Original Article

Page 1 of 13

Association between electronic health record use and quality of

care in high Medicaid nursing homes
Ganisher Davlyatov1^, Justin Lord2^, Akbar Ghiasi3, Robert Weech-Maldonado4^
1
Department of Health Administration and Policy, Hudson School of Public Health, University of Oklahoma Health Sciences Center, Oklahoma
City, OK, USA; 2Department of Health Administration, College of Business, Louisiana State University, Shreveport, LA, USA; 3Department of
Health Administration, H-E-B School of Business and Administration, University of the Incarnate Word, San Antonio, TX, USA; 4Department of
Health Services Administration, School of Health Professions, University of Alabama at Birmingham, Birmingham, AL, USA
Contributions: (I) Conception and design: R Weech-Maldonado, G Davlyatov, J Lord; (II) Administrative support: R Weech-Maldonado; (III)
Provision of study materials or patients: R Weech-Maldonado; (IV) Collection and assembly of data: G Davlyatov, A Ghiasi, J Lord; (V) Data analysis
and interpretation: G Davlyatov; (VI) Manuscript writing: All authors; (VII) Final approval of manuscript: All authors.
Correspondence to: Ganisher Davlyatov, PhD, MD. Department of Health Administration and Policy, Hudson School of Public Health, University of
Oklahoma Health Sciences Center, 801 Northeast 13th Street, Oklahoma City, OK 73126, USA. Email: gdavlyat@ouhsc.edu.

Background: Nursing homes operating in resource-constrained environments typically have lower

professional staffing and worse quality. Electronic health records (EHRs) have been utilized as an effective
tool to improve the quality of care in nursing homes. This study examines the association between EHR use
and the quality of care in high Medicaid nursing homes.
Methods: The study used primary and secondary data from Brown University’s Long-Term Care Focus,
Nursing Home Compare, Area Health Resource File, and Medicare Cost Reports for the years 2017–2018.
The primary survey data was collected through a national mailer to Directors of Nursing (DONs) in high-
Medicaid nursing homes. The dependent variable, nursing home quality, was conceptualized using Nursing
Home Compare Five-Star Quality Rating System where the higher score represents better quality (1 to 5).
The independent variable, EHR score, was a composite measure developed from 23 items. Ordered logistic
regression was used to model the relationship between the average EHR score and the quality star rating in
high-Medicaid nursing homes.
Results: There was a significant positive relationship between the average EHR score and the five-star
quality rating. For a one unit increase in the average EHR score, the odds of being in a higher star rating
category increases by 50%. Additional factors, such as, being a not-for-profit, having higher occupancy rate,
and being located in a higher per capita income county were significantly associated with higher quality.
Conclusions: We found that EHR use in high-Medicaid nursing homes was positively associated
with improvements in quality. This finding provides additional support to the promising role of EHR in
improving quality of care among resource-constrained nursing homes. These under-resourced nursing
homes face challenges as it relates to quality, the adoption and use of EHRs may facilitate improvements in
quality of care.

Keywords: Electronic health record (EHR); quality of care; nursing home; health disparities

Received: 20 May 2020; Accepted: 06 November 2020; Published: 25 September 2021.

doi: 10.21037/jhmhp-20-64
View this article at: http://dx.doi.org/10.21037/jhmhp-20-64

^ ORCID: Ganisher Davlyatov, 0000-0001-9410-9696; Justin Lord, 0000-0002-4557-955X; Robert Weech-Maldonado, 0000-0002-5005-
0909.

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Page 2 of 13 Journal of Hospital Management and Health Policy, 2021

