American Journal of Infection Control xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

American Journal of Infection Control

journal homepage: www.ajicjournal.org

Major Article

The longer the catheter, the lower the risk of complications: Results
of the HERITAGE study comparing long peripheral and midline
catheters
Adam Fabiani MNS, RN a,b, Nicola Aversana MNS, RN c, Marilena Santoro RN b,
⁎
Dario Calandrino RN d, Paolo Liotta RN b, Gianfranco Sanson PhD, RN e,
a
Department of Biomedicine and Prevention, University of Rome Tor Vergata, Rome, Italy
b
Cardiothoracic-Vascular Department, Azienda Sanitaria Universitaria Giuliano-Isontina, Trieste, Italy
c
School of Nursing, University of Trieste, Trieste, Italy
d
Internal Medicine Department, Azienda Sanitaria Universitaria Giuliano-Isontina, Trieste, Italy
e
Department of Medicine, Surgery and Health Sciences, University of Trieste, Trieste, Italy

Key Words: Background: Although widely used in clinical practice, long peripheral (LPCs) and midline catheters (MCs)
Catheter-failure are often misclassified because of their similar characteristics. Comparative studies on these devices are
Catheter-related bloodstream infection lacking. This study aimed to explore complications risks in polyurethane LPCs and MCs.
Catheter-related thrombosis
Methods: Prospective cohort study. Catheter-failure within 30 days was the primary outcome, catheter-
Fibroblastic sleeve
related bloodstream infection (CR-BSI), thrombosis, and fibroblastic sleeve were secondary outcomes. The
Long peripheral catheter
average number of drugs infused per day was computed to measure the overall intensity of catheters’ use.
Results: The catheter-failure incidence was 5.7 and 3.4/1,000 catheter-days for LPCs and MCs, respectively.
MCs were associated with an adjusted lower risk of catheter-failure (hazard ratio 0.311, 95% confidence
interval 0.106-0.917, P = .034). The daily number of drugs infused was higher for MCs (P < .001) and was
associated with a greater risk catheter-failure risk (P = .021). Sensitivity analysis showed a decreased ca­
theter-failure risk for MCs starting from day-10 from positioning. The incidence of CR-BSI (0.9 vs 0.0/1,000
catheter-days), thrombosis (8.7 vs 3.5/1,000 catheter-days), and fibroblastic sleeve (14.0 vs 8.1/1,000 ca­
theters-days) was higher for LPC catheters.
Conclusions: Despite more intensive drug administration, MCs were associated with a longer un­
complicated indwelling time.
© 2024 The Author(s). Published by Elsevier Inc. on behalf of Association for Professionals in Infection
Control and Epidemiology, Inc. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).

BACKGROUND such a catheter, being limited by nature to very superficial visible or

palpable veins, can sometimes be difficult because of the patient’s
Between 59% and 70% of hospitalized patients need venous access.1,2 characteristics (eg, edema, obesity, and poor superficial venous “heri­
In most cases, traditional, short peripheral catheters (24-60 mm length, tage”) representing a “difficult intravenous access” (DIVA) condition.5,6
26-14 G internal diameter) are used, because they represent the safer Moreover, patients may require a medium- or long-term continuum use
(very low complication rates), simpler (placement requiring limited ex­ of the vascular device (weeks or months), which could cause a pro­
pertise), and most cost-effective (potential to administer—with a few gressive reduction of the superficial venous pool.7 All such patients are
exceptions—most medications) choice.1,3,4 The opportunity to position exposed to the risk of suboptimal care, being undergoing multiple can­
nulation attempts, or the administration of medications through in­
appropriate devices (eg, extremely thin catheters, veins of legs or feet),
leading to a high risk of complications (eg, extravasation, phlebitis, in­
⁎
Address correspondence to Gianfranco Sanson, PhD, RN, Department of Medicine, fections, and failure to administer drugs or to obtain blood samples) and
Surgery and Health Sciences, University of Trieste, Strada di Fiume 447, 34148 Trieste,
to avoidable suffering.
Italy.
E-mail address: gsanson@units.it (G. Sanson). A valid alternative consists in positioning a longer venous ca­
Conflicts of interest: None to report. theter in deep veins of the upper limbs under ultrasound guidance

https://doi.org/10.1016/j.ajic.2024.06.019
0196-6553/© 2024 The Author(s). Published by Elsevier Inc. on behalf of Association for Professionals in Infection Control and Epidemiology, Inc. This is an open access article
under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
2 A. Fabiani et al. / American Journal of Infection Control xxx (xxxx) xxx–xxx

