Background
Cardiovascular diseases (CVD) are among the leading causes of morbidity and mortality worldwide, with over 80% of CVD deaths occurring in low- and middle-income countries (LMICs) [
1]. In the middle-income country of Brazil for example, in 2013, 4.2% (6.1 million) of people 18 years of age or older had a diagnosis of some form of CVD [
2].
Cardiac rehabilitation (CR) – a comprehensive outpatient program of secondary prevention and lifestyle changes [
3] – can mitigate this burden. Robust evidence demonstrates positive effects of CR participation, including reductions of mortality up to 25% as well as decreases in hospitalizations [
4]. Reduction in risk factors, as well as increase in quality of life and functional capacity are also reported in studies undertaken in LMICs [
5,
6]. Dose-response associations are observed [
7], hence it is not only important that patients enroll, but that they adhere and complete programs to achieve these benefits.
Despite consequent clinical practice guideline recommendations to refer CVD patients [
8,
9], CR programs are highly unavailable and under-utilized, particularly in LMICs [
10]. CR is only available in approximately 25% of LMICs [
6,
10], with Brazil for example, having a density of 1 program per 4.9 million inhabitants [
10]. The barriers are multifactorial, and include health system [
11], referring physician, program and patient-level factors [
12‐
16].
While complex, there are very few studies which consider these multi-level barriers concurrently [
14,
17], and hence enable a fulsome understanding of the context of CR under-utilization, so that effective strategies to overcome them can be identified and implemented. There is even less data from LMICs [
18] (only 13 studies identified, most not multi-level), which is a major omission considering [
1] this is where the need for CR is greatest but availability is lowest, and [
2] the context is considerably different than that of high-income countries (i.e., often private and public systems; low availability of primary healthcare). In South America there are only some discrete data on healthcare administrator perceptions of CR barriers [
19], CR programs [
20‐
22], as well as those among patients [
23,
24]. Therefore, the aim of this study was to concurrently assess barriers to CR delivery at the healthcare system (including funding source), CR program, and patient (inclusive of barriers to not only enrolment, but also to adherence and completion of the program by enrollees) levels, in a low-resource context.
Results
Respondents characteristics
Ninety-one institutions providing cardiology services were identified across the state of Minas Gerais; of these, 47 (51.6%) were publicly-funded. Healthcare administrators from 32 (35.2%) institutions responded: 24 (75.0%) from public (14 hospitals, and 10 outpatient clinics) and eight (25.0%) from private (5 hospitals, and 3 outpatient clinics) institutions. Only five (15.6%) respondents opted for the mailed printed survey. Of the 19 hospitals, 16 (84.2%) had intensive care, and five (31.2%) had a CR program. The characteristics of the healthcare administrators are shown in Table
1.
Table 1
Characteristics of healthcare administrators, cardiac rehabilitation providers, and cardiac patients
HEALTHCARE ADMINISTRATORS | N = 32 |
Sex |
Male | 20 (62.5%) |
Highest Educational Attainment |
Post-Secondary | 20 (62.5%) |
Post-Graduate | 12 (37.5%) |
Professional position |
Clinical Director | 13 (40.6%) |
Manager | 9 (28.1%) |
General Director | 2 (6.3%) |
Cardiology Coordinator | 2 (6.3%) |
Other | 6 (18.7%) |
Cardiac rehabilitation providers | N = 16 |
Sex |
Female | 10 (61.5%) |
Highest Educational Attainment |
Post-Graduate | 16 (100.0%) |
Healthcare Profession |
Physiotherapist | 7 (43.7%) |
Physician | 3 (18.7%) |
Exercise specialist | 3 (18.7%) |
Dietitian | 1 (6.3%) |
Nurse | 1 (6.3%) |
Other | 1 (6.3%) |
CARDIAC PATIENTS | N = 805 |
Sociodemographic |
Age, years (mean ± SD) | 62.85 ± 12.42 |
Sex, n (%) |
Male | 442 (54.9%) |
Marital status |
Single | 121 (15.0%) |
Married | 488 (60.6%) |
Divorced | 79 (9.8%) |
Widowed | 117 (14.5%) |
Highest Educational Level |
Elementary School | 447 (55.7%) |
High School | 198 (24.7%) |
Post-Secondary | 138 (17.2%) |
Post-Graduate | 20 (2.5%) |
Clinicala |
Cardiac History |
Coronary Artery Disease | 500 (61.4%) |
Myocardial Infarction | 337 (41.4%) |
Percutaneous Coronary Intervention | 267 (32.8%) |
Heart Failure | 92 (11.3%) |
Arrhythmia | 181 (22.5%) |
Valve Disorder | 83 (10.3%) |
Risk Factors |
Hypertension | 646 (79.4%) |
Dyslipidemia | 410 (50.4%) |
Smoking history | 410 (50.4%) |
Diabetes | 233 (28.6%) |
Cardiac rehabilitation (enrollees) | n = 305 |
Wait time in months (mean ± SD) | 4.03 ± 5.74 |
Number of missed sessions in last month (mean ± SD) | 1.60 ± 1.82 |
Forty-one CR programs were identified, of which nine (21.9%) were publicly-funded. Responses from providers at 16 (39.0%) programs were received. Their characteristics are shown in Table
1.
