 2016: 6(2):e008751.
10. Tocco S. Human factors engineering can
improve patient safety. Am Nurse Today.
2013;8(7): 34-35.
11. Burton K, Pendergrass T, Byczkowski T, et
al. Impact of simulation-based extracorporeal
membrane oxygenation training in the simula-

260
32

ECMO Emergencies

Micheal L. Heard, RN, FELSO, Monika F. Cardona, RN, BSN

Introduction programs use a designated ECMO Specialist to

monitor the ECMO patient, other programs are
Mechanical complications occur commonly on structured in alternate fashions and use a registered
ECMO. A review of the ELSO Registry database nurse or respiratory therapist as the bedside care-
found that mechanical failures occurred in approxi- giver. Certified clinical perfusionists provide expert
mately 40% of all ECMO courses.1,2 Catastrophic consultative services for the ECMO patient, and in
complications in ECMO are attributed to the long- some cases, the managing attending physician may
term utilization of disposable components and high- be the deemed expert. This frontline caregiver must
tech equipment. The Air Force pilots who fly the have specialized education in the management of
B-52 bombers for long hours at a time describe the extracorporeal emergencies. Effective emergency
process of flying that aircraft as consisting of “hours management involves the employment of many
and hours of boredom interspersed by moments of techniques. Therefore, recognition must be given to
stark terror.” The provision and delivery of ECMO the universal application of these techniques despite
can often present in a similar fashion. The ECMO the differences that exist in ECMO equipment.
Specialist must not be lulled into a false sense of Every ECMO Center must have a comprehen-
security by periods of relative calm and minimal sive education plan for both the original certification
changes in care patterns. Minuscule changes such of personnel as well as continuing annual educa-
as a rise in the circuit pressure, a slow formation tion.3 Didactic and online education with hands-on
of a clot on the face of the membrane lung, or the practice employing water drills or simulations must
finding of air in the arterial line are examples that be included. Practice sessions should be timely,
can portend impending disaster. Therefore, it is im- consistent and comprehensive, and deliberately
perative that the care delivery model surrounding an designed to establish competency of the ECMO
ECMO patient involve a specially trained caregiver Caregiver.
who remains at the bedside at all times. The presence
of this caregiver facilitates the prompt recognition, Off-Bypass Procedure
management, and prevention of disastrous scenarios.
The purpose of this chapter is to familiarize the The first and most important aspect of emergen-
reader with the assessment of commonly encoun- cy management is the knowledge of how to safely
tered mechanical emergencies during an ECMO remove a patient from bypass. Upon recognition of
course and successful management strategies that an emergency, the patient must be separated from
can be employed. the ECMO circuit to eliminate further detriment
to the patient and establish a line of delineation.
The ECMO Caregiver There are a variety of ECMO circuits and pumps for
delivery of ECMO support. In the case of a roller
The individual clinician trained to assess, man- head pump with a bridge in line, place a clamp on
age, and intervene with the ECMO circuit may vary the Venous line first, open the Bridge, and then
from center to center. While many pediatric ECMO lastly clamp the Arterial line. This stops drainage

261
Chapter 32

to the pump first. The acronym V-B-A may be components including the raceway, pump head,
used for ‘Very Bad Accident’, and establishes an and oxygenator. The security of connectors
easy association for the caregiver to remember the must be assessed for potential disconnection
proper procedure in a time of immense stress. With as well, ensuring appropriate placement of tie
a centrifugal pump and no bridge, place a clamp on bands. All cannulae must be intact and secured
the Arterial line first then the Venous line. This stops to the patient with sutures. Circuit tubing should
both passive and retrograde flow to the pump head. also be secured to the bed with towel clamps
If the circuit contains a bridge, simply add opening or other means.
the bridge in between the two lines (“A-B-V”). Once • The alarms on all equipment and their appropri-
ECMO support is removed, the patient will require ate function are verified. The bubble detector,
hand bagging or emergency ventilator settings to if present, should be attached appropriately,
ensure adequate respiratory support. All involved engaged, and functioning properly.
caregivers must be prepared to provide advanced • All pressure monitors must be appropriately
cardiopulmonary support during an unplanned inter- connected to the circuit and alarm limits set
ruption in ECMO. However, the ECMO Caregiver’s per policy.
focus is the management of the emergency, and • The heater should be appropriately connected to
returning the patient to extracorporeal circulation the heat exchanger and the water level kept at
quickly and safely. capacity. The temperature probe, if applicable,
The ability of the caregiver to proceed in a time- is connected to the appropriate site within the
ly and accurate manner with emergency procedures circuit. The temperature set-point is maintained
is invaluable and may prevent secondary injury. The in conjunction with the patient temperature, per
use of a timer during practice drills may seem puni- policy.
tive; yet the patient benefits when the caregiver can • The gas module must be connected to the appro-
perform any off-bypass emergency in time frame priate blender, wall sources, and/or tank sources.
of three minutes or less. Assuming patients have The gas line is secured and without leaks to the
no native cardiac or lung function; rapid return of oxygenator gas inlet port.
extracorporeal circulation may prevent irreversible • All power cords must be securely plugged into
brain injury.4,5 the appropriate receptacle, deemed the safest
or as backup emergency by the organization.
The ECMO Circuit Check • The ECMO pump cart and the patient bed
should be in a “braked” position to prevent
Most emergency ECMO events can be prevent- unwanted movement of the patient or ECMO
ed with early recognition and the appropriate appli- equipment.
cation of available technology. The circuit check is
the main responsibility of the ECMO Caregiver and Additionally, the ECMO Caregiver must assure
brings with it the purpose of emergency prevention. that they have all potential emergency supplies close
A circuit check includes a comprehensive, detailed at hand. Items that should be carefully considered
examination of all ECMO circuitry, equipment, include:
system safety alarms, fluids, and the patient. ECMO • Tubing clamps must be available on the pump
circuit checks should be completed at the beginning cart at all times. Guarded clamps are preferred
of every shift and with every subsequent hour of to further solidify safety and preservation of
support. It consists of the following: ECMO circuitry. The number of clamps may
• Close scrutiny of all circuit components from vary, however a minimum of six is strongly
the venous drain to the arterial return line for recommended. This allows the clinician to
clots, air, leaks, and fibrin strands. Assessing the safely separate the patient from ECMO and also
function and integrity of individual components minimize blood loss if circuit interventions are
should also occur; carefully visualizing all necessary.

262
 ECMO Emergencies

• A fully stocked ECMO supply cart or room must 1. A ‘Time Out’ should be performed prior to
be available within a short distance of the pa- beginning any procedure to review every
tient care environment. The cart should contain participant’s role in the emergency procedure
all disposable components that may need to be and ensure clarity of purpose (Table 32-2). As-
replaced during an ECMO course. surance is given that all essential and potential
• A second ECMO pump cart or individual pieces supplies are available, and that every step of the
of equipment must be available in close proxim- procedure is reviewed. No components can be
ity to replace failing equipment. omitted. A checklist assures that no surprises
• A volume expander, (e.g. 5% albumin or crys- occur and no step is left uncovered.
talloid), should always be accessible on the 2. The patient and circuit are prepared appropri-
pump cart. ately for the procedure including preparing a
• An emergency blood product supply should be sterile field.
maintained in the Blood Bank or in an appro- 3. The patient is taken off ECMO as per protocol
priately designated refrigerator in the patient and ECMO support is temporarily interrupted/
care area. stopped. (Table 32-3)
• Additional supplies for performing procedures 4. The affected component is isolated utilizing the
collectively referred to as an ‘Emergency Kit,’ ‘double clamp’ method. (Table 32-4)
should also be readily available. (Table 32-1) 5. The damaged component is then cut out and
• Backup personnel must be ready to aid in the discarded.
event of the emergency. These clinicians pro- 6. The new component is placed airlessly using
vide expertise and assistance during an emer- a saline drip method. This ensures that all air
gent, off-pump event and may be in house or
on call. The personnel may include physicians,
perfusionists, coordinators, or other ECMO Table 32-2: ECMO Time Out - Planned ECMO Procedure
***Assure Right Patient and Right Procedure before
Caregivers. beginning.
Right People MD Team Lead Role Assignment
ECMO Attending completed and
Basic Emergency Management Physician understood by
ECMO Specialist participants
ECMO Name tags?
Every center should have specific procedures for Primer/Perfusionist
RNs (documenter, meds,
changing out each specific component of the ECMO misc.)
circuit. While there are differences in preparing dis- RRT
Right Review appropriate Out loud Review with
posable parts for replacement, the general approach Plan/Procedure Policy/Procedure Roles Assigned
to changing parts should be consistent with minimal Right Prep Tubing prep
Volume: Blood and
Give time to prep
Discuss specific
variation, including the total circuit change: Crystalloid amounts/types
Medications Code drugs (?x
Patient sedated/paralyzed rounds)
Table 32-1:‘Emergency Kit’ Components Medications on ECMO As required
side? Do they need to be
• Sterile scissors or #10 blade Ventilation specifics moved to pt?
• Sterile tubing clamps Ambu-bag or
mechanical
• Betadine swabs or Chlorhexidine ventilation?
Gluconate (CHG) prep swabs Right Review procedure and Where is secondary
Equipment assure all equipment and equipment or supplies
• Needles or blunt tips supplies available, with available?
• Various sizes of syringes backup as necessary
• Sterile and non-sterile 4 x 4 gauze pads Right Cuff B/P present and QRS audible
Monitoring cycling Is length needed if
• 1-liter bag normal saline at bedside IV Access for under drapes?
• 1-60cc syringe with 60cc NS-changed q meds/volume
Timer for procedure
24 hours-dated, timed and initialed Right Rescue Code Cart Open and available
• Hand crank available on pump Code Sheet with code sheet on top
Defibrillator
• Safety goggles or mask with shield Right Attitude “Does anyone have any State out loud with
• Sterile and non-sterile gloves questions or concerns?” time for response

263
Chapter 32

bubbles are removed from the newly made Teaching every ECMO Caregiver how to double
connection. clamp, cut tubing, and fill tubing airlessly is invalu-
7. Leaving the patient OFF ECMO, all procedural able. Tubing clamps can be difficult to manipulate
clamps are removed from the circuit. the first time an inexperienced person operates them.
8. Blood flow is reestablished. If a bridge is pres- Practicing opening and closing clamps securely on
ent, it is preferable to establish flow across the tubing of various sizes ensures a secure occlusion
bridge before returning the patient to ECMO of the tubing and prevents fluid loss. Additionally,
to assure no leaks, clamps, or other identified the use of heavy duty scissors or a scalpel requires
problems remain. If no bridge is present, take a practice to assure that the cut is sure and even and
few seconds to assess the circuit for these issues avoids ragged edges. The ability to make an airless
before returning to ECMO. connection on the ECMO circuit takes practice,
9. The patient is returned to ECMO as per protocol. practice, practice! It is absolutely essential that an
(“V-B-A” or “V-A”) ECMO Caregiver has the ability to make air free
connections. This is best accomplished when utiliz-
ing additional care team members to drip saline in
Table 32-3: Off-Bypass Procedure with a Stopcock Bridge
Taking a patient off bypass (A-B-V) the de-aeration and reconnection phases.
1. Clamp Arterial line above the bridge
2. Open Arterial then open Venous bridge stopcock Thrombosis
3. Clamp Venous line above the bridge
4. If necessary, stop flow (or decrease to < 200cc/min to
flow through Stopcock Bridge) Clots in the oxygenator are the most common
Returning a patient to bypass (V-B-A) mechanical complication during ECMO (51 - 58%).1
1. Open Venous line clamp Even excellent anticoagulation for the circuit can-
2. Turn Venous then Arterial bridge stopcock off to
circuit
not completely prevent clot formation. In fact, the
3. Open Arterial line clamp presence of ‘normal’ clots should be recognized and
4. Return to previous ECMO flow rate taught to the ECMO Specialist. These clots are small
Eliminate the B if there is no bridge present. in size and usually have no potential to cause harm

Table 32-4: Double Clamp and Cut Procedure

1. If it is a large component to be removed, such as an oxygenator:
• Using 2 clamps for above the component, place 1st clamp closest to the component to be
removed, then the 2nd clamp ~ 2 – 3 inches distally from the 1st clamp.
• Using 2 clamps for below the component, repeat as above.
• Using heavy scissors or a scalpel blade, cut closest to the clamp that is nearest to the
component being removed. This will assure that there is ample tubing left to make a new
connection.
2. If it is a small component to be removed, such as a connector or pigtail:
• Only using 2 clamps, place each clamp on either side of the component.
• If only a pigtail is to be removed, the distance of the two clamps from the site should only be
such that blood loss is prevented.
• Once the clamps are in place, the pigtail can be safely removed, and a new one placed
airlessly.
3. When removing a connector, assure that each clamp is positioned far enough away from the
connector to allow ample room to make a cut and then a connection. The primary purpose of these
clamps is to prevent blood loss.
4. Cut closest to the connector, leaving as much tubing as possible on the side to be connected to the
new connector.
5. Follow the saying “Cut closest to the thing you want to get rid of”. This will assure that you
always have enough tubing to make new connections.
6. Maintaining aseptic technique and cleaning the tubing prior to cutting should be incorporated into
all policies pertaining to circuit procedures.
Prior to placing 2nd clamp, if you squeeze the tubing, then place 2nd clamp, it will cavitate the fluid
within the tubing and decrease the likelihood of blood spray when the tubing is cut.

264
 ECMO Emergencies

to the patient or circuit. Normal clots occur wherever also be caused by clot formation within the cannula
there is stagnation that is not preventable, such as or vessel itself.
at the dependent corners of the Maquet Quadrox Once clot formation has been identified, and is
iD Adult membrane lungs. (Maquet Getinge Group, causing a component failure, tubing occlusion or
Rastatt, Germany) This area may show dark clots is at risk of embolizing to the patient, the caregiver
after a few hours or days into the ECMO course. If must remove or change the component or the cir-
the clots do not encroach upon the blood inlet or gas cuit. This may be done electively before complete
outlet or cause pressure changes within the mem- failure occurs, often in a more controlled, planned
brane, they are usually not considered dangerous. procedure.
Clot assessment can be difficult for the ECMO
Specialist. The use of a bright halogen flashlight Air Embolism and Circuit Pressures
assists in the discovery, visualization, and analysis
of any clots. The ECMO Specialist must assess the Air in the circuit occurs in 46% of all adult
ECMO circuit with a 3-dimensional approach. The ECLS run according to the ELSO Registry.1 It can
‘top’ of the tubing may be clear, but the ‘bottom’ range from small bubbles visualized in the venous
or dependent side can be riddled with clots. Clot side of the circuit to a massive air embolus to the
formation within tubing and connectors is often patient. The ECMO circuit has both negative and
very easy to see and can be defined as dark clots or positive pressure areas within the system. Nega-
white, fibrin strands. tive pressure is measured on the venous drain side
Turbulence, defined as the spinning and tossing of the circuit, originating from the venous cannula
of blood cells, causes lysing of cells. Red blood and terminating in the bladder, if one is employed
cells or white platelets are most commonly affected within the circuit. Positive pressure begins at the
by this phenomenon. Depending on which cells are point where blood starts to move forward though the
present, the clots will be colored accordingly. Most pump head, into the oxygenator through the arterial
often seen streaming from connectors, fibrin strands return cannula. The appropriate placement and use
result as an effect of having fibrin wrap around lysed of pressure monitors that servo regulate the pump
cells while attempting to form a solid clot. Every provide increased safety for the patient. 6
ECMO Caregiver must review the clotting cascade Centrifugal pumps are preload and afterload
and understand its activation occurs with any expo- dependent. ECMO flow with a centrifugal pump
sure to biomaterials. This exposure and activation is achieved with adequate preload or venous drain-
causes an attempt to ‘clot off’ the ECMO circuit as age to the pump head, and may be increased with
an innate protective strategy. Dark clots often form decreased afterload or resistance to the pump head.
at sites of stagnation, including the oxygenator, the Many centers forgo monitoring circuit pressures
bladder or the centrifugal pump head. The ECMO when centrifugal pumps are used. However, these
Caregiver is tasked with locating these clots and pumps can exert very high negative pressures on
monitoring them for any potential change through- the venous limb of the ECMO circuit. If any port
out the ECMO run. is opened on this side of the circuit, air can be en-
The use of pressure monitoring aids in the trained very quickly that will travel swiftly to the
assessment of clot formation and potentially may first air trap–usually the bladder or oxygenator.
prevent worsening of the component failure with The Roller Head pump must have a servo-
the early warning. For example, clot formation regulated pressure monitoring system. Without it, a
within the oxygenator may cause occlusion of kink or occlusion of the venous limb of the circuit or
blood flow and increased premembrane pressure. loss of venous return can have catastrophic results
The postmembrane pressure aids in monitoring as this pump type will generate significant negative
any occlusion downstream from the oxygenator. pressure. As a result, a phenomenon known as cavi-
Increased postmembrane pressure may result from tation occurs as air is pulled out of solution. Without
kinking or movement on the patient side, but may safety mechanisms to stop the pump in the event of
loss of blood return, large amounts of air can travel

265
Chapter 32

into and throughout the circuit and ultimately into from the cerebral circulation. Once the patient has
the patient. been stabilized, the use of a hyperbaric chamber
Positive pressure may also become excessive, for decompression should be considered. If air has
contributing to the failure of a component or causing entered the coronaries and caused acute cardiac
a separation of tubing. Without a servo-regulated decompensation, high dosages of inotropic drugs
pressure monitor to assess pressures within and may be necessary.7
beyond the membrane, circuit rupture can result if Still, the discovery of air within the circuit could
an occlusion were to occur. be benign. Air entrainment on the venous side from a
Another potential source of air occurs when the loose connection or dislodgement of a side hole of a
partial pressure of postmembrane oxygen (mea- drainage cannula may result in the entrainment of air
sured by a blood gas) in the blood is very high or bubbles. These bubbles are small and easily moved
supersaturated, potentially forcing oxygen out of by raising the tubing and allowing the air to rise and
solution. Hitting the membrane or operating the move forward with ECMO flow. Where the bubbles
circuit in a low ambient pressure environment (such can be trapped depends on each individual circuit
as in flight in a nonpressurized cabin) may produce configuration. If present, the bladder is considered
foam at the top of the oxygenator. Prevention of the first air trap within a circuit. The next air trap
this phenomenon is easily achieved by keeping the is the oxygenator. Once the bubbles are isolated in
postmembrane pO2 < 600 mm Hg. the bladder, they can be easily removed. In circuits
Additionally, the gas egress or exhaust should without a bladder, they may become trapped in the
never be occluded during the ECMO run. While the centrifugal cone, the oxygenator, or momentarily
Specialist is encouraged to assess for a condensate in the raceway. Bubbles present in the cone may
as well as for signs of pink-tinged fluid, they should be removed by stopping blood flow and allowing
never fully occlude the port with their finger. Even the air to move out of the cone into the oxygenator.
this brief occlusion can cause a precipitous rise Most ECMO programs use a Quadrox oxygenator.
in the gas phase pressure and risk a rupture of the The presence of the deairing port on the top venous
membrane. side allows air to easily transverse through the hy-
The hourly circuit checks performed by the drophilic membrane and out of the oxygenator. Air
ECMO Specialist permit the prompt recognition bubbles that are large may overwhelm the capacity
and discovery of an air embolus, potentiating an of the Quadrox, especially the Pedi-Quadrox, requir-
appropriate response and resolution to the finding. ing immediate intervention for removal through the
However, the chance of an air embolus occurring postmembrane pigtail.
at the exact moment the ECMO Specialist actively Finally, the use of an air detector that servo
assesses the circuit would be incredibly rare. Yet, regulates the pump must be considered as the ul-
every Specialist should become an expert at the timate preventive measure. While air embolus is a
continuous visual checks of the circuit. If a bubble is rare complication, the use of this simple device on
seen traveling to the patient, or if the bubble detector the tubing before the return cannula can prevent any
alarms, stopping pump flow first will immediately air from entering the patient.
prevent the air from traveling further due to the for-
ward flow of the ECMO pump. It is then imperative Membrane Oxygenator Failure
that the clinician clamp the arterial limb of tubing
close to the patient to prevent further movement The majority of ECMO Centers use a polymeth-
of air upwards toward the patient. The bridge is ylpentene (PMP) oxygenator. These oxygenators
opened, the venous limb of tubing clamped, and are extremely effective and not prone to failure as
the patient is separated from ECMO. If, however, was seen with previous silicone membrane oxygen-
air has already entered the patient, additional pro- ators. Oxygenator failure can be defined with gas
tective measures must be taken. Once the patient is exchange alterations on blood gases and with pres-
off ECMO, the head is lowered relative to the body sure measurement changes. Gas alterations are seen
as much as possible in order to move any air away when oxygen or carbon dioxide transfer is decreased.

266
 ECMO Emergencies

A significant increase, whether acute or gradual, in nection be tie-banded to further ensure maintenance
premembrane pressure measurements without ac- of the connections, even manufactured or glued
companying rise in postmembrane pressure, known connections. This prevents blood leakage from
as an increasing transmembrane pressure, can reflect any weak points that may occur in the event of
clot formation within the oxygenator, leading to a over pressurization of the blood path. The use of a
need to replace it. Blood leaks from the gas exhaust servo-regulated system also prevents the accidental
port are another troublesome sign and usually war- increase in pressure, avoiding accidental tubing
rant the removal of the oxygenator. separation.
Membrane replacement is considered an ad- Raceway rupture is a potential emergency, and
vanced procedure because it occurs uncommonly. every ECMO Specialist should be trained to perform
While the bedside Specialist would be responsible the timely changing of a raceway. The prevention of
for recognizing the signs and symptoms of failure rupture is easily accomplished by routine ‘walking
and alerting the care team, it is beneficial to have of the raceway.’ This procedure involves advancing
a core group of Specialists or Perfusionists trained the section of raceway tubing that has been in the
change the oxygenator. Careful surveillance of an roller pump head out of the pump head, allowing
ECMO circuit facilitates a planned change out pro- for a new segment of raceway tubing to remain in
cedure instead of an acute emergent event. Every contact with the roller, redistributing pressure due to
ECMO center must design their specific policy for tubing wall stress. Each ECMO Center should have
identifying oxygenator failure and indications for a defined policy for the frequency of this procedure.
change. The procedure should also be individual- In the event that a raceway accidentally ruptures,
ized for the center encompassing the type(s) of the procedure for changing it out is straight forward.
oxygenators employed. All supplies must always be readily available. The
Circuits that have an incorporated pump head patient must be removed from ECMO immediately
and oxygenator such as the Maquet Cardiohelp HLS, and emergency ventilation and patient manage-
do not readily allow the oxygenator or any compo- ment strategies employed. The actual procedure
nent to be changed individually. The entire circuit for changing out the damaged section of tubing
must be changed in the event of oxygenator failure. may vary across institutions. An example of how
to perform it is provided in Table 32-5.
Tubing Rupture
Centrifugal Pump Head Failure
Tubing rupture is a rare mechanical complica-
tion due to the diligence of the ECMO Specialist, as Although the newest generation of centrifugal
well as the development of Super Tygon (S95E or pumps has minimized the many historically experi-
S65HL) (Norton Performance Plastics, Inc., Akron, enced risks of failures, they still require a disposable
OH) raceway tubing. However, polyvinyl chloride pump head that has the propensity to fail in a variety
tubing (PVC), used for all other tubing requirements of ways. The most important step for preventing
aside from the raceway, can become damaged or centrifugal pump failure is ensuring proper seating
cracked. The Specialist assesses the entire ECMO of the pump head into the drive unit. This can be
circuit, assuring that tubing is not caught under accomplished through a locking pin incorporated
wheels, kinked, or worn. The use of guarded tubing into the pump itself to secure placement and prevent
clamps is a safer strategy as they do not allow tubing accidental dislodgement, or through a “twist and
to become ‘bitten’ or damaged within the jaws of turn” locking procedure. Some centrifugal pumps
the clamps. This decreases the damage that may be have a small lip upon which the disposable pump
done to the bridge and other tubing pieces. Careful head resides to maintain proper placement. While
use of nonpenetrating towel clamps will also prevent visual inspection of the pump head is critical in di-
accidental piercing of tubing. agnosing pump head dislodgement, auditory clues
Tubing connectors are used throughout the are also available to clinicians. Often, a “grinding”
ECMO circuit. It is recommended that every con- noise is generated when the disposable pump head

267
Chapter 32

is not properly engaged with the pump, alerting the be visualized. However, if a probe is not present, the
ECMO Specialist that an intervention is needed. clinician must quickly recognize whether the level
However, clinicians must remember that mainte- of support has altered by assessing changes in circuit
nance of proper position of the pump is crucial to pressures and patient vital signs. Replacement of the
prevent loss of flow, hemolysis, or damage to the conduction paste is not always necessary. If there
pump itself.6 is verified flow to the patient, you may be able to
Other potential centrifugal pump failures de- continue with the SIG error. If outflow cannot be
serve consideration by ECMO clinicians. Some confirmed, then the paste must be replaced quickly
models require a conduction paste on the outflow to minimize interruptions in ECMO support.
side of the pump head in order to communicate and Another potential area of failure resides within
deliver a flow signal to the main console. Absence, the design of the pump head itself. Certain models
age, and amount of the paste can lead to an unin- use a low friction bearing to drive the propellers of
tended flow signal disruption or “SIG” error, causing the centrifugal pump head. Fibrin and clots collect
the ECMO Specialist to falsely believe that flow has in this area of the head and can create an occlusive
been disrupted. If an independent, secondary flow body that prevents any inflow or outflow from the
probe is used on the arterial return limb of the circuit, head. Symptoms include no outflow from the pump
then confirmation of continued ECMO support can head even though a visual inspection of the head
Table 32-5: Raceway Rupture Procedure
A. Equipment:
1. Sterile section of 1/4" or 3/8" or 1/2” raceway tubing (Super Tygon-S95E or S65) with
connectors attached of appropriate size
2. Sterile scissors or sterile scalpel blade #10
3. 4 tubing clamps
4. 60cc syringe filled with Normal Saline
5. Betadine or CHG swabs

B. Procedure:
1. Call for help, Turn off pump.
2. Come off bypass as per protocol
3. Clamp inlet and outlet tubing of pump head
4. Open tubing gate clamps and remove raceway from pump head
5. Double clamp each side of the raceway section out leaving 2" between clamps
6. Ideally, wipe tubing between clamps with Betadine or CHG swab briskly and allow to
dry
7. Using sterile scissors or scalpel blade, cut between double clamps and remove raceway
8. The new raceway section should first be connected to end of circuit tubing closest to
bladder
9. Release clamp on circuit tubing closest to bladder and allow volume from bladder to fill
tubing by gravity. (drop tubing level down to fill)
10. Once the tubing is filled, clamp tubing on fluid filled section
11. Using a 60cc syringe with needle, fill remaining raceway section with normal saline
12. Tap gently as it fills to remove air bubbles
13. Make airless connection to other end of circuit tubing on the pre-membrane side.
14. REMOVE ALL CLAMPS
15. Place tubing into pump head, clockwise, following lay of tubing
16. Secure the inlet side tubing gate first.
17. Turn pump head forward several turns and ensure adequate seating of the raceway, then
secure the outlet tubing clamp.
18. The bladder and/or pre and post pressure readings will be alarming!
19. Give volume (saline, albumin or pRBCs) until bladder alarm is not sounding.
20. Turn on pump flow and recirculate through bridge, check for clamps, air, etc.
21. Bladder alarm may sound again as pressure equalizes.
22. Go on bypass as per protocol
23. Check ACT, Hct, replace volume if necessary
24. Tighten connections and tie band