and using a guide wire. Based on their length, these catheters have Catheter placement procedure
been classified as long peripheral catheters (LPCs: length > 6 cm
and ≤ 15 cm) and midline catheters (MCs: length > 15 cm).8 All catheters were placed by nurses with advanced education and
As patients with a DIVA condition may be present in many health high expertise with regard to vascular access. According to the most
care settings, in recent years, an increasing number of health care updated recommendations,15,16 the catheter insertion procedure was
professionals—especially nurses—have been trained and have de­ standardized as follows:
veloped clinical experience in the placement of these devices.
Unfortunately, the spread in their use has not been accompanied by – Preprocedural systematic ultrasound examination of the veins of
a parallel increase in research. Consequently, comparative studies on the arms and the cervical-thoracic districts according to the
the safety, reliability, and cost-effectiveness of the different devices “RaPeVA” and “RaCeVA” protocols,17,18 and clear identification of
are lacking,9-11 with the existing studies often misclassifying the the median nerve and brachial artery.
devices.8 – Identification of the best vein to cannulate, paying attention to (1)
Therefore, the aim of the study was to explore the differences in placing the catheter’s exit site within the Dawson’s green zone19;
terms of safety and reliability between LPCs and MCs as used in daily (2) choosing—whenever possible—a vein having an inner size
clinical practice. large enough in respect of a catheter-to-vein ratio of 33% or less,
both at the insertion point and at the expected catheter’s tip level.
– Antiseptic handwashing, use of maximal barrier precautions,
METHODS
and skin antisepsis with 2% chlorhexidine in 70% isopropyl al­
cohol.
Study design, setting, and population
– Ultrasound-guided catheter positioning via the short axis or out-
of-plane approach, by adopting a “modified Seldinger tech­
The HERITAGE (long peripHERal and mIdline caTheters
nique” for MCs and a “simplified Seldinger technique” for LPCs.
uncomplicAted dwellinG timE) study was a prospective cohort study
– Final ultrasound assessment of the catheter position up to the tip
carried out in Cardiac Surgery, Cardiology, and Internal Medicine
location.
Departments of the Trieste University Hospital, Italy, where the ultra­
– Catheter sutureless securement, protection of the exit site using
sound-guided positioning of venous catheters was part of the routine
cyanoacrylate glue, and semipermeable transparent polyurethane
daily nursing clinical practice. All consecutive adult patients receiving
membrane dressings with chlorhexidine-based gel (3M Tegaderm
either a LPC or an MC—because of a DIVA condition5,6 or because of the
CHG), application of a needle-free connector to each catheter hub
need for a medium-term intravenous therapy (> 7 days)4,7,12,13—were
to be covered with a port protector (3M Curos Port Protectors
considered for inclusion in the study. Catheters positioned in a same
Caps) while the catheter is locked.
patient after removing of the existing one were considered as well. Pa­
tients with immediate life-threatening conditions, transferred to a clin­
ical setting where continuing the catheter follow-up was not possible, or
Catheter management and surveillance
who refused to participate in the study, were excluded.
Based on the primary study endpoint, a minimum required
All intravenous devices were managed by hospital or community
sample size of 182 catheters was calculated based on the estimated
clinical nurses based on the same, shared policy. The modality to
difference between the overall complication rates of LPCs (19.7%)10
access the catheter for drug administration or blood sampling and to
and MCs (5.3%-5.9%),14 enabling a type-I probability error of 5% and a
manage the catheter while locked was standardized according to the
desired statistical power of 80%. Expecting the risk of 5% of the pa­
most updated best practice.4,15,16 Briefly, the use of a port protector
tients being excluded due to relevant missing data, an enrollment of
to cover the needle-free connector was strongly recommended, al­
at least 200 patients was planned for.
ternatively a vigorous rubbing (5-15 times) of the needle-free con­
The study was conducted in accordance with the ethical principles
nector was performed before accessing the catheter. Before and after
enshrined in the Helsinki Declaration. The study protocol was approved
each access, the catheter was flushed with 10 to 20 mL of normal
by the Independent Regional Ethics Committee (n.2826/2023). All en­
saline using the “stop and go” technique. The catheter dressing, in­
rolled patients were asked to sign an informed consent form.
fusion lines, and needle-free connectors were replaced weekly, un­
less damaged or contaminated, except for the infusions of solutions
Characteristics of the catheters containing lipids, when infusion lines and needle-free connectors
were replaced daily.
The following devices were involved: The infusion of parenteral nutrition (osmolarity > 900 mOsm/L)
through these catheters was forbidden, while all medications asso­
– PowerGlide PRO polyurethane LPC catheter (Becton Dickinson), ciated with a high risk of endothelial damage (eg, pH of < 5 or > 9,
available in the following sizes (internal area, external diameter, osmolarity > 600 mOsm/L)20 were administered only after an ap­
and length): (1) 20 G (1.15 mm), 8 cm; (2) 18 G (1.35 mm), 10 cm. propriate dilution.
– PowerMidline polyurethane MC (Becton Dickinson), available in Each catheter was regularly assessed according to the following
the following sizes (internal area, external diameter, and length): protocol: daily exit-site inspection; ultrasound evaluation of the
(1) 18 G (single-lumen), 4 Fr (1.32 mm), 20 cm; (2) 19/21 G (dual- catheter position and possible ongoing complications every 3 to
lumen), 4 Fr (1.40 mm), 20 cm; (3) 18/18 G (dual-lumen), 5 Fr 5 days; immediate evaluation in the case of any malfunction or
(1.67 mm), 20 cm. adverse events (persistent withdrawal occlusion, subocclusion,
complete occlusion or extravasation from the exit site, and phle­
Since this study aimed at exploring the endpoints in the context bitis). No routine catheter replacement was planned. The catheters
of daily clinical practice, the decision whether to place a LPC or an were removed as soon as they were no longer needed (ie, when no
MC was based on the individual bedside nurse’s clinical judgment, further intravenous administration or blood sampling was expected)
according to the characteristics and expected duration of the in­ or in the presence of any major catheter-related complication (see
travenous therapy and her or his confidence and skill with one or the below). Conversely, in patients presenting minor complications,
other type of catheter. possible catheter removal was considered when the catheter
 A. Fabiani et al. / American Journal of Infection Control xxx (xxxx) xxx–xxx 3

