Article Text

other Versions

Cohort study
Postangiography patients who complete cardiac rehabilitation have reduced risk of mortality or hospitalisation compared to non-completers
  1. Rosemary Olive Higgins,
  2. Michael Richard Le Grande
  1. Heart Research Centre, Melbourne, Victoria, Australia
  1. Correspondence to : Dr Rosemary Higgins
    Heart Research Centre, The Royal Melbourne Hospital, Box 2137, Post Office, VIC 3050, Australia; rosemary.higgins{at}

Statistics from

Commentary on: OpenUrlAbstract/FREE Full Text

Implications for practice and research

  • Cardiac rehabilitation (CR) underutilisation remains as a problem and requires action.

  • CR needs to be funded as a standard component of cardiac care.

  • Strategies to enhance CR attendance in women and elderly patients need to be tested.

  • Flexible models of care are needed to enhance the reach of CR services.


CR is a multidisciplinary, multimodal health intervention. CR comprises exercise training, health education, behavioural counselling and self-management support. Programmes provide a springboard for secondary prevention of cardiovascular disease.

This study makes a sound contribution to the body of evidence supporting CR as an integral component of cardiac care. It addresses some limitations of previous research by making substantial efforts to control for participation bias. A key strength of the study is the large, unselected sample, which enhances understanding of the benefits of CR in diverse patient groups.


This Canadian prospective cohort study conducted between 1996 and 2009, of 5886 individuals referred to CR after coronary angiography is undoubtedly one of the largest CR cohorts ever studied. Furthermore, the study authors were able to link data from several large databases to provide detailed sociodemographic and clinical data on every participant. To minimise bias in pretreatment characteristics between groups, propensity score matching was used to control a very extensive list of clinical and sociodemographic variables. The comparison of CR completers, non-completers and non-attenders provided more detail than the simple comparison of attenders with usual care in controlled trials. Outcomes of interest over a median follow-up period of 5 years included mortality, hospital re-admissions and emergency room visits. Dose–response relationship was assessed using survival models which included the number of exercise sessions attended as a covariate.


The baseline findings confirmed the underutilisation of CR reported in other studies.1 Of the 5886 referred to CR, only 49% completed the programme, 9% dropped out and 42% did not enrol. Patients who completed CR were more likely to be male and younger. CR completion was significantly associated with reduced hospitalisation and mortality compared with CR non-enrolment or non-completion. The association remained strong after adjustment and propensity score matching. Among those who did not complete the CR programme, there was a 1% reduction in mortality for each additional exercise session attended.


The findings of this study add to the compelling weight of evidence supporting the benefits of CR attendance. The health system benefits of reduced hospital admissions appear to be substantial. Demonstration of the economic sense of CR is important in these times of rationalisation of health services. The study identifies substantial benefits of CR completion in those patients lucky enough to be referred to CR. These include mortality benefits over an extended time period in CR completers. However, many eligible patients are not referred to available programmes, denying them the choice of this evidence-based intervention. Of particular interest is the under-referral among specific patient groups, despite the WHO, among others, recommending automatic referral of all eligible patients. Although this study indicates substantial benefits from CR completion for both female and elderly patients, these patients are less likely to be referred to CR.1 This indicates either deliberate or inadvertent discrimination against such patients or, more kindly, a lack of systems to ensure equity of access.2 Such systemic factors should be identified and rectified to ensure an equal likelihood of referral for all who could benefit from this evidence-based intervention. Automatic referral practices and assertive follow-up will substantially remove attendance biases.3

Alongside the evidence on the benefits of CR attendance, the paper documents the use of CR. Rightly, the authors point out that referral to CR is insufficient as a substantial proportion of referred patients do not attend. The 49% completion rate might represent, in part, the preference of some patients for more flexible models of care.4 Moving away from a ‘one-size-fits-all’ philosophy, through endorsement of flexible models, would extend the reach of CR.

View Abstract


  • Competing interests None.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.