268
 ECMO Emergencies

reveals continued spinning. Another sign is the change out as individual component cannot be
presence of a consumptive circuit coagulopathy changed.
that may indicate clot formation. Additionally, in A final consideration for centrifugal pump
circuits with a bladder, the bladder (venous line) failure is when a complete loss of communication
pressure will grow increasingly positive as it cannot occurs between the console and the drive unit. This
empty into the pump head. Patient decompensation can be quickly determined through careful examina-
can rapidly ensue. Therefore, once the suspicion of a tion of the console itself, specifically pursuing the
clot is confirmed, the ECMO Specialist must begin pathway of the cord that connects the drive unit
the change out procedure to replace the pump head to the console and facilitates all communication
quickly. (Table 32-6) There are centrifugal pumps between the two pieces of equipment. If a strong,
where you are unable to fully visualize the pump secure connection remains in place and the com-
head and other critical components of the circuit. munication failure persists, the clinician should
With these systems, diagnosis of pump failure (i.e. initiate manual support of the patient per center
loss of forward flow) requires a complete circuit policy, minimizing the interruption in ECMO sup-

Table 32-6: Centrifugal Pump Head Change Out Procedure

A. Equipment:
1. Replacement Centrifugal Pump Head
2. Sterile Scissors or sterile scalpel blade #10
3. 6 Guarded Tubing Clamps
4. Betadine or Chlorohexidine Swabs
5. 2 – 60 cc syringes filled with 0.9% NS

B. Procedure:
1. Call for help, Turn off pump.
2. Come off ECMO as per protocol –
3. The patient should be separated from ECMO support through the “A-B-V” or “V-A”
process
4. Clamp inlet and outlet tubing of pump head
5. Place a second clamp on each side of the pump head leaving 2" between clamps
6. Ideally, wipe tubing between clamps with Betadine or CHG swab briskly and allow to
dry
7. Using sterile scissors or scalpel blade, cut between double clamps and remove pump head
8. The new pump head should first be connected to end of circuit tubing closest to bladder
9. Attach a 60 cc syringe of 0.9% NS to a pigtail above the bladder (if present) or to a
pigtail above the inlet side of the pump head
10. Release clamp on circuit tubing closest to bladder and simultaneously push volume into
the circuit from the 60 cc syringe of 0.9% NS. This will allow volume from bladder to fill
tubing and pump head.
11. Move the pump head directionally to allow the air to exit through the outlet as the 0.9%
NS fills the pump head
12. Tap gently as it fills to remove air bubbles
13. Once the pump head is filled, re-clamp above the inlet of the pump head and maintain the
pump outlet in an upward position to prevent fluid loss
14. Make airless connection to other end of circuit tubing on the pre-membrane side by using
a 60cc syringe (without a needle) to drip in 0.9% NS to complete the connection.
15. REMOVE ALL CLAMPS near pump head but have PATIENT REMAIN OFF ECMO
16. Seat centrifugal pump head into the drive unit and ensure new pump head is securely in
place
17. Be aware that all pressure readings will be alarming!
18. Volume (saline, albumin or pRBCs) may need to be administered to the ECMO circuit
until bladder alarm is not sounding – if applicable
19. Turn on pump flow and recirculate through bridge, check for clamps, air, etc.
20. Monitored pressures may sound again as pressure equalizes within the circuit
21. Place patient back on ECMO as per protocol - through the “V-B-A” or “V-A” process
22. Check ACT, Hct / Hgb and replace volume if necessary
23. Tighten and tie band all newly made connections

269
Chapter 32

port. Simultaneously, the ECMO Specialist should contribute to LV failure and cardiac stun. In addition,
quickly reboot the entire system to try to reestablish the cannula can be improperly placed across the
the communication. If communication cannot safely aortic valve, causing aortic insufficiency. Finally, the
be reestablished or maintained, a new system should cannula can be inserted against the LV endothelium
be brought to the bedside and the former system with the potential for LV disruption or perforation.
replaced. Finally, Specialists must remember that Assuring proper suturing technique lessens the
hematuria indicates the presence of hemolysis and chance of accidental cannula kinking. The Special-
merits further investigation. Hematuria is a sign ist may also use a blanket or towel bundle placed
that the heat generated within the pump head is underneath the cannula to support the weight of
excessive, causing hemolysis. This is most often at- the tubing connections and decrease likelihood of
tributed to high RPM settings on the pump. Plasma dislodgement.
free hemoglobin levels should be drawn and trended
daily as it is an excellent trending tool to depict Accidental Decannulation
levels of hemolysis that may be present.7
Accidental decannulation occurs uncommonly;
Cannula Problems but can have devastating consequences because
of hemorrhage and loss of ECMO support, and is
ECMO cannulas are inserted during a sterile usually preventable. One strategic, preventative
surgical procedure and care must be undertaken to measure that must be undertaken by all ECMO
avoid vascular injury. Even after appropriate venous Specialists is appropriate securing of the cannulas
cannula placement, cannula obstruction due to kink- to a fixed object, (i.e. the mattress or side of the
ing is a common problem; therefore, adequate flow bed). Relying on sutures and/or tape on the surgical
is extremely dependent on cannula positioning and site is ill advised as these can become loose over
fixation. Caution must be used by the surgeon as time on ECMO. One way to assure the placement
sutures have been known to kink a cannula. of the cannulas is maintained is for the Specialist
If fluoroscopy is not used during insertion, to observe and record the depth of insertion of
venous and/or arterial cannula position must be each neck and/or groin cannula at the incision site.
assessed by chest radiograph or echocardiogram to Comparing measurements with those from previ-
confirm placement. Initially, the practitioner will be ous shifts, the ECMO Specialist will be able to
able to determine effective placement by the ability identify placement changes and notify the ECMO
to achieve calculated flows ranging from 120–150 physician. The use of towel clamps, unique tubing
cc/kg/min in pediatric patients or 3-6 LPM for adults. holders made for ECMO tubing, or other methods
An inability to achieve full flow on ECMO may be of securement must be carefully evaluated by each
attributed to a misplaced venous cannula and may center for their efficacy.
sometimes be remedied by the surgeon at the time Additionally, the ECMO Specialist must protect
of cannulation. High arterial return pressures may their environment at all times. Parents, nursing
indicate a problem with the arterial return cannula. staff, and other personnel must be made aware of
In children, difficulties with arterial cannulation the placement of cannulas and advised not to touch
can also arise from the catheter being inserted too far the attached tubing and circuitry without assistance
into the ascending or descending aorta, compromis- from the Specialist. The Specialist also assures that
ing coronary and cerebral oxygenated blood flow, both the ECMO pump and patient bed are appro-
or from being misplaced into the subclavian artery. priately placed in the braked position to prevent
If the returned blood flow can selectively enter the inadvertent moving of these objects away from
right subclavian artery, the right upper extremity each other. Finally, the nursing staff must assure that
can be infused with hyper-oxygenated blood flow, patients are restrained in order to protect themselves
while the rest of the body is hypoxic and cyanotic. A from harm. Even with all precautions in place, ac-
cannula in the ascending aorta can cause increased cidental decannulation still may occur. For example,
afterload to left ventricular (LV) outflow and may tissue fatigue or loss of the sutures with long-term

270
 ECMO Emergencies

cannulations may allow cannulas to be dislodged. that every Specialist should know and have access to.
High pressure situations without a servo-regulated In the event of complete failure, a suitable replace-
system may cause arterial back pressure and allow ment must be a readily available, and the Specialist
the cannula to displace. During transport, the bed is responsible for appropriately managing such a
may become separated from the ECMO pump by change-out procedure.
too much distance. Every ECMO program must Loss of electrical power, while unlikely, may
consider the potential for this emergency and have a occur. There have been many accounts of this oc-
policy in place to manage it. An example of a policy curring during a natural disaster, such as Hurricane
for this emergency is in Table 32-7. Katrina in New Orleans. The loss of power to the
pump head can be managed with an uninterruptible
Equipment Failure battery source. The battery should engage immedi-
ately with any power loss and ensure maintenance
The potential for malfunctioning equipment of pump flow. However, in the event that the battery
always exists. Each manufacturer describes a fails or is fully discharged, manual hand cranking
troubleshooting regime in their equipment manuals must be performed. Every Specialist should practice
this procedure during their water drills and be famil-
Table 32-7: Accidental Decannulation Management iar with assessing the efficacy of their hand cranking
Procedure when there are no patient monitors available.
A. Diagnosis:
1. Air in venous limb of circuit for venous
decannulation Miscellaneous Circuit Components
2. Venous bladder alarm, pump stopped
3. Blood loss at cannulation site
4. Cannula visible out of incision site
Individual circuit pieces may also crack or fail
from frequent use and fatigue, including the pigtails
B. Management: and stopcocks. Prevention is the best method to
1. Turn off pump flow first! Call for help, including
surgeon!
prevent emergent failure. Avoiding over tightening
2. Take patient off bypass as per protocol and circulate of stopcocks onto pigtails, and pigtails onto con-
through the bridge nectors will lessen the chance of cracking the luer
3. Have bedside RN place direct pressure on site
4. Prepare to immediately give available volume (NS).
connections. Using gauze to protect pigtails when
5. Call for emergency blood from Blood Bank clamping them prevents the tubing clamps’ ‘teeth’
6. Stop ECMO flow from indenting and cracking the miniature tubing.
7. Attach volume to side of bridge pigtail opposite of
decannulated site, OR,
Additionally, routine of changing frequently used
8. Attach volume to pigtail closest to arterial return stopcocks, such as the venous sampling port, pre-
cannula or on venous limb vents the stopcock from becoming ‘sticky’ or stiff
9. Clamp bridge, and unclamp line of remaining
cannula
with use and eventually cracking. Stopcocks that
10. Push volume are used on the stopcock bridge may also fail with
11. Continue to give volume as necessary improper positioning of the bridge, causing torque
12. Every 5 minutes, stop pushing volume, clamp line,
unclamp bridge and recirculate briefly
and stress on these parts. Assuring that the bridge
13. The first recirculation will require removal of air in is secured without inadvertent twisting and tension
venous limb through bladder minimizes the potential of failing.
14. If volume is still required, return to procedure,
otherwise, recirculate
Summary
• In the event of an accidental decannulation on a patient
with a double lumen catheter, the patient will have to be
managed without the benefit of ECMO catheters to give
The possibility of encountering a mechanical
volume. complication during an ECMO run is ever present.
• Volume should not be administered through a cephalad The Boy Scouts’ motto, “Be Prepared,” is a fitting
catheter. one for any ECMO Specialist to embrace. The bet-
• These management techniques are to be only used with
two cannulation site VV ECMO or VA ECMO. ter prepared and educated the ECMO Specialist is,
the more likely that an encountered emergency will

271
Chapter 32

be handled quickly, efficiently, and safely; ensuring

the best possible outcome for the patient. Therefore,
every ECMO center must develop and maintain
comprehensive guidelines or policies and proce-
dures for managing potential failures in the realm
of ECMO circuitry and equipment.8
The Authors would like to acknowledge Joseph
B. Zwischenberger who stated: “You can’t die on
VA bypass!”- but found that complications could
overrule that. He invented many of the ways to
manage circuit complications, mainly because they
happened to him first.

272
 ECMO Emergencies

References

1. ECMO Registry of the Extracorporeal Life Sup-

port Organization (ELSO), Ann Arbor Michigan,
August, 2017.
2. Fleming GM, Gurney JG, Donohue JE, Reme-
napp RT, Annich GM. Mechanical component
failures in 28,171 neonatal and pediatric extra-
corporeal oxygenation courses from 1987 to
2006 Ped Crit Care Med. 2009;10(4):439-444,
doi: 10.1097/PCC.0b013e318198b275.
3. The ELSO Guidelines for Training and Continu-
ing Education of ECMO Specialists. elso.org.
https://www.elso.org/Resources/Guidelines.
Version February 2010. Accessed November
11, 2017.
4. Manley G, Knudson MM, Morabito D, Damron
S, Erickson V, Pitts L. Hypotension, hypoxia,
and head injury frequency, duration, and con-
sequences. Arch Surg.2001;136(10):1118–1123,
doi:10.1001/archsurg.136.10.1118.
5. Howard RS, Holmes PA, Koutroumanidis MA.
Hypoxic-ischaemic brain injury. Practical Neu-
rology. 2011;11(1):4-18.
6. Toomasian JM, Vercaemst L, Bottrell S, Hor-
ton SB. The Circuit. In Brogan TV, Lequier L,
Lorusso R. et al, eds. Extracorporeal Life Sup-
port: The ELSO Red Book. 5th ed. Ann Arbor,
MI Extracorporeal Life Support Organization,
2017:49-80.
7. Heard ML, Lynch JE, Zwischenberger JB. Me-
chanical Complications. In Short BL, Williams
L. eds. ECMO Specialist Training Manual Third
Edition. Ann Arbor, MI Extracorporeal Life
Support Organization, 2010:99-111.
8. The ELSO Guidelines for ECLS Centers v1.8.
elso.org. https://www.elso.org/Resources/
Guidelines. Version March 2014. Accessed
November 11, 2017.

273
33

Patient Troubleshooting

Barbara Haney, RN, MSN, RNC-NIC, CPNP-AC, FELSO, Patricia English, MS, RRT-NPS, Gillian Wylie,
RN, RSCN, BSc

Problem Causes Solutions

Arterial Line Unclear etiology Echocardiogram (Echo) if sudden loss of
ejection in previously ejecting patient to
Tracing Flat exclude early tamponade/LA distention
High/full VA ECMO support May be appropriate with full VA non-
Patient well perfused,
pulsatile support - no intervention needed
narrow or no pulse
Decreased cardiac output Support cardiac output as needed (e,g.
pressure
fluid bolus, inotropic support)
Pressure transducer Flush, zero, replace pressure transducers
malfunction as needed
Cardiac stun  Full cardiac support with VA-
ECMO
 Allow time for myocardial
recovery
 If cannulated via carotid artery,
evaluate arterial cannula tip to
ensure not directed on the aortic
valve
 Evaluate/decompress left heart for
overload

275
Chapter 33

Problem Causes Solutions

Bleeding ACT/PTT high  Decrease ACT/PTT parameter goal
 Consider checking heparin level
 Check coagulation profile
 Consider holding or decreasing anticoagulation
Platelets low  Administer platelets
 Consider heparin-induced thrombocytopenia (HIT) if
persistent thrombocytopenia occurring after repeated heparin
exposure
DIC  Other blood product replacement as indicated by laboratory
values
 Change circuit if suspect circuit DIC (consumptive
coagulopathy without patient source)
Sepsis Sepsis evaluation and treatment
Recent surgery or procedure  Consider holding anticoagulation
 Consider aminocaproic acid/aprotinin/tranexamic acid infusion
or Novo7 administration
 Consider thromboelastography to guide management
 Local control if isolated site (sutures to cannula, Surgicel,
StatSeal, Quick Clot, etc.)
 Evaluate and consider stripping chest tubes
 Consider surgical reexploration/washout
Agitation  Calm or sedate patient, increase dose or consider an additional
agent
 Consider paralysis if cannulation or surgical site bleeding
 Provide distractions, child life therapy, PT/OT, oral
sucrose/pacifier
 Encourage family/spouse participation
 Evaluate for delirium
Hypertension  Minimize/concentrate infusions
 Diuresis
 Decrease vasopressors
 Reduce VA ECMO pump flow
 Treat agitation/pain
 Anti-hypertensive medication
Inadvertent decannulation  Prevention:
o Good communication between team members with all
patient care and transports
o Frequent assessment of intact sutures /suture
security/radiographic or ultrasound placement verification
o Display at bedside last known position of cannulae (i.e. if
high or low)
o Prevent tension on circuit that could promote cannula
dislodgement
 Firm pressure at cannulation site to limit patient blood loss
 Clamp circuit to limit circuit blood loss/minimize air in circuit
to preserve for continued use
 Prevent circuit contamination
 STAT page surgical team
 Transfuse/volume resuscitation as needed

276
 Patient Troubleshooting

Problem Causes Solutions

Cardiac Arrest Any cause of cardiac arrest  Initiate CPR if on VV ECMO
 Maximize ECMO pump flow if on VA ECMO (may
require volume)
 Initiate CPR if on VA ECMO and unable to maintain
adequate flow
Bleeding See “BLEEDING” section
Tamponade  Drain pericardium
 Aggressively strip chest tubes
Tension pneumothorax  Evacuate pneumothorax
 Strip chest tubes
 Decrease ventilator settings if adequate ECMO support
to allow healing of air leak
Electrolyte imbalance Check and correct electrolyte imbalance
Hypoxia See “DECREASING PATIENT PO2” section
Circuit malfunction or  Initiate CPR
disruption or inadvertent  Address circuit malfunction/disruption
decannulation
 See “INADVERTENT DECANNULATION” section
Arrhythmia  Evaluate and correct potential causes (refer to
ACLS/PALS protocols)
 Consider antiarrhythmic medications as indicated

Problem Causes Solutions

Decreased Hypovolemia/hypotension Volume or pressor support
Capillary leak syndrome  Volume or pressor support
Urine Output
 Diuretics
Poor cardiac output  Increase pump flow
 Stimulate cardiac output and renal blood flow
 Consider vasodilatation to reduce afterload
Ischemic renal disease Hemofiltration, dialysis
Inadequate ECMO flow Increase ECMO flow
Urinary catheter mal-  Assess for position, function, & patency of urinary
positioned or occluded catheter
 Evaluate for clot in bladder

Problem Causes Solutions

Decreasing/Low Sweep gas flow too high Decrease sweep gas flow
Sweep gas CO2/Carbogen Increase or add sweep gas CO2/Carbogen
Patient PCO2 too low
Over ventilation Wean ventilator
Improving respiratory Consider ECMO weaning
function
Hypothermia Adjust ECMO warmer/radiant warmer

277
Chapter 33

Problem Causes Solutions

Decreasing  Pneumothorax, atelectasis,  Check ETT position
Patient PO2 ventilator/ETT problem  Suction/bronchoscopy
 Pericardial tamponade  Obtain chest radiograph or ultrasound
With cyanosis,  Hemothorax  Treat pneumothorax, hemothorax, effusion or tamponade
acidosis, poor  Pleural effusion  Aggressively strip chest tubes in the postoperative cardiac
perfusion,  Deteriorating native lung patient
lethargy, function  Consider echocardiogram
increasing lactic  Inadequate ECMO flow  Increase ECMO pump flow
acid  Excessive recirculation  Reposition catheter(s) or patient (see below)
Hypovolemia Administer volume
Sweep gas line to oxygenator Tighten sweep gas line connections or replace if needed
loose, disconnected, cracked or
blocked
Sweep gas oxygen tank depleted Replace gas tank
Sweep gas blender malfunction Evaluate blender settings, replace if needed
or set incorrectly
Oxygenator failure /high  Increase ECMO FIO2 to increase post oxygenator PaO2
thrombus burden/condensate on  Check post-oxygenator blood gas
oxygenator fibers  Assess transmembrane pressure
 Sigh/burp the oxygenator PRN
 Replace oxygenator as indicated
Seizures  Treat seizures
 Evaluate neuro status
Sepsis Sepsis evaluation and treatment
Agitated patient  Calm or sedate patient
 Provide distractions, child life therapy, PT/OT
 Encourage family participation
 Evaluate for delirium
Hypervolemia, increased  Evaluate blood volume and correct as needed
pulmonary perfusion prior to  Administer diuretics
pulmonary recovery
Decreased cardiac output  Consider vasopressor support if VV-ECMO
especially on VV-ECMO  Evaluate PDA blood flow in the neonate
Significant recirculation on VV-  Evaluate cannula position
ECMO  Change patient position
 Increase Hgb/HCT/
 Consider Pneumothorax / Tamponade
 Decrease ECMO blood flow
 Consider conversion to VA
Decreased patient hematocrit Transfuse if hematocrit/hemoglobin low
Consider structural cardiac Obtain echocardiogram
defect
Change in patient preload or Interventions to increase preload or decrease afterload as
afterload causing inadequate or clinically indicated
altered blood flow
On VA-ECMO improving  Increase ventilator FiO2 and/or other ventilator settings to
cardiac output with poor native optimize native lung function
lung function  If cardiac function adequate convert to VV-ECMO
On VV-ECMO increased  Aggressive sepsis treatment
cardiac output and increased  Vasoactive drugs
tissue oxygen extraction caused  Decrease metabolic rate
by super infection/sepsis

278
 Patient Troubleshooting

Problem Causes Solutions

Decreasing Improving cardiac output  VA-ECMO: Consider weaning ECMO pump flow
with poor lung function, Tolerate lower SpO2/pO2 if adequate perfusion
Patient PO2 resolving cardiac stun

 Optimize hematocrit, ventilator support
With patient well  Increase sweep gas FiO2 if <1.0
appearing Deteriorating native lung  Obtain chest radiograph
function  Check ETT position
 Suction/bronchoscopy
 Optimize PEEP/ventilator management

Problem Causes Solutions

Excessive Error/inconsistency in ACT Review sampling technique, repeat test
technique, amount of blood
Anticoagulation used
New heparin lot/ direct Consider replacement of heparin/direct thrombin inhibitor
High ACT
thrombin inhibitor infusion infusion
High PTT
Error in calculation or Use commercially manufactured pre-mixed heparin/direct
High antiXa
mixture of heparin/direct thrombin inhibitor infusions
thrombin inhibitor infusion
Infusion pump malfunction Check infusion pump programming, consider replacement
or pump set incorrectly
Alterations/malfunction/error  Inspect sampling site
of sampling site/pigtail (e.g.  Change adapter, stopcocks, pigtails if needed
clots, contamination)
 Separate ACT/PTT blood sampling site from
anticoagulant infusion site
 Prevent specimen contamination (line draw, etc.)
 Repeat test
ACT instrument or  Review sampling technique
tubes/cartridges malfunction  Review/perform instrument Quality Control
 Verify appropriate instrument temperature ranges
 Review expiration dates/storage procedures of
tubes/cartridges
 Replace tubes/cartridge or instrument as needed
 Repeat test
Low or decreasing platelet Check platelet count and other coagulation parameters, correct as
counts, coagulopathy indicated
Check heparin level
Sample drawn in a Repeat test using non-heparinized syringe
heparinized syringe
Heparin from another Look for heparin administered in other sources (minimal
source (e.g. TPN, line amounts by continuous infusions usually will not cause ACT
flushed) alterations)
Vitamin K deficiency Replace vitamin K, correct other coagulation parameters as
indicated.
DIC due to circuit Correct cause of DIC (e.g. circuit change, sepsis)
coagulopathy
DIC due to sepsis Sepsis evaluation and treatment
Alteration in device Review/perform temperature checks per manufacturer’s
temperature (device specific) recommendations. Replace instrument as needed.
Decreased urine output Assess for changing urine output and address as clinically
indicated

279
Chapter 33

Problem Causes Solutions

Hemolysis Inaccurate sampling Repeat test - draw slowly, send specimen stat to lab
Draw from pigtails with syringe, do not draw through a
Plasma free needle
hemoglobin Pump over-occluded, raceway  Check roller pump occlusion
greater than 100 tight in pump, inappropriate
 Optimize RPMs to achieve flow
mg/dl; pink/tea high/low RPMs, centrifugal
colored urine;  Check centrifugal pump head position
pump head not correctly seated
renal dysfunction  Change circuit or circuit components if indicated
Water bath malfunction or  Evaluate heater, turn down water bath temperature,
temperature too high replace if indicated
 Clots in patient  Check for clots or kinks in circuit, hemofilter, or
 Kinks in system patient
 Decrease ECMO flow if tolerated
Centrifugal pump, especially Consider pump cone replacement if centrifugal pump or
at low flows in small patients use roller pump
Small cannula for desired or  Consider upsizing or adding additional cannula
high flow  Lower RPM/blood flow if possible

Unique patient characteristics  Evaluate for clinical contributing factors (eg:

adenovirus, sickle cell, etc)
Any cause Increase urine output; consider plasmapheresis or circuit change
to remove plasma free hemoglobin while resolving mechanical
causes

Problem Causes Solutions

High SvO2 SvO2 monitor malfunction  Assess/correct SvO2 monitor
 Calibrate SvO2 monitor
Inaccurate SvO2 measurement  Assess cannula position
due to recirculation on VV  Identify optimal ECMO flow to minimize recirculation
ECMO
 If bridge in place, ensure fully occluded
Inaccurate SvO2 measurement Utilize other indicators of adequate O2 delivery, such as lactic
due to mixing cardiac lesion acid levels, patient venous blood gas
Excessive ECMO flow Wean pump rate

Problem Causes Solutions

Hypertension Fluid overload Diuretics or hemofiltration
Pain Treat pain
Agitation  Calm or sedate patient
 Provide distractions, child life therapy, PT/OT
 Encourage family participation
 Evaluate for delirium
Idiopathic Anti-hypertensive medication
Improved cardiac output Decrease ECMO flow
High pump flow Decrease ECMO flow
Recent inotropic or steroid Wean inotropic or steroid support as appropriate
administration
Renal dysfunction Optimize renal support

280
 Patient Troubleshooting

Problem Causes Solutions

Hypotension Decreased cardiac output Support cardiac output as needed with inotropic support and
increased ECMO flow
Hypovolemia Volume
Capillary leak syndrome Identify patient specific cause and treat as indicated
Massive hemorrhage Transfusion, evaluate coagulation status, surgical consult. See
“BLEEDING” section
Sepsis Sepsis evaluation and treatment
Low pump flow (VA) Increase pump flow if adequate right atrial volume
Tamponade Assess for other signs of tamponade. See “CARDIAC
TAMPONADE” section

Problem Causes Solutions

Inadequate Error/inconsistency in ACT Review sampling technique, repeat test
technique, amount of blood
Anticoagulation used
New heparin lot/new direct Consider replacement of heparin/direct thrombin inhibitor
Low ACT
thrombin inhibitor syringe infusion
Low PTT
Error in calculation or Use commercially manufactured pre-mixed heparin/direct
Low anti Xa
mixture of heparin/direct thrombin inhibitor infusions
thrombin inhibitor infusion
Infusion pump malfunction Check infusion pump programming; including weight,
or pump set incorrectly concentration, and dose/rate, consider replacement if needed
Alterations/malfunction/error  Inspect sampling site
of sampling site (e.g. clots,  Change adapter, stopcocks, pigtails if needed
contamination)
 Prevent specimen contamination (line draw, etc.)
 Repeat test
ACT instrument or  Review sampling technique
tubes/cartridges malfunction  Review/perform instrument Quality Control
 Verify appropriate instrument temperature ranges
 Review expiration dates/storage procedures of
tubes/cartridges
 Replace tubes/cartridge or instrument as needed
 Repeat test
Recent platelet transfusion Consider increasing heparin infusion/ during platelet transfusions
due to heparin inactivation
Recent Factor VII (Novo7) Check coagulation parameters, including antithrombin III level,
transfusion and correct as indicated
Alterations/malfunction of  Change adaptors, pigtail, stopcocks, or tubing as needed
heparin infusion (e.g. tubing  Check infusion pump
clamped, not connected,  Check heparin infusion site
stopcock off, clots,
obstructed flow)
Patient diuresis, large Assess for changing urine output and address as clinically
increase in urine output indicated
ACT low / PTT Check heparin level, if therapeutic consider reducing ACT
normal/acceptable parameter
Incorrect anti-Xa level Try to minimize hemolysis
(falsely lowered) due to high
plasma free hemoglobin
(uninterpretable if PHg>250)

281
Chapter 33

Problem Causes Solutions

Inadequate Inadequate upper body  Assess oxygenation using right arm SpO2 and
perfusion due to mixing on right arm blood gases
cerebral VA-ECMO with peripheral
oxygenation  Optimize native lung oxygenation
(femoral) cannulation.
 Full VA-ECMO support
(“North-South Perfusion of the lower body is
from ECMO support and the  Consider alternative cannulation strategy
syndrome”)
upper body perfusion is from
native cardiac output.
Head/face cyanosis with
adequate
truncal/extremity
perfusion on VA-
ECMO

Problem Causes Solutions

Inadequate Cannula impeding circulation  Insert distal perfusion cannula
to lower extremity Measure flow to lower extremity
perfusion of 
cannulated lower  Monitor lower extremity perfusion/pulses, color,
temperature. Consider using near infrared
extremity
spectroscopy (NIRS) monitoring on lower
extremity
 Consult surgeons for compartment syndrome

Problem Causes Solutions

Inadequate Inadequate dose – due to  Increase dose
tachyphylaxis or increased Consider alternative agents or alternative sedative
sedation/Agitation Volume/Distribution of

strategies
ECMO circuit
 Encourage family/spouse participation
 Provide distraction opportunities such as music,
child life, etc.
Medication sequestration in  Increase dose
ECMO circuit/oxygenator  Administer a bolus with circuit changes
 Consider alternative sedative strategies
Delirium  Promote day/night sleep cycles
 Consider alternative sedative strategies
 Daily sedation “holiday”
 Mobilization
 Utilize delirium assessment tools
 Consider antipsychotics if needed
Boredom/inactivity  Encourage family /friend participation
 Provide distraction opportunities such as music,
child life, etc.
 Mobilization
 Pet therapy
 Patient preferred hobbies

282
 Patient Troubleshooting

Problem Causes Solutions

Increasing Patient Decreased native cardiac output  Increase ECMO flow
with same ECMO flow Evaluate causes
PO2 
Hypovolemia Administer volume
With cyanosis, acidosis,  Pneumo/hemothorax/  Obtain chest radiograph, ultrasound
poor perfusion, effusion  Evacuate pneumo/ hemothorax/effusion
increased lactic acid  Pneumo/hemopericardium
Cardiac stun or inadequate VA Increase ECMO flow
flow
Cardiac tamponade  Drain pericardium
 Aggressively strip chest tube in the postoperative
cardiac patient
Sepsis with peripheral shunting  Sepsis evaluation and treatment
 Vasopressors as indicated

Problem Causes Solutions

Increasing Patient Improving respiratory function Consider ECMO weaning
PO2 Altered ventilator or sweep gas Adjust ventilator or sweep gas FiO2
FiO2
With patient well-
Cardiac stun on adequate VA Continue full ECMO flow
appearing
flow

Problem Causes Solutions

Increasing Patient  Sweep gas leak/  Attach/troubleshoot sweep gas
PCO2 disconnected  Restore gas supply
 Gas source empty or  Check post oxygenator blood gas
With tachypnea, disconnected  Sigh/burp oxygenator PRN –especially at low
acidosis, agitation,  Oxygenator failure or loss sweep flows
hypertension of efficiency/increased  Evaluate transmembrane pressure
clot burden or condensate  Replace oxygenator if indicated
on oxygenator fibers
Sweep gas CO2/Carbogen too Decrease or remove CO2/Carbogen in sweep gas
high
Sweep gas flow too low Increase sweep gas flow
Patient agitation  Calm or sedate patient
 Provide distractions, child life therapy, PT/OT
 Encourage family participation
 Evaluate for delirium
Underventilation Increase ventilation
ETT problem Fix ETT problem
Pneumothorax/hemothorax/  Evacuate pneumothorax/ hemothorax/effusion
effusion  Decrease ventilator settings if adequate ECMO
support to allow healing of air leak
Atelectasis, worsening chest  Sepsis evaluation and treatment
radiograph  Pneumonia evaluation and treatment
 Bronchoscopy for secretion management and
infection diagnosis
Hyperthermia  Assure appropriate temperature management.