functioning was compromised, or in the presence of phlebitis or Unadjusted survival analysis was conducted by comparing
exit-site infections. Kaplan-Meier curves, while the differences between the catheters’
survival rates were calculated using the Mantel-Cox log-rank test.
Catheter outcomes Multivariate Cox proportional-hazard analysis with stepwise for­
ward selection was run to estimate the time-to-event effect of the
The primary study outcome was related to catheter-failure, different catheters on the risk of a catheter-failure, adjusted for
which was defined as catheter removal because of loss of its us­ baseline confounders showing to be significantly related to the oc­
ability (due to any complication) and requiring a catheter replace­ currence of catheter-failure in bivariate analyses. The results were
ment. Catheter-related bloodstream infection and symptomatic presented as a proportional-hazard ratio (HR) with a corresponding
catheter-related thrombosis were considered as major complica­ 95% confidence interval (CI) and cumulative survival-adjusted
tions, while the presence of asymptomatic catheter-related throm­ curves. Moreover, after observing that virtually no catheter-failure
bosis, fibroblastic sleeve, or exit-site infection or phlebitis were occurred before the 10th day in both subgroups, in order to examine
deemed to be minor complications. All catheter-related complica­ the potential impact of survival bias among the catheter groups, we
tion incidences were reported as the number of cases per 1,000 compared the 9-day and 10- to 30-day catheter outcomes separately
catheter-days. for sensitivity analysis.
For all tests, the statistical significance was set at an alpha level of
Study endpoints P < .05. Statistical analyses were performed using the software SPSS
Statistics for Windows, version 24.0 (IBM Corp).
The primary study endpoint was the risk of catheter-failure,
calculated as the interval (days) between placement and removal.
The observations were censored at 30 days, based on the observed RESULTS
median indwelling times of 13 (interquartile range: 7-21) and 22
(interquartile range: 13-33) days for LPCs and MCs, respectively. During the study period, a total of 298 potentially eligible pa­
tients were considered for inclusion. After applying the exclusion
Other collected variables criteria (in 2 patients, the catheter placement failed, in 12, a poly­
ethylene LPC was placed, and 15 were lost to follow-up), 269 sub­
Data were collected both during the hospitalization and after jects (138 LPCs, 51.3%; 131 MCs, 48.7%) constituted the final study
hospital discharge for patients discharged with a catheter still in population. All patients received antiaggregant or anticoagulant
place. drugs.
The patient’s sex, age, and comorbidity condition—computed by Table 1 shows the main characteristics of the patients and ca­
the Charlson Comorbidity Index21—were collected to describe her or theters. At the insertion point, a statistically significant greater
his basic characteristics at hospital admission. percentage of LPCs exceeded the 33% vein-to-catheter thresholds
Data on the catheter type (length, internal and external diameter, compared with MCs, while overcoming of the 45% threshold was
and number of lumens) and cannulation procedure (insertion site, uncommon and similar between the catheter groups. At the tip level,
number of attempts, internal diameter of the cannulated vein at both the 33% and 45% thresholds were exceeded more frequently in
insertion, and tip levels) were documented, and the derived ca­ LPCs than in the MC subgroup. A significantly higher average
theter-to-vein diameter ratios were computed. number of drugs was infused daily via MCs compared with LPCs.
The type and amount of any fluids or medications administered Overall, the catheters’ indwelling time was 20.8 ± 19.2 days and
through the catheters was registered. Accordingly, the overall average was significantly (P < .001) longer for MCs (26.6 ± 23.8 days, max
daily number of different medications infused through a catheter during 184 days) than LPCs (15.4 ± 11.1 days, max 66 days).
its overall indwelling time was computed. For example, for a catheter left The incidence of catheter-failure was 5.7 and 3.4 per 1,000 ca­
in place for 20 days in a patient receiving 3 drugs for 7 days, 1 drug for theter-days for LPCs and MCs, respectively. A higher incidence of all
10 days, and no medications in the remaining 3 days, the average was explored outcomes was documented for LPCs compared with MCs
1.6 per day (3 × 7 + 1 × 10 + 0 × 3)/20. This variable was created to (Table 2). Overall, a catheter-failure was associated with the onset of
quantify the theoretical amount of “stress” caused to the venous en­ major complications in a minority of cases (catheter-related blood­
dothelium due to the administration of multiple medications. The same stream infection: 8.3%, symptomatic catheter-related thrombosis:
index was also computed by considering only the infusion of high-risk 8.3%). Other complications determining a catheter-failure were
medications. persistent withdrawal occlusion (41.7%), total occlusion (25%), ca­
The type and the date of detection of any catheter-related com­ theter breakage at the extravascular tract (8.3%), extravasation
plication, as well as the date and reason for catheter removal, were (4.2%), and exit-site inflammation (4.2%), often associated with the
documented. Moreover, any antiaggregant or anticoagulant medi­ presence of asymptomatic catheter-related thrombosis or fibro­
cation administered to the patient was recorded as they were con­ blastic sleeve.
sidered to be potential protective factors against venous thrombosis. A catheter-failure was significantly associated with the male sex,
catheter-to-vein ratio > 33% at the tip level, and with a higher
Statistical analysis average number of drugs infused daily through the catheter
(Table 3). Compared with LPCs, MCs showed a lower failure risk (log-
The continuous variables were presented as means and standard rank test 5.020, P = .025) in crude survival analysis (Fig. 1A) and in
deviations, according to the normality of the data distribution as univariable Cox statistics. This finding was confirmed (HR 0.311, 95%
assessed by the Kolmogorov-Smirnov test. The difference between CI 0.106-0.917, P = .034) when the model was adjusted for the
the means was analyzed using a parametric nonpaired t test, after baseline variables found to be associated with catheter-failure (sex,
determining whether equal variance could be attributed to the catheter-to-vein ratio > 33% at the tip level, Fig. 1B), where neither
subgroups according to Levene’s test. The nominal variables were the patient’s sex nor the catheter-to-vein ratio > 33% at the catheter
described as a number and percentage, and analyzed though con­ tip resulted in being statistically significant in the model. As ex­
tingency tables and the Pearson’s χ test or Fisher exact test, as ap­ pected, the sensitivity analysis showed a decreased risk for catheter-
propriate. failure for MCs only starting from the 10th day from catheter
4 A. Fabiani et al. / American Journal of Infection Control xxx (xxxx) xxx–xxx