Introduction systems. Under-resourced nursing homes may be even less

likely to adopt EHRs as compared with other facilities with
The need for long-term care is booming as the projected
adequate levels of resources. Given the benefits that EHR
number of people requiring nursing home care is expected
can have on resident quality, we wanted to explore the
to jump from 1.3 million in 2010 to 2.3 million by 2030 (1).
adoption and use of EHR as it related to quality of care in
The rapid growth of the aging population has placed an
under-resourced nursing homes.
increased burden on the long-term care industry (2). High-
Medicaid nursing homes may face even more pressure
due to their payer-mix and low level of resources. High- Conceptual framework
Medicaid nursing homes are described as having a high
Using tenets from Donabedian’s Structure-Process-
proportion of Medicaid residents (85% or higher), a lower
Outcome (SPO) model and knowledge-based view of
percentage of private pay residents (10% or less), and a low
the firm, we examine the relationship between the EHR
percent of residents on Medicare (8% or less) (3). Quality
use and quality of care in high-Medicaid nursing homes.
of care is a prevalent issue facing all nursing homes, yet this
issue may be worse for high-Medicaid nursing homes as Donabedian’s SPO model is one of the most widely used
they are often characterized as having lower professional frameworks in examining the factors that influence quality
staffing and worse quality (3). of care (18). Structure captures the resources that a nursing
There have been many efforts to address the challenges home has in place to deliver care. Structural elements have
of delivering high quality care in nursing homes. The use been conceptualized as the presence of EHR, number of
of electronic health records (EHR) may be one effective nurses per patient, and even leadership styles. Process refers
tool in improving the quality of care in long-term care to the activities health care providers utilize to deliver
settings (4). Unlike traditional paper-based documentation, care, such as, the appropriate use of catheters, providing
which is often illegible, inconsistent (5), error-prone, and fall-related guidance, and other health related activities.
difficult to update (6), EHRs help provide accurate, up- Outcomes are the results of health care provider activities.
to-date, and complete information at the point of care; For example, whether the residents received their flu
improve efficiency and productivity; and enable quick access shots, and/or how many residents fell and so on. EHR
to patient records (7,8). Further, facilities can access and implementation and use, is an integral part of the nursing
share patient information from different organizations, homes structure and has been found to help organizations
which improves care coordination and patient safety, while deliver better care by minimizing errors and improving
potentially leading to reduced costs (9). However, there safety (16,19). This study used the SPO model to examine
are barriers to EHR implementation in nursing homes the relationship among the structure and outcomes,
especially those that are under-resourced. through the use of EHRs in high-Medicaid nursing homes
The adoption and implementation of EHR systems can and resident quality outcomes.
be costly. Unlike hospitals and ambulatory care providers, According to the knowledge-based view, knowledge
nursing homes were excluded from federal incentives, such is a source of a competitive advantage (20). Knowledge
as the Health Information Technology for Economic and management is defined as the process of creating or locating
Clinical Health (HITECH) Act that have been instrumental knowledge and managing the dissemination of knowledge
in helping providers secure public funds to offset EHR within and between organizations (21,22). EHRs allow for
adoption costs (10,11). As such, EHR adoption rates have information/knowledge to be collected, communicated,
been around 65% (10), while hospitals and ambulatory a n d a c t e d u p o n m o r e e f f i c i e n t l y a n d e f f e c t i v e l y.
care providers have EHR adoption rates over 80% (12,13). Knowledge managmenet has been conceptualized using
Additional barriers to nursing home EHR adoption are three dimensions: knowledge acquisition; knowledge
the costs associated with EHR training, infrastructure, and dissemination, and knowledge responsiveness (23,24).
maintenance (14,15). Studies have found that the adoption Knowledge acquisition pertains to the processes used
and use of EHR has had a mostly positive effect on the to locate, create, or discover knowledge. This may also
quality of care in nursing facilities (16,17). Despite the include systems, like an EHR, that helps nursing homes
benefits attributed to EHRs, nursing homes lag behind collect and use data. Knowledge dissemination pertains
other healthcare settings in adoption and use of these to how knowledge is distributed and applied throughout

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Journal of Hospital Management and Health Policy, 2021 Page 3 of 13

the nursing homes. For example, the accessibility of Variables

resident data that an EHR provides can facilitate staff
The main dependent variable was the five-star quality rating
care. Knowledge responsiveness relates to the way the
obtained from The Centers for Medicare and Medicaid
organization utilizes knowledge and includes activities. The
Services’ (CMS) Nursing Home Compare Five-Star Quality
responsiveness can be seen as how quickly the organization
Rating System. The Nursing Home Compare website
responds to residents or implement changes based on
rates nursing homes on a scale between 1 (nursing homes
information/knowledge. Knowledge management processes
with quality below average) and 5 (nursing homes with
can improve organizational outcomes and quality because
quality above average). The five-star quality rating include
it supports better decision-making (22). EHR use can help
three sources to determine a nursing home’s overall rating:
nursing homes process, disseminate, and react to knowledge
health inspections, staffing, and quality measures. The
more quickly thus improving quality outcomes, therefore, it
health inspection rating captures information on standard,
is hypothesized:
as well as, complaint surveys. The staffing rating has
H1: Greater EHR implementation will have a positive impact
information about the number of hours of care provided on
on resident care quality in high-Medicaid nursing homes.
average to each resident daily by nursing staff. The quality
We present the following article in accordance with
measures rating has information on nine different clinical
the SURGE reporting checklist (available at http://dx.doi.
measures, such as, number of hospitalizations, number of
org/10.21037/jhmhp-20-64).
outpatient emergency department visits, bladder catheter
use, antipsychotic medication use, pressure ulcers, urinary
Methods tract infection, flu shot, pneumonia vaccine, and changes in
resident mobility.
Data
The main independent variable—EHR score—was
The study used primary and secondary data sources for comprised of 23 items from dimensions of administrative
the years of 2017–2018. The primary survey data was functions, documentation, order entry, results viewing, and
collected through a national mailer to Directors of Nursing clinical tools (Table 1) (25). The administrative functions
(DONs) in high-Medicaid nursing homes. Primary survey included administrative processes and reporting such as
data of nursing home administrators was collected through scheduling systems and clinical task assignments. The
three rounds of mailed and online surveys. A cover letter documentation included health information and data such as
outlining the purpose of this survey and signed by the resident demographics and medical history. The order entry
Principle Investigator of this project (#1R01HS023345-01) had order management information including medication
was sent to all participants. As an incentive for study order entry. The results viewing had data to help manage
participation, the respondents of the survey were sent a $25 results such as routing, managing, and presenting test
gift card. The first round of surveys was sent to all nursing results to clinical personnel for review. The clinical tool
homes (n=1,518) who had a 85% or higher Medicaid had decision support system and telemonitoring/telehealth
census. Additional criteria were applied to the sample size data. Each item had four response options (0= not available,
that excluded nursing homes with more than 10% of private 1= paper only, 2= paper and electronic, 3= fully electronic)
pay and greater than 8% supported by Medicare (3), which (Table 2). The composite score was the average of 23 items.
led to a sample size of 1,050. In the end, we had received Control variables included organizational-level
391 responses for a response rate of 37%. (ownership, chain affiliation, size, occupancy rate, Medicare
Survey data were merged with secondary datasets and Medicaid payer mix, use of nurse practitioners/
including Brown University’s Long-Term Care Focus physician assistants, Acuity Index, and race/ethnicity), and
( LT C F o c u s ) , N u r s i n g H o m e C o m p a r e , a n d A r e a county-level factors (Medicare Managed Care Organization
Health Resource File. LTCFocus data provides nursing market penetration, per capita income, educational level,
home organizational, demographic, quality, and market unemployment rate, poverty level and competition/
information. The Nursing Home Compare data provides Herfindahl-Hirschman Index (HHI), location, and percent
quality of resident care and staffing information. The Area of individuals over 65). Ownership was a categorical variable
Health Resource File provides market and demographic that identified whether a nursing home was for-profit (0=
information for the county. for-profit), not-for-profit (1= not-for-profit) or government-

