Journal of Racial and Ethnic Health Disparities (2019) 6:1–11

https://doi.org/10.1007/s40615-018-0478-x

PERSPECTIVE ARTICLE

Disparities in Cardiac Rehabilitation Among Individuals from Racial

and Ethnic Groups and Rural Communities—A Systematic Review
2,5
Luis R. Castellanos 1,2,3 & Omar Viramontes 1,4 & Nainjot K. Bains 1,2 & Ignacio A. Zepeda

Received: 3 October 2017 / Revised: 26 January 2018 / Accepted: 27 February 2018 / Published online: 13 March 2018
# W. Montague Cobb-NMA Health Institute 2018

Abstract
Purpose Despite the well-described benefits of cardiac rehabilitation (CR) on long-term health outcomes, CR is a resource that is
underutilized by a significant proportion of patients that suffer from cardiovascular diseases. The main purpose of this study was
to examine disparities in CR referral and participation rates among individuals from rural communities and racial and ethnic
minority groups with coronary heart disease (CHD) when compared to the general population.
Methods A systematic search of standard databases including MedlLine, PubMed, and Cochrane databases was conducted using
keywords that included cardiac rehabilitation, women, race and ethnicity, disparities, and rural populations. Twenty-eight
clinical studies from 1990 to 2017 were selected and included 478,955 patients with CHD.
Results The majority of available clinical studies showed significantly lower CR referral and participation rates among individ-
uals from rural communities, women, and racial and ethnic groups when compared to the general population. Similar to
geographic region, socioeconomic status (SES) appears to directly impact the use of CR programs. Patients of lower SES have
significantly lower CR referral and participation rates than patients of higher SES.
Conclusions Data presented underscores the need for systematic referrals using electronic health records for patients with CHD in
order to increase overall CR referral and participation rates of minority populations and other vulnerable groups. Educational
programs that target healthcare provider biases towards racial and ethnic groups may help attenuate observed disparities.
Alternative modalities such as home-based and internet-based CR programs may also help improve CR participation rates among
vulnerable populations.

Keywords Cardiac rehabilitation . Race and ethnicity . Rural communities . Socioeconomic status . Health disparities

Introduction
* Luis R. Castellanos
lrcastellanos@ucsd.edu Cardiovascular diseases affect over 80 million Americans and
close to half of all cardiovascular deaths are due to coronary
* Ignacio A. Zepeda
zepedai@nychhc.org heart disease (CHD) [1]. For over a quarter of a century, car-
diac rehabilitation, a multifaceted program consisting of car-
1
Department of Medicine, The University of California San Diego, diac risk factor modification, prescriptive exercise therapy,
School of Medicine, La Jolla, CA, USA and medication adherence has been shown to be effective in
2
The University of California San Diego Division of Cardiovascular secondary prevention of cardiovascular diseases [2–4]. In par-
Medicine and Sulpizio Family Cardiovascular Center, La Jolla, CA, ticular, participation in cardiac rehabilitation (CR) has been
USA
associated with significant reductions in overall mortality, re-
3
The University of California San Diego, Altman Clinical and current myocardial infarcts, emergency department visits,
Translational Research Institute, 3rd Floor 9452 Medical Center
rehospitalizations, and depressive symptoms in patients with
Drive, MC 7411, La Jolla, CA 92037, USA
4
CHD [5–8].
The University of California Los Angeles, Geffen School of
Patients with CHD are encouraged to participate in CR
Medicine, Los Angeles, CA, USA
5
shortly after a cardiovascular procedure or a myocardial in-
Department of Medicine, Albert Einstein College of Medicine,
Jacobi Medical Center, Building 3, 3N21 1400 Pelham Pkwy, S,
farct. CR focuses on mobilization of the patient, education of
Bronx, NY 10461, USA cardiovascular risk factors, and establishes a plan for when the
2 J. Racial and Ethnic Health Disparities (2019) 6:1–11