Table 1
General characteristics of the study population and differences according to the catheter type

All catheters (n. 269) LPCs (n. 138) MCs (n. 131) P value

Patient characteristics
Age (y) 70.3 ± 13.3 72.3 ± 11.7 68.1 ± 14.5 .010
Sex (male) 154 (57.2%) 73 (52.9%) 81 (61.8%) .139
Charlson Comorbidity Index 3.3 ± 2.2 3.2 ± 2.0 3.5 ± 2.4 .309
Department
Cardiac surgery 143 (53.2%) 66 (47.8%) 77 (58.8%) .003
Cardiology 51 (19.0%) 21 (15.2%) 30 (22.9%)
Internal medicine 75 (27.9%) 51 (37.0%) 24 (18.3%)
Cannulated vein
Basilic 168 (62%) 75 (44.6%) 93 (55.4%) < .001
Brachial 69 (26%) 32 (46.4%) 37 (53.6%)
Cephalic 32 (12%) 31 (96.9%) 1 (3.1%)
Measures at the insertion level
Inner caliber of the vein (mm) 4.7 ± 1.4 4.4 ± 1.4 5.0 ± 1.4 .001
Depth of the vein (mm) 11.0 ± 4.0 10.3 ± 3.8 11.6 ± 4.1 .005
Catheter-to-vein ratio (%) 30.9 ± 9.6 32.2 ± 10.5 29.5 ± 8.3 .023
Catheter-to-vein ratio > 33% 81 (30.0%) 51 (37.0%) 30 (22.9%) .012
Catheter-to-vein ratio > 45% 14 (5.0%) 9 (6.5%) 5 (3.8%) .318
Measures at the tip level
Inner caliber of the vein (mm) 7.3 ± 2.8 5.9 ± 2.6 8.8 ± 2.3 < .001
Catheter-to-vein ratio (%) 21.8 ± 10.6 26.5 ± 12.4 16.7 ± 4.7 < .001
Catheter-to-vein ratio > 33% 34 (13.0%) 33 (23.9%) 1 (0.8%) < .001
Catheter-to-vein ratio > 45% 10 (4.0%) 10 (7.2%) 0 (0.0%) .002
Average number of drugs infused per day
Any medication 1.0 ± 0.7 0.7 ± 0.5 1.3 ± 0.8 < .001
High-risk medications only 0.2 ± 0.3 0.1 ± 0.2 0.3 ± 0.4 < .001

LPCs, long peripheral catheters; MCs, midline catheters.

Table 2 Table 3
Incidence per 1,000 catheter-days of the study outcomes Variables associated with differences in catheter-failure rates

All catheters LPCs MCs Catheter-failure P value

Catheter-failure 4.3 5.7 3.4 No Yes

CR-bloodstream infections 0.4 0.9 0.0
Patient characteristics
Catheter-related thrombosis* 5.4 8.7 3.5
Age (y) 70.2 ± 13.3 71.1 ± 13.6 .752
Symptomatic 0.4 0.5 0.3
Sex (male) 145 (59.2%) 9 (37.5%) .040
Asymptomatic* 5.1 8.2 3.2
Charlson Comorbidity Index 3.3 ± 2.2 3.1 ± 2.4 .593
Fibroblastic sleeve* 10.3 14.0 8.1
Department
CR, catheter-related; LPCs, long peripheral catheters; MCs, midline catheters. Cardiac surgery 131 (53.5%) 12 (50.0%) .728
*
n. 254. Cardiology 45 (18.4%) 6 (25.0%)
Internal medicine 69 (28.2%) 6 (25.0%)
Cannulated vein
positioning both in bivariate (log-rank test 4.219, P = .040) and Basilic 156 (63.7%) 12 (50.0%) .349
multivariate (HR 0.335, P = .051) analysis. Brachial 60 (24.5%) 9 (37.5%)
No statistically significant between-catheter difference was Cephalic 29 (11.8%) 3 (12.5%)
