Cardiac Rehabilitation in Patients with Coronary Heart Disease – Provision, Attendance, and Outcomes: Results from the INTERASPIRE Survey from Fourteen Countries Across Six WHO Regions

Introduction

Although the age-specific incidence of coronary heart disease (CHD) has declined in many regions of the world (1), demographic changes increase the societal burden of CHD and costs. In patients with established CHD the risk of recurrent events remains high; therefore secondary prevention strategies are essential and it has been shown that multi-component cardiac rehabilitation (CR) programmes results in better outcomes (234). Unfortunately CR is not offered to the majority of patients who suffer from a clinical CHD event. This has been clearly demonstrated in various European countries in the EUROASPIRE surveys (56) and also more recently in Norway (7) where only 14% of all patients who had been discharged with an acute myocardial infarction (AMI) during the period 2018–2021 participated in CR. The authors found only slightly lower CR activity during and after the COVID-19 pandemic (13.7%) compared to the years before (14.6%). At an international level less information is available on the provision, attendance, and effects of CR programmes. Some international CR registries have reported the provision of services using different methodologies and content but all depend on self reports with no direct contact with patients using, or not, such services (8).

INTERASPIRE is an international study of secondary prevention of coronary heart disease (CHD), undertaken in partnership with the World Heart Federation in 14 member countries drawn from all six WHO regions: Africa (Kenya, Nigeria and Tanzania), Americas (Argentina and Colombia), Eastern Mediterranean Region (Egypt and UAE), Europe (Poland and Portugal), South-East Asia (Indonesia), and Western Pacific Region (China, Malaysia, Philippines, Singapore) (910). The study was conducted in 2020–2023 through national societies of cardiology. The objective was to determine if international guideline targets for secondary prevention of CVD are being achieved and CR services provided together with their outcomes. The specific objectives of this study of CR were to determine:

• To what extent were patients referred to a CR facility following hospital discharge and which patient-specific factors are associated with referral?

• To what extent did such patients participate in a CR programme, if advised to do so, and which patient-related factors are associated with participation?

• Is participation in a CR programme associated with achieving lifestyle target and risk- factor control, cardioprotective drug use, and adherence?

Material and methods

A detailed description of the study protocol for INTERASPIRE has been published (910). Within each country several geographical areas with a population greater than 500,000 people were selected and all public hospitals serving these populations were identified. A convenience sample of one or more hospitals from each region was taken. In brief, 4,548 consecutive patients, 18–80 years of age, with a first or recurrent coronary hospitalisation (elective or emergency coronary artery bypass surgery (CABG), elective or emergency percutaneous coronary intervention (PCI), AMI, and unstable angina/acute myocardial ischaemia) were identified. They were invited to a study visit at one of the 88 participating hospitals across 14 countries from the six WHO regions between six months and two years after the index hospitalisation (except for Africa where the timeframe was extended to three years to allow for lower incidence of the disease).

Results

A total of 4,548 patients participated in the survey (21% women), varying between countries from 126 in Kenya to 408 in Portugal. They were examined a year (median 1.05 year) after the index event. In Table 1, numbers of participating hospitals, numbers of recruited patients, and numbers and proportions of referrals to CR programmes are presented (overall and by country). In total, 34.4% of all patients reported having been advised to participate in a CR programme, with such reports varying by country from 4.0% in Kenya to 69.6% in Poland.

In Table 2 the associations between patient characteristics and referral to CR programmes in the 1,484 patients who reported being referred are described. There were no differences in CR referral rate by sex. Older patients aged ≥65 years were less frequently referred. Patients with elective PCI as the recruiting event were considerably less likely to be referred by comparison with patients following elective CABG. Patients with a lower educational level and those who were unemployed were also less likely to be referred.

Patients with a previous history of an AMI were more likely to be referred to a CR programme, but a personal history of other CV events, such as stroke or peripheral artery disease (PAD), was not associated with referral.

From multivariate analysis including age, recruiting events, educational level, and a personal history of AMI, referral rates to CR programmes were significantly lower in older patients, in those with elective PCI or unstable angina as the recruiting event, and in patients with a low educational level.