patient is discharged. Outpatient CR consists of three ambu- to individuals of areas with a population of fewer than 50,000
latory stages that continue to emphasize exercise training, people as defined by the U.S. Census Bureau, or more than a
medical adherence, and reduction of cardiac risk factors in 30-min drive to the nearest hospital [19, 21]. The inclusion of
an effort to control symptoms and enhance the patient’s psy- at least one of these diverse groups was necessary in order to
chosocial and vocational status [9]. allow for a comparison with white patients or the general
Despite the benefits of CR, less than 50% of eligible pa- population. In order to avoid reporting bias and limiting the
tients with CHD participate in outpatient CR [10]. analysis of available data, we analyzed peer-reviewed pub-
Furthermore, a limited number of studies have shown that lished articles that included randomized controlled trials, ob-
referral and participation rates are lower among patients from servational studies, and cross-sectional studies performed in
rural areas, as well as individuals from diverse groups North America that contained at least 60 participants and ≥
[11–13]. In particular, women, older patients, and racial and 20% women of patients studied. These different study designs
ethnic minority patients with heart disease tend to have lower may affect the quality of the data presented; however, analyses
CR participation rates when compared with the general pop- of potential bias were discussed in the relevant sections of this
ulation [11, 14, 15]. Economic and geographic factors have review. Study selection excluded studies conducted outside of
been shown to reduce participation rates in CR programs [16]. North America given differences in population demographics,
Suaya et al. found a significant deterrent effect on CR partic- race/ethnicity, and healthcare systems when compared to
ipation among Medicare patients when there was a greater those in the USA. After reviewing all the searched articles
distance from the patient’s residence to a CR facility [10]. using the parameters mentioned above, a total of 28 studies
Therefore, the objective of this review article is to systemati- were selected (Fig. 1).
cally review and analyze peer-reviewed published literature
available over the past quarter of a century that describes fac- Cardiac Rehabilitation Eligibility and Participation Criteria All
tors that may contribute to current disparities in cardiac reha- studies reviewed used pre-specified CR inclusion criteria in-
bilitation referral and participation in patients with CHD from cluding coronary artery disease (CAD), acute myocardial in-
racially diverse and vulnerable groups when compared to the farct (AMI), acute coronary syndrome (ACS), coronary artery
general population. bypass grafting (CABG), or percutaneous coronary interven-
tion (PCI) to define the cohorts evaluated (Table 1). Potential
confounding variables of some studies included (1) require-
Methods ment of uninterrupted enrollment in fee-for-service payment
and lack of information on severity of illness [10], (2) entirely
This is a systematic review reported according to the PRISMA urban population and cost-containment policies unique to the
statement [17] that evaluated cardiovascular outcomes includ- geographical area [14], (3) lack of in-depth insurance cover-
ing CR referral rates and CR participation rates among vul- age information and self-report based information [15], (4)
nerable groups. A literature search was performed by statistics limited to one CR center, (5) limited information on
searching MedlLine, PubMed, and Cochrane databases using patients’ comorbidities [37], (6) unrepresentative samples for
the following keywords: Bcardiac rehabilitation,^ Breferral,^ patient groups studied, (7) unpredictable psychosocial vari-
Bparticipation,^ Bdisparities,^ Bminorities,^ Bwomen,^ Brace ables [23], and (8) inclusion of patients with coronary heart
and ethnicity,^ Belderly,^ Brural populations,^ and disease with secondary comorbidities that included congestive
Bsocioeconomic status.^ All peer-reviewed articles published heart failure, valvulopathies, and arrhythmias.
between January 1, 1990 and June 1, 2017 were critically
evaluated using standards from the PRISMA statement [17].
Inclusion criteria for this study consisted of (a) sample size of
≥ 60 participants, (b) cohort population ≥ 20% women, and (c) Findings and Discussion
inclusion of one or more of the following diverse groups:
racial and ethnic minorities (African American, Hispanic/ Characteristics of Studies A total of 28 clinical studies were
Latino, Asian American, Native American), elderly (> evaluated in this study and the majority of these were conduct-
65 years), individuals living in rural areas, and patients of ed in the USA (Table 1). Most studies evaluated disparities in
low socioeconomic status (SES). Racial and ethnic classifica- CR referral and participation with respect to age, gender, SES,
tions used in this study, including the terms white and non- race/ethnicity, and were compared to the general population.
white were based on prior published literature that used these The number of participants in the studies identified ranged
terms [10, 14, 15, 18]. Studies that measured SES were heter- from 87 to 267,427 patients [10, 34]. Sampling frames varied
ogenous and used different definitions that included house- across studies, such that Suaya et al. included only Medicare
hold income, Medicaid status, and level of education as de- beneficiaries who had an index hospitalization in 1997 in a US
fined per Federal standards [19, 20]. Rural status was assigned non-federal, acute care hospital with coronary diagnoses or
J. Racial and Ethnic Health Disparities (2019) 6:1–11 3

Fig. 1 Analysis of cardiac

rehabilitation studies using a

Identification
Records idenfied through Addional records idenfied through other
PRISMA flow diagram database searching sources (references inside other arcles)
(n = 676) (n = 4)

Records aer duplicates removed

(n = 541)

Screening
Records screened Records excluded based
(n = 541) on tle or abstract
(n = 468)

Full-text arcles assessed Full-text arcles excluded,

Eligibility
for eligibility based on eligibility criteria
(n = 73) (n = 45)

Studies included in
qualitave synthesis
Included

(n = 28)

procedures [10], while Gregory et al. included cardiac patients groups [41]. Among Medicare beneficiaries, white patients
from three community hospitals in Maryland [14] (Table 1). were 33% more likely to undergo CR participation than non-
white patients after adjusting for age and gender [10].
Race and Ethnicity Studies that evaluated CR referral and par- Overall, race and ethnicity remain a significant factor when
ticipation rates based on race and ethnicity mainly presented evaluating low referral and participation rates in CR [14].
information using white, African American, or non-white Furthermore, studies that included racial and ethnic groups
groups; specific data on Hispanics and Asian Americans was have revealed that the inability to speak English was a strong
limited (Tables 1 and 2). Patients’ race and ethnicity are factors indicator for low CR referral rates for minority patients (RR
in referral and participation rates in cardiac rehabilitation pro- 9.56, 95% CI 2.18–41.99) [28, 36]. While referral and enroll-
grams [14]. The disparity between white and non-white pa- ment rates in CR vary between the studies analyzed; in gen-
tients in CR referral and participation was identified in several eral, the referral rates for non-white patients are significantly
studies (Table 2). Findings published by Gregory et al. and lower when compared with the general population and white
Roblin et al. displayed a lower referral rate for non-white pa- patients [14].
tients compared with white patients (4.7 vs. 11.5 and 17.3 vs.
26.9, respectively, p < 0.05 for both) [14, 41]. Furthermore, Coronary Heart Disease in Women Clinical studies have
Mazzini et al. found that 39% of Hispanics compared with shown that women with CHD benefit as much as men from
58% of white patients were referred to CR programs (p < CR programs, yet CR referral and participation rates involving
0.01) [37]. Interestingly, in hospitals with high overall CR re- women are consistently lower than those in men (Table 3).
ferral rates (> 90%), Aragam et al. reported that there was no Halm et al. concluded that women were less likely to be re-
difference in CR referral between white vs. non-white patients ferred to phase II CR programs when compared to men (48 vs.
[24]. Yet, there was a statistically significant difference in re- 66%, respectively, p < 0.05), despite women having higher
ferral rates for non-white patients when compared with white eligibility rates than men [34]. Furthermore, Colbert et al.
patients in hospitals with low overall CR referral rates (< 10%) found that women were less likely to be referred to CR than
[24]. Additional studies have reported consistently lower CR men (31.1 vs. 42.2%, p < 0.0001), and less likely to complete
participation rates for non-white patients compared with white a CR program (50.1 vs. 60.4%, p < 0.0001), despite women
patients [10, 13]. Prince et al. found that white patients were having a greater mortality benefit compared with men [29].
more than twice as likely (77%) to initiate CR than non-white Aragam et al. showed that in medical centers with the lowest
 4

Table 1 Cardiac rehabilitation studies evaluated

Author, year Study design Primary aim and patient population Study location and period n Women Mean age White AA Hispanic
(%) (y)† (%) (%) (%)