283
Chapter 33

Problem Causes Solutions

Intracranial Altered cerebral blood flow,  Midline head positioning
prematurity, hypoxia, Elevate head of bed 30%
Hemorrhage hypotension, reperfusion injury

 Optimize anticoagulation
 Consider Antifibrinolytics
 Avoid hypertension/hypotension
 Utilize VV ECMO when possible
 Consider using cephalad jugular venous
drainage cannulation
 Consider decannulation

Problem Causes Solutions

Low HCT Bleeding  See “BLEEDING” section above
 Look for bleeding source
 Lower HCT goal
 Transfusion when indicated
Laboratory sampling Minimize laboratory sampling once stable

Problem Causes Solutions

Low SvO2 Increased O2 consumption or  Increase ECMO flow
decreased O2 delivery  Evaluate reasons for increased O2
consumption: fever, shivering, nutritional
status
SvO2 monitor malfunction  Assess/correct SvO2 monitor
 Calibrate SvO2 monitor
Inaccurate SvO2 measurement Evaluate sources of inaccurate SvO2 measurements,
due to interference such as methemoglobinemia or methylene blue

Problem Causes Solutions

Patient Cold Patient treated with therapeutic Evaluate re-warming plan
hypothermia
Heater unit malfunction Replace water heater if indicated
Water heater not plugged in or Plug in, turn on water heater and adjust temperature
turned on
Temperature set point too low Adjust set temperature on water bath or overhead
heater
Temperature LED malfunction Check and reset temperature set point - turn heater
after power interruption - reading off/on, then reset temperature set point
characters rather than temperature
Water shut off valves turned off Open water shut off valves
Water hoses not connected to Verify water hoses are connected, unobstructed and
oxygenator or kinked, occluding shut off valves are open
water flow
Assisted heating device (radiant Replace if needed
warmer) malfunction
Heater unit set in FLUID mode Check appropriate mode set on heater
without inline temperature probe
Large amount of cold water Consider isolating patient from heater until water
added to water heater reservoir warms
Water reservoir empty Fill with water as directed by manufacturer

284
 Patient Troubleshooting

Problem Causes Solutions

Patient Hot Heater unit malfunction Replace water heater if indicated
Radiant warmer or patient  Change patient temperature probe
temperature probe malfunction Adjust set temperature on overhead warmers
Inaccurate patient temperature  Evaluate and standardize temperature
measurement assessment site
Troubleshoot thermometer/ temperature probe
placement
Water temperature set too hot Adjust set temperature on water heater
Radiant warmer  Troubleshoot water heater settings, hoses,
overcompensating due to: valves, water wheel turning, etc.
o water heater set point  Replace water heater
too low
o water heater turned off
 Evaluate skin flushing/core temperature
o water shut off valves relationship
turned off or hoses
kinked
Water heater malfunction
Consider patient may have  Sepsis evaluation, antipyretic medications
fever

Problem Causes Solutions

Pulmonary Edema VA ECMO: Blood returning to  ECHO to confirm non-contracting left
left heart with poor/no left ventricle
ventricular ejection  Increase pump flow
 Atrial septostomy
 Left ventricular vent
VA or VV ECMO: Massive  Diuresis
fluid resuscitation with  CVVH
capillary leak and decreased
 Increased PEEP
renal function

Problem Causes Solutions

Pulmonary Suction trauma  Measured depth suctioning
Hemorrhage  Suction only PRN
 Consider reducing anticoagulation
Pulmonary over-circulation  Increase PEEP
 Consider echocardiogram
 Maximize VA-ECMO support
 Consider VV conversion to VA
Spontaneous/unknown cause  Increase PEEP
 Suction PRN
 Bronchoscopy to identify source and airway
clearance
 Broaden diagnostic considerations
 See “BLEEDING” section

285
Chapter 33

Problem Causes Solutions

Seizures Ischemic brain injury Anticonvulsants
 Cerebral edema  Utilize VV ECMO when possible
 Infarction/stroke  Treat as indicated based on reason for ECMO,
 Intracranial hemorrhage time course of ECMO, and underlying cause
of seizure
 Obtain EEG, head ultrasound, CT scan as
indicated for diagnosis

Problem Causes Solutions

Stroke  Inadequate circuit  Prevention:
anticoagulation o Optimize circuit anticoagulation
 Intermittent low flow o Monitor and intervene for arterial line
states clots on VA-ECMO
 Circuit embolism  Optimize ECMO blood flow
Any cause  Identify signs and symptoms quickly
 Minimize anticoagulation
 Close monitoring of neurologic status

Problem Causes Solutions

Thrombocytopenia Exposure to extracorporeal Platelet transfusion
circuit
Circuit DIC Circuit change
Sepsis Sepsis evaluation and treatment
Heparin Induced  Test for HIT
Thrombocytopenia (HIT)  Utilize direct thrombin inhibitor instead of
heparin
 Utilize non-heparin bonded circuit
 Remove heparin from all infusions

286
 Patient Troubleshooting

References poreal Life Support: The ELSO Red Book. 5th

ed. Ann Arbor, MI: Extracorporeal Life Support
1. Bhombal S, Sheehan AM, Van Meurs KP. Medi- Organization; 2017:49-80.
cal management of the neonate with respiratory 11. Kirk C, Abel EE, Muir J, Dzierba AL. Strate-
failure on ECLS. In Brogan TV, Lequier L, gies for medication management. In Brogan TV,
Lorusso R, MacLaren G, Peek G, eds. Extracor- Lequier L, Lorusso R, MacLaren G, Peek G, eds.
poreal Life Support: The ELSO Red Book. 5th Extracorporeal Life Support: The ELSO Red
ed. Ann Arbor, MI: Extracorporeal Life Support Book. 5th ed. Ann Arbor, MI: Extracorporeal
Organization; 2017:183-199. Life Support Organization; 2017:795-808.
2. Bridges BC, Ranucci M, Lequier LL. Anti- 12. Dalton H, Reeder R, Garcia-Filion P, et al. Fac-
coagulation and disorders of hemostasis. In tors associated with bleeding and thrombosis in
Brogan TV, Lequier L, Lorusso R, MacLaren children receiving extracorporeal membrane
G, Peek G, eds. Extracorporeal Life Support: oxygenation (ECMO). Am J Respir Crit Care
The ELSO Red Book. 5th ed. Ann Arbor, MI: Med. 2017;196(6):762-782.
Extracorporeal Life Support Organization;
2017:93-103.
3. Jenks CL, Tweed J, Gigli KH, Venkataraman R,
Raman L. An international survey on ventilator
practices among extracorporeal membrane oxy-
genation centers. ASAIO J. 2017;63(6)787-792.
4. Mosier JM. Extracorporeal membrane oxy-
genation (ECMO) for critically ill adults in the
emergency department: history, current applica-
tions, and future directions. Crit Care. 2015;19:
431.
5. Prine KB, Goracke K, Rubarth LB. Extracor-
poreal membrane oxygenation in the NICU.
Neonatal Netw 2015; 34(3):183-188.
6. Roberts J, Keene S, Heard M, McCracken
C, Gauthier TW. Successful primary use of
VVDL+V ECMO with cephalic drain in neona-
tal respiratory failure. J Perinat. 2016;36(2):126-
131.
7. Sell LL, Cullen ML, Lerner GR, Whittlesey
GG, Shanley CJ, Klein MD. Hypertension
during extracorporeal membrane oxygen-
ation: cause effect, and management. Surgery.
1987;102(4):724-730.
8. Short BL. Extracorporeal membrane oxygen-
ation. In: Avery G, Mhairi G, Macdonald M,
Seshias M, Mullett M, eds. Avery’s Neonatol-
ogy 6th ed. Philadelphia, PA: Lippincott Wil-
liams & Wilkins; 2005:622-33.
9. Sidebotham D. Troubleshooting adult ECMO.
J Extra Corpor Technology. 2011; 43(1):27-37.
10. Toomasian JM, Vercaemst L, Bottrell S, Horton
SB. The circuit. In Brogan TV, Lequier L, Lo-
russo R, MacLaren G, Peek G, eds. Extracor-

287
34

Case Simulations

Rodrigo Diaz, MD, Rocio Agliati, RN, Peter Chi Keung Lai, RN, Kollengode Ramanathan, MD, Archana
Dhar, MD, Mark T Ogino, MD, Bishoy Zakhary, MD

Introduction Simulation-based education and training have

proven effective in multiple areas: improving the
Simulation has been defined as the creation of acquisition of medical knowledge, enabling commu-
something that is not real, but is authentic, allowing nication and teamwork, facilitating the development
the student to be exposed to active learning with of technical skills, reducing procedural stress, and
repetition, feedback, and reflection.1 Gaba further improving clinical outcomes.6 For optimal success,
adds that simulation replaces the real experience crisis resource management principles suggest that
with a controlled or guided situation, providing the resuscitation teams function best when a team leader
capacity to replicate the fundamental clinical aspects is identified, individual roles are assigned, and com-
in an interactive manner.2 In medical education, munication is prioritized.7
simulation teaches the anticipation of planning for Simulation-based ECMO education has been
future events, and to identify potential problem areas shown to increase knowledge and confidence in
in the clinical setting. addressing ECMO emergencies.8,9 Moreover, it is
Simulation-based medical education follows a more effective than traditional water-drill-based
series of consecutive stages. The brief first stages fo- education, the benefit is maintained over the long-
cus on creating a psychologically safe environment term, and the skills transfer to novel scenarios not
for learners while the instructor clarifies expecta- encountered during initial training.10
tions, objectives, roles, and limitations. A fictional ELSO has published educational guidelines for
contract, based on a commitment by the participant ECMO practitioners.11 To develop and implement
to act as though the experience were real, is also simulation-based ECMO education, clinical sce-
established in this phase.3 The second phase is the narios with well-defined learning objectives need
scenario itself, in which the programmed clinical to be designed. The following chapter provides ex-
case is executed. The third phase is the essential amples of ELSO Comprehensive Simulation-Based
element of debriefing, consisting of an interactive ECMO Course scenarios. ECMO centers looking to
process where a formative evaluation is carried out.4 expand their educational offerings are invited to use
Clinical simulation creates an ideal environment these examples to run their own simulations and as
for learning how to recognize, understand, and be templates for further scenario development.
immersed in all aspects of ECMO care and manage-
ment. In particular, high-risk low-frequency events, Simulation Scenario Format
such as life-threatening ECMO emergencies, can be
practiced so that participants learn both how to man- The simulation scenarios are divided into sev-
age emergencies and to anticipate issues before they eral subsections.
occur. Scenarios can be designed to be predictable,
consistent, standardized, and safe thereby provid- Simulation Scenario Overview
ing the ideal setting for ECMO education. Effective
learning is ensured when training is consolidated This section lists the learning objectives that the
with appropriate debriefing techniques.5 scenario attempts to achieve. It is important to have

289
Chapter 34

scenario objectives well-defined before creating and the standard response under Change in Case. The
running a simulation. The roles required to run the listed actions include those that are desirable as well
simulation, divided into participants and confeder- as those that are not. Standardizing the response
ates, are also listed. ensures a consistent experience for participants.
While an attempt is made to capture as many likely
Simulation Scenario Critical Actions interventions as possible, simulation instructors
must also be prepared to respond to unanticipated
This section expands on the scenario objectives interventions.
and lists the important participant actions for sce-
nario resolution. The actions are divided into three Simulation Scenario Investigations
categories: cognitive, technical, and behavioral.
This section is to be referenced while running the This section lists the laboratory testing and im-
simulation and can subsequently guide scenario de- aging results to be used in the scenario if requested
briefing. Additionally, scenario stop points, either a by participants. Again, having standard laboratory
desired final action or a time limitation, are defined. values and imaging results allows for a consistent
experience for participants.
Simulation Scenario Setup

This section outlines the components of the

simulation scenario and defines the initial state of
scenario setup. The importance of this section is
that is allows for a consistent teaching experience
for each participant group independent of instructor
team and equipment.
Brief Case Background: Standardized introduc-
tion to scenario and simulated patient.
Mannequin Set-up: Mannequin preparation with
listing of associated equipment.
Initial Vitals: Mannequin vital signs at scenario
start.
Physical Exam: Mannequin physical exam find-
ings at scenario start.
Medications: Active medications at scenario
start.
Ventilator Settings: Mechanical ventilator set-
tings at scenario start.
ECMO Settings: ECMO circuit settings and
pressures at scenario start. Note is also made of the
presence or absence of a color differential between
the access and the return cannulae. Since this is
difficult to simulate, this information is offered if
participants inquire.

Scenario Flow

This section lists anticipated participant inter-

ventions and the standard response to these actions.
The table lists the intervention under Event and

290
 Case Simulations

References

1. Decker S, Sportsman S, Puetz L, Billings L. The

evolution of simulation and its contribution to
competency. J Contin Educ Nurs. 2008; 39(2):
74-80.
2. Gaba DM. The future vision of simulation in
health care. QualSaf Health Care. 2004:13
(suppl 1): i2-i10.
3. Rudolph JW, Raemer DB, & Simon R. Establish-
ing a safe container for learning in simulation:
the role of the presimulation briefing. Simul
Healthc. 2014;9(6):339-349.
4. Rudolph JW, Simon R, Raemer DB, Eppich WJ.
Debriefing as formative assessment: closing
performance gaps in medical education. Acad
Emerg Med. 2008;15(11):1010-1016.
5. Cheng A, Eppich W, Grant V, Sherbino J, Ze-
ndejas B, Cook DA. Debriefing for technology
enhanced simulation: a systematic review and
meta-analysis. Med Educ. 2014; 48(7): 657–666.
6. Okuda Y, Bryson EO, DeMaria S Jr, et al. The
utility of simulation in medical education:
what is the evidence? Mt Sinai J Med. 2009;76
(4):330-343.
7. Sancho R, Maestre JM, Del Moral I. Manejo de
las crisis. Papel de la simulación en la seguridad
del paciente. Rev Esp Anestesiol Reanim. 2014:
59(Supl 2):S53-S59.
8. Alinier G, Hassan IF, Alsalemi A et al. Addressing
the challenges of ECMO simulation. Perfusion.
2018 1:267659118777194.
9. Brum R, Rajani R, Gelandt E, et al. Simulation
training for extracorporeal membrane oxygen-
ation. Ann Card Anaesth. 2015; 18(2):185-190.
10. Zakhary BM, Kam LM, Kaufman B S, Felner
KJ. The utility of high-fidelity simulation for
training critical care fellows in the manage-
ment of extracorporeal membrane oxygenation
emergencies: a randomized controlled trial. Crit
Care Med 2017:45(8): 1367-1373.
11. ELSO Guidelines for Training and Continuing
Education of ECMO Specialists. Available at:
http://www.elso.org/Portals/0/IGD/Archive/
FileManager/97000963d6cusersshyerdocum-
entselsoguidelinesfortrainingandcontinuinged-
ucationofecmospecialists.pdf.

291
Chapter 34

Left Ventricular Distention on Peripheral VA ECMO

Simulation Scenario Overview
Objectives
Understand and communicate the risks of a non-pulsatile arterial tracing on peripheral VA ECMO
Manage left ventricle distension on VA ECMO

Roles
Participant 1: ECMO specialist
Participant 2: Physician trainee
Confederate: Critical care nurse

Simulation Scenario Critical Actions

Cognitive
 Identify inappropriately elevated systemic arterial pressure
 Understand the risks of left ventricular distension
 Associate increased afterload and lower pulse pressure with risk for left ventricular distension

Technical
 Request echocardiography
 Reduce vasopressors
 Attempt decrease in circuit blood flow

Behavioral
 Inform team members regarding left ventricular distention concerns
 Closed loop communication of management priorities

Stop Points
 Request mechanical device or surgical procedure for left ventricular venting
 5 minutes have elapsed

Simulation Scenario Setup

Brief Case Background
A 55-year-old male patient was admitted with cardiogenic shock after anterior wall myocardial infarction. Following
catheterization and stent placement, he was cannulated for VA ECMO (right femoral venous access and left femoral
arterial return).
Mannequin Set-Up
Endotracheal tube with pink foam
Mechanical ventilator
ECMO pump and circuit
Right femoral vein access cannula
Left femoral artery return cannula
Right internal jugular pulmonary artery catheter
Left internal jugular triple lumen central venous catheter
Right radial arterial line
Urinary catheter

292
 Case Simulations

Initial Vitals

Temperature Heart Rate Blood Pressure Respiratory Rate Pulse Oximetry

37.0 ˚C 80/min 80/73 mmHg 14/min 100%

CVP PA Pressure Wedge Pressure SvO2

10 mmHg 20/18 mmHg 20 mmHg 70%

Initial Physical Exam

General Intubated on mechanical ventilation and ECMO

Neurologic Sedated, pupils 3mm bilaterally, reactive to light
Cardiovascular Regular rhythm, no murmurs, capillary refill 3 secs, pulseless
Pulmonary Symmetrically diminished breath sounds with basilar rales
Extremities Warm and well perfused

Medications
Norepinephrine 0.1 mcg/kg/min
Dobutamine 5 mcg/kg/min
Midazolam 0.1 mcg/kg/hr
Fentanyl 50 mcg /hr
Heparin 1000 units/hr

Ventilator Settings
PC PEEP Respiratory Rate FiO2
Pressure Control 14 cmH2O 10 cmH2O 10 bpm 0.4

ECMO Settings
Blood flow 3000 4.0 L/min Pre-pump Pressure -40 mmHg
RPM
Gas flow 0.7 FO2 4.0 L/min Pre-oxygenator Pressure 210 mmHg
Color differential Yes Post-oxygenator Pressure 190 mmHg

Simulation Scenario Flow

Scenario Branch Points

Event Change in Case

No intervention for 1 min Increase in PA Pressure to 33/25 and in Wedge Pressure to 26 over 3
minutes. Increase in endotracheal tube pink foam.
Reduction in ECMO flow Reduction in BP by 10 mmHg in both systolic and diastolic with
reduction in SvO2 by 10%. Persistent pulmonary edema.
Increase in ECMO flow Increase in BP by 10 mmHg in both systolic and diastolic with increase in
SvO2 by 5%. Persistent pulmonary edema.
Reduction in vasopressors Reduction in BP by 5 mmHg in both systolic and diastolic but persistent
pulmonary edema.
Increase in inotropes Increase in pulse pressure by 3 mmHg but persistent pulmonary edema.
Request for mechanical or Scenario stop
surgical LV vent

293
Chapter 34

Simulation Scenario Investigations

Laboratory Testing
Complete Blood Count
WBC 14 x103/mL
Hb 12 g/dL
Plt 240 x103/mL

Basic Metabolic Panel

Na+ 135 mmol/L
K+ 4.4 mmol/L
Cl- 110 mmol/L
HCO3- 20 mmol/L
BUN 14 mmol/L
Cr 1.2 mg/dL
Glucose 100 mg/dL

Coagulation
aPTT 60 s
INR 1.3
Anti-Xa 0.2

Pre-Oxygenator Gas
pH 7.38
PaCO2 45 mmHg
PaO2 50 mmHg

Post-Oxygenator Gas
pH 7.4
PaCO2 38 mmHg
PaO2 300 mmHg

Patient Gas
pH 7.40
PaCO2 40 mmHg
PaO2 280 mmHg

Imaging
 Chest X-ray: Pulmonary edema
 Echocardiography: Dilated left ventricle, Closed aortic valve

294
 Case Simulations

Elevated Post-Oxygenator Pressure on VV ECMO

Simulation Scenario Overview
Objectives
Recognize and treat elevated post-oxygenator pressure as a result of a kinked circuit
List possible causes of elevated post-oxygenator pressure

Roles
Participant 1: ECMO specialist
Participant 2: Physician trainee
Confederate: Critical care nurse

Simulation Scenario Critical Actions

Cognitive
 Relate inadequacy in systematic oxygenation with drop in ECMO blood flow
 Recognize a lack of improvement in blood flow with increases in ECMO pump speed
 Identify elevation in post-oxygenator pressure as a result of a problem downstream of the oxygenator
 Differentiate between diagnoses that can cause a rise in post-oxygenator pressure

Technical
 Attempt to increase ECMO pump speed to maintain blood flow
 Perform circuit check
 Release kinking in return tubing

Behavioral
 Assess patient vital signs
 Closed loop communication of management priorities

Stop Points
 Identify and release kinking in return tubing
 5 minutes have elapsed

Simulation Scenario Setup

Brief Case Background
A 30-year-old male presents with H1N1 pneumonia and the acute respiratory distress syndrome requiring
VV ECMO. Today is Day 2 of VVECMO. He remains intubated and sedated. The nurse just finished
patient repositioning. You come to assess the patient after morning sign-out.

Mannequin Set-Up
Endotracheal tube
Mechanical ventilator
ECMO pump and circuit
Right femoral venous access cannula (23 Fr, marking 49)
Right internal jugular return cannula (17 Fr, marking 23)
Left internal jugular triple lumen central venous catheter
Right radial arterial line
Urinary catheter

295
Chapter 34

Initial Vitals
Temperature Heart Rate Blood Pressure Respiratory Rate Pulse Oximetry
37.0 ˚C 100/min 100/60 mmHg 10/min 90%

Initial Physical Exam

General Intubated
Neurologic Sedated (RASS –3)
Cardiovascular Stable on vasopressor
Pulmonary Reduced breath sounds
Extremities Cool

Medications
Fentanyl 200 mcg/hr
Dexmedetomidine 0.6mcg/kg/hr
Norepinephrine 0.18mcg/kg/min
Heparin 1000 units/hr

Ventilator Settings
PC PEEP Respiratory Rate FiO2
Pressure Control 10 cmH2O 12 cmH2O 10/min 0.4

ECMO Settings
Blood flow 3000 4.0 L/min Pre-pump Pressure -40 mmHg
RPM
Gas flow 1.0 FO2 4.0 L/min Pre-oxygenator Pressure 130 mmHg
Color differential Yes Post-oxygenator Pressure 110 mmHg

Simulation Scenario Flow

Scenario Branch Points

Event Change in Case

On case start, confederate turns patient and Immediately, ECMO blood blow decreases to 1.0 L/min.
secures ECMO tubing with forceps Pre- and post- oxygenator pressures increase to 240 mmHg
(kinking return tubing) and 220 mmHg, respectively.

Over 1 min, heart rate increases to 135/min and blood

pressure to 140/90mmHg while pulse oximetry drops to
65%.
If kink in circuit not released Over 4 mins, heart rate decreases to 40/min and BP to
60/30 mmHg

296
 Case Simulations

Simulation Scenario Investigations

Laboratory Testing

Complete Blood Count

WBC 14 x103/mL
Hb 9.2 g/dL
Plt 240 x103/mL

Basic Metabolic Panel

Na+ 131 mmol/L
K+ 4.4 mmol/L
Cl- 110 mmol/L
HCO3- 16 mmol/L
BUN 14 mmol/L
Cr 1.4 mg/dL
Glucose 96 mg/dL

Coagulation
aPTT 61 s
INR 1.3
Anti-Xa 0.35

Pre-Oxygenator Gas
pH 7.33
PaCO2 50 mmHg
PaO2 45 mmHg

Post-Oxygenator Gas
pH 7.40
PaCO2 30 mmHg
PaO2 480 mmHg

Patient Gas
pH 7.38
PaCO2 35 mmHg
PaO2 65 mmHg

Imaging
 Chest X-Ray: Diffuse bilateral infiltrates. No pneumothorax. Normal heart size.
 Echocardiography: Normal left and right ventricles. Normal valves.