The sample of cardiac participants consisted of 805 respondents. Their sociodemographic and clinical characteristics are shown in Table
1. In total, 495 (61.5%) participants were from public, and 310 (38.5%) participants were from private institutions. Overall, 305 (37.9%) patients were enrolled in CR; sex and age for enrollees and non-enrollees are shown in Table
4.
Perceptions of healthcare administrators
The majority of respondents (n = 23, 71.9%) stated that CR programs should be funded by the Ministry of Health and 15 (46.9%) by private health plans. Also, most of the healthcare administrators (n = 21; 65.6%) considered CR as a good use of public healthcare resources.
Seventeen (53.1%) respondents agreed that acute care institutions are responsible for providing patient connections to outpatient services for continuity of care. Nine (28.1%) encouraged physicians and residents to refer participants to CR, but without systematization. Seven (21.9%) institutions had systematic CR referral. In five (15.6%), referral was hardly or never discussed at meetings.
All (100.0%) respondents indicated their institutions did not have sufficient resources for CR, and lacked capacity to provide care to referred patients, but the healthcare administrators affirmed that they perceived their institution would provide more support if more financial resources were available (again 100.0%).
Table
2 shows mean scores on the knowledge, perceptions and attitudinal items. Overall, the healthcare administrators had satisfactory to good knowledge about CR. Their perceptions towards CR were very positive, and attitudes moderately positive.
Table 2
Healthcare administrators’ knowledge, perceptions and attitudes regarding cardiac rehabilitation, N = 32
KNOWLEDGEa |
My knowledge of what CR entails | 2.75 ± 1.34 |
Rates of participation in CR at the institution where I am employed | 2.09 ± 1.11 |
The location of the nearest CR program | 2.00 ± 1.29 |
Level of knowledge about CR of my colleagues | 1.71 ± 0.85 |
PERCEPTIONSb |
The importance of CR for outpatient care | 4.37 ± 0.55 |
The role of CR access programs in reducing patient length of stay | 4.18 ± 0.64 |
The role of CR programs in reducing re-admissions | 4.15 ± 0.76 |
The importance of care of patients with other vascular conditions in CR | 4.00 ± 0.76 |
Perceptions of your institution about the importance of CR | 3.81 ± 0.85 |
ATTITUDESc |
CR programs provide benefits beyond what primary care providers can offer | 4.28 ± 0.72 |
CR programs promote sustainedbehavioral changes that improve patient outcomes | 4.09 ± 0.92 |
It is likely that government funding for CR programs will be sustained over time | 4.06 ± 0.80 |
It is the hospital’s responsibility to provide all eligible inpatients with the information they need to begin CR | 3.87 ± 1.00 |
The government should provide more funding for CR | 3.87 ± 0.65 |
Government ministry funding models are a financial disincentive to CR provisiond | 3.68 ± 1.09 |
Patients and their families should be responsible for their own health behavior changes and risk reduction self-management posthospitalizationd | 3.46 ± 1.31 |
We do not have enough space to run a CR program at my institutiond | 3.40 ± 1.26 |
The closest available CR program is of good quality | 3.15 ± 0.84 |
CR services are generally one of the first programs to be cut back when we make budget reductionsd | 2.65 ± 1.00 |
Scarce healthcare money should not be spent on outpatient care at the expense of acute cared | 2.25 ± 1.13 |
Health care providers on the cardiac floor have other more important clinical duties than to refer patients to CRd | 1.90 ± 0.77 |
I am skeptical about the benefits of CR programsd | 1.84 ± 0.76 |
Government health insurance should not cover CR services for cardiac patients post-hospitalizationd | 1.56 ± 0.50 |
Perceptions of CR providers
The perceptions of CR providers regarding CR delivery are shown in Table
3. It is the perception of CR staff that referring physicians are not sufficiently aware of the benefits of CR and do not refer. Respondents were highly supportive of providing fully comprehensive CR.