Ades, 1992 [22] Prospective cohort Determine predictors of CR participation University of Vermont Medical Center. 226 43 70 – – –
in older patients with AMI or CABG.
Allen, 2004 [23] Prospective cohort Determine predictors of CR referral and Baltimore, Maryland. One academic 234 100 66 57 43 –
enrollment in women. Women with center and two community hospitals.
MI, PCI or CABG. 2001–2002.
Aragam, 2011 [24] Retrospective cohort Examine CR referral rates and assess for Michigan. 31 hospitals. 2003–2008. 145,661 34 64 86 9 –
racial and gender disparities. Patients
with PCI.
Beckie, 2010 [25] Randomized clinical trial Determine the influence of a Southeastern United States. Outpatient 252 100 63 72 17 13
motivationally enhanced, CR facilities. 2004–2008.
gender-tailored CR program on atten-
dance. Women with AMI, Angina, PCI
or CABG.
Bhuyan, 2013 [26] Retrospective cohort Examine rural and urban differences in Nebraska Hospital Association discharge 12,764 39 – – – –
AMI survival outcomes. Patients with data. 2005–2009.
AMI.
Bittner, 1999 [27] Prospective cohort Determine referral rates and predictors of Alabama. 5 county hospitals. 1996–1997. 995 35 62 68 30 –
referral to CR. Patients with MI,
Angina, PCI, or CABG.
Brady, 2013 [28] Retrospective cohort Describe rates of, and examine factors Ontario, Canada. 18 Cardiac Care 3739 24 – – – –
affecting, referral to CR after Network hospitals. 2011–2012.
revascularization. Patients with PCI or
CABG.
Colbert, 2015 [29] Retrospective cohort Examine sex differences in long-term Alberta, Canada. 1996–2012. 25,958 24 65 – – –
mortality, based on CR referral rates
and attendance patterns in CAD.
Dunlay, 2014 [30] Retrospective cohort Examine the association between CR Minnesota. Olmstead Medical Center and 1569 41 67 – – –
participation and outcomes, including Mayo Clinic. 1987–2010.
readmissions and death after MI.
Patients with AMI.
Grace, 2008 [31] Prospective cohort Examine physician and patient-level fac- 97 Ontario cardiology practices 1490 28 – – – –
tors affecting verified CR enrollment
Grace, 2012 [32] Prospective cohort Compare rates of referral, enrollment, and Ontario, Canada. 11 community and 1680 27 65 80 – –
participation following systematic academic hospitals. 2006–2008.
versus nonsystematic CR referral.
Patients with ACS and PCI or CABG,
and patients with CHF or arrhythmia.
Grace, 2016 [33] Single-blind, 3 Compare program adherence and Ontario, Canada. 3 CR facilities. 169 100 63 90 – –
parallel-arm, pragmatic functional capacity between women 2009–2013.
randomized controlled referred to supervised mixed-sex,
trial. supervised women only, or
home-based CR.
Gregory, 2006 [14] Cross-sectional Determine whether racial disparity exists Baltimore, Maryland. Three community 1933 50 65 55 45 –
in the referral to CR. Patients with hospitals. 1995–1997.
AMI, CHF, arrhythmia, angina, and
atherosclerosis.
J. Racial and Ethnic Health Disparities (2019) 6:1–11
Table 1 (continued)

Author, year Study design Primary aim and patient population Study location and period n Women Mean age White AA Hispanic
(%) (y)† (%) (%) (%)

Halm, 1999 [34] Comparative, prospective Determine CR referral and completion St. Paul, Minnesota. St. Paul Heart Clinic. 87 53 70* 78 – –
cohort rates for men and women, and reasons
for nonparticipation. Patients with
angina, MI, PCI, or CABG.
Harlan, 1995 [20] Prospective cohort Evaluate correlates of nonparticipation in Durham, North Carolina. Duke Medical 393 24 – 91 – –
CR. Patients with CABG. Center. 1987–1989.
Heid, 2004 [35] Comparative, prospective Determine factors that influence women’s North Central Texas. One urban 202 42 65 87 7 2
cohort decision to enroll or not enroll in CR. community hospital. 2001.
Patients with angina, atherosclerotic
heart disease, and MI.
Johnson, 1998 [12] Prospective cohort Identify the factors that influence the use Four hospitals in Montana and Nevada 254 33 64 98 – –
J. Racial and Ethnic Health Disparities (2019) 6:1–11

of cardiac rehabilitation services by

rural residents. Adults hospitalized for
MI, CABG, or PCI.
King, 1999 [36] Retrospective cohort Examine the impact of demographic and Western Canada. One Tertiary care center 1328 23 63 80 – –
health variables on CR referral and and two associated CR programs.
attendance. Patients with MI, PTCA, 1996–1997.
or CABG.
Mazzini, 2008 [37] Retrospective cohort Evaluate the effect of an American Heart Boston, Massachusetts. Boston Medical 780 35 63 82 8 6
Association Get with the Center. 2001–2003.
Guidelines-based clinical pathway on
referral and enrollment into CR.
Patients with AMI.
Mochari, 2006 [15] Cross-sectional Determine if referral and barriers to CR New York and North Carolina. Three 304 100 62 48 34 15
differ by ethnic status among women. academic medical centers.
Women with CAD, AMI, UA, stable
angina, PCI, or CABG.
Nguyen, 2013 [38] Retrospective cohort Assessment of barriers to CR referral after Ontario, Canada. 2000–2007. 3338 33 64 – – –
hospitalization.
Parashar, 2012 [39] Prospective observational Examine the prevalence of, and factors United States. 2003–2004 1568 30 60 11 – –
associated with, CR participation 1 and
6 months after AMI.
Prince, 2014 [40] Retrospective cohort Examine predictors of initiation and Bronx, New York. Montefiore Medical 822 39 61 49 – –
adherence. Patients with CAD, AMI, Center. 2001–2011.
PCI, CABG, CHF, valvular disease.
Roblin, 2004 [41] Prospective cohort Assess referral and enrollment rates for Atlanta, Georgia. One Kaiser Permanente 783 26 – 75 25 –
outpatient CR. Patients with MI, Hospital. 1997–1999.
PTCA, CABG.
Shanmugasegaram, Cross-sectional Examine barriers to cardiac rehabilitation Ontario, Canada. Secondary analysis of 1809 25 65 83 – –
2013 [19] by socioeconomic status and rurality. CRCARE. Eleven hospitals.
Patients with ACS, PCI, CABG, valve 2006–2008.
surgery, CHF.
Suaya, 2007 [10] Retrospective cohort Identify predictors of, and geographic United States. Medicare’s National 267,427 44 65 92 – –
variations in CR use. Medicare Claims History File. US non-federal
beneficiaries with AMI or CABG. acute care hospitals. 1997.
Thomas, 1996 [13] Cross-sectional United States. 163 CR programs. 1990. 2740 48 64* 93 – –
5
6 J. Racial and Ethnic Health Disparities (2019) 6:1–11