297
Chapter 34

Intra-hospital Transport of Neonate on VA ECMO

Simulation Scenario Overview

Objectives
Understand the importance of and implement a standard pre-ECMO transport checklist
Delegate tasks effectively during the transport process
Comprehend the importance of and implement a standard post-ECMO transport checklist

Roles
Participant 1: ECMO specialist
Participant 2: Physician trainee
Confederate: Critical care nurse

Simulation Scenario Critical Actions

Cognitive
 Review decision for transport to CT scanner
 Weigh the risks and benefits of holding heparin with low circuit blood flows
 Consider interventions to reduce risk of clot formation while off heparin (such as adding a partial
bridge to allow for an increase in circuit blood flow)

Technical
 Utilize standard pre- and post- ECMO transport checklists

Behavioral
 Delegate roles prior to ECMO transport
 Review transport route and identify potential difficulties

Stop Point
 Patient wheeled back to ICU and post-ECMO transport checklist completed
 5 minutes have elapsed

Simulation Scenario Setup

Brief Case Background

A 3.0-kg infant born by normal vaginal delivery developed meconium aspiration syndrome with
respiratory distress necessitating transfer to the intensive care unit, intubation, and mechanical ventilation.
CXR post intubation showed bilateral patchy infiltrates. The baby continued to decompensate on
sedation, cisatracurium, epinephrine, and dopamine. A cranial ultrasound (CUS) revealed no
abnormality. Coagulation panel was normal. The infant was initiated on VA ECMO for persistent
hypoxemia [PaO2 48 mm Hg]. Subsequent blood gas was satisfactory [pH 7.33, PaCO2 36, PaO2 66,
lactate 3] and paralytic agent was stopped. An hour later, the baby developed seizures and bedside CUS
showed a hyperechoic region in the frontal brain suspicious for intracranial hemorrhage. Heparin has
been switched off and a CT brain has been ordered.

298
 Case Simulations

Mannequin Set-Up
Endotracheal tube
Mechanical ventilator
ECMO pump and circuit
Right internal jugular access cannula
Right common carotid return cannula
Left femoral triple lumen central venous catheter
Right radial arterial line
Urinary catheter

Initial Vitals
Temperature Heart Rate Blood Pressure Respiratory Rate Pulse Oximetry
37 ˚C 124/min 67/38 mmHg 10/min 93%

Initial Physical Exam

General Weight, Length: 3 kg, 40 cm
Neurologic Sedated, Intubated & Ventilated
Cardiovascular S1, Pan systolic murmur, S2 loud
Pulmonary Bronchial breath sounds bilaterally, occasional wheeze
Extremities Warm well perfused

Medications
Morphine 20 mcg/kg/hr
Midazolam 1 mcg/kg/min
Heparin 15 U/kg/hour
Epinephrine 0.01 mcg/kg/min
Dopamine 5 mcg/kg/min.

Ventilator Settings
PC PEEP Respiratory Rate FiO2
Pressure Control 10 cmH2O 10 cmH2O 10/min 0.3

ECMO Settings
Blood flow 500 RPM 300 mL/min Pre-pump Pressure -30 mmHg
Gas flow 0.3 FO2 300 mL/min Pre-oxygenator Pressure 110 mmHg
Color differential Yes Post-oxygenator Pressure 100 mmHg

Simulation Scenario Flow

Scenario Branch Points

Event Change in Case

No intervention for 1min No hemodynamic changes
ECMO specialist prompts to start transport to CT

299
Chapter 34

Simulation Scenario Investigations

Laboratory Testing

Complete Blood Count

WBC 14 x103/mL
Hb 9.2 g/dL
Plt 240 x103/mL

Basic Metabolic Panel

Na+ 131 mmol/L
K+ 4.4 mmol/L
Cl- 110 mmol/L
HCO3- 16 mmol/L
BUN 14 mmol/L
Cr 1.4 mg/dL
Glucose 96 mg/dL

Coagulation
aPTT 52 s
INR 1.3
Anti-Xa 0.35

Pre-Oxygenator Gas
pH 7.3
PaCO2 55 mmHg
PaO2 44 mmHg

Post-Oxygenator Gas
pH 7.34
PaCO2 42 mmHg
PaO2 344 mmHg

Patient Gas
pH 7.32
PaCO2 46 mmHg
PaO2 63 mmHg

Imaging
 Chest X-ray: Bilateral patchy infiltrates.
 Electrocardiogram: Sinus tachycardia.
 Echocardiography: Poor biventricular function. Elevated pulmonary artery pressures with right to left
shunting at the ductal level and at the foramen ovale.

300
 Case Simulations

Recirculation in Pediatric Patient on VV ECMO

Simulation Scenario Overview
Objectives
To identify and manage recirculation on VV-ECMO

Roles
Participant 1: ECMO specialist
Participant 2: Physician trainee
Confederate: Critical care nurse

Simulation Scenario Critical Actions

Cognitive
 Identify recirculation as the cause for hypoxemia
 Describe the impact of volume, cannula position, cardiac output, and pump flow on
recirculation
 Understand the limitations of SVO2 in VV ECMO

Technical
 Perform circuit check
 Reduce pump flow
 Recognize need for cannula position check

Behavioral
 Share concern for recirculation as the cause for hypoxemia with team members
 Closed loop communication of management priorities

Stop Point
 Order echocardiography to confirm cannula displacement
 5 minutes have elapsed

Simulation Scenario Setup

Brief Case Background
A 17-year-old 60 kg male presented with pneumonia and ARDS. He was placed on VV-ECMO after
failure of high frequency oscillatory ventilation (HFOV) and inhaled nitric oxide (iNO). Cannulation is a
23 Fr right femoral drainage cannula and a 21 Fr right internal jugular return cannula. You are called to
evaluate the patient due to progressive hypoxemia.

Mannequin Set-Up
Endotracheal tube
Mechanical ventilator
ECMO pump and circuit
Right internal jugular 21 Fr
Right femoral 23 Fr
Left femoral triple lumen central venous catheter
Right radial arterial line
Urinary catheter

301
Chapter 34

Initial Vitals
Temperature Heart Rate Blood Pressure Respiratory Rate Pulse Oximetry
37.0 ˚C 100/min 110/65 mmHg 18/min 75%

SVO2
72%

Initial Physical Exam

General Intubated on mechanical ventilation
Neurologic Sedated, pupils are 3mm bilaterally and reactive
Cardiovascular No murmur, capillary refill 3 secs, pulses 2+
Pulmonary No breath sounds auscultated
Extremities Warm and well perfused

Medications
Fentanyl 100 mcg/kg/hr
Epinephrine 0.05 mcg/kg/min

Ventilator Settings
PC PEEP Respiratory Rate FiO2
Pressure Control 10 cmH2O 10 cmH2O 10/min 0.3

ECMO Settings
Blood flow 3200 4.8 L/min Pre-pump Pressure -45 mmHg
RPM
Gas flow 1.0 FO2 2.4 L/min Pre-oxygenator Pressure 160 mmHg
Color differential Yes Post-oxygenator Pressure 150 mmHg

Simulation Scenario Flow

Scenario Branch Points

Event Change in Case

No intervention for a minute HR: 110, BP: 90/58, RR:16, SpO2: 65 SVO2: 80. Pronounced duskiness
Initiate emergency ventilator HR: 120, BP: 100/60, RR: 18, SpO2: 75, SVO2: 75. Patient cyanosis
settings
Increase the pump flow HR: 110, BP: 110/60, RR:18, SpO2: 70, SVO2: 90. More pronounced
duskiness
Administer fluid bolus HR: 100, BP: 110/65, RR:16, SpO2: 80, SVO2: 70

302
 Case Simulations

Simulation Scenario Investigations

Laboratory Testing

Complete Blood Count

WBC 14 x103/mL
Hb 12 g/dL
Plt 240 x103/mL

Basic Metabolic Panel

Na+ 135 mmol/L
K+ 4.4 mmol/L
Cl- 110 mmol/L
HCO3- 20 mmol/L
BUN 14 mmol/L
Cr 1.2 mg/dL
Glucose 100 mg/dL

Coagulation
aPTT 80 s
INR 1.3
Anti-Xa 0.4

Pre-Oxygenator Gas
pH 7.38
PaCO2 45 mmHg
PaO2 280 mmHg

Post-Oxygenator Gas
pH 7.4
PaCO2 38 mmHg
PaO2 300 mmHg

Patient Gas
pH 7.2
PaCO2 55 mmHg
PaO2 50 mmHg

Imaging
 Chest X-ray: Advancement of right IJ cannula deep into the RA.
 Echocardiography: The right IJ cannula is now at the RA-IVC junction with approximation of
drainage and return cannulae.

303
35

Configuration Nomenclature for Extracorporeal Membrane Oxygenation

L. Mikael Broman, MD, PhD, Steven A. Conrad, MD, PhD, Fabio S. Taccone, MD, PhD

Introduction Configuration

Uniform nomenclature and clear abbreviations ECMO configuration describes the number of
for complex systems minimize risk for confusion cannulae and their respective insertion sites and
and misinterpretation. In the medical field, this is placement. The following classification for con-
necessary for routine and urgent communication figuration is based on four levels of complexity,
as well as for research purpose. This concept is with increased precision of ECMO configuration
exemplified by the IUPAC terminology for organic description per level used.
chemistry,1 the TNM classification for malignant
tumors,2 and nomenclature to describe human muta- Flow Direction
tions.3 The same complexity arises in summarizing
the various configurations used today for extracor- The most basic description of cannulae con-
poreal life support (ECLS).4 figuration denotes the direction of extracorporeal
In this chapter, we provide a pragmatic descrip- blood flow described in a left to right convention.
tion focusing on nomenclature and abbreviations The drainage side is always on the left and the return
for peripheral cannulation in ECMO patients. The of oxygenated blood to the patient is on the right.
description is based on the consensus statement for Since the description concerns ECMO, a membrane
ECLS, The Extracorporeal Life Support Organiza- lung (ML, i.e. artificial lung, oxygenator) is always
tion Maastricht Treaty for Nomenclature in Extra- included in the to provide gas exchange. The ML,
corporeal Life Support,5 as well a second part, which placed between the drainage and return cannulae,
is in submission, The Extracorporeal Life Support is expressed with a hyphen (-). As an example, the
Organization Maastricht Treaty for Abbreviations simplest expression for a VV or VA mode configu-
for Cannulation Configuration in Extracorporeal ration with two single lumen cannulae would be
Life Support. V-V or V-A.

The ELSO Classification for Peripheral Cannula- Level One: Cannula Hierarchy
tion Configurations
Any cannula contributing to the major flow (i.e.
Modes drainage or return), is expressed with an upper case
letter including (V) for cannula placed in a vein,
The ECMO modes include venovenous (VV), (A) for in an artery and, more uncommonly, (P) for
venoarterial (VA) and the hybrid mode venoveno- placement in the pulmonary artery (i.e. V-P ECMO).
arterial (VVA). Nevertheless, mode denotes only As previously mentioned, the hyphen represents
the type of ECMO support but relates no details on the ML separating the drainage and return cannulae.
where, how or how many cannulae that have been In certain cases, (i.e. when the rated flow of the ML
employed. is exceeded4-5) dual oxygenators may be used. The
nomenclature indicates two lungs in parallel with an

305
Chapter 35

equal sign (“=”), and when in series with a plus sign letter. The letter describing the minor flow catheter
(“+”). Cannulation for VV ECMO using a double- is positioned after the respective major flow cannula
lumen cannula (DLC) is indicated with a “dl” ([dl] on the side to which it belongs. Such a catheter may
V-V). For greater precision “dl” may be changed to be a placed proximally (i.e., groin) denoted by “a”,
“ca” for a cavo-atrial DLC, or “bc” for a bi-caval or distally (i.e., in the posterior tibial or dorsal pedal
DLC. A recent design DLC for right ventricular artery) and is denoted by “d” for distal perfusion
support, draining from the right atrium (RA) of the cannulated leg. The catheter may provide
and returning flow to a hooked tip placed in the cephalad drainage cranial to the venotomy of a
mean pulmonary artery is abbreviated (dl)V-P. cannulated internal jugular vein (“c”), or venting
of a cardiac atrium or ventricle (“v”). Accordingly,
When an additional cannula is used for im-
V-Aa is V-A with a cannula for distal perfusion in
proved venous drainage (V) or return arterial flow,
the arterially cannulated leg, while Vv-A is a V-A
the letter denoting the cannula is placed after the side
with cardiac venting catheter.
(venous or arterial) of the catheter. Adding a drain-
Figure 35-1 illustrates the nomenclature while
age cannula in a V-A configuration becomes VV-A.
Table 35-1 displays several of cannulation strategies.
If the additional cannula returns blood to the venous
circulation the abbreviation would read V-AV. Ac-
Level Two: Cannulation Site
cordingly, starting from V-V ECMO, adding a return
cannula into an artery is V-VA, etc. In verbal com-
To provide information on cannulation site, an
munication, the possibility of confusion still exists
indexed lower case letter is added directly after the
especially concerning VV-A or V-VA where it is
cannula or catheter to which applies. The cannula-
necessary to say, “Vee-and-Vee-to-Ay”, or “Vee-to-
tion site may be exemplified by the following: a
Vee-and-Ay”, respectively. Thus, the abbreviations
femoro-femoral flow direction V-V configuration
also provide information on sequence of cannulation
would be Vf-Vf, and Vf-Af is a femoro-femoral V-A
during the ECMO course. Physiologically, V-VA
ECMO. When, for example, adding a venous return
and V-AV are similar. A DLC can be identified in
cannula placed via the internal jugular vein to the
the configuration where an additional return cannula
former V-A configuration it would be represented
is inserted to the arterial side: (dl)V-VA.
as Vf-AfVj. The same notation is used for a DLC.
VV-VA indicates two single lumen drainage
When inserted via the jugular vein: (dl)Vj-V, or to
cannulae on the venous side with one return can-
specify right jugular vein, (dl)Vjr-V. The side left
nula on the venous and one on the arterial side
(l) or right (r) directly follows the vessel notation.
of the circulation. The hyphen indicates that the
The rare case of a jugular bi-caval DLC with an
oxygenator is placed between the 2 drainage and 2
additional femoral arterial cannula with a distal
return cannnulae.
perfusion cannula is (dl)Vj-VAfa . In this context,
A minor flow cannula, such as a distal limb
the application of a chimney graft (“g”) for inser-
reperfusion cannula, is written with a lowercase

Figure 35-1. The rationale for the nomenclature

for configuration abbreviations at the respective Figure 35-2. Sites for major and of minor flow
level of precision. cannulae, and placement of cannula tips.

306
 Configuration Nomenclature for Extracorporeal Membrane Oxygenation

tion of the return cannula in certain configurations age cannula tip in the RA, and Vfru-VjAfrd describes
is placed after the vessel (and side). 6-7 Thus, Vj-Asrg a venovenoarterial configuration with the drainage
, displays drainage via a jugular cannula with the cannula implanted via the right femoral vein and
tip in the RA and the return via a chimney graft in tip in upper IVC with a primary (time sequential)
the right subclavian artery. From the examples, the jugular venous return, and a right femoral arterial
user may adapt partial or fully detailed abbreviation cannula. A perfusion catheter is adapted for distal
on level two. perfusion of the right foot.
A low flow cannula, such as a venting cannula,
Level Three: Cannula Tip Position may also be provided a position index for increased
information. A venting catheter placed in the left
Cannula tip position is described using an index atrium (“a”) in a V-A application is be expressed
lowercase letter positioned alone or after the level as Vva-A, and if the vent is in the left ventricle
two index letter/s. Knowledge of tip position (in the (chamber, “c”) the abbreviation would be Vvc-A.
context of cannula design) may aid in understand-
ing causes of hypoxemia or oxygenation deviations. Level Four: Cannula Dimensions
Examples include differential hypoxemia in V-A,8-9
and recirculation during V-V ECMO.10 Thus, Vja-Vf Cannula diameter is given in French units. One
is a jugulo-femoral V-V configuration with the drain- French (Fr) equals 1/3 mm in outer diameter of the

Mode of ECMO VV VA VVA

Flow direction “from”  V-V, V-V V-A, VV-A V-VA V-VA V-AV
ML “to” VV-V from V-V from V-A
VV-VV
Configuration SLC DLC SLC SLC from DLC SLC from
V-V V-A

Level 1: Hierarchy V-V (dl)V-V V-Aa Vv-VAa (dl)V-VA V-AaV

Major flow: Vein, Artery, Vc-V (ca)V-V (dl)VV-Ad VV-AdV
Pulmonary artery (bc)V-V
Minor flow: cephalad, distal (dl)Vc-V
(ankle), distal (groin),
cardiac venting
Misc: dual-lumen, cavo-
atrial, bi-caval
Level 2: Cannulation site Vf-Vj (bc)Vfl-V Vj-Af Vf-VjAf (dl)Vj- Vj-AfVf
Vessel: femoral, jugular, VjVf-Afa VA
subclavian Vj-Asrg
Side: left, right
Misc: chimney graft
Level 3: Tip position Vu-Va Vja-Afldva Vf –VfuAf (bc)VfV- Vfa-AflVfl
Position: carotic artery, Vflac-Vf Vja-Asrg Aflg
atrium, upper VCI, left
ventricle (chamber), Vena
cava superior
Side: left, right
Level 4: Cannula dimension V25/25- (dl31)V- Vfa29-Afa V25/38j – (dl27)V- V17ja-
Cannula size: X Fr/ Y cm A17/18 V Vf VA17/18f A15aV15
(length will not be printed V18j- A19/18fa la /50
without diameter) A14V15
/50
SLC=two single lumen cannulae configuration; DLC=dual lumen cannula; VCI=vena cava inferior
Note: A lower case “c” can only be a cephalad cannula; an indexed lower case “c” after a lower case “v” (vc)
can only be vent in left cardiac ventricle

Table 35-1. Examples of abbreviations for different cannula configurations for the different modes for extra-
corporeal membrane oxygenation (ECMO).

307
Chapter 35

cannula. The length is expressed in centimeters (cm). for contemporary use and future innovation. The
Of note, there is no consensus among manufacturers user may apply the nomenclature to the degree of
on definition for cannula length. Most manufactures specification and within each level there is room for
refer to the maximal insertion length, however, increasing degrees of precision. This nomenclature
some refer to total cannula length. A “23/25” should aid communication, ease center collaboration
cannula (23Fr/25, 23Fr/25cm), is of an outer and facilitate compilation of registry, research and
diameter of 23 Fr and 25 cm long. In the nomen- quality data from local to global level.
clature, the user may use only the diameter or
both. If the diameter is not used, then the length
should be omitted. To exemplify: V14-A12;
V23/25f-Af ; V25/38ja-A17/19srg. The size of a
cavo-atrial DLC will be expressed (ca32)V-V,
etc. In this case the length is not given due to
the limited products yet available on the market.
As this configuration information may be
transmitted via email or cellular phones, which
would not allow for indexing, indexed letters
should then be put into brackets, i.e. V(jr)-
A(srg) for Vjr-Asrg .

Classification for Central Cannulation Configu-

ration

In this text, we focus on the nomenclature for

peripheral ECLS/ECMO. The nomenclature as a
whole expands to cover central cannulation configu-
rations and the combined use of both peripheral and
central cannulation. For the complete ELSO nomen-
clature please see The Extracorporeal Life Support
Organization Maastricht Treaty for Nomenclature
in Extracorporeal Life Support,5 and The Extracor-
poreal Life Support Organization Maastricht Treaty
for Abbreviations for Cannulation Configuration in
Extracorporeal Life Support (subm.).

Summary

This chapter summarizes the consensus on

the nomenclature for peripheral ECMO cannula-
tion published ahead of print. This nomenclature
provides a concrete framework for abbreviations
concerning cannulation configuration including flow
direction, number and placement of cannulae and
MLs, additional ECLS-related catheters, cannula-
tion sites, cannula tip position, and cannula dimen-
sions. The nomenclature is flexible and adaptable

308
 Configuration Nomenclature for Extracorporeal Membrane Oxygenation

References

1. Favre A, Powell WH. Nomenclature of Or-

ganic Chemistry. IUPAC Recommendations
and Preferred Name 2013. Cambridge, UK:
The Royal Society of Chemistry; 2013. ISBN
978-0-85404-182-4.
2. Sobin LH, Gospodarowicz MK, Wittekind C.
TNM Classification of Malignant Tumours,
7th ed. Oxford: Wiley-Blackwell; 2009. ISBN
978-1-4443-3241-4.
3. den Dunnen JT, Antonarakis SE. Mutation
Nomenclature Extensions and Suggestions to
Describe Complex Mutations: A Discussion.
Human Mutation 2000;15:7-12.
4. Brogan TV, Lequier L, Lorusso R, MacLaren
G, Peek G. Extracorporeal Life Support: The
ELSO red book. 5th ed, Ann Arbor, MI, USA:
Extracorporeal Life Support Organization; 2017.
5. Conrad SA, Broman LM, Taccone FS, et al.
The Extracorporeal Life Support Organization
Maastricht Treaty for nomenclature in extra-
corporeal life support: a position paper of the
Extracorporeal Life Support Organization. Am
J Respir Crit Care Med. 2018; doi:10.1164/
rccm.201710-2130CP.
6. Javidfar J, Brodie D, Costa J, et al. Subclavian
artery cannulation for venoarterial extracor-
poreal membrane oxygenation. ASAIO J.
2012;58:494-498.
7. Biscotti M, Bacchetta M. The “sport model”:
extracorporeal membrane oxygenation us-
ing the subclavian artery. Ann Thorac Surg.
2014;98:1487-1489.
8. Hou X, Yang X, Du Z, et al. Superior vena cava
drainage improves upper body oxygenation
during veno-arterial extracorporeal membrane
oxygenation in sheep. Crit Care. 2015;19:68.
9. Lindfors M, Frenckner B, Sartipy U, Bjällmark
A, Broomé M. Venous cannula positioning in
arterial deoxygenation during veno-arterial
extracorporeal membrane oxygenation-a simu-
lation study and case report. Artif Organs.
2017;41:75-81.
10. Palmér O, Palmér K, Hultman J, Broman LM.
Cannula design and recirculation during veno-
venous extracorporeal membrane oxygenation.
ASAIO J. 2016;62:737-742.

309
36

Case Scenarios

Chapter 2
A 35 year-old woman with severe ARDS has been placed on venovenous ECLS for respiratory sup-
port. She has been well supported with a blood flow of 3 LPM, sweep gas blood flow of 4 LPM, FDO2
of 100% and has a pre-oxygenator saturation of 78%. The ventilator has been weaned to rest settings and
she is receiving ultra lung protective ventilation. However, on ECLS day #3, she develops fever and her
ABG returns with the following results: pH: 7.28 / PaCO2: 55 / PaO2: 50 / SaO2: 81%. She has developed
a new increase in serum creatinine and her lactate has increased from 1 to 4 mg/dL. Use this information
to answer the following questions:

1. The best method to increase DO2 in this patient:

a. Increase the ventilator FIO2 from 40 to 70%
b. Increase blood flow from 3 to 3.5 LPM
c. Increase sweep gas blood flow from 4 to 5 LPM
d. Convert the patient to venoarterial ECLS to bypass the lungs

2. CO2 clearance across the membrane oxygenator is determined by all of the following except:
a. Sweep gas flow rate
b. Pre-inlet PaCO2
c. Membrane surface area
d. ECLS blood flow rate

3. This parameter that affects DO2 the least:

a. Hemoglobin concentration
b. Cardiac output
c. FIO2
d. Dissolved O2

4. The patient is most likely developed which of the following conditions:

a. Sepsis from a hospital-acquired infection
b. Acute blood loss anemia
c. Circuit-related hemolysis
d. Hypothermia

311
Chapter 36

5. All of the following are true about oxygen consumption (VO2) except:
a. DO2 normally exceeds VO2 by a by about 400%
b. DO2 is approximately equal to VO2 during resting conditions
c. At increased VO2, the SvO2 may decrease substantially
d. The amount of CO2 produced can be estimated by a known VO2

Answers
1. b
2. d
3. d
4. a
5. b

A 4-year old boy has been placed on VA-ECMO via internal jugular and common carotid cannulation
for pneumonia with concerns for poor cardiac output. He weighs 20 kg. After two days of ECMO support,
you are called by the specialist who is concerned as his pulse oximetry saturations are low (86%). The
venous saturations in the circuit are 70%. His pump flows are 1.2 L/min. The sweep gas flow is 1.2 L/min
and the FiO2 is 0.50. A new non-invasive cardiac monitor estimates his native cardiac output at 1.2 L/min.
An arterial blood gas is obtained and shows a pH of 7.32, PCO2 of 49 mm Hg and a PaO2 of 58 mm
Hg. The venous gas shows a pH of 7.25, PCO2 of 56 mm Hg and a PaO2 of 33 mm Hg while the post –
oxygenator gas is pH of 7.37, PCO2 of 44 mm Hg and a PaO2 of 200 mm Hg. The specialist increases the
FiO2 of the sweep gas to 1.0.

1. What do you think the next arterial saturation will be?

a. 100%
b. 86%
c. 98%
d. 88%
e. 95%

2. After increasing the FiO2, the specialist then turns the flow up to 2.0 L/min. The non-invasive cardiac
monitor now shows that the native cardiac output is 0.4 L/min. What do you think the arterial satura-
tion will be now?
a. 75%
b. 86%
c. 98%
d. 88%
e. 92%

Answers
1. d
2. c

Discussion:

The blood exiting from the oxygenator is essentially fully saturated. Thus by increasing the FiO2 the
specialist will only augment the dissolved fraction of oxygen (equation 1) which will result in a very mode

312
 Case Scenarios

increase in the patient’s arterial saturation and PaO2, (answer c). However when the specialist increases
the flow (without an obvious increase the total cardiac output) there will be a much larger increase in
arterial saturations (equation 6).

Chapter 3
A 16-year-old otherwise healthy male (65 kg) was admitted yesterday with hypoxia and shortness of
breath requiring endotracheal intubation for acute hypoxic and hypercarbic respiratory failure with a chest
radiograph showing diffuse, dense bilateral pulmonary infiltrates. The morning labs revealed a pH: 7.10/
PaCO2: 64 mm Hg/PaO2: 60 mm Hg. The central venous oxygen saturation was 40% and the serum lactate
8.1 mmol/L. At that time, you increased ventilator settings (PEEP, rate, and FiO2) ordered neuromuscular
blockade, inhaled nitric oxide at 20 ppm, and placed the patient in a prone position.
Two hours later, the PCR turns positive for H1N1 influenza. Upon re-evaluation of the patient, he has
mottled lower extremities to the thighs, sluggish capillary refill, bulging neck veins and coarse bilateral
breath sounds. The repeat lab values worsened to a pH: 7.04/ PaCO2: 67 mm Hg/PaO2: 58 mm Hg and
[hemoglobin]: 13.4 g/dL/hematocrit: 40%. The creatine kinase (CK) is elevated at 3548 U/L along with
an elevated troponin I.  The central venous oxygen saturation remains at 40%, but the serum lactate has
climbed to 9.2 mmol/L. Hemodynamic monitoring shows an arterial blood pressure of 85/55, heart rate
of 150, CVP of 14, and a SpO2 of 85%. A cerebral oximetry is placed and the value is 48%. Mechanical
ventilator settings are PEEP: 18, rate: 25/min, FiO2: 1.0, and tidal volume: 330 mL.

Questions:

What data in the above scenario reveal a problem with global oxygen content and/or delivery?
In the above scenario, what markers illustrate poor perfusion?
What do the hemodynamic data suggest regarding cardiac function?
What other data or tests might you request in order to make your next clinical decisions?
Would you consider ECMO at this time? If so, what type?

Discussion:

The patient has severe acute hypoxic respiratory failure from influenza pneumonia and possible in-
fluenza myocarditis. His arterial oxygen content is low. While the hemoglobin and hematocrit are normal,
the arterial pO2 and oxygen saturation are low despite high ventilator settings.With the data above, we
cannot calculate oxygen delivery since we do not know the cardiac output. However, the low central ve-
nous oxygen saturation and high lactate suggest that the patient does not have adequate oxygen delivery.
Hemodynamically, the low arterial blood pressure indicates hypotension despite high central venous pres-
sure, potentially indicating cardiac dysfunction.
With a reversible underlying disease, persistent hypoxia and impaired global oxygen delivery despite
high ventilator settings, ECMO should be considered. While vasopressors and inotropes will likely be
used (if not already), you should next evaluate cardiac function. This can be accomplished with a Swan-
Ganz catheter and/or a bedside echocardiogram. The results of the cardiac output data will help you to
decide whether VA or VV-ECMO would be preferential.  If the patient has severe contractile dysfunc-
tion, VA-ECMO may be required because it provides circulatory support and permit myocardial rest. In
some cases, simply improving oxygenation with VV-ECMO may be sufficient to improve mild cardiac

313
Chapter 36

dysfunction especially if it is predominantly right ventricular dysfunction. Both VA and VV-ECMO permit
lung protective ventilation and right ventricle protective ventilation.