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Page 4 of 13 Journal of Hospital Management and Health Policy, 2021

Table 1 Electronic Health Records (EHR) Functionalities and Components

EHR Functionalities Components Mean (SD) Cronbach’s Alpha

Administration Minimum data set 1.80 (0.41) 0.82

Assessments for residents, protocols, care area 1.36 (0.77)

Financial management 1.42 (0.57)

Quality improvement and reporting 1.02 (0.66)

Patient care planning 1.35 (0.71)

Task list 1.08 (0.88)

Documentation Patient demographics 1.51 (0.61) 0.92

Advance directives 0.70 (0.67)

Medical history 0.90 (0.73)

Clinical notes 1.19 (0.82)

Problem list 1.17 (0.82)

Allergy list 1.37 (0.71)

Medication administration record 1.24 (0.89)

Treatment administration record 1.22 (0.89)

Summary reports (transfer, discharge, consult) 1.04 (0.73)

Orders Medication order entry 1.24 (0.82) 0.95

Other order entry 1.23 (0.79)

Results viewing Labs 0.94 (0.66)

0.90
Radiology 0.87 (0.69)

Diagnostic tests 0.79 (0.66)

Consults 0.66 (0.64)

Clinical tools Clinical decision support 0.91 (0.72) 0.91

Telemonitoring/telehealth 0.96 (0.77)

owned (3= government-run). Chain affiliation reflected and who require assistance with ambulation or transfers.
whether the nursing home was part of a chain (0= free- Race/ethnicity was the proportion of nursing home residents
standing; 1= chain affiliated). Size captured the total number who were Black, Hispanic, and other.
of beds within the nursing home. Occupancy rate was the With regards to market-level factors, Medicare Advantage
percentage of occupied nursing home beds. Payer mix (MA)/managed care market penetration was calculated as
identified the proportion of the facilities residents who were the proportion of all Medicare beneficiaries in the county
on Medicaid and Medicare. Presence of nurse practitioners/ who were enrolled in a MA plan. Per capita income is a
physician assistants simply indicated if the facility had a measure of the average wealth of individuals in a county.
nurse practitioner/physician assistant or not. The Acuity Educational level was percent of individuals in a county with
Index was an average measure of the resident’s level of care a high-school degree or better. Unemployment rate was the
needed. This measure was based on the number of residents percent of individuals in the county who were unemployed.
needing various levels of assistance with mobility, activities Poverty level was the percent of persons and families below
of daily living (ADL), special treatments, as well as, the poverty threshold in the county, as defined by the Office of
proportion of residents that are bedfast, exhibit dementia, Management and Budget’s Statistical Policy Directive 14.

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Journal of Hospital Management and Health Policy, 2021 Page 5 of 13

Table 2 Electronic Health Record Questionnaire Table 2 (continued)

Administrative functions •  Fully electronic

A. Minimum Data Set Assessment/Resident Assessment •  Not available/not applicable

Protocol/Care Area Assessments H. Advance directives
•  Paper only (no automation) •  Paper only (no automation)
•  Paper and electronic •  Paper and electronic
•  Fully electronic •  Fully electronic
•  Not available/not applicable •  Not available/not applicable
B. Assessments other than Minimum Data Set I. Medical history
•  Paper only (no automation) •  Paper only (no automation)

•  Paper and electronic •  Paper and electronic

•  Fully electronic •  Fully electronic

•  Not available/not applicable •  Not available/not applicable

C. Financial management J. Clinical notes

•  Paper only (no automation) •  Paper only (no automation)