Measures at the insertion level
found when the secondary outcomes (catheter-related bloodstream
Catheter-to-vein ratio (%) 30.8 ± 9.7 31.8 ± 7.7 .620
infection, catheter-related thrombosis, and fibroblastic sleeve) were Catheter-to-vein ratio > 33% 70 (28.6%) 11 (45.8%) .079
explored (Fig. 1C-E). Catheter-to-vein ratio > 45% 13 (5.3%) 1 (4.2%) .810
Measures at the tip level
Catheter-to-vein ratio (%) 21.5 ± 10.5 24.1 ± 11.7 .264
DISCUSSION
Catheter-to-vein ratio > 33% 27 (11.0%) 7 (29.2%) .011
Catheter-to-vein ratio > 45% 8 (3.3%) 2 (8.3%) .210
In the study population, compared with LPCs, MCs showed an overall Average number of drugs infused per day
lower incidence of complications per 1,000 catheter-days and an 85% Any medication 0.9 ± 0.7 1.3 ± 0.8 .021
lower adjusted proportional risk of catheter-failure. Interestingly, the risk High-risk medications only 0.2 ± 0.3 0.3 ± 0.4 .102

seemed superimposable during the first 9 days, while from that mo­
ment, the risk clearly increased for the LPCs (Fig. 1). This impression was
confirmed by sensitivity analysis (although with a P value of .051 in attained less frequently for LPCs (Table 1). This result is consistent
multivariable regression, probably because of the exclusion of a number with the normal anatomy of the venous system, since the vein size—
of subjects from this analysis). Although confirming previous literature and consequently its blood flow—is expected to gradually increase
results,8 this finding was somewhat surprising since we compared ca­ while approaching the large thoracic vessels. In bivariate analyses,
theters made of an identical biomaterial (and produced by the same we found that exceeding the recommended catheter-to-vein ratio
company), having their length (20 cm for MCs and 8-10 cm for LPCs) as was associated with a catheter-failure, confirming previously re­
the sole differentiating feature. Precisely this different length seems to ported literature results.22
offer a plausible explanation for their markedly different degree of re­ Furthermore, we found that a higher average number of drugs
liability. infused daily through the catheter was associated with a higher risk
The aim of respecting a catheter-to-vein ratio less than or equal for catheter-failure. However, despite theoretically experiencing a
to 33% at the tip level was essentially always achieved for MCs, being higher vein stress (Table 1), overall MCs had a lower catheter-failure
 A. Fabiani et al. / American Journal of Infection Control xxx (xxxx) xxx–xxx 5