In Table 3 participation in a CR programme is presented by country among patients who reported being referred to participate in CR. Overall 57% of those referred participated in ≥50% of all sessions, 15% in less than 50%, and 27% did not participate at all. These proportions varied greatly between countries. Only 19.6% (847/4314) of all patients recruited to the study attended ≥50% of sessions (234 patients answered ‘Don’t know’ to the question were they referred to CR).

In Table 4 the patient characteristics associated with attendance of at least half of the sessions of a CR programme are presented. Good attendance did not differ by sex but was higher in older patients, those with elective CABG as the recruiting event, better educated patients, those retired or employed, and those with a higher income level.

The content of the CR programmes as reported by patients is presented in Table 5. A supervised exercise programme was the most frequently reported component, but the heterogeneity in reported content is the most striking observation between countries.

The prevalence of lifestyle risk factors according to attendance at CR programmes are shown in Table 6. Smoking, persistent smoking, and physical inactivity were less frequent in patients who participated in ≥50% of all sessions as compared to those who did not, or attended <50% of sessions.

In Table 7 (a) the patients reporting attendance at CR programmes if referred compares non-participants (none or <50% of sessions) and participants (≥50% of sessions) regarding blood pressure levels, LDL-cholesterol levels and variables related to diabetes. In Table 7 (b) only those on prescribed drug treatment for hypertension, dyslipidemia or diabetes in the two groups are compared. For all patients, as well as for those on drug therapy blood pressure and LDL-cholesterol was better controlled among those attending ≥50% of the sessions, but a majority of patients did not achieve any of the guideline risk factor targets.

In Table 8 non-participants are compared with participants in the use of cardioprotective drugs and medication adherence. Participants report higher prescription rates for beta-blockers, ACE inhibitors/ARBs, lipid-lowering drugs and the ‘four drug pillars of secondary prevention’ defined as follows: (1) antiplatelets or anticoagulants including NOACs, (2) beta-blockers, (3) angiotensin-converting enzyme inhibitors (ACE inhibitors) or angiotensin II receptor blockers (ARBs), (4) lipid-lowering drugs. Medication adherence is also significantly higher for lipid-, blood-pressure-, and glucose-lowering drugs in particpants compared to non-participants.

The prevalence of an abnormal HADS depression score ≥11 was significantly higher in the non-participants (8.8%) compared with the participants (4.0%) (p = 0.0002). For the abnormal HADS anxiety score ≥11 the results were: 10.8% in non-participants vs 8.0% in participants (p = 0.07).

In Table 9 non-participants are compared with participants for the INTERASPIRE Guideline Target Score, a published 10-point summary score based on lifestyle, risk factors, and medications (10). The distribution of the score is significantly better in participants compared with non-participants.

The Guideline Target Score is from 0 to 10 and is defined as follows. For each of the 10 items listed below either 1 or 0 is awarded. Lifestyle: Non-smoker or ex-smoker validated with breath CO < 10 ppm; BMI < 25 kg/m²; Performing regular physical activity ≥30 minutes on average 5 times a week; Risk factors: BP < 140/90 mmHg; LDL-C < 1.8 mmol/L; HbA1c < 6.5% (48 mmol/mol) and FG < 5.5 mmol/l (and in patients with diabetes HbA1c < 7% (53 mmol/mol) and FG < 7 mmol/l); Therapeutics: antiplatelet therapies or anticoagulants; beta blockers; RAAS inhibitors; lipid-lowering drugs.

Discussion

This international study of cardiac rehabilitation (CR) across all six WHO regions shows that despite a Class 1 recommendation in all cardiovascular prevention guidelines the majority of patients with CHD were not referred and did not attend CR. Most strikingly there is enormous heterogeneity between countries with an up to 16-fold difference (4.0% to 69.6%) in referral between countries. The reasons for this enormous variation in service provision are complex and likely to be very different both between countries, and between regions and hospitals within countries, but the startling fact is that in every country not all patients are referred to CR, and in most countries it is only a minority.

CR is a multicomponent model of care including exercise training and physical activity promotion, health education, cardiovascular risk factor management, and psycho-social support, personalised to the individual needs of patients diagnosed with heart disease (11). A wealth of scientific evidence supports CR as a clinically and cost-effective intervention for patients with CHD. CR results in better control of CVD risk factors and improved quality of life as demonstrated in randomised controlled trials as well as meta-analyses of observational cohort studies (4111213141516). An updated Cochrane Review on exercised-based CR supports the conclusion that CR provides important benefits to people with CHD, including reduced risk of MI, a likely small reduction in all-cause mortality, and a large reduction in all-cause re-hospitalisation, along with associated healthcare costs, and improved health-related quality of life up to 12 months’ follow-up (4).