•  Paper and electronic •  Paper and electronic

•  Fully electronic •  Fully electronic

•  Not available/not applicable
•  Not available/not applicable
K. Problem list
D. Quality improvement and reporting
•  Paper only (no automation)
•  Paper only (no automation)
•  Paper and electronic
•  Paper and electronic
•  Fully electronic
•  Fully electronic
•  Not available/not applicable
•  Not available/not applicable
L. Allergy list
E. Patient care planning
•  Paper only (no automation)
•  Paper only (no automation)
•  Paper and electronic
•  Paper and electronic
•  Fully electronic
•  Fully electronic
•  Not available/not applicable
•  Not available/not applicable
M. Medication administration record
F. Task list (e.g., CNA workflow)
•  Paper only (no automation)
•  Paper only (no automation)
•  Paper and electronic
•  Paper and electronic
•  Fully electronic
•  Fully electronic •  Not available/not applicable
•  Not available/not applicable N. Treatment administration record
Documentation •  Paper only (no automation)
G. Patient demographics •  Paper and electronic
•  Paper only (no automation) •  Fully electronic
•  Paper and electronic •  Not available/not applicable
Table 2 (continued) Table 2 (continued)

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Page 6 of 13 Journal of Hospital Management and Health Policy, 2021

Table 2 (continued) Table 2 (continued)

O. Summary reports including transfer, discharge, and consults •  Not available/not applicable

•  Paper only (no automation) •  Clinical tools

•  Paper and electronic V. Clinical decision support

•  Fully electronic •  Paper only (no automation)

•  Not available/not applicable •  Paper and electronic

Order entry •  Fully electronic

P. Medication order entry •  Not available/not applicable

•  Paper only (no automation) W. Telemonitoring/telehealth

•  Paper and electronic •  Paper only (no automation)

•  Fully electronic •  Paper and electronic

•  Not available/not applicable •  Fully electronic

Q. Other order entry •  Not available/not applicable

•  Paper only (no automation) For each function listed below, please indicate the level of
automation (or computerization) currently in use at your facility.
•  Paper and electronic

•  Fully electronic

•  Not available/not applicable Competition was conceptualized using the HHI, which is
measured as the sum of the squared of the market shares
Results Viewing
(based on beds) for nursing homes in a county. HHI is a
R. Labs continuous variable that ranges from 0 to 1 with lower values
•  Paper only (no automation) associated with higher competition—a HHI score close to
•  Paper and electronic zero would represent perfect competition. The location
variable was included to capture the difference as it related
•  Fully electronic
to different markets. It had urban and rural categories where
•  Not available/not applicable the urban was the reference category. Number of individuals
S. Radiology over the age of 65 was the proportion of all individuals who
•  Paper only (no automation)
were 65 and older to the total population.

•  Paper and electronic

•  Fully electronic
Analysis

•  Not available/not applicable To adjust for potential non-response bias of nursing homes
not participating in the survey, we included propensity score
T. Other diagnostic tests
weights in the regression analysis (26). The propensity score
•  Paper only (no automation)
weights were calculated as the inverse of the propensity
•  Paper and electronic scores for nursing homes that participated in the survey. To
•  Fully electronic estimate the propensity score, we used a logistic regression
model where we regressed respondence status (respondent
•  Not available/not applicable
=1, non-respondent =0) on the control variables: size,
U. Consults ownership status, chain affiliation, payer mix, acuity index,
•  Paper only (no automation) occupancy rate, race/ethnicity, registered nurse staffing mix,
•  Paper and electronic registered nurse hours per resident day, licensed practical
nurse hours per resident day, certified nursing assistant
•  Fully electronic
hours per resident day, Medicare MCO market penetration,
Table 2 (continued) per capita income, poverty, unemployment, education,

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Journal of Hospital Management and Health Policy, 2021 Page 7 of 13

Table 3 Descriptive statistics of the nursing home sample (N=391) Table 3 (continued)
Mean/ Standard Mean/ Standard
Variable Variable
frequency deviation, % frequency deviation, %

Dependent variable County-level control variables

Star rating Medicare advantage penetration 29.4 14.9

(%)
* 38 10%
Per capita income $43,332 $13,800
** 72 18%
Educational level (with high 84.6 6.4
*** 93 23%
school)
**** 91 24%
Unemployment rate 5.8 1.8
***** 97 25%
Poverty level 17.8 6.5
Independent variable
Competition (Herfindahl- 0.2 0.3
Electronic Health Record Implementation Hirschman index)

Administrative function 1.3 0.5 Location (urban)