Fig. 1. (A, B) Crude and adjusted Kaplan-Meier survival curves of catheter-failure risk for the compared catheter groups. (C, D, E) Crude Kaplan-Meier survival curves of the
secondary study outcomes for the compared catheter groups.

risk compared with LPCs: in other terms, MCs showed a greater Conversely, the length of LPCs allows the tip to be positioned within
reliability, despite being used longer and more intensively. Inter­ the cannulated vein (eg, basilic, brachial, and cephalic) or, con­
estingly, the rate of catheter-failure was similar when considering sidering that all catheters were inserted at the Dawson green zone,
only the infusion of high-risk medications as a vein stress factor at most within the very proximal segment of the axillary vein: at this
(Table 3): this could be explained by considering that, in our popu­ level, the drugs’ dilution will only rely on the blood flow of the
lation, all high-risk medications were administered only after ap­ cannulated vein itself (Fig. 2A).
propriate dilution. These results are consistent with the previous In summary, the MC tip tends to be placed at a level where a
literature,22,23 and seem to suggest that the risk was associated with higher catheter-to-vein ratio is easier to ensure, allowing for a faster
the number of medications and the continuity with which they were and less-turbulent blood flow, thus enabling a quicker dilution and
administered, rather than their chemical characteristics. transport, and a shorter contact of any administered medications
Based on these findings, we may speculate that the reason for with the vessel wall, consequently reducing the risk of thrombo­
such a greater MC reliability lies in its greater length and, thus, in the genesis.25,26 This reasoning is consistent with previous research,
different blood flow present at its tip. Indeed, the MC tip predictably demonstrating that the further the venous segments are from the
reaches the axillary vein at its distal thoracic segment, immediately vena cava, the higher and earlier the risk of thrombosis.27
before the conjunction with the subclavian vein, where a greater We believe a brief discussion of the results of the secondary
blood flow is expected as, at this level, the axillary vein collects the study endpoints might be useful. The incidence of catheter-related
blood drained from other tributary veins (Fig. 2B). Interestingly, it bloodstream infections was extremely low. However, while no cases
has been demonstrated that almost 60% of people, irrespective of sex were documented with regard to MCs, 2 cases of catheter-related
and body side, have an accessory axillary vein originating from the bloodstream infections have occurred in the LPC subgroup. After
lateral brachial, the common brachial or the deep brachial vein in a analyzing the clinical documentation, we noticed that the compli­
substantially similar proportion of cases. The presence of this ac­ cations that occurred after the patients were transferred to clinical
cessory vessel contributes to the venous drainage of the upper limb settings different from those in which the catheter was positioned,
and can act as an important collateral circulation path in the event of where the actual adherence to the protocol for catheter access and
an axillary vein obstruction, for example, due to a thrombosis.24 maintenance was difficult to assess and monitor. In addition,
6 A. Fabiani et al. / American Journal of Infection Control xxx (xxxx) xxx–xxx

Fig. 2. Schematic example of the different blood flows expected the tip level for long peripheral (A) and midline (B) catheters. Dotted green line: Dawson zone. Dotted red circle:
tip catheter area.