Chapter 4
A 15 yo male has been referred for ECMO due to severe cardiopulmonary failure, refractory to cur-
rent therapeutic interventions, from severe methicillin resistant Staphylococcus aureus sepsis and pulmo-
nary hemorrhage secondary to H1N1 influenza infection. He is coagulopathic (INR: >3, platelet count:
< 25,000/uL). His hematocrit is 37% but he continues to have significant pulmonary hemorrhage. He is
being prepared for placement onto VA-ECMO. Please answer the following questions with regards to the
anticoagulation management of this patient.

1. What would be the best plan for initial UNFH dosing during initiation of anticoagulation?
a. Give full dose of 100 U/kg
b. Give half dose of 50 U/kg or do not give any due to severe coagulopathy and just start UNFH
infusion.
c. Do not use UNFH, start on a direct thrombin inhibitor as patient has HIT
d. None of the above

2. The patient was successfully placed on VA-ECMO. Despite increasing the UNFH dose, neither the
ACT, nor the antiXa are increasing. What are the next steps?
a. Do an AT level
b. Just keep increasing UNFH dose until there is a change.
c. Measure an aPTT
d. Stop measuring antiXa and ACT as UNFH dose is in the range required.
e. Switch anticoagulants as UNFH isn’t working.

3. Patient begins bleeding from chest tubes, cannula site, IV sites and arterial line site. How do you
control this?
a. Wean off ECLS and prepare family patient will not survive.
b. Explore the chest for bleeding
c. Switch to a direct thrombin inhibitor
d. Stop the UNFH
e. Check anticoagulation/hemostasis bloodwork, optimize clotting factors (platelets, cryoprecipitate,
etc), reduce UNFH infusion to lowest range, replace blood loss with PRBCs, consider antifibrinolytic
as fibrinolysis (check a TEG or thromboelastometry) may be occurring.

Answers and Discussion:

1. b. In most cases a 50 U/kg dose would be given despite severe coagulopathy as there can be increased
clot formation with disseminated intravascular coagulopathy (DIC), resulting in clot formation within
the cannulae during placement. If the coagulopathy is severe enough some groups may choose to
withhold all anticoagulation at time of cannulation.
2. a. UNFH effect is indirect and requires AT. If the AT level is low then administration of AT is required
to achieve heparinization and appropriate anti-Xa activity.

314
 Case Scenarios

3. e. Determination of bleeding causes on ECLS is imperative for its management. When bleeding oc-
curs from multiple sites this can result from excessive anticoagulation, loss of clotting factors and
platelets, or fibrinolysis. Supportive measures and diagnosis of cause are the first line. If fibrinolysis
is determined then addition of an antifibrinolytic is warranted. Large clot in the chest cavity can also
stimulate fibrinolysis and therefore consideration for surgical exploration may be required if control
cannot be achieved.

Chapter 5
A 6 year old girl with a history of acute myeloid leukemia and recent chemotherapy treament presents
to the Emergency Department with fever. Labs there reveal WBC: 3.2 K/dL (ANC 780), hemoglobin: 7.2
g/dL, hematocrit: 22%, platelet count 80 K/dL, and blood type Type A, RhD positive. She is admitted
for observation, but quickly deteriorates. Positive for influenza A, she develops respiratory failure with
maximum mechanical support. The intensive care unit team wishes to cannulate her for VV venovenous
ECLS but to do so, she needs blood products.

1. What special considerations come to mind when transfusing immunocompromised patients, such as
those receiving chemotherapy?

2. Assuming she needs a RBC transfusion, would you order “emergency issue” or “standard issue” red
blood cells?

Answers and Discussion:

1. Immunocompromised patients are more susceptible to transfusion-associated infections and graft-vs-

host disease. As such, if time allows, all products issued to immunocompromised patients should be
CMV-safe (or negative) and irradiated. Each transfusion service should have a policy or procedure
outlining how they define “immunocompromised” and which modifications blood products should
undergo.

2. It depends on how urgent her cannulation is needed. The intent of emergency-issue blood is to issue
blood quickly. It skips all pretransfusion compatibility testing and issues universally compatible Group
O RBCs and Group A or AB plasma and platelets. Therefore, all products will be largely compatible
for ABO, but may or may not be compatible for other minor blood groups. Standard-issue blood is
tested for compatibility (i.e., type and screen plus a crossmatch) and is modified (i.e., CMV-safe,
washed, irradiated, etc.) prior to issuing the blood to the bedside. This process ensures that the safest
blood is issued, but can take anywhere from one hour to several days and is therefore inappropriate
for patients who bleed briskly or who need blood urgently.

315
Chapter 36

Chapter 6
A 5-year old male receives ECLS cannulation for acute respiratory distress syndrome secondary to
submersion injury while swimming in a local lake. He weighs 20kg. His serum creatinine is 0.5mg/dL
and he has good urine output.

1. He was brought to your center from an outside hospital where he received sedation with fentanyl,
currently at a rate of 2 mcg/kg/hr (standard starting dose). He has received numerous bolus doses
and the team is concerned that he will not safely remain cannulated due to agitation. What would you
recommend?
a. Continue fentanyl but increase slightly to 3 mcg/kg/hr
b. Add a continuous propofol infusion to help improve sedation
c. Switch from fentanyl to morphine–use higher starting dose of morphine
d. Add a midazolam infusion

2. The team would also like to start a dexmedetomidine infusion and requests an initial dosing recom-
mendation. What dosing would you recommend? The standard starting dose of dexmedetomidine is
0.2mcg/kg/hr.
a. Start at 0.2mcg/kg/hr and titrate slowly to effect.
b. Give a loading dose of 100-500mcg/kg/dose and start at 2mcg/kg/hr
c. Start at 0.1mcg/kg/hr to avoid over sedation
d. Start at 0.4mcg/kg/hr – consider loading dose if extremely agitated

3. Prior to arrival at your center he was started on broad-spectrum antibiotics including vancomycin. The
outside hospital started with standard dosing for age: 15mg/kg IV q6h. Urine output remains good.
What dosing recommendation might you make?
a. Decrease vancomycin dose empirically because there is a risk of decreased clearance while on
ECLS
b. Continue current dose – follow vancomycin levels closely to assess clearance and make patient-
specific adjustments as indicated.
c. Wait one week to assess vancomycin level to make sure vancomycin is truly at steady-state
d. Continue current dose – no need to check levels as patient’s clinical status is stable

4. The infectious disease (ID) service recommends antifungal coverage due to the presence of lake water
in his lungs. The ID fellow suggests that voriconazole would provide good broad-spectrum coverage.
What information might you provide to your team to help guide their decision?
a. Voriconazole has a high log P value and volume of distribution which will make it challenging to
achieve an effective dose – other agents should be considered
b. Voriconazole can be started at standard dosing (~4mg/kg/dose) and then adjusted per levels
c. Voriconazole should only be administered once daily to avoid toxicity
d. Voriconazole is the best choice for patients on ECLS

Answers:
1. c
2. d
3. b
4. a

316
 Case Scenarios

Chapter 7
Neonatal Respiratory ECMO:

A 2.9kg LGA male infant is born at 36 weeks after prolonged labor and maternal fever. He is breath-
ing at delivery but grunting and in obvious distress. Arterial saturation remains at 50% at 10 minutes of
life, he is intubated with improvement in saturations to 90% on and FiO2 of 1.0. CXR is hazy bilaterally
with poor expansion. Nitric oxide and dopamine are started with a transeint response, but he declines over
the next 24 hours and is transferred for possible ECMO. Echo reveals pulmonary hypertension but good
cardiac function with normal anatomy.

1. Calculate the oxygenation index for Baby C who is currently on HFOV with a MAP of 16, Amp of
42, Hz of 8 and 100%. Blood gas results as 7.16/52/32/-8
a. 32
b. 45
c. 50
d. 64

2. This patient is a reasonable candidate for ECMO, but has several risk factors that increase the odds
of morbidity and mortality. Which of the following is NOT a risk factor?
a. Acidosis
b. History of CPR
c. Gestational age
d. Chronological Age
e. Diagnosis

3. What survival rate would you predict for this baby?

a. 40%
b. 51%
c. 70%
d. 92%

Answers and Discussion:

1. c. OI = (MAP x FiO2)/PaO2. (16x1x100)/32=50.

2. d. Late prematurity, history of CPR and acidosis are all risk factors for poor outcome, as is his diag-
nosis—which is currently idiopathic PPHN but may also be pneumonia or “other”
3. c. Although prediction is difficult, most of the diagnostic category this baby fits into have survival
rates of 60-80%, gestational age will also affect that 92% is more typical for MAS, while 51% is more
typical for CDH.

Pediatric Respiratory ECMO:

You are in a busy pediatric intensive care unit with 3 patients on ECLS. The first is a 2 month old with
RSV bronchiolitis and superimposed bacterial pneumonia. The second is a 7 year old with severe status
asthmaticus in response to environmental allergens. The third is a 12 year old with leukemia and fungal
pneumonia who has been on ECLS for 24 days.

317
Chapter 36

1. Which patient on ECLS in the intensive care unit described above has the highest likelihood of survival
to hospital discharge?
a. RSV bronchiolitis
b. Bacterial pneumonia
c. Status asthmaticus
d. Fungal pneumonia

2. Which patient on ECLS in the intensive care unit described above has the lowest likelihood of survival
to hospital discharge?
a. RSV bronchiolitis
b. Bacterial pneumonia
c. Status asthmaticus
d. Fungal pneumonia

3. In the busy pediatric ICU described above, how common would you expect co-existing medical condi-
tions to be in patients who are placed on ECLS for respiratory failure?
a. Approximately 10% of patients have comorbidities
b. Approximately 20% of patients have comorbidities
c. Approximately 30% of patients have comorbidities
d. Approximately 50% of patients have comorbidities

4. Which statement is most accurate about prolonged courses of ECLS for pediatric respiratory failure
similar to the third patient described?
a. Survival in patients on ECLS for greater than 21 days is similar to survival in patients on ECLS
for fewer than 21 days.
b. Survival on ECLS declines as duration of ECLS increases but there are a fair number of patients
who survive a prolonged course, which makes it difficult to know when to withdraw ECLS support
in the absence of severe complications.
c. Mortality is highest for patients with prolonged ECLS around 21 days but if patients can survive
these 3 weeks, then they have a high likelihood of survival.
d. Mortality is so high in patients who are on ECLS for 21 days or greater that ECLS should not be
continued for more than 3 weeks.

Answers
1. c
2. d
3. c
4. b

Adult Respiratory ECMO:

A 32-year-old obese woman (87 kg) with hypoxic and hypercarbic respiratory failure secondary to H1N1
influenza is in the ICU. She has been intubated for 2 days. She is sedated, paralyzed, hemodynamically
supported with phenylephrine. She is 12 liters positive since admission 3 days prior. She is on FiO2 100%,
PEEP 24 cm H2O, plateau airway pressure of 36 cm H2O. Chest x-ray shows bilateral patchy infiltrates
without obvious effusions. Her ECHO is hyperdynamic without valvular disease and left ventricle appears
normal with ejection fraction 60%. Arterial PaO2 is 56 mmHg.

318
 Case Scenarios

1. What is her estimated mortality risk?

a. 25%
b. 50%
c. 70%
d. >80%

2. You choose to support her with ECMO. Which mode of ECMO should be initially offered?
a. Venoarterial
b. Venovenous

3. She is supported with Venovenous ECMO via a dual lumen cannula in the right internal jugular vein.
Her ECMO flows are 5.2 liters per minute. Her hemoglobin is 11.3 gm/dL after 2 units of blood. Ven-
tilator settings were changed to respiratory rate of 10, FiO2 of 30%, PEEP 10 cm H2O, peak airway
pressure is 20 cm H2O and tidal volumes are150 cc. Sedation and paralytics were discontinued and she
was beginning to wake up. Her mixed venous saturation is 67%, measured with a pulmonary artery
catheter. Her arterial saturations are 82% and her PaO2 is 54mmHg. What do you do next?
a. Add a second venous drainage cannula to increase ECMO flows to 7 liters per minute
b. Increase FiO2 to 100%, PEEP to 20 cm H2O
c. Re start sedation and paralytics
d. Go have a cup of coffee

Answers and Discussion:

1. d. This patient has ARDS. Using the Berlin definition this is severe ARDS. Her APPS score would be
7-8. Murray Lung Injury score would be 3-4. Her mortality risk would be >80% and ECMO would
be indicated.
2. b. She has primary respiratory failure. Venovenous ECMO should provide sufficient support. While
she requires hemodynamic support with phenylephrine, her hemodynamic compromise is likely a
consequence of high intrathoracic pressure, which impedes cardiac return, combined with hypotensive
effects of sedation.
3. d. Your patient is adequately supported on VV-ECMO as demonstrated by acceptable arterial satura-
tions, mixed venous saturations with appropriate hemoglobin levels. “Rest ventilator” settings are ap-
propriate with intention of minimizing ventilator related injury. Discontinuing sedation and paralytics
is important so that neurologic status can be assessed.

Chapter 8
A newborn boy with hypoplastic left heart syndrome is admitted to the cardiac intensive care unit
(CICU). He is maintained on a prostaglandin infusion for several days while he undergoes diagnostic
work-up. At 5 days of age, he undergoes a Norwood procedure with a Blalock-Thomas-Taussig shunt. He
has a prolonged cardiopulmonary bypass (CPB) time and aortic cross-clamp time. Initially, on the first at-
tempt to wean from CPB, his echocardiogram reveals poor right ventricular (RV) contractility and he has
hypotension with poor oxygen saturation. For this reason, he is placed back on CPB and tentative plans
are made to transition to ECLS. However, before transitioning to ECLS, another attempt to wean him off
CPB is made and proves successful. He remains intubated with an open chest as he is brought back to the
CICU. His vasoactive infusions include milrinone (0.75mcg/kg/min) and epinephrine (0.1mcg/kg/min).
After arrival to the CICU, he has borderline low blood pressures, a persistent lactic acidosis, and no urine

319
Chapter 36

output. A bedside echocardiogram bedside reveals moderate RV dysfunction but cannot visualize the BTT
shunt, pulmonary arteries or aortic arch. Over the course of several hours, he continues to develop signs
of poor systemic perfusion with elevated heart rates, no urine production, increasing lactic acidosis and
increasing epinephrine requirements to maintain his blood pressure. Several hours after he has been in the
cardiac ICU, he has a PEA arrest.

1. Is this patient an ECMO/ECPR candidate?

2. What type of ECLS (VA vs VV) should he receive?
3. How should he be cannulated?
4. Qualitatively, how much ECLS flow may this patient require?
5. What should be considered when discussing how to ensure that this patient is able to wean off of
ECLS?

Chapter 9
Pediatric IH Case:

An infant with transposition of the great arteries (TGA) underwent complete repair this week and was
extubated today. In the intensive care unit, he remains fussy, febrile, tachypneic, and tachycardic. The
family is at the bedside and is holding him for the first time since surgery. He becomes acutely limp and
unresponsive. The pulse oximetrer cannot detect a saturation, and the nurse sees a wide complex brady-
cardia on the monitor with no palpable pulse. The code bell is activated and the infant is placed back in
bed. The nurse immediately begins CPR.

The ICU team is providing good quality compressions, the trachea has been re-intubated, intravenous
epinephrine has been delivered. The arterial blood pressure and end-tidal CO2 show wave forms indica-
tive of a good response to CPR measures. The ICU team is systematically reviewing possible causes of
the bradycardic arrest. For 7 min, he has ongoing pulseless electrical activity in spite of attempts to pace,
and to exclude tamponade, bleeding, anemia, hypovolemia, and pneumothorax. The team must consider
the following: opening the chest (via sternal surgical site) to improve cardiac function and/or address any
coronary ischemia and/or whether to use ECMO to provide return of circulation (ROSC) to enhance ce-
rebral and myocardial perfusion and oxygenation. ECMO will also allow time to investigate and address
the cause of the cardiac arrest. The team would need to decide whether to use an open sternal approach
or neck cannulation for ECMO. The team must also consider the timely requirement for diagnostic and
interventional catheterization and or preparation for surgical review. Ideally the interval of time to achieve
either ROSC or ECMO flow should be as short as possible. Separately, in this case the bedside team may
have a different threshold to define ‘refractory’ to resuscitation measures to minimize ischemia.

Adult OOH Case:

A 42 year old was participating in mini-marathon to raise funds for his community’s charity when he
falls to the ground among the crowd. A volunteer sees him go down and runs to his side. Another volunteer
joins him with the AED while EMS is being called. Bystanders are providing CPR, the AED identified
VF and shocked him twice, when EMS arrive. They transition to Advanced life Support (ALS) CPR, call
the coordinating center with medical support, and are directed to the regional cardiac center’s emergency
room. It’s now 25 min since the start of bystander CPR, and while ongoing CPR measures (including
repeated defibrillation) are delivered, they are directed to transport the patient to catheterization suite to

320
 Case Scenarios

investigate and address the source of refractory VF. A team prepared for interventional cardiology and
for ECMO cannulation is waiting ready for them. The next steps will involve deciding when to secure
vascular access to initiate ECPR while ongoing therapies for VF are being delivered or continue ongoing
advanced therapies to address possible acute coronary syndrome.

Chapter 10
Trauma

A 25-year-old motorcyclist hit by a car on the highway subsequently presented with blunt chest trauma,
a pelvic fracture and a penetrating injury to the right lower extremity. His Glasgow Coma Scale score
was 8 at the scene. His heart rate was 130 beats/min, his blood pressure was 70/40 mmHg, his respiratory
rate was 12 breaths/min and his oxygen saturation was 83%. He was intubated in the ED and mechani-
cally ventilated with the FiO2 of 1.0. Chest tube thoracostomy was performed to treat bilateral tension
pneumothoraces. Coagulopathy persisted following vascular reconstructive surgery of the femoral artery
and pelvic vascular intervention. The patient required increasing doses of vasopressors and increasing
mechanical ventilation settings.

The arterial blood gas analysis and ventilator settings were as follows:

Arterial O2 of 68 mm Hg, CO2 of 110 mm Hg, and pH of 7.12. The tidal volume was less than 150
mL, although the “delta” P was > 20 cm H2O, PEEP was 15 cm H2O and inspiratory pressure was 38 cm
H2O with an FiO2 of 1.0.

1. What is your interpretation of the results?

2. What is happening and why? Recommend a treatment option

Answers and Discussion:

1. The patient experienced severe hypoxemia, hypercapnia and respiratory acidosis secondary to decreased
lung compliance and acute, severe lung failure.
2. The treatment of severe hypoxemia included veno-venous extracorporeal life support (VV ECLS).
VV ECLS is a feasible lung support option in case of severe lung failure after major trauma. In cases
of reduced lung compliance, VV ECLS can facilitate adequate gas exchange and be used to prevent
ventilator-induced lung injury, allowing the lungs to rest, e.g., on ultra-protective ventilation during
spontaneous breathing.

Oncology

A previously normal and healthy 3 year old girl with newly diagnosed acute lymphoblastic leukemia
(ALL) presents with a WBC of 170,000/μL, Hgb of 6 g/dL, 12,000/μL platelets, and acute hypoxemic
respiratory failure from influenza. She has received broad spectrum antibiotics, influenza anti-virals, and
has not received any chemotherapy. She is supported with a high frequency oscillating ventilator, with
FiO2 of 1, MAP 32, delta P of 96, and frequency of 5.5 Hz generating an arterial blood gas of 7.32/52/56
and an oxygenation index of 57. She requires no inotropes/vasopressors, and has no evidence of other
organ injuries.

321
Chapter 36

1. Would you offer ECLS therapy for this patient? Describe positive and negative risk factors for this
patient.
2. What do you expect to happen to the WBC immediately after starting ECLS?

Answers and Discussion:

1. Offering ECLS would be reasonable in this patient. Favorable prognostic factors include that she has
a cancer with a very high 5 year survival rate (>90%), and long term cure of this cancer would be
expected. She has not yet received chemotherapy and so has not had a chance to be treated. The un-
derlying condition requiring ECLS is influenza, which is known to be responsive and reversible with
ECLS with reasonable success rates. Negative prognostic factors include increased risks of bleeding
due to thrombocytopenia, thrombosis due to active cancer, potential increased risk of secondary infec-
tions due to immunosuppression, and altered pharmacology of administering chemotherapy while on
ECLS.
2. The white blood cell (WBC) count will drop, often precipitously, after ECLS cannulation even without
chemotherapy. This effect increases with decreasing patient size because of both hemodilution (the
ECLS prime contains no WBC) and due to binding of the WBC to the ECLS circuit and oxygenator.

Chapter 11
A patient is accepted for admission from an outside hospital with a suspicion of cardiomyopathy. The
supervising physician asks for activation of the ECMO system to be prepared for ECPR in the event of
a cardiac arrest should this patient require endotracheal intubation as part of the management. The call
goes to the ECMO Specialist Primer. What information is desirable in order to prepare for the initiation
of ECLS? Select all that apply:

a. Age/weight of patient
b. Primary diagnosis
c. Admission room
d. Time of arrival
e. Have orders been placed in the EHR for:
Blood products
Infusions
Priming medications (ECMO Pack)
f. Plan of care/ECMO candidacy
g. Family members

Discussion:

All of this information is desirable when planning for cannulation, with the exception of knowing if
family members will be present. The point of the scenario is to emphasize that planning and time manage-
ment is key to successful activation.

322
 Case Scenarios

Chapter 12
A 58-year-old male had a witnessed cardiac arrest with immediate cardiopulmonary resuscitation
by a bystander. The patient was brought quickly to a Tertiary Care hospital with an ambulance equipped
with an automatic cardiac massage device. Upon arrival, he has no recovery of spontaneous circulation,
despite multiple attempts at defibrillation. The time from witnessed cardiac arrest and arrival at ED is 40
minutes. The decision by the ED doctors is to proceed with VA-ECMO cannulation and then catheterize
the patient for myocardial ischemia evaluation. After ECMO flow begins, he converts to sinus rhythm but
has myocardial ischemia on ECG.

1. The patient receives peripheral VA-ECMO support as part of his eCPR resuscitation. Transthoracic
echocardiogram demonstrates minimal LV contractility and LV distention. A possible next step to
address this is:
a. Addition of inotrope
b. Atrial septostomy
c. LV vent via left thoracotomy
d. A, C
e. A, B, C
2. To avoid Harlequin Syndrome, cannulation for VA-ECMO support should be configured with arterial
inflow in
a. Superficial femoral artery
b. Axillary artery
c. Aorta
d. B, C
e. None of above

Answers:
1. e
2. d

F.C. is a 2.1 kg neonate (2nd day after birth) with meconium aspiration syndrome unable to maintain
adequate gas exchange (severe hypoxia with increasing lactic acidosis) despite maximal medical therapy.
He is on HFOV with FiO2: 1.0, MAP: 17 cm H2O, DP: 65; Frequency: 9 Hz, iNO: 25 ppm.
The neonatologist calls the ECMO team to provide ECMO.

Which kind of ECMO will be used to manage this child?

a. VV
b. VA
c. ECCO2 R
d. None of the above

Which kind of pump flow will be choosen?

a. 80 ml/kg/min
b. 100 ml/kg/min
c. 60 ml/kg/min
d. A pump flow between 100-200 ml/kg/min according to gas exchange

323
Chapter 36

Answers
1. b
2. d

Chapter 13
A five months old 4.2 kg boy born at 32 weeks gestation underwent primary reconstructive surgery
for esophageal atresia, with unsatisfactory results. RSV pneumonitis and Klebsiella pneumonia caused
severe hypoxemic respiratory failure requiring ECMO treatment. ECMO was initiated at the referring
hospital by a mobile ECMO team. No cardiac failure was detected thus a 13 Fr Dual lumen cannula was
placed for VV-ECMO. The position of the cannula was confirmed by a chest radiograph. The patient was
transported on ECMO by fixed-wing transport.

During the first hours in your institution his oxygen consumption was rather high suggesting subop-
timal sedation. The ECMO blood flow was 120 mL/kg/min. After a few hours, simultaneous blood gas
samplings were undertaken from three different locations; the radial artery, and pre- and post-membrane
lung (ML), (Table 1).
Patient Pre-ML Post-ML
pH 7.35 7.36 7.44
SatO2 % 31 67 100
pO2, mmHg [kPa] 25 [3.3] 40 [5.3] 394 [52]

1. What is your interpretation of the results?

2. What would be your next step?
3. What would you do now?

Answers and Discussion:

1. Recirculation on VV-ECMO
2. Echocardiography was immediately performed which showed a native cardiac output of almost 2
L/min! The ECMO cannula was visualized far up in the right atrium and color Doppler indicated
recirculation. A plain chest radiograph may show the suboptimal cannula position. It may also reveal
a pneumothorax, or any volume expansion affecting cardiac filling with secondary effects on native
cardiac output. A reduced CO increases the recirculation fraction.
3. Adjust the cannula position, preferably under visualization with echocardiography. The cannula was
pushed down 10 mm and within a few minutes the peripheral oxygen saturation increased to 88%.

Chapter 15
A 65 year-old man with a history of COPD presents to the Emergency Department (ED) with a 2-day
history of worsening shortness of breath which started following a recent viral infection. In the ED, his
oxygen saturation is 88% on room air. He has been working hard to breath and speaks only in short sen-
tences. On exam, he has diffuse wheezes with prolonged expiratory phase. His chest radiograph reveals
a new focal right infiltrate. An arterial blood gas (ABG) shows pH: 7.30, PaCO2: 80 mm Hg, HCO3-: 39
mmol/L, PaO2: 62 mm Hg, SpO2: 92%. FiO2: 0.5. The patient had a NIV trial in PSV with PS 14 cm H2O,

324
 Case Scenarios

PEEP 8 cm H2O and FiO2 0.5 but after initial improvement his ABG again worsened: pH: 7.25, PaCO2:
89 mmHg, HCO3-: 37 mmol/L, PaO2: 83 mmHg, SpO2: 96%, FiO2 = 0.5.

What would you do?

NIV plus ECCO2R
WOB decreased and NIV was discontinued, pH and PaCO2 improved
ABG results: pH: 7.41, PaCO2: 56, HCO3-: 38 mmol/L, PaO2: 85 mmHg, SpO2: 94%, FiO2: 0.4

Chapter 18
A slim 19-year old woman develops severe acute respiratory failure due to influenza B complicated
by methicillin sensitive Staphylococcus aureus. Her oxygenation worsens despite increasing ventilator
support and she is placed on VV-ECMO with a 31-Fr double lumen catheter. On the evening of cannula-
tion the pump flow is 4.3 L/min. Her arterial saturation is 87% while the ECMO venous saturation is 68%.
Her hemodynamics are uncompromised. On the following morning, the specialist worries that the patient
has suffered severe neurologic injury as the saturation of the venous blood returning to the oxygenator is
now 90% and the pulse oximeter on her left hand reads 66%.

Her pupils are small and when you look at her cannula you see that if has flipped itself in a position
where the venous limb lies above (anterior) to the arterial limb (the patient is supine) and the skin sutures
are stretched. You recognize that the catheter is backward with the arterial jet positioned with flow away
from the tricuspid valve. On further investigation you notice that the tubing from the pump lies naturally
with the cannula in a backward position. You realize that when the catheter was attached to the tubing,
the later was torqued (or rotated) and during the night the tubing simply rotated the catheter 180 degrees
from its normal position. The solution is to rotate the catheter back to its correct position and to consider
cutting the tubing to reconnect it in a relaxed position.