Documentation 1.1 0.6 Urban 369 94%

Order entry 1.2 0.8 Rural 22 6%

Results viewing 0.8 0.6 Percent of population over 65 15.1 3.3

Clinical tools 0.9 0.7

Organizational-level control variables

Ownership (for-profit) competition (HHI), location, and percent of individuals

For-profit 265 68% over 65. Then we calculated the inverse of the propensity
Not-for-profit 79 20% score, the propensity score weight, to include in the models.
Given the ordered nature of the dependent variable (star
Government-run 47 12%
rating), ordered logistic regression was used to model the
Chain affiliated (yes) relationship between average EHR score and quality star
Yes 226 58% rating. As a sensitivity analysis, we ran a separate model
No 165 42%
where we recoded EHR score into low, medium, and high
tertiles. Stata 16 was utilized for data management and
Size (number of total beds) 103.3 70.0
analysis, and statistical tests were evaluated at the 0.05 level
Occupancy rate (%) 85.3 13.1 of significance. The study was conducted in accordance
Payer-Mix: Medicaid (%) 88.3 7.2 with the Declaration of Helsinki (as revised in 2013). The
study was approved by the Institutional Review Board of
Payer-Mix: Medicare (%) 4.7 4.4
the University of Alabama at Birmingham (IRB-140828005)
Payer-Mix: other (%) 6.9 5.7 and informed consent was taken from all the survey
Employ nurse practitioners/ 153 39.3% participants.
physician assistants

Acuity Index (%) 11.8 2.4

Results
Percent of White residents (%) 61.3 31.8
The measure of internal consistency, Cronbach’s alpha,
Percent of Black residents (%) 19.3 26.4
among the EHR items was 0.86 (Table 1). Table 3 provides
Percent of Hispanic residents (%) 5.3 15.1 descriptive statistics of the high Medicaid nursing home
Percent of other race/ethnicity 14.1 21.0 whose DONs completed the survey instrument. Most of the
nursing homes were for-profit and chain-affiliated. These
Table 3 (continued)
nursing homes had 88% Medicaid and 4.7% Medicare

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Page 8 of 13 Journal of Hospital Management and Health Policy, 2021

payer-mix. The average occupancy was around 85%. These timely access to resident data (29). Further, because of
under resourced nursing homes had a resident-mix of 19% their complex chronic care needs, nursing home residents
Black, 5.3% Hispanic, and 14.1% other non-White. They frequently transition between their homes and different care
were more likely to be in urban environments with higher settings (30). Therefore, EHR adoption is vital in nursing
levels of competition. homes as it may facilitate successful transitions of care (31).
The proportional odds model (Table 4) yielded a As such, EHR may be an important structural component
statistically significant positive relationship (OR =1.5, of nursing home resident care, which may facilitate
P<0.05) between the average EHR score and five-star knowledge management and better processes of care, and
quality rating, thus, our hypothesis was supported. For a one may ultimately be associated with better outcomes of care,
unit increase in EHR score, the odds of being in a higher such as the nursing homes’ five-star quality ratings.
star rating category would increase by 50%. Sensitivity Due to the cross-sectional nature of the data, the findings
analysis (Table 5) revealed that nursing homes with high from this study are limited to providing associations
EHR implementation had higher odds of being in a higher between nursing home quality and average EHR score.
star rating category (OR =1.75, P<0.05), as compared with Additionally, this study was focused on under-resourced
low EHR category. Additionally, not-for-profits, occupancy nursing homes and these findings may not be applicable to
rate, and per capita income were significantly associated the nursing home population as a whole. Nevertheless, this
with higher quality. Specifically, nursing homes that were is the first study that examined the relationship between
not-for-profit, compared to for-profit nursing homes, had EHR scores and quality of care in resource-constrained
higher odds (OR =2.2, P<0.01) of being in a higher star High Medicaid nursing homes.
rating category. For a one percent increase in the occupancy
rate, the odds of moving to a higher star rating category
Policy implications
increased (OR =1.0, P<0.001). Moreover, an increase in per
capita income was positively associated with a higher star The HITECH Act enabled healthcare providers to
rating category (OR =1.0, P<0.05). access EHR incentive payments when they demonstrated
meaningful use of health information technology in
the forms of improved quality, safety, and effectiveness
Discussion
of patient care (32). While this incentive program has
Due to the growing number of aging adults, demand for helped many providers adopt/upgrade health information
long-term care services is expanding (27). For instance, technologies, long-term care facilities including skilled
there were 15,600 nursing homes serving 1.3 million nursing facilities and assisted living homes were considered
residents in 2016 (28). Considering the fact that majority of ineligible for incentives (33). Consequently, the long-term
those services are covered by Medicare and Medicaid, it is care facilities lag behind in EHR adoption (33). Still, due to
in the public interest to explore the strategies that can help promising evidence of the potential benefits of EHR use,
improve the quality of care and save costs. The expanded its adoption increased from 3% in 2010 (34) to 66 percent
use of technology such as EHRs in healthcare may help in 2017 (35). However, our findings complement existing
healthcare providers deliver higher quality services at an research to illustrate the importance of EHR adoption in
affordable price. long-term care facilities. Further Federal and State policies
Utilizing constructs of the SPO and the knowledge-based and funding could help more nursing homes adopt EHRs,
view of the firm, the purpose of this study was to examine which may provide another tool to help nursing homes
the relationship between EHR average scores and resident improve the delivery of resident care.
quality of care, as captured by Nursing Home Compare’s Nursing homes operate in a competitive and highly
Five-Star Quality Rating in high Medicaid nursing homes. regulated environment, mainly due to recent federal and
In line with the findings of previous studies (16,17), we state regulations, changes in reimbursement policies,
found that use of EHR in high Medicaid nursing homes and quality reporting requirements (36). Particularly,
indeed was positively associated with improvements in high Medicaid nursing homes are struggling to remain
quality. financially viable. This study suggests that with the use of
EHR use has numerous benefits such as improved EHR, nursing homes can improve their quality of care.
workflow, fewer medical errors, fewer duplicate tests, and Investing in quality improvement initiatives has shown to