considering the responsible isolated microorganisms (ie, technique” can make this choice advantageous when less-complex
Staphylococcus epidermidis and Enterococcus fecalis), we think that or short-term therapeutic plans have to be managed in DIVA pa­
both catheter-related bloodstream infections were due to a con­ tients. According to our results, an LPC would be appropriate to be
tamination ab extrinseco of the devices due to diminished health used for patients with anticipated intravenous requirements for no
care providers’ attention. more than 9 to 10 days, whereas an MC would be the better choice
Thrombosis is a common complication associated with vascular when a more prolonged therapy is expected.
access, caused by endothelium damage of the vein. It can remain
asymptomatic, manifest itself locally with inflammation or catheter Limitations
malfunction (persistent withdrawal occlusion, sub/total occlusion,
and extravasation), or, although rarely, lead to more serious systemic The main limitation of this study is related to its observational
complications such as pulmonary embolism.28,29 In our population, design, which intrinsically exposes to the risk of bias (eg, we en­
the risk of catheter-related thrombosis was similar between MCs rolled a sample of patients in which confounding factors could not
and LPCs, although a trend toward a reduced risk in the case of MC be controlled a priori, explaining the presence of several baseline
seems to emerge by observing the crude Kaplan-Meier curves, which differences in the 2 cohorts) and limits the generalizability of the
did not achieve statistical significance, probably due to the low results. Therefore, particular caution is required when considering
sample size. the reported associations between the catheter type and the ex­
Fibroblastic sleeve is a phenomenon completely different from plored outcome, which should be demonstrated by randomized
catheter-related thrombosis. The high incidence and the early pre­ controlled trials. Moreover, there is no evidence that the average
sentation of fibroblastic sleeve found in the present investigation for number of different medications infused might correlate with ve­
both types of catheter seems to confirm its pathophysiology, that nous endothelial stress, therefore, this index should be considered
describes it as a “foreign body reaction” inducing the deposition of only as representing the more or less-intensive use of the catheter
fibronectin, a circulating protein produced by the liver, on the ex­ during its indwelling time. Finally, it was not possible to perform a
ternal surface of the catheter. Fibronectin attracts blood macro­ regular ultrasound assessment in the case of patients who were
phages that differentiate into smooth muscle cells and fibroblast, discharged from the hospital, this may have led to an under­
and starts to produce collagen,28,30 leading to the development of a estimation of some complications such as fibroblastic sleeve and
sleeve made of connective tissue around the catheter surface. We asymptomatic catheter-related thrombosis, which may have ap­
think to be of interest having documented that polyurethane MCs peared at a later time.
and LPCs were equally associated with fibroblastic sleeve incidence,
irrespective of their length and the potentially associated blood flow CONCLUSIONS
at the tip level.
Based on our results, nurses should choose to place a MC when a Considering the same biomaterial and equal terms of manage­
medium- to long-term therapeutic plan is expected, especially when ment, the length of the catheter seems to be the characteristic
the administration of multiple medications is planned, or when they having the greater impact on the risk of complications. In the pre­
have the need to use dual-lumen devices: the reduced risk of ca­ sent study, the use of MCs was associated with a longer un­
theter-failure associated with these devices may increase the prob­ complicated catheter indwelling time, suggesting this should be the
ability that the therapeutic plan be completed without first choice for patients needing more complex or longer-duration
complications. It should be noted however, that positioning a MC therapies. Conversely, LPCs should be reserved for use with patients
may require longer time and greater skill, impacting on health care with DIVA conditions and needing for a shorter therapeutic plan.
organization. Accordingly, we believe that LPCs may play an im­ Further studies are needed to confirm these results in different
portant role in daily clinical practice: the greater simplicity and populations and taking into account catheters of different lengths
speed of LPC positioning allowed by the simplified “Seldinger and made of different biomaterials.
 A. Fabiani et al. / American Journal of Infection Control xxx (xxxx) xxx–xxx 7