ACS acute coronary syndrome, AMI acute myocardial infarction, CABG coronary artery bypass grafting, CR cardiac rehabilitation, CVD cardiovascular disease, CHF congestive heart failure, CAD
Hispanic
CR referral rates, there was a 26.7% decrease in CR referrals

–
for women when compared with men [24]. Yet, in centers with
(%)
(%) the highest CR referral rates for both men and women, there
AA

was only a 0.7% relative decrease in referral rates for women

–
compared to men; nevertheless, the difference was still statis-
White

tically significant [24].

90
(%)

Women from racial and ethnic groups are significantly af-

fected by gender disparities in CR programs. Mochari et al.
Mean age

found that only 17% of racial and ethnic minority women

coronary artery disease, MI myocardial infarction, PCI percutaneous coronary intervention, PTCA percutaneous transluminal coronary angioplasty, UA unstable angina
66

were instructed to use and attend CR programs, compared

(y)†

with 27% of white women [15]. After controlling for baseline

between-group differences, white women were twice as likely
Women

to be referred to CR programs, compared to non-white women

(%)

34

(OR 2.1, 95% CI 1.2–3.7) [15]. Similarly, Allen et al. reported

that African American women were half as likely to be re-
ferred to CR, and half as likely to enroll in CR programs when
450

compared with white women [23].

n

The study by Weingarten et al. found that upon receiving a

CR referral, enrollment rates for women and men were similar
hospitals in an academic medical
Olmstead County, Minnesota. Two

[42]. Although Heid et al. reported similar CR referral rates for

men and women, 44.1 and 41.7%, respectively, CR participa-
Study location and period

tion rates were 69.2 and 37.1% for men and women, respec-
tively (Table 3) [35]. Parashar et al. also found that CR partic-
center. 2005–2007.

ipation after being referred was lower in women compared with

men (OR 0.61 95% CI, 0.55–0.86) [39]. When comparing CR
participants to non-CR participants, Harlan et al. concluded that
women were more frequently in the non-participant group (26
vs. 12%, p < 0.02) [20]. Additionally, Grace et al. reported that
once enrolled in CR; women attend only half of the prescribed
Compare the enrollment rates of men and
enrollment rates. Patients with PTCA,

women to CR after referral. Patients

sessions regardless of the rehabilitation modality (mixed-

Primary aim and patient population

with PCI, Stable Angina, CABG.

gender or women only) [33]. The gender disparity was also

Examine gender differences in CR

observed among the elderly. Among patients > 65 years, wom-

en were less likely than men to participate in CR (14.3 vs.
22.1% respectively, p < 0.01) [10].
Evidence suggests that gender disparity can be mitigated
CABG, MI.

by healthcare providers since a CR referral from a physician

directly influences choices made by patients, especially wom-
en [15, 25, 35]. Studies have shown that only 22% of women
that meet criteria for CR referral receive physician instructions
to enroll in CR programs [15, 25]. Beckie et al. reported that
92% of women would enroll in CR if referred by their physi-
cian. These findings emphasize the importance of systematic
Retrospective cohort

referral of patients with CHD to CR programs by their health

*Mean age in years for female patients
Study design

providers and highlight obstacles women might face in order

to participate in CR programs [15, 35].
Mean age for entire cohort

Elderly Men and Women Age has been shown to influence CR

Weingarten, 2011 [42]

referral and participation rates among patients with CHD.

Table 1 (continued)

Overall, there is an inverse relationship between the utilization

of CR programs and the age of patients [10, 38, 39]. With each
Author, year

year of increased age, CR referral rates tend to decline by 3%

[27]. Regardless of the type of cardiac procedure, older pa-
tients (> 65 years) have significantly lower CR referral and
†
J. Racial and Ethnic Health Disparities (2019) 6:1–11 7

Table 2 Referral and participation rates of patients with CHD to cardiac rehabilitation programs based on race and ethnicity

CR referral (%) CR participation (%)

Author, year, country Overall Non-white White P value Overall Non-white White P value

Gregory, 2006 [14], USA* 8.5 4.7 11.5 < 0.01 71 80 69.1 0.184
Roblin, 2004 [41], USA* 24.4 17.3 26.9 < 0.05 7.1 10 6.9 < 0.05
Prince, 2014 [40], USA† – – – – 59.4 54.4 65.2 < 0.01
Thomas, 1996 [13] USA† – – – – 14 8.5 15 < 0.01