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Journal of Hospital Management and Health Policy, 2021 Page 9 of 13

Table 4 Ordered logistic regression of the relationship between nursing home EHR implementation and star rating (N=391)
Variables OR P value 95% CI

Provider characteristics

Electronic health record implementation 1.50 0.039* 1.02 2.20

Location

Urban Ref

Rural 0.76 0.596 0.28 2.07

Chain affiliation

No

Yes 1.05 0.823 0.69 1.59

Ownership

For-profit Ref

Not-for-profit 2.16 0.004** 1.28 3.64

Government 0.84 0.634 0.41 1.73

Employ nurse practitioners/physician assistants

No Ref

Yes 0.68 0.067 0.45 1.03

Number of total beds 1.00 0.246 1.00 1.00

Resident characteristics

Payer mix

Other Ref

Percent of Medicaid 1.01 0.518 0.98 1.05

Percent of Medicare 0.98 0.518 0.93 1.04

Acuity index 0.97 0.533 0.90 1.06

Occupancy rate 1.03 0.000*** 1.02 1.05

Race/ethnicity

White Ref

Black 0.99 0.329 0.98 1.01

Hispanic 1.01 0.450 0.99 1.02

Other race/ethnicity 1.00 0.516 0.99 1.01

Community characteristics

Medicare advantage penetration rate 1.00 0.964 1.00 1.02

Per capita income 1.00 0.032* 1.00 1.00

Poverty level 1.04 0.180 0.98 1.10

Unemployment rate 0.87 0.069 0.75 1.01

Education (with high school diploma) 0.98 0.419 0.93 1.03

% of people over 65 1.04 0.362 0.96 1.12

Competition (Herfindahl-Hirschman index) 1.75 0.272 0.65 4.72

*, P<0.05, **, P<0.01, ***, P<0.001.

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Page 10 of 13 Journal of Hospital Management and Health Policy, 2021

Table 5 Sensitivity analysis

Variables OR P value 95% CI
Provider characteristics
EHR implementation
Low Ref
Medium 1.22 0.418 0.75 1.99
High 1.75 0.026 1.07 2.86
Location
Urban Ref
Rural 0.72 0.515 0.27 1.94
Chain affiliation
No
Yes 1.04 0.868 0.68 1.57
Ownership
For-profit Ref
Not-for-profit 2.11 0.005 1.25 3.55
Government 0.85 0.665 0.41 1.76
Employ nurse practitioners/physician assistants
No Ref
Yes 0.69 0.068 0.46 1.03
Number of total beds 1.00 0.233 1.00 1.00
Resident characteristics
Payer mix
Other Ref
Percent of Medicaid 1.01 0.539 0.98 1.05
Percent of Medicare 0.98 0.477 0.92 1.04
Acuity index 0.98 0.583 0.90 1.06
Occupancy rate 1.03 0.000 1.02 1.05
Race/ethnicity
White Ref
Black 1.00 0.388 0.98 1.01
Hispanic 1.01 0.424 0.99 1.02
Other race/ethnicity 1.00 0.426 1.00 1.01
Community characteristics
Medicare advantage penetration rate 1.00 0.983 0.98 1.02
Per capita income 1.00 0.029 1.00 1.00
Poverty level 1.03 0.252 0.98 1.09
Unemployment rate 0.87 0.069 0.76 1.01
Education (with high school diploma) 0.98 0.350 0.93 1.03
% of people over 65 1.04 0.359 0.96 1.12
Competition (Herfindahl-Hirschman index) 1.86 0.216 0.70 4.99

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Journal of Hospital Management and Health Policy, 2021 Page 11 of 13