References 17. Brescia F, Pittiruti M, Spencer TR, Dawson RB. The SIP protocol update: eight
strategies, incorporating Rapid Peripheral Vein Assessment (RaPeVA), to mini­
1. Hopkinson SG, Green AK, Hewitt BE, Grace SA. Short peripheral catheter dwell time mize complications associated with peripherally inserted central catheter inser­
and associated complications: a systematic review. J Infus Nurs. 2020;43:200–207. tion. J Vasc Access. 2022;25:5–13.
2. Zingg W, Pittet D. Peripheral venous catheters: an under-evaluated problem. Int J 18. Spencer TR, Pittiruti M. Rapid Central Vein Assessment (RaCeVA): a systematic,
Antimicrob Agents. 2009;34:S38–S42. standardized approach for ultrasound assessment before central venous cathe­
3. Gorski LA. The 2016 infusion therapy standards of practice. Home Healthc Now. terization. J Vasc Access. 2019;20:239–249.
2017;35:10–18. 19. Dawson RB. PICC Zone Insertion MethodTM (ZIMTM): a systematic approach to
4. Loveday HP, Wilson J, Pratt RJ, et al. epic3: national evidence-based guidelines for determine the ideal insertion site for PICCs in the upper arm. J Assoc Vasc Access.
preventing healthcare-associated infections in NHS hospitals in England. J Hosp 2011;16:156–165.
Infect. 2014;86:S1–S70. 20. Manrique-Rodríguez S, Heras-Hidalgo I, Pernia-López MS, et al. Standardization
5. Fields JM, Piela NE, Au AK, Ku BS. Risk factors associated with difficult venous and chemical characterization of intravenous therapy in adult patients: a step
access in adult ED patients. Am J Emerg Med. 2014;32:1179–1182. further in medication safety. Drugs RD. 2021;21:39–64.
6. Bahl A, Johnson S, Alsbrooks K, Mares A, Gala S, Hoerauf K. Defining difficult 21. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying
intravenous access (DIVA): a systematic review. J Vasc Access. prognostic comorbidity in longitudinal studies: development and validation.
2021;17:11297298211059648. J Chronic Dis. 1987;40:373–383.
7. Pittiruti M, Van Boxtel T, Scoppettuolo G, et al. European recommendations on 22. Fabiani A, Santoro M, Sanson G. The catheter-to-vein ratio at the tip level, not the
the proper indication and use of peripheral venous access devices (the ERPIUP catheter type, as a risk factor for a catheter failure. A retrospective comparative
consensus): a WoCoVA project. J Vasc Access. 2021;24:165–182. study of polyurethane midline and long peripheral catheters. Heart Lung.
8. Fabiani A, Aversana N, Santoro M, Sanson G. Complications associated to midline-and 2023;60:39–44.
long peripheral catheters in adults. Systematic review of literature and proposal for a 23. Bahl A, Hijazi M, Chen NW. Vesicant infusates are not associated with ultrasound-
standardized model for data collection. Thromb Res. 2024;236:117–126. guided peripheral intravenous catheter failure: a secondary analysis of existing
9. Badger J. Long peripheral catheters for deep arm vein venous access: a systematic data. PLoS One. 2022;17:e0262793.
review of complications. Heart Lung. 2019;48:222–225. 24. Felix VB, dos Santos JAB, Fernandes KJ, et al. Anatomical study of the accessory
10. Fabiani A, Dreas L, Sanson G. Ultrasound-guided deep-arm veins insertion of long axillary vein in cadavers: a contribution to the axillary surgical approach. J Vasc
peripheral catheters in patients with difficult venous access after cardiac surgery. Bras. 2016;15:275–279.
Heart Lung J Acute Crit Care. 2017;46:46–53. 25. Nifong TP, McDevitt TJ. The effect of catheter to vein ratio on blood flow rates in a
11. Fabiani A, Eletto V, Dreas L, Beltrame D, Sanson G. Midline or long peripheral simulated model of peripherally inserted central venous catheters. Chest.
catheters in difficult venous access conditions? A comparative study in patients 2011;140:48–53.
with acute cardiovascular diseases. Am J Infect Control. 2020;48:1158–1165. 26. Sharp R, Carr P, Childs J, et al. Catheter to vein ratio and risk of peripherally
12. Qin KR, Pittiruti M, Nataraja RM, Pacilli M. Long peripheral catheters and midline inserted central catheter (PICC)-associated thrombosis according to diagnostic
catheters: insights from a survey of vascular access specialists. J Vasc Access. group: a retrospective cohort study. BMJ Open. 2021;11:e045895.
2021;22:905–910. 27. Zhang X, Chen H, Jing W, et al. The clinical topography of peripherally inserted
13. Qin KR, Ensor N, Barnes R, Englin A, Nataraja RM, Pacilli M. Long peripheral ca­ central catheter-related thrombosis in cancer patients: a prospective and long­
theters for intravenous access in adults and children: a systematic review of the itudinal observational study based on ultrasound scans every two days. Thromb
literature. J Vasc Access. 2021;22:767–777. Res. 2023;229:232–242.
14. DeVries M, Lee J, Hoffman L. Infection free midline catheter implementation at a 28. Passaro G, Pittiruti M, La Greca A. The fibroblastic sleeve, the neglected compli­
community hospital (2 years). Am J Infect Control. 2019;47:1118–1121. cation of venous access devices: a narrative review. J Vasc Access.
15. Buetti N, Marschall J, Drees M, et al. Strategies to prevent central line-associated 2021;22:801–813.
bloodstream infections in acute-care hospitals: 2022 update. Infect Control Hosp 29. Pinelli F, Balsorano P, Mura B, Pittiruti M. Reconsidering the GAVeCeLT Consensus
Epidemiol. 2022;43:553–569. on catheter-related thrombosis, 13 years later. 2020;22:501–508.
16. Gorski LA. A look at 2021 infusion therapy standards of practice. Home Healthc 30. Wilson CJ, Clegg RE, Leavesley DI, Pearcy MJ. Mediation of biomaterial–cell in­
Now. 2021;39:62–71. teractions by adsorbed proteins: a review. Tissue Eng. 2005;11:1–18.