Chapter 19
A baby boy is a 41week gestational age, 3.8 kg infant born at an outlying community hospital fol-
lowing emergent Caesarean section for poor fetal heart rate variability. Moderate meconium stained fluid
was noted at delivery and initially there was no respiratory effort and severe bradycardia. The heart rate
improved quickly with positive pressure support but the baby continued to have poor respiratory effort and
was intubated. Following transfer to the NICU, he required escalation of support for respiratory failure to
include 100% oxygen, high frequency oscillatory ventilation (HFOV) and 20 ppm of inhaled nitric oxide
(iNO). At 2 hours of age, pre- and post-ductal oxygen saturations were 86% and 75%, respectively, and
a UAC blood gas was 7.10/68/39/-9 with a calculated OI of 55. A chest radiograph was consistent with
meconium aspiration with diffuse lung injury but no air leak. By 4 hours of age, these numbers did not
improve so transport to an ECMO center was arranged.

1. Based on the above information, is Baby M an ECMO candidate?

2. What additional evaluation should be done prior to placing this baby on ECMO?
3. Should VV or VA support be used? What are the benefits and risks associated with both approaches?

Following arrival at the ECMO center, an echocardiogram was performed that showed normal structural
anatomy and cardiac function but was worrisome for severe PPHN. A cranial ultrasound was done that

325
Chapter 36

showed no abnormalities or evidence of bleeding. Additional laboratory testing noted a normal platelet
count and coagulation function. Despite good cardiac function, the baby continued to have severe hypoxic
respiratory failure with OIs 55 or greater, so a decision was made to go on VV-ECMO following discus-
sion and consenting of the family.

1. What size VV cannula would be appropriate for this baby?

2. How much blood is needed to prime the circuit and does it need to be washed or hemofiltered?
3. What ECMO pump flow should be targeted once on ECMO?
4. Where should the tip of the cannula be located for optimal flow? How would you confirm it is in good
position?

Washed, type specific blood (200 mL) was obtained from the blood bank for circuit priming. A 13-Fr
double lumen VV cannula was placed percutaneously by surgery in the right internal jugular vein and
secured at a depth of 7 cm. The tip of the cannula was confirmed in good position in the mid to lower right
atrium by echocardiogram with arterial flow directed at the tricuspid valve. ECMO flow was advanced
gradually over 20 minutes to 120 ml/kg/min without generating excessive negative venous pressure and
patient pre-ductal oxygen saturations improved to 94%. Ventilator support was adjusted to rest settings
with FiO2 of 0.40.

On Day 2 of ECMO, Baby M’s oxygen saturations acutely dropped into the 70’s following positioning
for his morning chest radiograph. A UAC blood gas was done and showed normal pH and PCO2 values,
but a low PaO2 of 30 mmHg. Additionally, the circuit SVO2 simultaneously increased from 78% to 94%
and there seemed to be a decreased color differential of the blood in the circuit pre- versus postoxygenator?

1. What are the possible causes for Baby M’s decreased oxygenation?
2. What tests should be done to confirm the problem?
3. What corrective actions should be taken?

Based on the above changes, the baby was suspected of having an increased recirculation fraction
secondary to malposition of his cannula. A chest radiograph showed no pneumothorax or pneumopericar-
dium but the cannula tip was several cm higher than previously. An echocardiogram confirmed that the
cannula tip to be high at the junction of the superior vena cava and right atrium. Surgery was contacted
and successfully repositioned the cannula in the right atrium, which immediately improved oxygenation
to the prior high levels.

By ECMO Day 4, he continued to have normal hemodynamic function without inotropic support with
minimal edema. His chest radiograph showed good inflation and he had improved lung compliance on
moderate ventilator support. A follow-up echocardiogram showed good cardiac function, minimal evidence
of PPHN, and a small PDA with predominantly left to right shunting. ECMO pump flow was 110 ml/kg/
min, sweep flow of 0.1 lpm and 0.35 FiO2. Circuit SVO2 was reading 85% and last patient UAC blood gas
had a PaO2 of 75 mmHg with a ventilator FiO2 of 0.40.

1. Is Baby M ready to trial off ECMO?

2. What is the procedure for trialing off VV support?
3. Does ECMO pump flow need to be further weaned?
4. What needs to be monitored closely during the trial off period?
5. If PPHN remained significant, what additional therapy could be reinitiated to help improve oxygenation
for the trial off?

326
 Case Scenarios

6. What constitutes a “successful trial” off ECMO support?

The patient underwent a trial off ECMO with “capping” of the sweep gas to the circuit oxygenator.
Ventilator settings were adjusted to provide adequate ventilation, and mean airway pressure and FiO2
were slightly increased to support patient oxygenation. After 1 hour off ECMO, a UAC blood gas was
7.42/48/90/-2 BE on HFOV settings of a MAP of 14 cm H2O, frequency of 9 Hz, ∆P 28 cm H2O and FiO2
of 0.50. The FiO2 was increased to 1.0 for 5 minutes and another UAC blood gas showed an increased
PaO2 to 260 mmHg. Based on these findings, the decision was made to decannulate him and discontinue
VV-ECMO support. He remained on the ventilator for an additional 72 hours following and was discharged
home on day of life 21.

Chapter 20
A 2 year old girl, weighing 15 kg, develops refractory respiratory failure post drowning. She is can-
nulated for VV-ECMO using a 16-Fr double lumen cannula inserted percutaneously through the right
internal jugular vein. You start with an FiO2 of 1.0 and sweep gas to pump flow ratio 1:1. You progres-
sively increase the ECMO centrifugal pump flow. The following are 4 different scenarios that may occur:
RPM Pump Flow pH PaCO2 PaO2 SaO2 Renal rSO2 *
1 2400 600 ml 7.41 41 55 75% 55%
2 2800 1000 ml 7.38 35 75 88% 85%
3 3200 1300 ml 7.42 38 100 97% 83%
4 3800 1400 ml 7.40 40 150 99% 85%
*Renal rSO2: Renal regional oxygenation measured by near-infrared spectroscopy

1. What is your optimal pump flow and why?

Choice 1: the pump flow is not optimal for adequate oxygenation
Choices 2 and 3 provide both adequate oxygenation with adequate kidney perfusion. Lower SaO2 could
be accepted as long as oxygen delivery is adequate.
Choice 2 may represent the best solution because at lower RPM, ECMO flow is achieving good organ
oxygenation.
Choice 4: the pump flow of 1400ml is achieved with high RPM and does not provide better oxygenation
than choices 2 and 3; thus it can lead to higher pressures and resistances in the circuit with consequent
hemolysis

Six hours after ECMO initiation, the patient demonstrates desaturations despite increased pump flow
and despite reducing oxygen consumption and red blood cell transfusion. You decide to perform prone
positioning. Prone positioning proves uneventful and oxygen saturation improves. But two hours later you
notice a progressive decrease in pump flow, and the arterial saturation drop again. Current assessment: BP:
93/55 (62) mm Hg, HR: 100/min, temperature: 36 C, urine output: 1ml/kg/h for the last two hours with
negative fluid balance. She is sedated and paralyzed.

What are the differential diagnoses and how do you rule them out?

Remember that pump flow is both preload and afterload dependent.

Perform a complete circuit check to rule out kink or occlusion in the venous or arterial limb;

327
Chapter 36

Result: Normal circuit check

Patient is not hypertensive or agitated; systemic vascular resistance is not elevated
Now you must rule out any cause of decreased preload and abnormal venous drainage:
Tension pneumothorax or cardiac tamponade. Perform chest radiograph and echocardiogram (possible
even in the prone position).
Also consider hypovolemia as prone positioning decreases RV afterload and enhances venous return. The
intravascular volume status may require optimization your after prone positioning with fluid infusion.

Chapter 21
A 7-day old, 4 kg infant with hypoplastic left heart syndrome returns to the Cardiac Intensive Care
(CICU) on VA-ECMO for low cardiac output following a Norwood Procedure (modified BT Shunt inser-
tion). Achieving hemostasis require prolonged effort post-operatively.
Current patient status:
Open sternum with central cannulation (14 Fr venous cannula RA and 10 Fr arterial cannula neoaorta).
Ongoing chest drain blood loss requiring replacement of 15-20 mL/hr. In the OR he received 20 mL/kg
of platelets, 2 units of cryoprecipitate and heparin was reversed with protamine.
ECLS: Centrifugal pump with a heparin coated circuit and PMP hollow fiber oxygenator set at 2400
RPMs with ECLS blood flow 600 mL/min (150 mL/kg/min). The venous pressure is -12 to -15 mmHg;
sweep gas flow is 0.3 L/min; circuit FiO2 is 0.30. Oxygenator transmembrane gradient is 15 mmHg. Heater
temperature set to 37°C.
Perfusion reports that no heparin is administered to the circuit or patient as ordered by the cardiac
surgeon. The CICU team orders target activated clotting times (ACTs) of 140-160 seconds and platelet
count of >100,000. As ECLS Specialist you complete your initial assessment, circuit check and routine
bloodwork. ACT is 130 seconds.

1. What is your initial concern?

The ACT is low, and there is risk that clot may start to form in the circuit even though the patient is bleeding
.
2. What are the risks when there is no anticoagulation for the patient on ECLS?
The main risk is circuit clotting – causing thrombosis (and potential embolization) with failure of oxygen-
ator and loss of support. Risk of clots increase with lower flows in neonates and children. Coatings
on circuits and cannulas may decrease anticoagulation requirements if used.

3. The order is for target ACTs of 140-160 seconds. What would your next steps be?
Start heparin infusion at 5-10 units/kg/hour immediately to achieve ACTs between 140-160 seconds.

Hematology lab results include platelet count: 63,000, fibrinogen level: 0.9 g/L, the standard heparin level
is 0.15 u/mL, and the anti-thrombin level is 0.37. The patient continues to require blood product re-
placement, a combination of frozen plasma (FP) and PRBCs, each infusing at 10 mL/hr into the circuit.

4. Should this patient receive additional blood products?

Transfuse 15-20 mL/kg of platelets to correct low platelet count in the post-operative bleeding patient. Ad-
minister cryoprecipitate (1 unit for every 10 kg of body weight) to maintain fibrinogen levels of >1.5 g/L.

Chest drain losses eventually taper off. The CVP rises from 5-6 mmHg to 18-20 mmHg and the sternal
patch appears to bulge outward. The low flow alarm rings on the ECLS console and measured ECLS

328
 Case Scenarios

blood flow is 200 mL/min. The negative venous pressure is fluctuating between -80 to -100 mmHg
and the circuit is chattering.

5. What is the most likely problem?

Loss of ECLS blood flow from decreased venous return and preload to the pump caused by cardiac tam-
ponade.

6. What steps will need to correct this problem and as the ECLS Specialist, what should you be prepared
for?
Additional volume may be required to ensure adequate preload and maintenance of ECLS blood flows.
Exploration and wash out of the mediastinum is indicated. The ECMO Specialist should be prepared
for further blood loss and order additional blood products. Expect ECLS blood flow to improve when
the sternum is evacuated.

Surgical exploration identifies and treats a small bleeding point around a suture line and following clot
evacuation ECLS flow improves. Chest drain losses fall to < 1 ml/kg/hr.
Over the next 4 hours the ACTs remain in the target 140-160 second range and heparin infusion is 10
units/kg/hour. Routine blood work demonstrates platelet count 78,000, fibrinogen 1.7 g/L, standard
heparin assay 0.28 units/mL and antithrombin level 0.38. ECLS Specialist notes new fibrin collec-
tions visible on connectors and the oxygenator in the circuit with a small clot appearing in the venous
drainage cannula.

7. Based on fibrin and clot formation in circuit, decreased bleeding and current bloodwork results, how
can anticoagulation be optimized?
Increase heparin infusion with an ACT target range to 180-200 seconds.
Consider antithrombin replacement with antithrombin concentrate or frozen plasma (see chapter 5). Some
centers routinely replace antithrombin (AT) to a level >0.5 which has improved anticoagulation man-
agement but not shown to prolong circuit integrity and can be expensive.
Discuss lowering platelet target now that bleeding has subsided to maintain >50-60,000 (but the low
range depend upon the institution.).

After discussion, the ACT target range was increased to 180-200 seconds so the Specialist increases the
heparin infusion to 20 units/kg/hour. Furosemide infusion commenced for diuresis.

8. What effect may diuresis have on the anticoagulation of this patient?

Heparin is cleared through renal excretion therefore an increase in urine output may result in a higher
heparin requirements.

The next day, the baby starts to awaken and move, requiring increased sedation for cannula security. Arte-
rial line trace shows improved pulsatility, systolic blood pressure is > 70 mmHg and the ECLS blood
flows have dropped to 100-120 mL/kg/min. The patient has palpable pulses with good perfusion. The
team considers performing a weaning trial.

9. What parameters can indicate that a weaning trial should be performed?

The patient should demonstrate significant improvement of the problem that necessitated ECLS support.
In this case, improved pulsatility on the arterial tracing implies augmented cardiac ejection. Cardiac
function can be assessed by echocardiography but findings are often difficult to interpret whilst on

329
Chapter 36

mechanical support. Clinical and biochemical (SV02/lactate) markers of adequate end organ perfusion
should correlate with other findings.

Upon direction of the medical team, the ventilation is increased and the Specialist decreases ECLS
pump flow to 75 mL/kg/min. The patient maintains acceptable mean blood pressure on epinephrine infusion
at 0.03 mcg/kg/min. After starting a separate heparin infusion to the patient, the ECMO circuit is clamped.
The patient maintained acceptable blood pressures. Echocardiography shows improved cardiac function.
The blood gases after the first 30 minutes of the weaning trial are acceptable with no increase in lactate
and the venous saturation is 78%. The trial is continued for 1 hour during which the patient demonstrated
no hemodynamic decline. The patient was decannulated.

Chapter 22
A 22 year old woman is on VV-ECMO for necrotizing pneumonia resulting in severe fibroproliferative
ARDS. She is cannulated via the right internal jugular and femoral veins. She has become progressively
hypotensive with hypoxia. A pulmonary artery catheter is inserted. The pulmonary artery pressure is 50/28
mm Hg (mean 38 mm Hg), CVP is 20 mm Hg. Cardiac index is 1.8 L/minute. The patient is on a milirone
infusion. ECMO settings include 5.0 L/min of flow with 3900 rpm, sweep gas flow of 10L/min. The nurse
tells you that the patient oxygen saturation is 80% and she is having episode of severe hypotension with
movement. Pressure control ventilator settings include a rate of 10/min, PEEP and plateaus pressures of
10 and 25cm H2O, respectively, achieving tidal volumes of 50cc. Oxygenation and hypotension worsen,
despite increasing flow, and with the increase in flow the cannula begins to “chug”. The patient is listed
for transplant.

Discussion:

1. What is a normal mean pulmonary artery pressure?

2. What is a normal cardiac output?
3. What is the patients underlying problem that is probably irreversible?
4. What drugs could be used to try to decrease pulmonary artery pressure?
5. Why would a double lumen cannula be inappropriate?
6. Should the patient be changed to VA-ECMO?
7. Can something be done with the ventilator to increase oxygenation?
8. Why would central cannulation right atrium to pulmonary artery be the appropriate treatment at this
time for this patient?

Chapter 23
A 55-year-old female is on day 6 of VA-ECMO for post myocardial infarction cardiogenic shock. She
is cannulated with a left femoral arterial catheter and a right jugular central line. Her ECMO parameters
include pump flow of 2.5L/min, RPM of 3890, sweep gas flow of 5 L/min and FiO2 of 0.40. Her heart has
been recovering slowly. She was extubated two days before without complication. This morning as she
is having breakfast she calls to describe shortness of breath. She appears dyspneic and cyanotic but her
saturation is 100% on the monitor. The intensivist asks you to draw an ABG and the result are similar to
the previous ones, normal with a PaO2 of 360 mmHg.

330
 Case Scenarios

1. What is happening to the patient?

This is a case of dual circulation or what is also called North/South syndrome or Harlequin syndrome.
The ABG samples have been drawn from a femoral arterial line and do not reflect cerebral oxygenation.
The pulse oximeter is probably on the left hand which explains why the arterial saturation of the patient
is 100%. The proper cerebral saturation will be shown if the nurse places the pulse oximeter on the right
hand and if the patient ABG is drawn from that hand too.

Chapter 24
Paul is a 2 month old former premature infant with RSV weighing 4.5kgs. He was cannulated to VVDL
ECMO with a 16Fr Origen cannula 18 days previously. His course has been complicated by episodes of
pulmonary hemorrhage and renal dysfunction requiring CVVH for fluid overload. Weaning has begun
and his ECMO settings on the centrifugal pump include flows of 500mls, RPM 3700, sweep gas flow 0.7
L/m and FiO2 of 0.55. Despite clearing of lung fields on the chest radiograph he has a persistent large
plural effusion on the left side which is thought to be precluding further weaning. A decision has been
made to insert a chest tube. What are the responsibilities of the ECLS specialist in preparing the patient
for the procedure?

Discussion

What should the platelet and fibrinogen parameters be for the procedure?
What ACT/aPTT/Xa levels should we target, should heparin infusion be paused?
What emergency drugs and volume are prepared, attaching IV line extension?
Which blood products should be available in ICU blood fridge or in cool box at the bedside?
Positioning of patient and placing of diathermy pad.
Ensuring adequate sedation and muscle relaxants given.
Ensure ventilation and ECLS settings are appropriate for procedure.
Ensure immediate access to circuit clamps / air removal box if ECLS emergency is declared.
Monitor patient and respond to any circuit issues during procedure.

The ECLS specialist checks the complete blood count and coagulation levels. Platelets are augmented
to 120,000/mm3 with 15mls/kg and fibrinogen to 2 g/L with cryoprecipitate 5mls/kg. A discussion about
heparin management with the physician team results in the heparin infusion being reduced from 40u/kg/
hr to 30u/kg/hr to achieve an ACT target of 160-180 and anti-Factor Xa level of 0.3-0.5. The heparin is
not paused as the circuit is 18 days old with some small areas of fibrin. All emergency drugs are prepared
and boluses of fentanyl and rocuronium are given. A unit of PRBC is placed in a cooler at the bedside, a
surgical time-out is performed before commencing. During the procedure the patients SpO2 drifts from
94% to 83% and so the sweep gas blender is increased to 100%. On tube placement 30mls of serosangui-
nous fluid is drained. Over the next few hours 10 – 15mls/hr is drained which becomes increasingly blood
stained despite ACT of 165 -172, but by hour 6 this has improved with nothing drained in the last 2 hours.
Paul becomes increasingly tachycardic to of 180, mean arterial pressure falls to 40, HCT is 25% despite
a blood transfusion of 20mls/kg. The circuit venous access pressure drops to -40. Patient saturations are
85% and venous saturations 84% with a sweep gas FiO2 of 100%.

Discussion

What do you think is causing this deterioration?

331
Chapter 36

Why is the reduction in chest drain losses significant?

Why has the blood transfusion not had the expected effect on HCT?
What is the significance of an increasing venous saturation, decreasing patient saturation, decreased perfu-
sion and drop in circuit venous pressures? Would this scenario present differently in the VA patient
If this situation is not resolved what effect might it have on circuit blood flow?
On examination, it is clear that the appearance of the chest is unequal and air entry reduced to the left side.
The chest drain is found to be blocked with clot and on clearing begins to drain again with improvement
in saturations. Further infusions of PRBC, platelets and cryoprecipitate are given which improve blood
pressure and HR decreases to 145. A loading dose of tranexamic acid is given and then an infusion of
45mg/kg/24hrs. With this management the bleeding resolves. Drainage of the effusion allows for more
rapid weaning of the sweep gas and Paul is ‘capped off’ 26hrs later for 2hrs and then decannulated.

Chapter 25
You are caring for a 54 year old man with a diagnosis of H1N1 Influenza A. He is day 9 on VV-ECMO
and is cannulated with a DL VV Cannula. He weighs 124 kg. ECMO flow is 4000 ml/min on 3600 rpm.
He is on CPAP/PS and is breathing spontaneously through a tracheostomy on FiO2 0.3. His tidal volumes
have improved from 150 ml to 500 ml over the last 3 days. His latest blood gas shows a pH 7.36, PO2 68
mmHg and a PCO2 of 48 mmHg. He is on no inotropes. His chest x-ray is shown below.

1. Is this patient suitable for trial off?

Yes-The expected duration for a patient with H1N1 Influenza A is approximately 10 days. Assuming
the diagnosis is correct then it would not be unreasonable to commence a trial off. His blood gas is good
although he remains on significant ECMO flow. He has shown improvement in chest compliance with im-
provement in TV to 500 ml. Although significant bilateral infiltrates remain the general course has been
one of improvement and the appearance would not deter a trial off.

2. How is this best achieved and what changes to ventilating settings are required?
Although he remains on 4000 ml/min ECMO flow he has shown signs of improvement and weaning
towards minimal flow of around 750 ml/min does not offer any advantage and could unnecessarily increase
time on ECMO with an increased risk of circuit thrombosis. He is breathing on CPAP/PS and a spontane-

332
 Case Scenarios

ous mode of ventilation is likely to maximize ventilation so this should be continued. To compensate for the
loss of ECMO the FiO2 should be increased to 0.5-0.6 and the trial off commenced. With VV-ECMO the
sweep gas is simply disconnected and if the peripheral arterial saturations remain stable then a blood gas
performed at 20 min. Decannulation should be performed with the aid of local anaesthetic and increased
sedation/paralysis avoided.

Chapter 26
Mary comes to your outpatient clinic for her routine follow-up visit at 8 years of age. Her medical
history reveals that she was born after an uncomplicated gestation of 41 weeks. Directly after birth, she
developed severe respiratory insufficiency due to meconium aspiration syndrome and was intubated.
Conventional mechanical ventilation failed. She was treated with VA-ECMO from the age of 8 hours.
The ECMO run was uncomplicated and decannulation occurred after 72 hours. She was extubated at 6
days of age and discharged home without supplemental therapy 11 days after birth. Prior to and during
ECMO serial cranial ultrasounds were normal. Also at discharge, no abnormalities were seen at a routine
cranial ultrasound.

Routine visits at 6 months, 1 and 2 years of age showed favorable outcomes; she grows well, has
no respiratory symptoms and she reaches normal neurodevelopmental milestones. At 2 years of age, her
score on the cognitive index of the Bayley Scales of Infant and Toddler Development was above average.
She scored above average on her routine IQ assessment at 5 years of age. She is a bright young girl, plays
sports and is healthy.

At her routine follow-up visit at 8 years, growth and lung function are normal. She tells the physical
therapist that she plays hockey twice a week and can keep up with peers well. Her motor function perfor-
mance and maximum exercise tolerance are both within the normal range. However, she has developed
severe memory deficits that hamper her daily life activities and school performance. The memory deficits
cause great distress for Mary as she fears people will think she is not smart. Her parents do not understand
why she cannot remember what she did last week or the homework that is due tomorrow. While she was
doing well in primary school as a younger child, both her parents and teacher notice that the schoolwork
has become increasingly difficult for Mary. For instance, she requires additional tutoring for math and
needs to work in a quiet, stimulus free environment as much as possible. Mary reports her school function-
ing to be below average. On neuropsychological assessment, verbal and visuospatial memory recall are
significantly below the norm, despite the fact that she has an above average IQ. On questionnaires, she
reports feeling less competent than others in school, but to have good self-confidence in general.

Chapter 27
A 20 year-old woman with a history of cystic fibrosis and multiple prior acute exacerbations presented
to an outside hospital with acute respiratory failure requiring urgent intubation. She was diagnosed with
H1N1 influenza-induced ARDS, intubated, and referred to your ECMO center due to refractory hypoxemia
and hypercapnia. On arrival, she was quickly cannulated for venovenous ECMO using a double-lumen
IJ cannula without complication.
Prior to her acute presentation, the patient had been physically active and regularly participated in
outpatient physical rehabilitation, including stationary biking, in preparation for eventual lung transplan-
tation. Approximately 24 hours after cannulation, she received a tracheostomy, without complication, in
anticipation of ambulation to allow for direct airway access in the unlikely event of a catastrophic pump

333
Chapter 36

event. The next day sedation was weaned, and she was fully awake and comfortable with only occasional
PRN analgesia. She was then able sit at the edge of the bed with assistance. With consultation among the
ICU team, the lung transplant team, the patient, and her family, the decision was made to list for lung
transplantation and to proceed with an active rehabilitation program, including ambulation, to optimize
her conditioning while awaiting transplantation. Two days later, with the help of a physical therapist, two
nurses, a respiratory therapist, and an ECMO specialist, the patient ambulated within the hospital room.
She continued to ambulate twice daily with increasing distance. Prior to transplantation, she was able to
ambulate over 700 feet with a walker in each single session. She remained on VV-ECMO until her suc-
cessful lung transplant on hospital day 16. The patient was discharged home on postop day #18.

Discussion

The importance of gradually weaning sedation, beginning activity, and increasing the amount of activity.

Chapter 28
A previously normal and healthy 12 yo, 50 kg girl develops acute respiratory failure from
influenza complicated by staphylococcal pneumonia. She has received acetaminophen and
ibuprofen for her fever, broad spectrum antibiotics including vancomycin for S. aureus, and
anti-influenza treatment. She is cannulated onto VVDL+V ECMO for refractory hypoxemia
after failing high frequency oscillating ventilation with FiO2 of 1.0, MAP 35 that generated an
oxygen index of 37. In her 24 hours of ICU admission prior to cannulation, she has had 4300
mL intake, and 3200 mL output. Her creatinine has risen from 0.4 mg/dL to 1.2 mg/dL.

1. Does this patient have acute kidney injury? What are modifiable risk factors that may help improve
her renal injury?
Yes, the rapid rate and degree of creatinine rise in this patient is consistent with acute kid-
ney injury and meets modern definitions. While not meeting the ELSO complication definition
of renal failure (creatinine > 1.5 mg/dL) literature in adults has demonstrated that a creatinine
rise as small as >= 0.3 mg/dL is associated with increased mortality. Consideration of add-
ing RRT to her ECLS circuit is reasonable at this point. One should also consider discussing
the medications she is receiving. The physican and pharmacy teams may need alter medicals,
change dosing or stop some medicatons completely in the face of kidney injury. In this case, a
discussion of potential alternative agents for vancomycin and ibuprofen should be entertained.

2. Does the patient have fluid overload at the time of ECMO initation? How much? How would you treat
it?
Yes, the patient has fluid overload at ECMO initiation. The formula for calculating percent
fluid overload is FO% = (total fluid in – total fluid out)/admission body weight. For this child,
that is 22% fluid overload at ECMO initiation. Options for treating fluid overload at this point
could include fluid restriction, pharmacologic therapy with diuretics, or mechanical removal
using one of the RRT techniques. Since the patient has evidence of co-existing AKI, many cen-
ters would consider initiation of CRRT.

334
 Case Scenarios

Chapter 29
A 10 year old, 34 kg, previously healthy girl presents with altered mental status. Her parents had noted
yellowing of her eyes 2 days prior and were unable to arouse her from sleep. She is emergently intubated in
the ED and brought to the ICU. Labs obtained are consistent with fulminant hepatic failure due to Wilson’s
disease. Throughout the day she has escalating ventilator settings with poor oxygenation, fluid overload
and decreasing urine output. The decision is made to place her on VV-ECMO with concomitant CRRT for
acute hypoxic respiratory failure refractory to conventional mechanical ventilation and acute kidney injury.