lead to better financial performance (37,38). EHRs may Practices that Influence Patient-Centered Health Care
provide high-Medicaid nursing homes additional benefits Delivery” was commissioned by the editorial office without
that address quality and financial concerns in resource any funding or sponsorship. Robert Weech-Maldonado
constrained environments. Further research needs to be serves as an unpaid editorial board member of Journal of
conducted on the financial impact that EHRs could provide Hospital Management and Health Policy from September
to these under-resourced nursing homes. Nursing home 2018 to August 2020. Dr. GD reports grants from Agency
administrators may need to take into account these benefits for Healthcare Research and Quality (1R01HS023345-01),
and realize that EHRs may represent a business cases to during the conduct of the study. Dr. JL reports grants
improve the quality of care (38). from Agency for Healthcare Research and Quality
(1R01HS023345-01), during the conduct of the study. Dr.
AG reports grants from Agency for Healthcare Research
Conclusions
and Quality (1R01HS023345-01), during the conduct of the
This finding supports the promising role of EHR in study. Dr. RWM reports grants from Agency for Healthcare
improving quality of care among nursing homes. Even Research and Quality (1R01HS023345-01), during the
though there may be barriers to the adoption and use of conduct of the study. The authors have no other conflicts of
EHR systems in nursing homes that operate in resource- interest to declare.
constrained areas, there are tangible benefits that can arise
from the use of EHRs. This paper illustrated how EHRs Ethical Statement: The authors are accountable for all
may help under-resourced nursing homes improve the aspects of the work in ensuring that questions related
quality of care. Providers and policy makers will need to to the accuracy or integrity of any part of the work are
consider strategies that ensure EHR adoption is promoted appropriately investigated and resolved. The study was
across a wider distribution of all nursing homes. conducted in accordance with the Declaration of Helsinki
(as revised in 2013). The study was approved by the
Institutional Review Board of the University of Alabama at
Acknowledgments
Birmingham (IRB-140828005) and informed consent was
Funding: Agency for Healthcare Research and Quality taken from all survey participants.
(1R01HS023345-01).
Open Access Statement: This is an Open Access article
distributed in accordance with the Creative Commons
Footnote
Attribution-NonCommercial-NoDerivs 4.0 International
Provenance and Peer Review: This article was commissioned License (CC BY-NC-ND 4.0), which permits the non-
by the Guest Editors (Naleef Fareed, Ann Scheck commercial replication and distribution of the article with
McAlearney, and Susan D Moffatt-Bruce) for the series the strict proviso that no changes or edits are made and the
“Innovations and Practices that Influence Patient-Centered original work is properly cited (including links to both the
Health Care Delivery” published in Journal of Hospital formal publication through the relevant DOI and the license).
Management and Health Policy. The article has undergone See: https://creativecommons.org/licenses/by-nc-nd/4.0/.
external peer review.

References
Reporting Checklist: The authors have completed the
SURGE reporting checklist. Available at http://dx.doi. 1. Mather M, Jacobsen L, Pollard K. Aging in the United
org/10.21037/jhmhp-20-64 States. Population Bulletin. Available online: https://
www.prb.org/wp-content/uploads/2016/01/aging-us-
Data Sharing Statement: Available at http://dx.doi. population-bulletin-1.pdf
org/10.21037/jhmhp-20-64 2. Stucki G, Bickenbach J, Gutenbrunner C, et al.
Rehabilitation: The health strategy of the 21st century. J
Conflict of Interest: All authors have completed the ICMJE Rehabil Med 2018;50:309-16.
uniform disclosure form (available at http://dx.doi. 3. Mor V, Zinn J, Angelelli J, et al. Driven to tiers:
org/10.21037/jhmhp-20-64). The series “Innovations and socioeconomic and racial disparities in the quality of

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Page 12 of 13 Journal of Hospital Management and Health Policy, 2021

nursing home care. Milbank Q 2004;82:227-56. Milbank Mem Fund Q 1966;44:Suppl:166-206.