Table 3
Perceptions of Cardiac Rehabilitation Staff on Delivery (N = 16)
FACILITATORS AND BARRIERSa |
CR participants understand the benefits of joining the program | 4.50 ± 0.51 |
Most physicians do not refer patients to CR | 4.31 ± 0.60 |
Most physicians are unaware of the benefits of CR | 4.06 ± 0.68 |
The rate of absenteeism in my program is very low | 3.75 ± 1.00 |
Delivering hybrid CRc could increase participation by patients | 3.62 ± 1.02 |
Participants enrolled in CR have difficulty staying in the program | 3.50 ± 1.03 |
Many patients are referred by doctors, but choose not to participate | 3.12 ± 1.14 |
Our program could serve a larger number of participants, but there is no demand | 2.93 ± 1.80 |
DELIVERY OF COMPREHENSIVE CRb |
Access to optimal medical therapy and reinforcement of the need to adhere to pharmacological treatments | 4.87 ± 0.34 |
The assessment and control of patient’s blood pressure, glucose and lipids | 4.81 ± 0.40 |
The inclusion of a comprehensive educational component within CR | 4.75 ± 0.44 |
Adequate physical space and resources to offer comprehensive CR | 4.43 ± 0.51 |
CR providers reported several additional barriers such as: high cost of CR programs for patients, lack of government initiative to create more CR programs, low educational level of patients (elementary school), lack of patient motivation to change habits, and lack of knowledge about CR by the non-medical professionals (e.g., nurses, dietitians).
Perceptions of cardiac patients
The CRBS item and subscales scores are shown in Table
4. The greatest barriers were related to lack of awareness and encouragement by physicians. Cardiac patients reported some additional barriers, which related to distance and transportation (items already assessed in CRBS).
Table 4
Mean Cardiac Rehabilitation Barrier Scale scores by funding source and CR participation status
Sex |
Male (%) | | 297 (60%) | 139 (45%) | 191 (63%) | 245 (50%) |
Age (mean ± SD) | | 60.8 ± 11.1 | 65.1 ± 13.7** | 65.4 ± 11.4** | 61.2 ± 12.7 |
CRBS item (number) / subscale a mean ± SD |
I didn’t know about CR [ 5] | 3.17 ± 1.54 | 3.14 ± 1.56 | 3.22 ± 1.52 | 1.97 ± 1.03 | 3.90 ± 1.34*** |
My doctor did not feel it was necessary [ 16] | 2.63 ± 1.43 | 2.63 ± 1.42 | 2.63 ± 1.46 | 1.65 ± 0.68 | 3.23 ± 1.45*** |
| 2.43 ± 1.41 | 2.60 ± 1.45*** | 2.16 ± 1.31 | 2.05 ± 1.09 | 2.66 ± 1.54*** |
| 2.42 ± 1.39 | 2.62 ± 1.45*** | 2.08 ± 1.21 | 2.09 ± 1.09 | 2.60 ± 1.50** |
Transportation problems [ 3] | 2.26 ± 1.31 | 2.45 ± 1.38*** | 1.93 ± 1.12 | 2.02 ± 1.00 | 2.40 ± 1.45 |
| 2.25 ± 1.29 | 2.08 ± 1.14 | 2.54 ± 1.45*** | 1.63 ± 0.66 | 2.64 ± 1.43*** |
| 2.13 ± 1.48 | 1.97 ± 1.10 | 2.40 ± 1.92** | 2.64 ± 1.80*** | 1.83 ± 1.15 |
Other health problems [ 14] | 2.13 ± 1.25 | 2.11 ± 1.24 | 2.13 ± 1.24 | 2.31 ± 1.31** | 2.00 ± 1.19 |
I find exercise tiring or painful [ 9] | 2.07 ± 1.14 | 2.09 ± 1.13 | 2.03 ± 1.14 | 1.78 ± 0.81 | 2.24 ± 1.26*** |
I can manage my heart problem on my own [ 18] | 2.04 ± 1.13 | 2.03 ± 1.09 | 2.06 ± 1.19 | 1.82 ± 0.87 | 2.18 ± 1.24** |
I already exercise at home, or in my community [ 7] | 2.01 ± 1.13 | 1.97 ± 1.05 | 2.09 ± 1.24 | 1.71 ± 0.69 | 2.21 ± 1.29*** |