1. According to ASFA (7th SI) guidelines (chapter reference 3), fulminant Wilson’s disease fits which
category indication and which type of therapeutic apheresis:
a. Category III, Therapeutic Plasma Exchange
b. Category III, Leukocytopheresis
c. Category I, Therapeutic Plasma Exchange
d. Category I, Leukocytopheresis
e. Category I, Extracorporeal Photopheresis

2. On a centrifugal VV-ECMO circuit the optimal arrangement for the CRRT filter and TPE device are:
a. In parallel, drainage limb post pump and return limb to the ECLS venous drainage line
b. In series, drainage limb post pump and return limb pre-venous ECLS circuit
c. In parallel, drainage limb pre pump and return limb to the ECLS venous drainage line
d. In series, drainage limb pre pump and return limb post pump/pre-oxygenator on ECLS circuit
e. In series, drainage limb pre-pump/pre-bladder and return limb pre-pump/post-bladder on ECLS
circuit

Answers
1. d
2. a

Chapter 30
A previously healthy 18-year-old female was admitted to hospital with dehydration secondary to diar-
rhea and vomiting with associated colicky abdominal pain that started 24 hours prior. Over the course of
the next 24 hours her biochemical indices demonstrated worsening hepatitis and acute kidney injury. Her
condition continued to deteriorate and she developed signs of encephalopathy. By report from friends, the
patient was foraging for edible mushrooms and herbs 2 days before admission. A mycological examina-
tion of her gastric washings was positive for Amanita phalloides spores. Given the presence of apparent
toxin mediated organ failure, five daily consecutive albumin dialysis procedures were performed in hope
of either supporting hepatic recovery or as bridging to liver transplantation. After the first liver albumin
dialysis, her symptoms and biochemical indices improved. Over the course of the remaining treatments
she progressively improved. Her liver and renal dysfunction resolved over the next 10 days.

Chapter 35
You have just been asked to summarize the cannulation configuration for the last patient put on. The
baby is a 1 day old boy (GA 42+1, 3.540 gr) with meconium aspiration syndrome but his CRP is rising
and circulation unstable, thus is started on a dopamine infusion. The patient is on VA ECMO, the venous

335
Chapter 36

cannula is a 12 Fr Bio-Medicus (lighthouse tip) cannula and the return cannula 10 Fr, also a Bio-Medicus
but with a blunt tip. The cannulae are both on the right side of the neck.

1. Simplest abbreviation described with the Maastricht Treaty nomenclature?

V-A (Vein to via the membrane lung to Artery)
2. What is then the most informative abbreviation for the same configuration?
V12jr-A10cr, (Vein12 Fr, jugular right, to Artery, 10 Fr, carotid right.

Niels is a Danish baker, heavy smoker, weighing 300 lbs (129 kg). He was placed on VV support for
four days but then an additional 19 Fr cannula was inserted into the left femoral artery for additional cardiac
support (echocardiography showed an estimated cardiac index of 2.1 L/min per m2). He already had an
arterial line for pressure measurement in his left dorsalis pedis artery and pulsatile flow was maintained
after cannulation via the ipsilateral femoral artery.

3. From the above information, how would you abbreviate the configuration?
V-VA19fl, Vein to Vein was all you knew from start. Then a 19 Fr return cannula was added in the left
femoral Artery. There is no distal perfusion catheter, just the pressure monitoring on the foot.

Since you are interested and have the ELSO Registry responsibility, you search for data to describe the
configuration in as much detail as you can. You find the following additional information: Both preceding
cannulae are in the contralateral right groin. The drainage cannula is a 29/50 with the tip high up kissing
the roof of the atrium. The return cannula has an outer diameter of 21 Fr.

4. Present configuration?
V29/50fra-V21frA19fl

336
37

Chapter Questions

Chapter 2 (Choose all that apply)

1. Oxygen content depends on which of the following?
a. Hemoglobin concentration
b. Patient temperature
c. Dissolved oxygen
d. PaO2
e. PaCO2

2. Which of the following is true of the arterial-venous oxygen difference (A-V DO2)?
a. Depends on oxygen delivery.
b. Is reflected in circuit venous saturation with VA-ECMO
c. Can be decreased by increasing sweep gas flow
d. Depends only on ECMO pump flow
e. May be decreased by cooling a patient

3. Factors that increase tissue unloading of oxygen include which of the following?
a. An increase in pH
b. An increase in 2, 3 bisphosphoglycerate
c. A decrease in patient temperature
d. A decrease in PCO2

4. Improved oxygen delivery of a patient receiving ECLS can occur with

a. Young circuit
b. Improvement in native lung function.
c. Increased pump flow with venovenous ECLS.
d. Increasing pump flow above the rated flow of the oxygenator

5. Carbon dioxide clearance by the ECLS pump in influenced by

a. Pre-oxygenator bicarbonate concentration
b. Sweep gas flow
c. Membrane surface area
d. Sweep gas O2 concentration
e. Sweep gas CO2

337
Chapter 37

Chapter 3 (Choose the best answer)

1. Which of the following is NOT a direct determinant of oxygen delivery to the tissues?
a. Cardiac output
b. Hemoglobin
c. Arterial oxygen saturation
d. Central venous pressure

2. Ejection of blood from the heart is considered afterload _________ and preload _________.
a. Dependent ; dependent
b. Sensitive ; dependent
c. Sensitive ; sensitive
d. Dependent ; sensitive

3. Which of the following can be used as a tool to assess adequacy of perfusion?

a. Tissue oximetry
b. Serum lactate concentration
c. Venous oxygen saturation
d. All of the above

4. VA-ECMO increases afterload to the:

a. Right ventricle
b. Left ventricle
c. Both
d. Neither

5. The cardiac output curve can be affected by afterload, contractility and heart rate. Assuming that
vascular resistance does not change, cardiac output is also related to:
a. Hemoglobin
b. Central venous pressure (CVP)
c. Vessel radius
d. PaO2

Chapter 5 (Choose the best answer)

1. In order to safely insert the venovenous cannulae, the surgical team wishes for her platelet count to
be above 30k/dL. What is the appropriate dose and modifications?
a. 20 ml/kg of irradiated platelets should increase her platelet count by 20,000/dL
b. 10 ml/kg of irradiated platelets should increase her platelet count by 30,000/dL
c. 1 unit of irradiated platelets should increase her platelet count by 50,000/dL
d. 10 ml/kg of non-irradiated platelets should increase her platelet count by 10,000/dL

2. Which of the following is the mechanism of irradiation performed on cellular blood products?
a. Breaks DNA in leukocytes so they cannot proliferate in an immunocompromised recipient
b. Kills cytomegalovirus (CMV) so that it cannot infect an immunocompromised recipient
c. Kills any bacteria in platelet units since they are at such high risk of being contaminated

338
 Chapter Questions

d. Removes cytokines thereby preventing febrile non-hemolytic transfusion reactions

3. What blood type of platelets will be compatible with this patient’s blood type?
a. Group O/RhD positive
b. Group B/RhD positive
c. Group O/RhD negative
d. Group AB/RhD positive

4. The PICU team also wishes to increase her oxygen carrying capacity with one 15ml/kg RBC transfu-
sion. The appropriate unit and dose are ordered and the unit is delivered. You notice that the unit was
collected one month ago and will expire in 10 days. What should you do?
a. Send the unit back since it is old and older blood has been shown to increase mortality
b. Send the unit back since it is old and older blood has been shown to increase transfusion reactions
c. Follow hospital transfusion policy to transfuse her the blood since older blood has not been shown
to change clinical outcomes
d. Throw the blood away and request that a fresh unit be delivered

Chapter 7 (Choose the best answer)

1. Calculate the oxygenation index for Baby C who is currently on HFOV with a MAP of 16, Amp of
42, Hz of 8 and 100%. Blood gas results as 7.16/52/32/-8
a. 32
b. 45
c. 50
d. 64

2. This patient is a reasonable candidate for ECMO, but has several risk factors that increase the odds
of morbidity and mortality. Which of the following is NOT a risk factor?
a. Acidosis
b. History of CPR
c. Gestational age
d. Chronological age
e. Diagnosis

3. What survival rate would you predict for this baby?

a. 40%
b. 51%
c. 70%
d. 92%

Chapter 8 (Choose the best answer)

1. A 7-year old female has just undergone a double switch operation for congenitally correct transposition
of the great arteries and now needs ECMO support for failure to wean from cardiopulmonary bypass.
Which of the following is least important in the management of this patient?
a. Ensure that there is adequate perfusion of end-organs while on ECMO support
b. Evaluate for potential residual lesions

339
Chapter 37

c. Evaluate for arrhythmias while on ECMO

d. Ensure left heart is adequately decompressed

2. A 5 day old male with hypoplastic left heart syndrome has just undergone the Norwood operation with
BTT shunt and is currently in the intensive care unit. He experiences a pulseless electrical activity
arrest and is placed on ECMO (ECPR). On ECMO, he has poor perfusion, rising lactates and no urine
output. The next step in his management is:
a. Increase ECMO flows to as high as 200mL/kg/min
b. Evaluate for left heart decompression
c. Partially clip the BTT shunt
d. A and C
e. B and C

3. A 14-year old male is admitted with increased work of breathing and poor perfusion after a 2 week
history of upper respiratory symptoms. An echocardiogram reveals a non-dilated heart with severely
decreased biventricular function. His clinical exam becomes progressively worse with worsening
perfusion, increasing lactate and increasing vasopressor requirement. Which of the following state-
ments is least likely to be true?
a. This is likely myocarditis and he has a high likelihood of clinical improvement and should be
supported with ECLS, if necessary
b. This is likely restrictive cardiomyopathy, which is associated with a good prognosis and should
be supported with ECLS, if necessary
c. The underlying mechanism of his clinical decline is decreased cardiac output from decreased
contractility and decreased ventricular compliance
d. This patient is at risk for cardiac arrest and may require ECPR

4. A 10-year old female presents in extremis with what appears to be a dilated cardiomyopathy. She is
cannulated through the femoral vessels for VA ECLS. Which of the following findings would suggest
that she may require decompression of her left heart?
a. She has a pulse pressure of 30 points in an arterial line in her right hand
b. Her left ventricle and left atrium are not dilated on echocardiogram
c. She has copious, pink frothy secretions from her endotracheal tube
d. The aortic valve opens regularly on echocardiogram

5. A 60-year old male with terminal, metastatic pancreatic cancer and a recent cerebral hemorrhage is
admitted with acute decompensated heart failure. Which of the following statements is most likely
to influence decision-making for ECLS?
a. He may not be a candidate for cardiac transplantation or ventricular assist device
b. He may require decompression of the left heart
c. He may have differential oxygenation with different saturations between his right upper extremity
and lower extremity
d. If he is cannulated via the femoral vessels, he may require a distal perfusion cannula to supply
blood to his lower extremity

6. Which of the following candidates is likely to wean from ECLS?

a. A neonate with total anomalous pulmonary venous return who has a pulmonary hypertensive crisis
before the operative repair

340
 Chapter Questions

b. A 16 year old female with idiopathic pulmonary hypertension with a two year history of slowly
worsening disease
c. A 13 year old female with viral myocarditis and elevated right ventricular pressures
d. A 2 year old with an unrepaired ventricular septal defect and concurrent pertussis infection

Chapter 9 (Choose the best answer)

1. What mode of ECMO is most often used for a patient undergoing ECPR for cardiopulmonary resus-
citation in the out-of-hospital setting?
a. Veno-venous, femoral
b. Veno-arterial, femoral
c. Veno-arterial-venous
d. Veno-venous, neck double lumen

2. Patients supported with ECPR in the context of cardiopulmonary arrest are more or less likely, or
similarly, likely to suffer neurologic complications compared to ECMO patients without a cardiopul-
monary arrest?
a. More
b. Less
c. Similarly
d. Unknown

3. Patients supported on ECMO following cardiopulmonary arrest are at risk of developing pulmonary
edema and hemorrhage associated with impaired left ventricular dysfunction and left atrial hyperten-
sion. Measures that should be considered to address this problem would include:
a. Increasing the FiO2 and PEEP on the mechanical ventilator
b. Decreasing the anticoagulation
c. Providing slow continuous ultrafiltration and diuretics
d. Undertaking a left atrial vent either by image guided atrial septostomy or surgical septectomy.
e. Provide time and allow the problem to self resolve

4. CPR quality is important in patients who undergo ECPR during which period?
a. Before launching ECPR
b. During period of cannulation for ECPR
c. Irrelevant once ECPR is launched
d. Before launching and during period of cannulation for ECPR

Chapter 10 (Choose the best answer)

1. ECLS devices offer an advanced technology and supplemental capacity during early resuscitation after
trauma that is superior to conventional modalities, especially
a. Lung-protective ventilation
b. Providing circulatory support
c. Volume infusion
d. Rewarming techniques to prevent or to correct hypothermia

341
Chapter 37

2. Trauma patients have a high risk of trauma-related complications. They are often hemodynamically
unstable and have a high risk of multi-organ failure. ECLS should not be considered in which of the
following clinical conditions:
a. Blunt chest trauma with injury of the trachea-bronchial tree and hemorrhagic shock
b. Spine and pelvic injuries because of the need for further surgical procedures
c. Traumatic brain injury with increasing intracranial pressure
d. In the presence of futile medical care and irreversible conditions known to be incompatible with
life

3. ECLS is a balancing act between the need for anticoagulation and the persistent risk of bleeding.
Heparin-free ECLS may be considered, except in the case of
a. The risk of intracerebral bleeding following traumatic brain injury
b. VV-ECLS in acute respiratory failure after pulmonary contusion in blunt chest trauma without
further co-injuries or coagulopathy
c. Injury of the trachea-bronchial tree with subsequent parenchymal bleeding resulting in inadequate
gas exchange
d. Profound hypoxemia and impairment in gas exchange related to mass transfusion and hemorrhagic
shock

4. A 29 year old woman with a history of asthma reports a 3 day history of flu like symptoms. She is
currently 25 weeks gestation with a singleton fetus. Soon after admission she collapses in the waiting
room. On review by the medical team her BP is 80/40, heart rate 150 bpm, respiratory rate is 35/min,
her oxygen saturations are 70% on 5L mask oxygen, her temperature is 39.5℃. A chest radiograph
is consistent with ARDS. Viral swabs are taken and she is transferred to ITU. Despite maximum
respiratory support she remains hypoxemic. What would you advise?
a. Nurse patient prone
b. Emergency Caesarean section
c. Discussion with the ECMO center for VA-ECMO
d. Await viral swab results before starting antiviral treatment
e. Commence a course of high dose steroids.

5. A pregnant patient on ECMO should be nursed

a. Prone
b. Sitting up with a tilt of 20 degrees
c. With a right lateral tilt of 20 degrees
d. With a left lateral tilt of 30 degrees
e. Supine

6. A patient on ECMO who is 33 weeks gestation has a blood pressure of 160/100 you should
a. Increase sedation as she is very restless.
b. Check urine for protein
c. Commence IV methyldopa
d. Arrange emergency delivery
e Increase heparin as her aPTT ratio is low

7. A patient on ECMO who is 23 weeks gestation is found to be diagnosed with fetal death soon after
commencing ECMO you should advise
a. Urgent caesarean section to deliver the fetus

342
 Chapter Questions

b. Caesarean section once the patient is stabilized

e. Administer mifepristone (antiprogesterone) run a heparin free circuit
d. Administer mifepristone (antiprogesterone) with a heparinised circuit
e. Wean off ECMO put on conventional ventilation until fetus is delivered
f. Continue ECMO until respiratory status improved then deliver fetus

8. Of patients with cancer who receive ECLS, the highest risk of death is those patients with:
a. Solid organ (lung, liver, etc.) cancers
b. Allogeneic hematopoietic stem cell transplantation
c. Hematologic (leukemia, lymphoma) cancers
e. Pediatric leukemia

Chapter 12 (Choose the best answer)

1. Patient is place on peripheral VA-ECMO support as part of his eCPR resuscitation. Transthoracic
echocardiogram demonstrates minimal LV contractility and LV distention. A possible next step to
address this is:
a. Addition of inotrope
b. IABP
c. LV vent via left thoracotomy
d. a and c
e. a, b, and c
f. a, b, c and d

2. To avoid Harlequin Syndrome, cannulation for VA-ECMO support should be configured with arterial
inflow in
a. Superficial femoral artery
b. Axillary artery
c. Aorta
d. b and c
e. None of above

Chapter 13 (Choose all that apply)

1. A one-year old 11 kg boy was put on VV-ECMO yesterday due to hypoxemic respiratory failure from
a pneumonia. After cannulation with a 16 Fr DLC (Cavo-atrial design) and commencement of ECMO
the patient was stable with good hemodynamics and oxygenation. Now, a day later severe pumping
problems occur more, or less constantly with drainage difficulties. What could be the cause?
a. Loss of tidal volumes, i.e. cephalad movement of the diaphragm
b. The patient is hungry
c. Hypovolemia due to fluid withdrawal
d. The ECMO pump rpm is increased

2. The same patient as above. Two weeks later still VV-ECMO, 16 Fr DLC, QEC 130 mL/kg per min.
ABG shows: pH 7.21, pCO2 45 mmHg (6 kPa), pO2 68 (9), BE -8, Hgb 1.2 g/dL, lactate 3.2 mmol/L.
A simultaneous PREML-BG: SPREO2 78%, SPRECO2 48 (6.3). He is communicating with his mom but

343
Chapter 37

a bit more tired though sedation is “low”. He is fed 50-50 with enteral and parenteral nutrition (EN/
PN), his sclerae appear more “yellowish”. What is your assessment now?
a. Inferior oxygen delivery to metabolic demand
b. Low difference in pCO2 over the ML, time for a Trial-off
c. Hemolysis
d. All of the above

3. The following factors influence oxygenation on V-V ECMO with no native lung function
a. Ventilator FiO2
b. Recirculation
c. Cardiac output
d. Hematocrit

4. An adult woman, 76 kg (167 lbs), with a primary viral infection and a secondary bacterial pneumonia
with complete opacification of both lungs (Murray* 3.75 [Murray JF, Matthay MA, Luce JM, Flick MR.
An expanded definition of the adult respiratory distress syndrome. Am Rev Respir Dis.1988;4(3):720–
723]), has now been on VV-ECMO for six days. Her tidal volumes have increased slightly but the
sweep flow remains high. Her fluid balance was slightly negative over the past few days. You observe
a decrease in her CRP and base excess but trends of increasing liver and kidney markers. A CT-scan
of the thorax is scheduled for the day. What other considerations would you entertain?
a. Recruit the lung and increase the ventilator settings
b. Increase fluid withdrawal, consider CRRT
c. Echocardiography to assess the heart with special attention to right ventricular function
d. Consider dual oxygenators since she has exceeded the rated flow

5. You are caring for a one year old with pneumonia on VV support. Pump flow is 1,100 mL/min. Sweep
gas is at 0.60 FiO2, 1.0 LPM. Ventilator settings are PIP 24, PEEP 6, Rate 10 FiO2 0.21. An ABG
drawn from the radial arterial line reveals pH 7.26, PaCO2 62 mmHg (8.3 kPa), PaO2 75 (10), HCO3
22.4 mmol/L and the cephalad venous saturation is 81% with a measured flow of 600 mL/min. What
would you do?
a. Increase bypass flow
b. Increase the FiO2 of the sweep gas
c. Increase the sweep gas flow
d. Increase the ventilator rate

Chapter 14 (Choose all that apply)

1. Which of the following may be associated with venoarterial ECMO support
a. Progressive pulmonary edema
b. Pulseless circulation
c. Normal Pulmonary blood flow
d. Hypoxia due to recirculation
e. Differential Hypoxia

2. Which of the following is/are true for venovenous ECMO support

a. Is always the first mode of support for severe pneumonia

344
 Chapter Questions

b. Provides oxygenation and ventilation to the pulmonary circulation and therefore prevents the
development of right ventricular failure
c. Is associated with different oxygen tensions around the systemic circulation
d. Maintains arterial oxygen saturations above 85% in severe respiratory failure
e. Recirculation is reduced with cavo-atrial circuit flow

3. Regarding the use of V-VA support for severe pneumonia, which are true
a. It is commonly required
b. Is a suitable mode if there is severe cardiac failure
c. Requires monitoring of flow in the arterial return
d. Relies on native cardiac output
e. Is associated with lower survival

4. 12 hours following the insertion of an implantable durable LVAD, your patient becomes progressively
hypotensive despite fluid and vasopressors and the LVAD continually alarms with “Low Flow”. Which
of the following are true
a. An echocardiogram can wait until morning. This is usual
b. If this is due to RV failure, then VA-ECMO would be the first choice for mechanical circulatory
support
c. The use of V-PA ECLS is a reasonable choice for RV support
d. The use of V-PA ECMO will prevent lung injury
e. The use of V-PA ECMO is indicated if there is associated respiratory failure

Chapter 15 (Choose all that apply)

1. All the following are indications to ECCO2R except:
a. Bridge to transplantation
b. COPD
c. Mild ARDS
d. Ultraprotective ventilation strategy

2. Indicate the two more frequent complication of ECCO2R

a. Bleeding
b. Tubing rupture
c. Membrane clotting
d. Infections

3. Which of the following apply to ECCO2R

a. Is a high flow extracorporeal device
b. Needs anticoagulation
c. Runs with blood flow ranging between 1500-3000 ml/min
d. Improves oxygenation

4. ECCO2R in COPD
a. Prevent NIV failure in patients at risk.
b. Avoid IMV when NIV has failed.

345
Chapter 37

c. Wean COPD patients from IMV.

d. All previous answers

Chapter 16 (Choose the best answer)

1. When evaluating a new ECMO console which of the following is not a factor to consider?
a. Programmatic goals
b. Staffing plan
c. Widespread console usage
d. Integrated safety features
e. Cost

2. Shunts in the ECMO circuit may facilitate all the following situations except:
a. Blood sampling
b. Increased ECMO flow to the patient
c. Retrograde ECMO flow
d. Decreased ECMO flow to the patient
e. Increased flow through the oxygenator

3. The following is true of surface coatings common in ECMO circuits:

a. Surface coatings increase the anticoagulation of the ECMO patient
b. Surface coatings may make the ECMO components more biocompatible
c. ECMO surface coatings should be heparin based for maximum effect
d. All of the above are true

Chapter 17 (Choose all that apply)

1. Which of the following factors must be considered when selecting the type and size of ECMO cannula.
a. The anticipated peak ECMO blood flow
b. The temperature to be set on the heater unit during ECMO support
c. The height of the patient in comparison to the ECMO pump device
d. Preexisting congenital cardiac pathology

2. Concerning VV-ECMO, which of the following is true:

a. Upon insertion, the effective position of the dual lumen Avalon Elite® cannula may be verified
by A-P chest radiograph
b. Increasing ECMO blood flow may be effective in providing a greater effective blood flow to the
patient, however, this strategy eventually reaches a limiting point where it is no longer effective in
raising arterial oxygen saturation
c. The Avalon Elite® dual lumen cannula is normally inserted via the right internal carotid artery
d. Recirculation does not occur in the setting of dual site femoral-internal jugular VV ECMO

3. The following are true concerning the centrifugal pump, except:

a. The rotating impeller or cone creates a constrained vortex which serves to provide the negative
pressure required to draw blood into the pump
b. The rotational speed of the impeller (RPM), is inversely proportional to pump flow

346
 Chapter Questions

c. Hemolysis is defined by ELSO as the appearance of a plasma free hemoglobin concentration in

excess of 50 mg/dl
d. The centrifugal pump is a non-occlusive pump

4. Concerning the ECMO Oxygenator, which of the following is/are true:

a. The set gas flow is inversely proportional to CO2 clearance
b. The set FiO2 is inversely proportional to the post-oxygenator PO2
c. CO2 clearance is relatively independent of blood flow
d. The “ideal flow” is achieved when the desired blood flow is greater than the rated blood flow for
the oxygenator

5. Which of the following is true, with regard to the components of the ECMO circuit:
a. TheBetter Bladder™ serves as a continuous renal replacement device in the setting of renal failure
b. The blood flow produced by a roller pump may be reduced by increased afterload
c. The use of high flow centrifugal pumps with ¼” circuits for pediatrics, may expose formed elements
to turbulence and shear stress, and contributes to blood cell damage and local thrombus formation
d. Centrifugal pumps are nonocclusive pumps and are not capable of producing a great enough nega-
tive pressure to cause cavitation

Chapter 18 (Choose the best answer)

A 23-year-old male is cannulated to VA ECLS via the right femoral vessels after suffering cardiac
arrest following a submersion injury.