4. Hitt LM, Tambe P. Health care information technology, 19. Ford EW, Menachemi N, Peterson LT, et al. Resistance
work organization, and nursing home performance. Ilr is futile: but it is slowing the pace of EHR adoption
Review 2016;69:834-59. nonetheless. J Am Med Inform Assoc 2009;16:274-81.
5. Cheevakasemsook A, Chapman Y, Francis K, et al. The 20. Eisenhardt KM, Santos FM. Knowledge-based view:
study of nursing documentation complexities. Int J Nurs A new theory of strategy. Handbook of Strategy and
Pract 2006;12:366-74. Management 2002;1:139-64.
6. O’Connell B, Myers H, Twigg D, et al. Documenting 21. Bennett R, Gabriel H. Organisational factors and
and communicating patient care: Are nursing care plans knowledge management within large marketing
redundant? Int J Nurs Pract 2000;6:276-80. departments: an empirical study. Available online:
7. HealthIt. Benefits of HER. 2017: Available online: https:// https://www.emerald.com/insight/content/
www.healthit.gov/topic/health-it-basics/benefits-ehrs doi/10.1108/13673279910288707/full/html
8. Tang P. Key capabilities of an electronic health record 22. Nonaka I. A dynamic theory of organizational knowledge
system. Washington, DC: Institute of Medicine of the creation. Organization Science 1994;5:14-37.
National Academies, 2003. 23. Darroch J. Developing a measure of knowledge
9. Adler-Milstein J, Jha AK. HITECH Act Drove Large management behaviors and practices. Available
Gains In Hospital Electronic Health Record Adoption. online: https://www.emerald.com/insight/content/
Health Aff (Millwood) 2017;36:1416-22. doi/10.1108/13673270310505377/full/html
10. Alvarado CS, Zook K, Henry J. Electronic health record 24. Orzano AJ, McInerney CR, Tallia AF, et al. Family
adoption and interoperability among US skilled nursing medicine practice performance and knowledge
facilities in 2016. Available online: https://www.healthit. management. Health Care Manage Rev 2008;33:21-8.
gov/sites/default/files/electronic-health-record-adoption- 25. Abramson EL, McGinnis S, Moore J, et al. A statewide
and-interoperability-among-u.s.-skilled-nursing-facilities- assessment of electronic health record adoption and health
in-2016.pdf information exchange among nursing homes. Health Serv
11. Vest JR, Jung HY, Wiley K Jr, et al. Adoption of Health Res 2014;49:361-72.
Information Technology Among US Nursing Facilities. J 26. Austin PC. An Introduction to Propensity Score Methods
Am Med Dir Assoc 2019;20:995-1000.e4. for Reducing the Effects of Confounding in Observational
12. Adler-Milstein J, Holmgren AJ, Kralovec P, et al. Studies. Multivariate Behav Res 2011;46:399-424.
Electronic health record adoption in US hospitals: the 27. Robison J, Shugrue N, Fortinsky RH, et al. Long-term
emergence of a digital "advanced use" divide. J Am Med supports and services planning for the future: implications
Inform Assoc 2017;24:1142-8. from a statewide survey of Baby Boomers and older adults.
13. The Office of the National Coordinator for Health Gerontologist 2014;54:297-313.
Information Technology. Office-based physician electronic 28. Harris-Kojetin LD, Sengupta M, Lendon JP, et al. Long-
health record adoption, Health IT Quick-Stat# 50. 2019: term care providers and services users in the United States,
Available online: https://dashboard.healthit.gov/quickstats/ 2015-2016. Available online: https://www.cdc.gov/nchs/
pages/physician-ehr-adoption-trends.php data/series/sr_03/sr03_43-508.pdf
14. Cherry B, Carter M, Owen D, et al. Factors affecting 29. Lieber HS. As Electronic Health Record Implementations
electronic health record adoption in long-term care Rise... So Will Disputes. Dispute Resolution Journal.
facilities. J Healthc Qual 2008;30:37-47. 2011;66:46.
15. Wang T, Biedermann S. Adoption and utilization of 30. The Office of the National Coordinator for Health
electronic health record systems by long-term care Information Technology. 2018. Available online: https://
facilities in Texas. Perspect Health Inf Manag 2012;9:1g. www.healthit.gov/resource/care-coordination-tool-
16. Kruse CS, Mileski M, Vijaykumar AG, et al. Impact of transition-long-term-and-post-acute-care
Electronic Health Records on Long-Term Care Facilities: 31. Burke RE, Juarez-Colunga E, Levy C, et al. Rise of post-
Systematic Review. JMIR Med Inform 2017;5:e35. acute care facilities as a discharge destination of US
17. Meehan R. Electronic Health Records in Long-Term Care: hospitalizations. JAMA Intern Med 2015;175:295-6.
Staff Perspectives. J Appl Gerontol 2017;36:1175-96. 32. Buntin MB, Burke MF, Hoaglin MC, et al. The benefits
18. Donabedian A. Evaluating the quality of medical care. of health information technology: a review of the recent

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64
Journal of Hospital Management and Health Policy, 2021 Page 13 of 13

literature shows predominantly positive results. Health Aff Interoperability-among-U.S.-Skilled-Nursing-Facilities-

(Millwood) 2011;30:464-71. and-Home-Health-Agencies-in-2017.pdf
33. Kruse CS, Mileski M, Alaytsev V, et al. Adoption factors 36. Lord J, Davlyatov G, Thomas KS, et al. The Role of
associated with electronic health record among long- Assisted Living Capacity on Nursing Home Financial
term care facilities: a systematic review. BMJ Open Performance. Inquiry 2018;55:46958018793285.
2015;5:e006615. 37. Weech-Maldonado R, Neff G, Mor V. Does quality of
34. Holup AA, Dobbs D, Meng H, et al. Facility care lead to better financial performance?: the case of
characteristics associated with the use of electronic health the nursing home industry. Health Care Manage Rev
records in residential care facilities. J Am Med Inform 2003;28:201-16.
Assoc 2013;20:787-91. 38. Weech-Maldonado R, Pradhan R, Dayama N, et al.
35. Henry J, Pylypchuk Y, Patel V. Electronic Health Record Nursing home quality and financial performance: is there
Adoption and Interoperability among US Skilled Nursing a business case for quality? INQUIRY: The Journal of
Facilities and Home Health Agencies in 2017. Available Health Care Organization, Provision, and Financing.
online: https://www.healthit.gov/sites/default/files/ 2019;56:0046958018825191.
page/2018-11/Electronic-Health-Record-Adoption-and-

doi: 10.21037/jhmhp-20-64
Cite this article as: Davlyatov G, Lord J, Ghiasi A, Weech-
Maldonado R. Association between electronic health record
use and quality of care in high Medicaid nursing homes. J Hosp
Manag Health Policy 2021;5:24.

© Journal of Hospital Management and Health Policy. All rights reserved. J Hosp Manag Health Policy 2021;5:24 | http://dx.doi.org/10.21037/jhmhp-20-64