Many people with heart problems don’t go, and they are fine [ 17] | 2.01 ± 1.07 | 1.99 ± 1.05 | 2.04 ± 1.11 | 1.88 ± 0.88 | 2.09 ± 1.17 |
Family responsibilities [ 4] | 2.00 ± 1.36 | 2.04 ± 1.52 | 1.95 ± 1.08 | 2.03 ± 1.04 | 1.99 ± 1.53 |
I don’t have the energy [ 13] | 1.99 ± 1.14 | 1.95 ± 1.11 | 2.05 ± 1.17 | 1.73 ± 0.82 | 2.14 ± 1.27** |
| 1.98 ± 1.11 | 1.96 ± 1.07 | 2.02 ± 1.16 | 1.68 ± 0.64 | 2.16 ± 1.28** |
Work responsibilities [ 12] | 1.94 ± 1.13 | 1.92 ± 1.11 | 2.00 ± 1.16 | 1.88 ± 0.94 | 2.00 ± 1.23 |
| 1.90 ± 1.03 | 1.96 ± 1.07 | 1.82 ± 0.95 | 1.90 ± 0.95 | 1.91 ± 1.08 |
I prefer to take care of my health alone, not in a group [ 21] | 1.86 ± 1.05 | 1.76 ± 0.96 | 2.01 ± 1.15** | 1.68 ± 0.71 | 1.96 ± 1.19 |
It took too long to get referred into the program [ 20] | 1.69 ± 0.82 | 1.76 ± 0.89 | 1.59 ± 0.68 | 1.87 ± 0.89*** | 1.58 ± 0.75 |
| 1.65 ± 0.86 | 1.65 ± 0.88 | 1.67 ± 0.82 | 1.59 ± 0.71 | 1.70 ± 0.94 |
I think I was referred, but the rehab prog didn’t contact me [ 19] | 1.62 ± 0.73 | 1.67 ± 0.79 | 1.53 ± 059 | 1.71 ± 0.70*** | 1.56 ± 0.73 |
SUBSCALES |
Subscale 1 Comorbidities / functional status | 1.98 ± 0.74 | 1.97 ± 0.72 | 1.99 ± 0.77 | 1.84 ± 0.58 | 2.07 ± 0.81*** |
Subscale 2 Lack of perceived need | 2.44 ± 0.84 | 2.43 ± 0.88 | 2.45 ± 0.77 | 1.80 ± 0.54 | 2.83 ± 0.75*** |
Subscale 3 Personal / family issues | 2.04 ± 0.80 | 2.04 ± 0.83 | 2.06 ± 0.77 | 1.85 ± 0.61 | 2.17 ± 0.88*** |
Subscale 4 Travel / work conflicts | 2.06 ± 1.03 | 1.96 ± 0.89 | 2.23 ± 1.20*** | 2.25 ± 1.10*** | 1.95 ± 0.96 |
Subscale 5 Access | 2.03 ± 0.79 | 2.16 ± 0.81*** | 1.84 ± 0.70 | 1.93 ± 0.67 | 2.10 ± 0.84** |
Total Score | 2.12 ± 0.57 | 2.12 ± 0.58 | 2.11 ± 0.54 | 1.89 ± 0.51 | 2.26 ± 0.55*** |
Differences in barriers by program funding source and CR enrolment status are also shown in Table
4. Respondents from public institutions considered distance, cost and transportation to be significantly greater barriers to CR participation than those from private institutions (and correspondingly the access subscale was significantly greater among patients from public institutions); Respondents from private institutions considered the following factors to be significantly greater barriers to CR participation than those from public institutions: lack of perceived need, travel (and correspondingly the travel/work conflicts subscale was significantly greater among patients from public institutions) and preferring to manage their chronic condition independently.
Cardiac patients who did not go to CR considered the following factors to be significantly greater barriers to CR participation than those who did: lack of awareness of CR, lack of physician encouragement, distance, cost, lack of perceived need, finding exercise tiring or painful, preferring to self-manage their chronic condition, already exercising at home or in their community, not having the energy, and time constraints (and correspondingly 4 of the 5 subscales were significantly greater among non-enrollees); CR participants endorsed travel, comorbidities, wait times, and lack of program follow-up as a barrier to a significantly greater degree than did non-participants (the travel/work conflict subscale was significantly greater in this group; Table
4).