1. Six hours post cannulation his right foot is noted to be dusky with very faint doppler pulses. What
intervention would you expect next:
a. Continue to monitor perfusion to the limb
b. Turn up the flows to improve perfusion
c. Placement of a reperfusion cannula
d. None of the above

2. On day 2 of ECLS the patient shows improved pulsatility on his arterial line but CXR shows bilateral
white out. Cerebral NIRS have trended down into the 40s (from the 60s) so an arterial blood gas is
obtained from the right radial arterial line. The PaO2 is 42 on the gas yet the pulse oximeter on the
left toe shows a SpO2 of 98%. What explains these findings:
a. The oxygenator is failing
b. The flow is inadequate to support this patient
c. Improved cardiac function has resulted in the patient developing harlequin (north-south) syndrome
d. None of the above

3. Common pressures measured on an ECLS circuit include:

a. Inlet or venous line pressure
b. Pre-oxygenator or internal pressure
c. Post-oxygenator or arterial line pressure
d. All of the above

347
Chapter 37

4. The oxygenator delta pressure:

a. Is the calculated difference between the pre- and post-oxygenator pressures
b. If increasing indicates obstruction in the oxygenator, usually related to thrombosis
c. Can be used independently to diagnose oxygenator failure
d. a and b
e. All of the above

5. Retrograde flow:
a. Occurs when the distal pressure exceeds the pump generated pressure
b. Is a risk associated with nonocclusive centrifugal pumps
c. Can be prevented with appropriately set low flow alarms
d. All of the above

6. The circuit SVO2:

a. On VA-ECLS reflects the patient SVO2
b. On VV-ECLS reflects the patient SVO2
c. On VV-ECLS can be a reflection of recirculation fraction
d. a and c
e. All of the above

7. Neuromonitoring strategies during ECLS include:

a. Cerebral NIRS monitoring
b. Intermittent or continuous EEG
c. CT scan
d. Cranial ultrasound in infants
e. All of the above

Chapter 19 (Choose the best answer)

1. What methods are used to “wean” or “trial off” with VV-ECMO?
a. Wean the FiO2 of the sweep gas
b. Disconnect the sweep gas
c. Wean ECMO flow to 100mL/minute and then clamp the patient off circuit
d. Extubate the patient
e. Both a and b

2. Your patient today is a 3 kg patient on ECMO for PPHN. Her blender FiO2 is 60%, the sweep gas
flow is 0.8 Lpm, and the ECMO flow is 300 ml/min. Her last blood gas showed a pCO2 of 30 mmHg
and a pO2 of 80 mmHg. What changes would you make?
a. Increase ECMO flow
b. Decrease ECMO flow
c. Increase sweep flow
d. Decrease sweep flow

3. Hypertension during ECMO may be associated with:

a. Increased risk of hemorrhage

348
 Chapter Questions

b. Vascular volume excess

c. Altered vascular volume regulation
d. Excessive ECMO pump flow
e. All of the above

4. Cardiac Stun:
a. Is a decrease in cardiac contractility
b. May be seen in VA ECMO within the first few hours/day
c. Will have a narrow pulse pressure
d. Will have a high patient PO2
e. All of the above

5. Intracranial hemorrhage during ECMO may be associated with:

a. Hypoxia prior to ECMO
b. Disordered autoregulation of cerebral blood flow
c. Increased cerebral venous pressure
d. Anticoagulation
e. All of the above

6. Which of the following usually has the greatest effect on tissue oxygen delivery during VA-ECMO?
a. pO2
b. Oxygen saturation
c. Cardiac output
d. ECMO blood flow rate
e. Ventilator FiO2
f. ECMO sweep gas FiO2

7. You are caring for a 3.2 kg baby on VA-ECMO for CDH. The physician is concerned that the baby
may not be ready for decannulation and wants to actually trial the baby off. What needs to be done?
a. Transfer infusions (TPN, Morphine, Midazolam, etc.) to the patient
b. Wean blood flow to 50-100 ml/min
c. Disconnect/turn off the sweep gas once actually clamped off ECMO
d. Increase ventilator settings
e. Place a bridge in the circuit
f. All of the above

Chapter 20 (Choose the best answer)

1. In case of hypoxia in a 3 year old boy supported with VV-ECMO, you need to rule out all following
EXCEPT:
a. Recirculation
b. Pain and agitation
c. Fever
d. Insufficient venous drainage
e. Differential hypoxia

349
Chapter 37

2. What is an acceptable arterial saturation for a child on VV-ECMO:

a. Must be at least 92%
b. Must be at least 85%
c. Must be at least 75%
d. Must be at least 70%, if urine output, cerebral oximetry, lactate level are adequate

3. Decreasing sedation but maintaining patient comfort should be the goal during VV-ECMO: a. True
b. False

4. What are the monitoring tools you should use at bedside:

a. Cerebral and somatic Near-Infrared Spectroscopy
b. Ultrasound for brain, lung and heart
c. ACT machine
d. Blood gases machine
e. All the above

5. Associated measures to enhance lung recovery during VV-ECMO include all the following, EXCEPT:
a. Prone positioning
b. Adding an additional arterial cannula
c. Renal replacement therapy
d. Spontaneous breathing

6. You are caring for a 15 year-old child, 50kg, supported with VV-ECMO for respiratory failure second-
ary to viral pneumonia. He has one drainage cannula inserted into the femoral vein and a reinfusion
cannula inserted into the right internal jugular vein. The maximum blood flow that you can obtain is
55ml/kg/min and the patient remains hypoxic despite all measures (sedation, paralysis, transfusion,
prone positioning, fluids). You decide to change the configuration of your circuit. What would you
do?
a. Use both of your cannulas to drain and you add an arterial cannula into the femoral artery; VA-
ECMO
b. Remove both cannulas, and place a double lumen cannula into the right internal jugular vein
c. Add an additional venous cannula into the contralateral femoral vein so you will drain from both
femoral veins and re-infuse into the cannula inserted into the right internal jugular vein

Chapter 21 (Choose the best answer)

1. A patient with transposition of the great arteries following arterial switch operation was found to have
compromised coronary circulation shortly after returning from the OR and required urgent VA ECMO
cannulation. At the time of cannulation, the surgeon also placed a LA vent cannula. The indication
for inserting a LA vent is:
a. To provide an additional source of venous drainage due to an inadequately sized venous cannula
b. To prevent distension of the right ventricle and impairment of myocardial recovery
c. To improve the capability to oxygenate and ventilate the patient
d. To prevent distension of the left atrium and ventricle, to allow myocardial recovery

350
 Chapter Questions

2. ECLS blood flow and patient mean arterial blood pressure targets should be influenced by patient
examination and markers of end organ perfusion that include which of the following:
a. Measured blood lactate levels
b. Urine output
c. Adequate mixed venous saturation
d. All of the above

3. Use of renal replacement therapy on ECLS may be indicated for early fluid removal as this can be
associated with improved outcomes. Fluid overload on ECLS is common due to:
a. Capillary leak and systemic inflammatory response related to CPB and exposure to ECLS circuit
b. Mechanical ventilation and anesthetic gas exposure in the operating room
c. Chronic renal failure from previous kidney injury
e. Distension of the left atrium and ventricle due to myocardial dysfunction

4. Bleeding in the early postoperative period on ECLS can be challenging. One strategy used to control
bleeding can be to reduce or stop anticoagulation transiently until bleeding is controlled. Neonates
and small children supported on ECLS are at a higher risk of circuit thrombosis when this occurs due
to:
a. Differences in developmental hemostasis between children and adults
b. Use of low blood flow rates and smaller diameter cannulas as compared with adults
c. Renal injury and fluid overload after exposure to CPB and the ECLS circuit
d. Dilution of clotting factors during CPB and volume replacement to manage losses

5. Contraindications for the use of ECLS to support neonates and children with cardiac disease include
all of the following except:
a. Cardiac disease is irreversible and the patient is not a candidate for transplant
b. Patient is too small to allow for peripheral cannulation
c. Previous discussions with family whose directives included no ECLS
d. Birth weight 2.1 kg

Chapter 22 (Choose the best answer)

1. In the patient on VV-ECMO for severe ARDS, a rising CVP and decreasing arterial pulse pressure
should raise concern for what cause of hemodynamic compromise:
a. Tachydysrhythmia
b. Right ventricular failure
c. Hypovolemic shock
d. ECMO circuit failure

2. The paralytic of choice for the patient on VV-ECMO is:

a. Cisatracurium
b. Rocuronium
c. Vecuronium
d. Succinylcholine

351
Chapter 37

3. In general, the patient who is placed on VV-ECMO for acute respiratory distress syndrome should
have a target SaO2 of:
a. 100%
b. 95%
c. 88%
d. 75%

4. Progressive oliguric or anuric renal failure in patients receiving VV-ECMO that no longer responds
to medical therapy will often require this therapy to optimize fluid status:
a. Continuous renal replacement therapy (CRRT)
b. Molecular adsorption recirculating system (MARS)
c. Aggressive intravenous fluid resuscitation
d. None of the above

Chapter 23 (Choose all that apply)

1. Oxygenation is regulated for a patient on VA-ECMO by
a. The sweep gas FiO2
b. The pump flow
c. The sweep gas flow
d. None of the above

2. Which complications are specific to VA-ECMO?

a. Recirculation
b. Limb ischemia
c. Harlequin syndrome
d. Hemolysis

3. Which strategies allows to treat pulmonary edema on VA-ECMO with poor left ventricular function?
a. Balloon pump
b. Impella
c. Atrioseptostomy
d. Central cannulation
e. All of the above

Chapter 24 (Choose all that apply)

1. The number and complexity of procedures carried out in patients on ECLS has:
a. Reduced over the past 10 years
b. Stayed largely the same
c. Increased significantly
d. Resulted in a significant increase in Mortality

2. Prior to undertaking a procedure on ECLS the specialist should

a. Correct any coagulopathy and increase platelet / fibrinogen parameters

352
 Chapter Questions

b. Reduce Heparin infusion to target ACT 160-180 and anti-factor Xa to 0.3-0.5

c. Increase circuit blood flows to >120 mls/kg
d. All of the above

3. Repositioning the patient prior to a procedure on ECLS may result in:

a. Changes to cannula position
b. A change in circuit blood flows
c. A change in RPM
d. Reduced sweep gas flow

4. During laparotomy and bowel resection the ECLS specialist may encounter:
a. Cavitation in the circuit
b. A decrease in TMP
c. A need for Antifibrinolytics
d. Bleeding necessitating blood product administration

Chapter 25 (Choose the best answer)

1. Which statement is true?
a. A patient must be at minimum flow for at least 6 hrs prior to commencing a trial off
b. Minimum flow is determined solely by patient weight
c. At minimum flow circuit thrombosis is very rare
d. ACTs should be doubled if at minimum flow
e. Minimum flow is mainly dependent on circuit configuration

2. Concerning trials off which statement is true?

a. It is NEVER appropriate to trial off on FiO2 0.8
b. PaO2 must be > 75 mmHg for the trial to be deemed successful
c. Trial off is unlikely to cause patient harm even if unsuccessful
d. The PaO2 is the sole determinant of a successful trial off
e. The patient should be sedated a muscle relaxed prior to commencing trial off

3. Concerning retrograde trial of VA-ECMO which statement is true

a. It is always possible to trial a patient off retrograde
b. It is only possible on patients in whom the diagnosis is pure respiratory
c. The maximum duration of trial off is 2 hrs
d. Increasing the pump rpm decreases retrograde flow
e. Decreasing the pump rpm decreases retrograde flow

4. A baby with left congenital diaphragmatic hernia (LCDH) is on VV-ECMO with a 13 fr DLVV can-
nula. He is improving on ECMO. Before trialing off
a. The pulmonary pressures should be less than 2/3rd systemic
b. The CDH must have been repaired
c. Keeping the PaCO2 <30 mmHg on ECMO is important to aid trial off
d. A long trial off >6hrs may be appropriate if blood gases are borderline
e. a and d

353
Chapter 37

5. A 62 year old lady is on ECMO for 3 weeks with pneumonia. No causative organisms have been
identified. Her Hb has dropped from 100 g/l to 78 g/l. She is noted to have a large rectal bleeding.
Despite correcting her clotting profile she continues to bleed heavily for >24 hrs. Appropriate action
would be.
a. Continue ECMO support and transfuse continually
b. Commence a trial off accepting higher PIP and FiO2
c. Take immediately to theater for total colectomy on ECMO
d. Trial off is not possible as an underlying illness has not yet been established
e. Decannulate immediately without any trial off

Chapter 26 (Choose all that apply)

1. John is 2 weeks old with a history of severe meconium aspiration syndrome with persistent pulmonary
hypertension, acute hypoxemic respiratory failure in the neonatal period requiring VA ECMO. During
the course of his ECMO run, he was noted to have seizures that were confirmed on EEG. A CT scan
of the brain showed no infarction or bleed. Following decannulation, he had a brain MRI that showed
thrombosis of the right carotid artery. Which of the following statements are true?
a. He should have regular neurology followup
b. He should have a further MRI brain scan as directed by his neurologist
c. There is no need for further followup if repeat MRI scan is normal
d. Hearing tests should be done pre-discharge
e. There is need for life long anticoagulation

2. John is now 3 years old, a bit behind on his speech and communication skills. His mother reports that
he has frequent temper tantrums. She has been trying to engage community services, but it has been
difficult. Which of the following statements are true?
a. He needs a full neurodevelopmental assessment
b. The primary health care doctor advises that repeat hearing is not needed as neonatal hearing screen
was normal
c. Age-appropriate questionnaires such as Ages and Stages Questionnaires filled by parents are help-
ful to complement a full neuropsychological assessment
d. He needs a referral to psychologist if problems identified on behavioral assessment
e. He needs only a referral to Speech and Language Therapy (SALT)

3. David is a newborn with antenatally diagnosed left Congenital Diaphragmatic Hernia (CDH), who
was supported on VA-ECMO for 7 days. He had normal EEG/Head Ultrasound Scan on ECMO. He is
now being planned for discharge home without supplemental oxygen and without sildenafil, bosentan,
or diuretics. Advice needs to be given to his parents before his discharge. Which of the following
statements are true?
a. He needs close monitoring of physical growth
b. Once operated, he is unlikely to have any feeding problems and reflux is uncommon
c. He will need pulmonary function tests in childhood and adolescence
d. He should have regular echocardiograms to check for pulmonary hypertension
e. He needs sequential followup right until transition to adult services

354
 Chapter Questions

4. Amelia is a previously well 4-year-old child who develops influenza A infection, complicated by
severe bacterial pneumonia with Staphylococcus aureus with severe hypoxemic respiratory failure.
She receives VA ECMO and then VV ECMO for a total of 5 weeks before she is successfully decan-
nulated. Her mother is concerned that this episode of critical illness may affect her future well-being.
What will you as an ECMO professional advise her? Which of the following statements are true?
a. She will need regular pediatrician followup
b. She needs a full neurodevelopmental assessment just before she starts school
c. She needs hearing assessment
d. Post-traumatic stress can manifest as behavioral problems
e. She does not need any follow-up as she was previously well

5. Tom is a 14-year-old boy who develops fever, myalgia and tummy upset. His local doctor treats him for
flu. However, he deteriorates rapidly and now develops fast heart rate, and presents cold and clammy
in a state of cardiovascular collapse in his local Emergency Department, and is noted to have arrhyth-
mia which progresses into an arrest with a total downtime of 6 minutes. He is resuscitated, intubated,
and commenced on ECMO. The mobile ECMO team transfer him to the tertiary ECMO center. He
is later diagnosed to have Parvovirus Myocarditis. CT brain shows cerebral edema, and a watershed
infarct in middle cerebral artery (MCA) territory. He makes significant cardiac recovery to be able to
come off ECMO. Which of the following statements are true?
a. He needs only cardiology/heart failure follow-up
b. He needs MRI Brain before discharge home
c. As a full grown teenager, he does not need neuropsychological assessment
d. Getting back to school may take some time and he may need time to catch up to his pre-illness
state
e. He will need physiotherapy and neurorehabilitation and may have exercise intolerance which will
need to be monitored

6. Grace is a 4-week-old neonate with congenital heart disease who undergoes complex cardiac surgery.
She develops severe low cardiac output state with hypotension, and a decision is made to put her on
chest cannulation VA ECMO. After a period of rest over 5 days, her heart recovers and she comes off
ECMO. Which of the following statements are true?
a. In addition to her cardiologist, she needs a general pediatrician to follow her at her local hospital
b. She does not need neuro-imaging as her pre-op head ultrasound scan was normal
c. Her parents need to be explained the importance of regular neurodevelopmental assessment
d. Hearing assessment is important
e. Sequential, longitudinal neurodevelopmental assessment up till adulthood should be part of her
long-term care package

Chapter 27 (Choose the best answer)

1. All of the following are important for ECMO ambulation EXCEPT:
a. A fully charged ECMO console/pump
b. Mobile phones
c. Suction equipment and supplies
d. An appropriate chair for the patient to rest as needed
e. Appropriate number of clinical staff

355
Chapter 37

2. Which of the following statements are true?

a. Pharmacologic sedation and neuromuscular blockade are well accepted risk factors for increased
morbidity and mortality in the critically ill population.
b. Ambulating ECMO patients has been demonstrated to decrease the risk for thromboembolism.
c. Family satisfaction is increased when patients are allowed to ambulate around a hospital.
d. Ambulating VV ECMO patients has been shown to decrease the need for lung transplantation.

3. All of the following steps should be taken prior to walking with an ECMO patient EXCEPT:
a. A clear plan of travel should be reviewed.
b. An overall goal for the ambulation should be discussed with the patients, care team, and the pa-
tient’s family (if applicable) prior to starting.
c. The patient should be provided an appropriate dose of an anti-anxiolytic prior to ambulation.
d. All necessary equipment should be available.
e. All members of the direct care team and the patient/parent should agree to proceed.

Chapter 28 (Choose the best answer)

1. Acute kidney injury in ECLS patients:
a. Is rare, and seen in <20% of patients
b. Is associated with worse mortality
c. Is caused by using CRRT
d. Occurs late in a patient’s ECLS course

2. Draining for CRRT from the distal venous limb of a centrifugal ECLS system:
a. Is the preferred site of drainage
b. Causes decreased recirculation
c. Often functions poorly due to local negative pressure
d. Adds to the risk of a “blood out” event of the circuit

3. When using the “in-line” technique of providing CRRT during ECLS:

a. Dialysis trained nurses are required for management
b. Recirculation is often a problem
c. You will not get adequate clearance of electrolytes
d. Hourly attention to fluid balance is essential to prevent complications

Chapter 30 (Choose the best answer)

1. Viable methods for mechanical liver support include all of the following except:
a. SPAD
b. CRRT
c. Clinical Ex-Vivo Liver perfusion (a biological non-mechanical approach)
d. Plasma Exchange

356
 Chapter Questions

2. As a therapy, SPAD has a drawback of which of the following?

a. Unable to regenerate albumin (one use per pass- expensive)
b. No evidence available for efficacy in toxin clearance
c. Requires specialty equipment specific for liver support
d. Complicated prescription requirements

3. Current indication(s) for use of MARS in the US include:

a. Toxin clearance
b. Bridge to liver transplantation
c. Treatment of hepatic encephalopathy
d. a and c

Chapter 32 (Choose all that apply)

1. An increase only in the pre membrane pressure may indicate:
a. A kinked arterial cannula
b. Reduced gas exchange
c. A blood leak
d. Clots in the membrane

2. The circuit pressures in the VA-ECMO patient you are caring for are rising. Both the pre and post
circuit pressures are 50 mmHg higher than they were at the start of your shift. ECMO pump flow has
not changed. Possible causes of this situation are:
a. Patient agitation
b. Kink pre oxygenator
c. Kink in the arterial return side catheter
d. Clot in the centrifugal pump head

3. Name the air traps in the ECMO circuit.

4. What is the proper procedure for taking a patient off bypass?

5. If you have an occlusion or impediment to the arterial return flow on the centrifugal pump, the RPMS
will remain unchanged and pump flow will:
a. Increase
b. Decrease

Chapter 35 (Choose the best answer)

1. Which of the following configurations would the least probable to ever experience in clinical practice?
a. VfVj-Va
b. Va-VfAsrg
c. Vj-Asrg
d. Vjrc-AcrAcl
e. Vjrava-AflaAfrd

357
Chapter 37

2. Patient with one cannula on the right side of the

neck, and one major cannula in the groin. There
is a thin tubing to a catheter directed downwards
towards the foot (?). What would it not be?
a. (dl31)V-VAa
b. Vj-Afla
c. V25/50-A21/18d
d. (dl23)VV-Aa
e. V25/25j=A21a

3. What is not true concerning the ELSO Maas-

tricht Treaty for Nomenclature in Extracorpo-
real Life Support?
a. I have always to apply to the flow direction
and hyphenation
b. When all terms of one level are completed
I easily continue with the next level
c. A may use whole or just part of a specific
level of the nomenclature
d. I may use level four, but skip level two and
three

358
 Chapter Questions

Answer Key Chapter 8 Answers

1. c
Chapter 2 Answers 2. d
1. a, d, e 3. b
2. a, b, e 4. c
3. b 5. a
4. b, c (There is usually a limit to how much 6. b
increasing pump flow improves oxygen de-
livery because at some point the increase in Chapter 9 Answers
recirculation that occurs with increases in 1. b
flow outweighs the improvement in forward 2. a
flow.) 3. d
5. b, c, e 4. d

Chapter 3 Answers Chapter 10 Answers

1. d 1. b
2. b 2. d
3. d 3. b
4. b 4. c
5. b 5. d
6. b
Chapter 5 Answers 7. d (Reasonable to aim for a vaginal birth
1. b while on ECMO, mifepristone takes 48 hrs
2. a to have a full effect then start prostaglan-
3. d (Table 5-3) dins at which time need a nonheparinized
4. c circuit when contractions start or delivery
is imminent.)
Chapter 7 Answers 8. b
1. c ( O I = ( M A P x F i O 2 ) / P a O 2 .
(16x1x100)/32=50) Chapter 12 Answers
2. d (Late prematurity, history of CPR and aci- 1. e
dosis are all risk factors for poor outcome, 2. f
as is his diagnosis—which is currently id-
iopathic PPHN but may also be pneumonia Chapter 13 Answers
or “other”) 1. a, c, d
3. c (Although prediction is difficult, most of 2. a, c
the diagnostic category this baby fits into 3. b, c, d
have survival rates of 60-80%, gestational 4. b, c
age will also affect that. 92% is more typi- 5. c
cal for MAS, while 51% is more typical for
CDH.) Chapter 14 Answers
1. a, b, e
2, e
3. b, c, d, e
4. c, e

359
Chapter 37

Chapter 15 Answers Chapter 21 Answers

1. c 1. d
2. a, c 2. d
3. b 3. a
4. d 4. b
5. d
Chapter 16 Answers
1. c Chapter 22 Answers
2. b 1. b
3. b 2. a
3. c
Chapter 17 Answers 4. a
1. a, d
2. b Chapter 23 Answers
3. b 1. b (Oxygenation is regulated by the pump
4. c flow for VV ECMO. The sweep gas flow
5. c regulates ventilation [CO2 clearance]).
2. b, c
Chapter 18 Answers 3. e
1. c
2. c Chapter 24 Answers
3. d 1. c
4. d 2. d
5. d 3. a, b
6. d 4. a, c, d
7. e
Chapter 25 Answers
Chapter 19 Answers 1. e
1. e 2. c
2. d 3. d
3. e 4. e
4. e 5. b
5. e
6. d Chapter 26 Answers
7. f 1. a, b, d
2. a, c, d
Chapter 20 Answers 3. a, c, d, e
1. e. 4. a. c
2. d 5. b, d, e
3. a 6. a, c, d, e
4. e
5. b
6. c

360
 Chapter Questions

Chapter 27 Answers
1. c (Suction equipment and supplies are not
often needed. Bringing excess equipment
and supplies may unnecessarily complicate
an ECMO ambulation.)
2. a (While options b, c, and d may be true,
definitive data are not available. The only
choice support by the medical literature is
a.)
3. c (Administering an anti-anxiolytic prior
to ambulation is not necessary, and it may
be harmful to administer a sedative prior
to ambulating a critically ill patient. If a
patient is overly anxious about walking
while on ECMO, the anxiety and its causes
should be addressed in advance of an at-
tempt to ambulate.)

Chapter 28 Answers
1. b
2. c
3. d

Chapter 30 Answers
1. c
2. a
3. d

Chapter 32 Answers
1. d
2. a, c
3. Tailored to your specific center’s equipment
and procedures.
4. Tailored to your specific center’s equipment
and procedures.
5. b

Chapter 35 Answers
1. d
2. c
3. b

361
 Index

A Central Line-Associated Bloodstream Infection policy

(CLABSI) 182
abbreviations 305 central venous pressure 35
Acute kidney injury 231 centrifugal pump 154
acute respiratory distress syndrome 195 cephalad jugular cannula 172
Afterload 32 cervical cannulation 117
Air/bubble detectors 164 chatter 198
ambulation 225 chest injuries 97
Analgesia 195 Circuit shunts 149
Anticoagulation 180,191,199 classification 305
Antifibrinolytic Therapy 50 Combined Acute Respiratory and Cardiac Failure 132
Apheresis 239 comorbidity 222
ARDS 68,140 Compartment syndrome 153
Arrhythmias 81 compliance 35
Assembly 107 complications 114
A-V oxygen difference 24 Complications of Transfusion 47
computed tomography (CT) 211
B
configuration 305
bacterial pneumonia 68 constrained vortex 154
Bladders 157 Continuous Renal Replacement Therapy
Bleeding 192 198,231,232,239,248
blood flow 169 contractility 32
Blood Flow Meters 164 contraindications 187
blood groups 44 COPD exacerbation 140
blood prime 107 Cryoprecipitate 47
brain injury 98 CT 166
bridge 123 CVL method 126
bronchoscopy 181,211
D
C
daily management 188
cancer 101 decannulation 110
cannulas 172,305 Delta Pressure (ΔP)/transmembrane pressure 163
Cannula Selection 152 Dexmedetomidine 179
Cannulation 124,187 dialysis 232,235,241
Cardiac catheterization 211 distal limb ischemia 153
cardiac output 31 Double Lumen Cannula 152
cardiogenic shock 79,113 Double lumen cannulae 177
CARDIOHELP 140 dressing 182
Catecholamines 32
CDH 67

363
Index

E indications 187
infection 100
ECLS Support Following Implantation of a Durable Infection Control 198
LVAD 135 inflow/negative pressures 157
ECLS survivors 221 Integrated circuits 149
ECMO Circuit Design 148 Integrated monitoring 148
ECMO Initiation 109 intracranial hemorrhage 169,171
ECMO Program Structure 147 Ischemia of the ipsilateral lower extremity 116
ECMO specialist team 19
ECMO transport 148 L
EEG 166
effective flow 126 laparotomy and bowel resection 211
Emergencies 191 left atrial (LA) vent 212
engraftment 102 Left Ventricular Distention 114,134
EOLIA Trial 201 leukapheresis 239
erythrocytapheresis 239 leukoreduced 108
ECCO2R 139 long-term followup program 222
ECPR 77 Long-term Support 149
eye assessment, eye care 180 loop 123
Low Cardiac Output State 77
F low frequency ventilation 139
lung biopsy 212
Failure to Wean 76 Lung improvement 178
Fick Principle 24 Lung Protective Ventilation 197
Flow Pressure Charts 152 lung transplantation 71,141,225
Fluid overload 190,231,235
followup program 222 M
Fractionated Plasma Separation and Adsorption (Pro-
metheus) 250 MARS 250
maternal 98
G meconium aspiration syndrome 67,167
Medical orders 107
gas exchange 23 mobilization 225
Glenn and Fontan 77 mode of ECLS 131
MODES 131
H
multiorgan dysfunction syndrome (MODS) 240
H1N1 pandemic 139 Myocarditis and Cardiomyopathy 78,118
Hand cranking 164
Harlequin Syndrome 153,165 N
Heat Exchanger Related Complications 160 Near-infrared spectroscopy (NIRS) 181
Hematopoietic stem cell transplant 102 Neonatal 67
Hemodialysis 248 neonatal ECLS 221
Hemofilters 232 neonates and children 117
Hemolung Respiratory Assist System 140 neurodevelopmental outcome 221
Hemolysis 126,168,172 neurologic complications 221
Hybrid modes 131 Neuromonitoring 166
Hypertension 189 NIRS 165
Hypoxemia during V-V ECMO 133 nomenclature 305
Novalung iLA Active 140
I
Nursing assessment 179
immune reconstitution 102
immunosuppressed 101 O
Impella 114 oncology 101

364
 Index

oral assessment 182 S

Oral Care 182
Oxygenators 157 sedation 195,225
oxygen consumption 24 Sedation/Pain Management 190
oxygen content 24,31 Seldinger 124
oxygen delivery 24,31,170 semi-percutaneous technique 124
oxygen extraction ratio, (DO2/VO2) 179 sepsis 100,240
Oxyhemoglobin Monitoring 164 septic shock 100
Single Lumen Cannula 153
P Single Pass Albumin Dialysis 249
Skin care 181
partial pressure 23 spontaneous breathing 178
pediatric critical illness 222 status asthmaticus 68
percutaneous cannulation 124,177 sternotomy 212
peripheral cannulation 305 subclavian arterial cannulation 116
Physical rehabilitation 226 surgical procedures 210
physical therapy 225 survival 116
Plasma 46 survival outcomes 221
plasma free hemoglobin 157 SVO2 126,164
Platelet 46 sweep gas 170
pneumonia 68 systemic vascular resistance 33
poly-methyl-pentene (PMP) membranes 158
positioning aids 181 T
Post oxygenator/arterial line pressure 163
Postpartum 100 therapeutic plasma exchange 239
pregnancy 98 thoracotomy/thoracoscopy 209
Preload 32 thrombocytopenia associated multiorgan failure
Preoperative Support 75 (TAMOF) 240
Pre-oxygenator/internal pressure 163 tracheostomy 226
pressure ulcers 181 transfusion 43
primer 107 Trauma 97
procedures 209 trial off 126
prone positioning 180,197 tumors 101
Prothrombin Complex concentrates 50
U
puerperium 98
pulmonary artery 114 Ultrasound dilution 126
Pulmonary hypertension 80 Upper body hypoxemia 114
Pulse oximetry 165
V
R
VA 305
recirculation 27,125,153 vasoconstrictor 33
Recombinant factor VIIa 49 Venous line/ inlet pressure 163
Red Blood Cell 44 ventilator induced lung injury 140
renal failure 235 VV 305
Respiratory electrodialysis 142 VVA 305
respiratory failure 67 V-VA ECMO (veno-venoarterial ECMO) 131
respiratory or cardiac ECLS 221
Retrograde flow 164 W
right ventricle (RV) failure 114
weaning 126,172,193
Risk Scores 116
roller pump 154
RV dysfunction 196
RV function 179

365