*Participation rates based on the total cohort

†
Participation rates are for the percentage of referred patients

participation rates than younger patients (≤ 65 years) [13], consistently lower than those of younger patients.
despite evidence demonstrating that older patients benefit as Furthermore, Hammill et al. reported that elderly patients with
much as younger patients from CR [43]. Roblin et al. reported CHD who completed 36 CR sessions had 47% lower risk of
similar findings among patients > 65 years with CHD. There death and 31% lower risk of MI, than those patients who only
were lower CR referral rates among patients > 65 years when completed 1 CR session [45]. Thus, clinical evidence shows a
compared with patients between 50 and 64 years (17.9 vs. strong positive correlation between the number of CR sessions
26.5%, p < 0.05) [41]. These results are in line with findings and favorable long-term outcomes in elderly patients with
from studies conducted by Thomas et al. and Dunlay et al., CHD.
who reported lower CR participation rates among older pa-
tients (> 65 years) with CHD when compared with younger Geographic Regions There appears to be an important relation-
patients (7 vs. 16% and 45 vs. 64% respectively, p < 0.01 for ship between the distance to the nearest CR facility and enroll-
both) [13, 30]. Among Medicare individuals (> 65 years) eval- ment as well as participation in CR programs. The distance
uated by Suaya et al., only 13.9% of patients with an acute MI between patients’ residence and the nearest CR facility is in-
and 31% of patients with CABG participated in CR despite versely related to participation rates in a CR program [31].
having a clear indication to enroll in a CR program [10]. Suaya et al. found that patients living farthest from a CR pro-
Additionally, findings published by Carroll et al. suggest gram were 71% less likely to participate in CR than those living
that community-based interventions such as telephone calls closest to a CR facility [10]. Furthermore, CR referral and
and home visits from advanced practice nurses, can promote participation rates differ significantly between patients living
greater participation in CR programs among elderly patients in an urban setting compared with patients living in rural areas.
with known CHD [44]. Clinical studies suggest that elderly Some studies have found that patients residing in rural areas
patients receive similar benefits from CR participation when have lower CR referral and participation rates than patients
compared to their younger counterparts [43], yet CR referral from urban areas [19, 46]. In addition, patients living outside
and participation rates involving elderly patients are metropolitan areas are 30% less likely to participate in CR

Table 3 Referral and participation rates of patients with CHD to cardiac rehabilitation programs for men and women

CR referral (%) CR participation (%)

Author, year, country Overall Men Women P value Overall Men Women P value

Ades, 1992 [22], USA* – – – – 21 25 15 NS

Aragam, 2011 [24], USA 60.8 61 58 < 0.0001 – – – –
Colbert, 2015 [29], CAN† – 42.2 31.1 < 0.0001 – 60.4 50.1 < 0.0001
Halm, 1999 [34], USA 56 66 48 < 0.05 – – – –
Heid, 2004 [35], USA† 43 44.1 41.7 NS 56 69.2 37.1 < 0.05
Parashar, 2012 [39], USA† – – – – 48.3 74.7 24.3 < 0.01
Roblin, 2004 [41], USA* 24.4 26.5 18.2 < 0.05 7.1 8.4 3.4 NS
Weingarten, 2011, USA† [42] – – – – 64 65 62 NS

NS non-significant difference between groups

*Participation rates based on the total cohort
†
Participation rates are for the percentage of referred patients
8 J. Racial and Ethnic Health Disparities (2019) 6:1–11

programs than individuals living in urban areas [46]. individuals from low SES backgrounds [20, 22, 39, 46].
Shanmugasegaram et al. reported that patients from rural areas Independent of environmental and other external factors, pa-
attended a significantly lower percentage of total CR sessions tients’ SES appears to be a strong predictor of CR referral and
offered when compared with patients from urban areas (76 vs. participation rates.
83%, p < 0.05). Some of the factors that may contribute to the Parashar et al. investigated several predictors for enrollment
differences observed include cost of transportation, distance in CR and findings revealed uninsured patients or those with
travel, family responsibilities, and severe weather [19]. economic burden were less likely to participate in CR pro-
Bhuyan et al. reported that patients from rural areas participated grams. However, in this particular study, having Medicaid/
in CR programs at a higher rate than patients from urban areas, Medicare insurance did not predict participation in CR. A num-
30.8 vs. 16.3%, respectively [26]. However, in that same study, ber of health-system related variables such as high overall cost
patients from rural areas had a higher 30-day in-hospital mor- of services, and copayment for outpatient services were respon-
tality rate than patients from urban areas. It has been postulated sible for lower CR participation [47]. Therefore, system-wide
that these results were in part due to differences in quality of changes such as increased reimbursements for CR, improve-
care and number of CR sessions attended [26]. These findings ment in patient education, effective communication between
illustrate external problems that may contribute to the higher the referring and CR providers, and development of new
mortality rates in patients from rural settings compared with models for the delivery of CR including home-based and
those from urban settings, in addition to the unique circum- community-based programs could potentially improve CR par-
stances that affect patients’ participation in CR. ticipation and ensure continuity of care, resulting in fewer
rehospitalizations and better long-term health outcomes [47].
Socioeconomic Status Similar to geographic region, SES ap-
pears to directly impact the use of CR programs. Patients of
lower SES have significantly lower CR referral and participa- Application to Practice and Future Directions
tion rates than patients of higher SES (Table 4) [1, 10]. Low
SES frequently adds another layer of complexity to enrollment In addition to the aforementioned challenges that vulnerable
and participation in CR. Women with low SES tend to partic- groups encounter, it is also important to recognize the role of
ipate less in CR than women of high SES; in fact, Allen et al. individual health providers in regards to CR referral disparities.
found that women with annual incomes of < $20,000 were Significant evidence has shown that physicians, much like the
two thirds less likely to be referred to CR, and 60% less likely general population, uphold implicit and explicit biases about
to enroll in CR programs than women with annual incomes of many stigmatized groups [48]. Thus, noting physicians’ biases
> $20,000 [23]. Furthermore, Suaya et al. reported that only are worth mentioning as they have been shown to be an impor-
5.2% of patients eligible for both Medicaid (an indicator of tant factor in patients’ cardiac referral and participation rates.
low SES) and Medicare enrolled and participated in CR, com- Blair et al. showed clinicians’ implicit bias significantly affect-
pared with 20.3% of patients with a higher SES [10]. This is ed their clinical relationship with African American patients
particularly important given that unlike Medicare and private [49]. Furthermore, Schulman et al. reported that internists re-
health insurances, most Medicaid programs do not routinely ferred women and African Americans to cardiac procedures
cover CR. Furthermore, multiple studies have shown that ed- significantly less than their male and white counterparts, re-
ucational background, cultural, and linguistic differences may spectively [18]. This was independent of clinical characteris-
contribute to the low referral and enrollment rates among tics, in this case the probability of coronary artery disease.