Discussion
In this first study to concurrently consider CR barriers in a low-resource setting from the perspective of healthcare administrators, CR providers, and cardiac patients, lack of resources and funding, lack of referral / physician encouragement, lack of patient awareness, and poor access for patients (i.e., distance, cost, transportation) were the main barriers identified. There is a clear incongruity between the recognition of the importance of CR and its effective implementation and use by healthcare administrators, and the low supply of CR programs and lack of resources to deliver services. Lack of referral is a key theme across all levels. The inter-relationship between the barriers at each level is evident – specifically without resources to offer programs, there are no programs to which physicians can refer (and hence inform and encourage patients to attend), and patients will continue to have barriers related to distance, cost and transport.
The healthcare administrators had low to moderate knowledge and attitudes about CR, but very positive perceptions. When compared to administrators in high-income countries [
32], their knowledge is much lower (means all above 3 in Canada), and their perceptions and attitudes are somewhat less positive. There was a notable discrepancy with regard to CR space, with Canadian administrators rating this as much less of an issue. In a survey of CR programs in Latin America [
20,
21] and the Arab world [
10] (of which many are LMIC), lack of space was also among the greatest barriers to CR provision.
Issues identified by programs included lack of physician referral, likely caused by the perceived lack of awareness of CR among physicians. They also reported lack of programs as a major issue; Indeed the low availability of CR programs in Minas Gerais has been previously established [
22]. In a review of national/regional surveys of CR 4barriers were human resources, financial resources and space. In a review on CR barriers in LMICs specifically [
18], the most frequently-reported barriers were lack of personnel and resources, as well as profitability.
With regard to patients, the major barriers among non-enrollees were lack of awareness and physician encouragement and barriers to program adherence among enrollees were travel, comorbidities, cost, distance, and family responsibilities. It is not appropriate to compare CRBS scores between studies except where data are shown by CR enrolment status (given the major differences in barriers, the proportion of enrollees and non-enrollees in the cohort would impact mean values), but scores were reported by enrolment in a sample of Brazilian cardiac patients [
23]. The findings herein are fairly consistent with that study, showing that the greatest barriers among non-enrollees were distance, lack of awareness and lack of encouragement, and among enrollees were travel and cost. In the review of CR delivery around the world [
34], patient-related barriers in LMICs were also lack of awareness, cost, transportation and time constraints.
Multi-level strategies to overcome these barriers in LMICs are forwarded in the review by Ragupathi et al. [
18]. In relation to the systems factors, for example, we can align incentives with service delivery and improve revenue streams, as well as CR delivery in community health service centers [
34], exploiting existing physical infrastructure (e.g., community exercise centers). We need more programs before we can promote and automate CR referral by physicians [
35]. CR programs themselves should be comprehensive, but simplified [
35], low-cost [
28] models have been forwarded, which should exploit unsupervised delivery modalities (e.g., smartphones) [
36]. Finally, evidence-based strategies that motivate the participation and adherence of patients such as counseling by clinicians should be applied [
37]. Studies evaluating the effect of such strategies in LMICs are scarce, and clearly this represents an important direction for future research.
This study has several limitations that need to be considered. First, as there is no official directory of cardiac care centers and CR programs in the state evaluated, it is possible that all programs were not identified (and in particular smaller ones), which may introduce selection bias. Second, the response rate was low for the healthcare administrators and CR providers, mainly in privately-funded services, suggesting that the results herein may be less representative of barriers in private care. Moreover, the response rate in patients was not captured, and hence there could be selection bias (e.g., higher socioeconomic status, more motivated patients represented in the sample than the average cardiac patient).
Third, the design was cross-sectional and therefore no causal conclusions should be drawn. Fourth, the cohorts were recruited from only one country state, thus results might not be generalizable to other states in Brazil, or to other low-resource settings more broadly. Fifth, the healthcare administrator survey was not professionally translated, nor was a formal process of cross-cultural adaptation applied. Therefore, the validity and reliability of that assessment is unknown. Sixth, CR-referring physicians were not directly surveyed; consideration of referral barriers were made indirectly through the CR provider and patient surveys. However, findings were consistent with other research. Finally, there were age differences in patients enrolling vs not in CR, which may be related to the barriers identified.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.