One of the major challenges facing CR is the heterogeneity in service provision for patients with CHD (17). In the INTERASPIRE study only a third of the patients reported at the interview having been referred to CR programmes. In a follow-up investigation of CR services, 12 out of the 88 participating hospitals (representing 496 patients or 17.5% of all those who did not report referral to CR programmes) reported they did not offer any CR programmes or refer patients to facilities offering CR. The impact of this observation on the final proportion being able to attend, but not doing so, increases the referral rate from 34.4% to 38.5% where CR services were available which is still a minority of patients. The low referral rates may have been influenced by the COVID-19 pandemic, but we were not able to establish with any certainty with the lead cardiologists at each hospital whether this was the case or not. Other factors will have included the attitude of cardiologists and other health care professionals about the value of CR and patient perceptions of such a service as well.

All patients should be eligible for CR and referral bias should not exist. However, older patients, those with a lower educational level and unemployed status were not referred. In addition, the event itself was associated with referral bias with patients following CABG being more likely to be referred. About a third of all patients were referred to a programme, but only 57.1% of these attended more than half of the sessions and almost a quarter did not attend at all. Attendance rates were higher in older patients, those with elective CABG as the recruiting event, those with a higher educational level, those retired or employed and in those with a self reported higher income level which will be a factor if CR services are charged for outside the public sector. In the principal results paper of INTERASPIRE we reported 9% attendance at CR across all 14 countries combined, but this was an age- and sex-standardized proportion, which explains this apparent difference with the present results.

The evolution of CR into a comprehensive programme encompassing both secondary prevention and rehabilitation is not reflected in the content of the CR programmes as reported by the patients. The top component was supervised exercise programmes (reported by three quarters of patients), followed by the provision of written educational material, diet/weight modification programmes, health promotion workshops, smoking cessation sessions for those who reported smoking at the time of the event, and finally stress modification and relaxation sessions (reported by only a third). The most striking observation is the enormous heterogeneity in reported programme content, even for supervised exercise sessions, between countries. In our systematic review and meta-analysis of contemporary randomized controlled trials of prevention and rehabilitation we found that comprehensive programmes (managing six or more risk factors) and those prescribing and monitoring medications within programmes to lower blood pressure and lipids both reduced cardiovascular events and total mortality (18).

Therefore, although there were some improvements in the lifestyle of those fully participating in CR the prevalence of persistent smoking still remained high at 40%. There was no impact on obesity and central obesity and, although physical inactivity was reduced, over half of these patients were still classified as inactive. The HADS depression score was also significantly lower in those participating in CR. Although there was no mention of managing risk factors—blood pressure, lipids or glucose—to defined targets, or prescribing and maximising adherence with cardioprotective medications in the patients description of programmes there were significant improvements in the control of blood pressure and LDL-C among patients attending more than half the CR sessions. The four drug pillars for secondary prevention of CVD—anti-platelets, beta-blockers, ACE inhibitors/ARBs and statins—were prescribed more frequently in those attending more sessions, and specifically this applied to beta-blockers, ACE inhibitors/ARBs and statins, but not high-intensity statins. Yet only a minority of patients had achieved the LDL-C target of <1.4 mmol/l. Medication adherence was also reported to be higher for blood pressure, lipid, and diabetes management in those attending more sessions. The INTERASPIRE Guideline Target Score, a 10-point score summarising achievement of lifestyle, risk factor and therapeutic targets, was also higher in those attending more sessions.

The results of INTERASPIRE are in accordance with other observational studies showing that access to CR remains poor with between 5–50% of all eligible patients with coronary disease receiving rehabilitation despite Class 1 guideline recommendations (111920212223). Global availability and inequity in CR provision was quantified by the International Council of Cardiovascular Prevention and Rehabilitation (ICCPR) audit (82324). The ICCPR study revealed that CR is available in only half of the countries of the world, and this geographical distribution of CR is negatively correlated with the incidence of ischaemic heart disease, according to the Global Burden of Disease study (25). The densities in CR provision were estimated to be only one CR place available for every 66 patients with CHD in low- and middle-income countries, compared with one place for every 3.4 patients in high-income countries (2324). CR practice faces major challenges globally in service provision and access, and as a consequence all too many patients have a poorer quality of life and a shorter life expectancy.