Table 4 Referral and participation rates of patients with coronary heart disease to cardiac rehabilitation programs based on socioeconomic status

CR referral (%) CR participation (%)

Author, year, country Overall Low SES High SES P value Overall Low SES High SES P value

Suaya, 2007 [10], USA* – – – – 18.7 5.2 20.3 < 0.001

Johnson, 1998 [12], USA* 41 39 46 NS – – – –
Shanmugasegaram, 2013 [19], CAN* – 61.4 68.1 <0.01 51.9 50.3 56.1 < 0.05
Parashar, 2012 [39], USA† – – – – 48.3 9.1 34.3 < 0.001
Ades, 1992 [22], USA* – – – – 21 13 41 < 0.001

NS non-significant difference between groups

*Participation rates based on the total cohort
†
Participation rates are for the percentage of referred patients
J. Racial and Ethnic Health Disparities (2019) 6:1–11 9

Further studies showed that physicians may have implicit amplified using these alternative CR modalities instead of
preference for white patients as well as stereotypes of African offering only standard hospital-based CR programs.
Americans as less cooperative with cardiovascular proce-
dures, such as thrombolysis for myocardial infarction [46,
54]. Race-concordant care might be an alternative to reducing Summary
these disparities. The study by Sabin et al. showed that
African American physicians did not show an implicit prefer- Participation in CR has been shown to improve cardiovascular
ence for blacks or whites patients, and women physicians health among patients with CHD. Despite significant evidence
showed significantly less bias than male providers [50]. The regarding CR benefits, racial and ethnic minorities, women,
findings by Green et al. also provide an example of how in- elderly, individuals from rural communities, and those of low
creasing physicians’ awareness of their susceptibility to im- SES have significantly lower referral and participation rates in
plicit bias may affect healthcare-related behavior. Overall, CR programs, when compared with the general population.
more research needs to be conducted in order to clarify the Barriers including financial obstacles, geographic constraints,
magnitude of physicians’ implicit biases on clinical decisions lack of transportation, language, and limited insurance cover-
and outcomes, short- and long-term, and whether interven- age have been found to contribute to the lower levels of par-
tions reduce these biases and improve the delivery of cardio- ticipation in CR by these diverse groups of patients with CHD.
vascular care, including CR referral and participation [18, 51]. Furthermore, the role of healthcare providers and the potential
Due to overall low referral and participation rates among magnitude of the physician’s implicit biases on clinical deci-
diverse racial and ethnic groups and individuals from rural sions and cardiovascular outcomes need further investigation.
communities, it is imperative to educate healthcare providers, Alternative methods must be utilized to address these dispar-
healthcare systems officers, patients, and their families on the ities in referral and participation in CR programs. Systematic
importance of CR [52]. Aragam et al. demonstrated that by CR referral algorithms for patients discharged after a cardiac-
increasing inpatient CR referral rates of hospitalized patients related hospitalization may increase enrollment in a CR pro-
with CHD, the referral rates among women and African gram. The implementation of alternative models of CR deliv-
Americans significantly increased as well [24]. Electronic sys- ery such as home-based, community-based, and internet CR
tematic referral protocols can also help close the gap in referral programs managed with the support of health professionals,
rates for diverse racial and ethnic groups and compensate for may help to increase CR participation rates of individuals with
unconscious bias from providers [24, 32, 53, 54]. Given the CHD; thus, improving patients’ cardiovascular health and
low percentage of patients with CHD who participate in su- helping to eliminate identified health disparities.
pervised hospital-based CR, alternative approaches to provide
CR have been recommended by cardiovascular societies [2, Acknowledgements The authors would like to thank UC San Diego,
55]. The convenience of a home-based CR program may in- Division of Cardiovascular Medicine and Mrs. Blanca Barba for general
administrative assistance. LRC also wishes to acknowledge with tremen-
crease participation rates in patients who have limited access dous gratitude Mrs. Lizzette Herrera esquire and LLC.
to traditional CR programs. Several CR models of delivery Funding This research did not receive any specific financial support from
services that include home-based programs, internet-based funding agencies in the public, commercial or not-for-profit sectors.
modules, and community-based group programs with guid-
ance by nurses and health professionals, provide alternative Compliance with Ethical Standards
paradigms that may increase patient participation in CR and
improve patients’ overall cardiovascular health. It is plausible Conflict of Interest The authors declare that they have no conflict of
that patients from rural areas, racial and ethnic minorities, interest.
elderly patients, and individuals from low SES who tend to
have lower participation rates in traditional hospital-based CR
programs, may benefit from different modes of delivering CR. References
Community- or home-based CR programs may also help to
remove obstacles such as distance, cost of transportation, and 1. Benjamin EJ, Blaha MJ, Chiuve SE, Cushman M, Das SR, Deo R,
et al. Heart disease and stroke statistics—2017 update: a report from
language and cultural barriers.
the American Heart Association. Circulation. 2017;135(10):e146–
Online/smartphone-based applications (OSPA) offer new 603.
methods to deliver CR information. Widmer et al. showed that 2. Wenger NK, et al. Cardiac rehabilitation as secondary prevention.
patients with OSPA display significant reductions in blood Agency for Health Care Policy and Research and National Heart,
pressure and weight, and lower rates of rehospitalizations Lung, and Blood Institute. Clin Pract Guidel Quick Ref Guide Clin.
1995;17:1–23.
and emergency department visits than the control group (re- 3. Oldridge NB, Guyatt GH, Fischer ME, Rimm AA. Cardiac reha-
ceiving standard CR care) [56]. Thus, secondary prevention bilitation after myocardial infarction: combined experience of ran-
strategies for the treatment of cardiovascular diseases can be domized clinical trials. JAMA. 1988;260(7):945–50.
10 J. Racial and Ethnic Health Disparities (2019) 6:1–11

4. Antman EM, Anbe DT, Armstrong PW, Bates ER, Green LA, Hand 24. Aragam KG, Moscucci M, Smith DE, Riba AL, Zainea M,
M, et al. ACC/AHA guidelines for the management of patients with Chambers JL, et al. Trends and disparities in referral to cardiac
ST-elevation myocardial infarction; a report of the American rehabilitation after percutaneous coronary intervention. Am Heart
College of Cardiology/American Heart Association Task Force on J. 2011;161(3):544–551.e2.
Practice Guidelines (Committee to Revise the 1999 Guidelines for 25. Beckie TM, Beckstead JW. Predicting cardiac rehabilitation atten-
the Management of patients with acute myocardial infarction). J dance in a gender-tailored randomized clinical trial. J Cardiopulm
Am Coll Cardiol. 2004;44(3):E1–e211. Rehabil Prev. 2010;30(3):147–56.
5. Jolliffe JA, et al. Exercise-based rehabilitation for coronary heart 26. Bhuyan SS, Wang Y, Opoku S, Lin G. Rural-urban differences in
disease. Cochrane Database Syst Rev, 2001(1): p. Cd001800. acute myocardial infarction mortality: evidence from Nebraska. J
6. Taylor RS, Brown A, Ebrahim S, Jolliffe J, Noorani H, Rees K, Cardiovasc Dis Res. 2013;4(4):209–13.
et al. Exercise-based rehabilitation for patients with coronary heart 27. Bittner V, Sanderson B, Breland J, Green D. Referral patterns to a
disease: systematic review and meta-analysis of randomized con- university-based cardiac rehabilitation program. Am J Cardiol.
trolled trials. Am J Med. 2004;116(10):682–92. 1999;83(2):252–5. a5
7. Milani RV, Lavie CJ. Impact of cardiac rehabilitation on depression 28. B r a dy S, P ur d h a m D, Oh P, G r a c e S . C l i n i c a l a n d
and its associated mortality. Am J Med. 2007;120(9):799–806. sociodemographic correlates of referral for cardiac rehabilitation
8. Heran BS, et al. Exercise-based cardiac rehabilitation for coronary following cardiac revascularization in Ontario. Heart Lung.
heart disease. Cochrane Database Syst Rev, 2011(7): p. Cd001800. 2013;42(5):320–5.
9. Rehabilitation, A.A.o.C.a.P., AACVPR cardiac rehabilitation re-
29. Colbert JD, Martin BJ, Haykowsky MJ, Hauer TL, Austford LD,
source manual. 2006, Champaign: Human Kinetics Publishers.
Arena RA, et al. Cardiac rehabilitation referral, attendance and
10. Suaya JA, Shepard DS, Normand SLT, Ades PA, Prottas J, Stason
mortality in women. Eur J Prev Cardiol. 2015;22(8):979–86.
WB. Use of cardiac rehabilitation by Medicare beneficiaries after
myocardial infarction or coronary bypass surgery. Circulation. 30. Dunlay SM, Pack QR, Thomas RJ, Killian JM, Roger VL.
2007;116(15):1653–62. Participation in cardiac rehabilitation, readmissions, and death after
acute myocardial infarction. Am J Med. 2014;127(6):538–46.
11. Brown TM, Hernandez AF, Bittner V, Cannon CP, Ellrodt G, Liang
L, et al. Predictors of cardiac rehabilitation referral in coronary 31. Grace SL, Gravely-Witte S, Brual J, Monette G, Suskin N,
artery disease patients: findings from the American Heart Higginson L, et al. Contribution of patient and physician factors
Association’s Get With The Guidelines Program. J Am Coll to cardiac rehabilitation enrollment: a prospective multilevel study.
Cardiol. 2009;54(6):515–21. Eur J Cardiovasc Prev Rehabil. 2008;15(5):548–56.
12. Johnson JE, Weinert C, Richardson JK. Rural residents’ use of 32. Grace SL, et al. Effectiveness of inpatient and outpatient strategies
cardiac rehabilitation programs. Public Health Nurs. 1998;15(4): in increasing referral and utilization of cardiac rehabilitation: a pro-
288–96. spective, multi-site study. Implement Sci. 2012;7:120.
13. Thomas RJ, Miller NH, Lamendola C, Berra K, Hedback B, 33. Grace SL, Midence L, Oh P, Brister S, Chessex C, Stewart DE, et al.
Durstine JL, et al. National survey on gender differences in cardiac Cardiac rehabilitation program adherence and functional capacity
rehabilitation programs. Patient characteristics and enrollment pat- among women: a randomized controlled trial. Mayo Clin Proc.
terns. J Cardpulm Rehabil. 1996;16(6):402–12. 2016;91(2):140–8.
14. Gregory PC, LaVeist TA, Simpson C. Racial disparities in access to 34. Halm M, Penque S, Doll N, Beahrs M. Women and cardiac reha-
cardiac rehabilitation. Am J Phys Med Rehabil. 2006;85(9):705– bilitation: referral and compliance patterns. J Cardiovasc Nurs.
10. 1999;13(3):83–92.
15. Mochari H, Lee JWR, Kligfield P, Mosca L. Ethnic differences in 35. Heid HG, Schmelzer M. Influences on women’s participation in
barriers and referral to cardiac rehabilitation among women hospi- cardiac rehabilitation. Rehabil Nurs. 2004;29(4):116–21.
talized with coronary heart disease. Prev Cardiol. 2006;9(1):8–13. 36. King KM, Humen DP, Teo KK. Cardiac rehabilitation: the forgotten
16. Wyer S, Joseph S, Earll L. Predicting attendance at cardiac rehabil- intervention. Can J Cardiol. 1999;15(9):979–85.
itation: a review and recommendations. Coron Health Care. 37. Mazzini MJ, Stevens GR, Whalen D, Ozonoff A, Balady GJ. Effect
2001;5(4):171–7. of an American Heart Association Get With the Guidelines
17. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. program-based clinical pathway on referral and enrollment into
Preferred reporting items for systematic reviews and meta-analyses: cardiac rehabilitation after acute myocardial infarction. Am J
the PRISMA statement. Ann Intern Med. 2009;151(4):264–9. w64 Cardiol. 2008;101(8):1084–7.
18. Schulman KA, Berlin JA, Harless W, Kerner JF, Sistrunk S, Gersh 38. Nguyen TN, Abramson BL, Galluzzi A, Tan M, Yan AT, Goodman
BJ, et al. The effect of race and sex on physicians’ recommenda- SG, et al. Temporal trends and referral factors for cardiac rehabili-
tions for cardiac catheterization. N Engl J Med. 1999;340(8):618– tation post-acute coronary syndrome in Ontario: insights from the
26. Canadian Global Registry of Acute Coronary Events. Can J
19. Shanmugasegaram S, Oh P, Reid RD, McCumber T, Grace SL. Cardiol. 2013;29(12):1604–9.
Cardiac rehabilitation barriers by rurality and socioeconomic status:
39. Parashar S, Spertus JA, Tang F, Bishop KL, Vaccarino V, Jackson
a cross-sectional study. Int J Equity Health. 2013;12:72.
CF, et al. Predictors of early and late enrollment in cardiac rehabil-
20. Harlan WR 3rd, et al. Importance of baseline functional and socio-
itation, among those referred, after acute myocardial infarction.
economic factors for participation in cardiac rehabilitation. Am J
Circulation. 2012;126(13):1587–95.
Cardiol. 1995;76(1):36–9.
40. Prince DZ, Sobolev M, Gao J, Taub CC. Racial disparities in car-
21. U.S. Census Bureau. Census urban and rural classification and ur-
diac rehabilitation initiation and the effect on survival. Pm r.
ban area criteria. 2010. Available at https://www.census.gov/geo/
2014;6(6):486–92.
reference/ua/urban-rural-2010.html. Accessed 27 July 2017.
22. Ades PA, Waldmann ML, McCann W, Weaver SO. Predictors of 41. Roblin D, Diseker RA III, Orenstein D, Wilder M, Eley M.
cardiac rehabilitation participation in older coronary patients. Arch Delivery of outpatient cardiac rehabilitation in a managed care or-
Intern Med. 1992;152(5):1033–5. ganization. J Cardpulm Rehabil. 2004;24(3):157–64.
23. Allen JK, Scott LB, Stewart KJ, Young DR. Disparities in women’s 42. Weingarten MN, Salz KA, Thomas RJ, Squires RW. Rates of en-
referral to and enrollment in outpatient cardiac rehabilitation. J Gen rollment for men and women referred to outpatient cardiac rehabil-
Intern Med. 2004;19(7):747–53. itation. J Cardiopulm Rehabil Prev. 2011;31(4):217–22.
J. Racial and Ethnic Health Disparities (2019) 6:1–11 11