Strengths and limitations

The strength of INTERASPIRE is that it uses standardised methodology to characterise patients with CHD and how they were managed. This investigation is comprehensive because it includes aspects of lifestyle together with anxiety and depression, risk factor management and prescriptions of cardioprotective medications. Even if the low referral rate to CR may partly be due to the COVID-19 pandemic, the results are consistent with the ICCPR report showing wide variation and low provision of CR globally. The selected regions and hospitals in each country may not be nationally representative and may result in selection bias. However any bias is likely to be conservative as less specialised hospitals are less likely to offer CR and sicker patients may be less willing to attend CR, and therefore, a truly representative picture of patient management is likely to be even poorer. The use of self-reported data and the measures to assess physical activity may be a limitation but the interviews were all conducted by centrally trained research workers using the same questionnaire and some of the lifestyle factors are also reflected in more objective measures using standardised instruments (e.g breath CO) by the interviewers.

In conclusion

INTERASPIRE provides a unique picture of CR provision using standardised methodology in patients with CHD. It demonstrates that referral to and attendance at CR varies enormously between countries despite CR being a Class 1 recommendation in all international guidelines. Only one third of patients reported being advised to attend any form of CR with just one in five patients attending 50% of the sessions. This resulted in poor patient outcomes in terms of lifestyle, risk factor control, and drug prescriptions. To improve patient outcomes national cardiology societies and patients associations should advocate for urgent investment in CR services.

Data Accessibility Statement

The data underlying this article will be shared on reasonable request to the corresponding author. The scientific protocol is publicly available through the website of the National Institute for Prevention and Cardiovascular Health http://www.nipc.ie. Aggregated patient data from INTERASPIRE will be made available after publication of principal results through the World Heart Observatory (world-heart-federation.org/world-heart-observatory/) of the World Heart Federation. Individual patient data will be returned to the participating national societies of cardiology for advocacy at a national level.

Additional File

The additional file for this article can be found as follows:

Competing Interests

GYHL is a National Institute for Health and Care Research (NIHR) Senior Investigator. He is Consultant and speaker for BMS/Pfizer, Boehringer Ingelheim, Daiichi-Sankyo, Anthos (No fees are received personally).

KR received unrestricted research grants to Imperial College London, Amgen, Sanofi, Regeneron, Daiichi Sankyo, and Ultragenix; consulting fees from Novartis, Daiichi Sankyo, Kowa, Esperion, Novo Nordisk, MSD, Lilly, Silence Therapeutics, AZ, New Amsterdam Pharma, Bayer, Beren Therapeutics, CLEERLY, EMENDOBIO, SCRIBE, CRISPR, VAXXINITY, Amarin, Regeneron, Ultragenix, Cargene, and Resverlogix; payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing, or educational events: Novartis, BI, AZ, Novo Nordisk, Viatris, Amarin, Biologix Pharma, Sanofi, Amgen, Esperion, Daiichi Sankyo, Macleod Pharma for CME and non-CME symposia at international meetings; stock or stock options: New Amsterdam Pharma, Scribe Therapeutics, and PEMI31

LR receives research grants from the Swedish Heart-Lung Foundation, Erling Persson foundation, Amgen, Boehringer-Ingelheim and Novo Nordisk. He receives lecture fees from Bayer and Novo Nordisk.

The authors have no competing interests to declare.

Authors Contributions

KK, DDB and GDB contributed to the conception, study design and methodology, literature search, tables, data analysis and interpretation, drafted, critically revised and finally approved the manuscript. CJ and DW contributed to the conception, study design and methodology, data interpretation, critically revised and finally approved the manuscript. JWM, LR, KR, GYHL, IE, SG, TV, AA, AA, WA, AMA, SC, JG, IG, HHA, SH, YH, PY, RMJ, JJ, YL, ASMZ, AM, ACM, LM, JLNE, OSO, ENO, AQB, MUS, MISL, JWCT, DT, MAUT, TJY critically revised and finally approved the manuscript.

Kornelia Kotseva and Dirk De Bacquer shared first authorship.

David Wood and Guy G. De Backer shared senior authorship.