43. Lavie CJ, Milani RV, Littman AB. Benefits of cardiac rehabilitation 51. Green AR, Carney DR, Pallin DJ, Ngo LH, Raymond KL, Iezzoni
and exercise training in secondary coronary prevention in the elder- LI, et al. Implicit bias among physicians and its prediction of throm-
ly. J Am Coll Cardiol. 1993;22(3):678–83. bolysis decisions for black and white patients. J Gen Intern Med.
44. Carroll DL, Rankin SH, Cooper BA. The effects of a collaborative 2007;22(9):1231–8.
peer advisor/advanced practice nurse intervention: cardiac rehabil- 52. Arena R, Williams M, Forman DE, Cahalin LP, Coke L, Myers J,
itation participation and rehospitalization in older adults after a car- et al. Increasing referral and participation rates to outpatient cardiac
diac event. J Cardiovasc Nurs. 2007;22(4):313–9. rehabilitation: the valuable role of healthcare professionals in the
45. Hammill BG, Curtis LH, Schulman KA, Whellan DJ. Relationship inpatient and home health settings: a science advisory from the
between cardiac rehabilitation and long-term risks of death and American Heart Association. Circulation. 2012;125(10):1321–9.
myocardial infarction among elderly Medicare beneficiaries. 53. Meillier LK, Nielsen KM, Larsen FB, Larsen ML. Socially differ-
Circulation. 2010;121(1):63–70. entiated cardiac rehabilitation: can we improve referral, attendance
46. Receipt of cardiac rehabilitation services among heart attack and adherence among patients with first myocardial infarction?
survivors–19 states and the District of Columbia, 2001. MMWR Scand J Public Health. 2012;40(3):286–93.
Morb Mortal Wkly Rep, 2003. 52(44): p. 1072–5. 54. LaBresh KA, Ellrodt AG, Gliklich R, Liljestrand J, Peto R. Get with
47. Balady GJ, Ades PA, Bittner VA, Franklin BA, Gordon NF, the guidelines for cardiovascular secondary prevention: pilot re-
Thomas RJ, et al. Referral, enrollment, and delivery of cardiac sults. Arch Intern Med. 2004;164(2):203–9.
rehabilitation/secondary prevention programs at clinical centers 55. Leon AS, Franklin BA, Costa F, Balady GJ, Berra KA, Stewart KJ,
and beyond: a presidential advisory from the American Heart et al. Cardiac rehabilitation and secondary prevention of coronary
Association. Circulation. 2011;124:2951–60. heart disease: an American Heart Association scientific statement
from the Council on Clinical Cardiology (Subcommittee on
48. Zestcott CA, Blair IV, Stone J. Examining the presence, conse-
Exercise, Cardiac Rehabilitation, and Prevention) and the Council
quences, and reduction of implicit bias in health care: a narrative
on Nutrition, Physical Activity, and Metabolism (Subcommittee on
review. Group Process INtergroup Relat. 2016;19(4):528–42.
Physical Activity), in collaboration with the American association
49. Blair IV, Steiner JF, Fairclough DL, Hanratty R, Price DW, Hirsh HK, of Cardiovascular and Pulmonary Rehabilitation. Circulation.
et al. Clinicians’ implicit ethnic/racial bias and perceptions of care 2005;111(3):369–76.
among black and Latino patients. Ann Fam Med. 2013;11(1):43–52. 56. Widmer RJ, Allison TG, Lerman LO, Lerman A. Digital health
50. Sabin JA. Physicians’ implicit and explicit attitudes about race by intervention as an adjunct to cardiac rehabilitation reduces cardio-
MD race, ethnicity and gender. J Health Care Poor Underserved. vascular risk factors and rehospitalizations. J Cardiovasc Transl
2009;20(3):896–913. Res. 2015;8(5):283–92.