Article Text


Cardiac rehabilitation and mortality reduction after myocardial infarction: the emperor’s new clothes?
  1. Robert Lewin1,
  2. Patrick Doherty2
  1. 1BHF Care and Education Research Group, University of York, York, UK
  2. 2Faculty of Health and Life Sciences, York St John University, York, UK
  1. Correspondence to Professor Bob Lewin, Department of Health Sciences, University of York, Area 4, Seebohm Rowntree Building, York, YO10 5DD, UK; bob.lewin{at}

Evidence in favour of cardiac rehabilitation


This piece highlights the strength of evidence in favour of cardiac rehabilitation (CR) and postulates that the emperor is indeed well dressed. The reason why a single negative trial, in the UK, has caused such hullabaloo in the literature and clinical practice is examined against overwhelming evidence from over 40 positive randomised controlled trials. The lack of motivation to promote lifestyle change and the role of patients in determining outcome is also explored. To conclude, we set the scene for the final chapter of this story by outlining what needs to be done to answer the question about the real-world effectiveness of CR.

  • Myocardial Ischaemia And Infarction (Ihd)

Statistics from

The most recent Cochrane review and meta-analysis of exercise-based cardiac rehabilitation (CR) includes 10 794 patients from 47 trials and reports an improvement in relative survival for those attending CR of 13% for all-cause mortality (relative risk (RR) 0.87, 95% CI 0.75 to 0.99) and 26% for cardiac mortality (RR 0.74, 95% CI 0.63 to 0.87).1 The systematic review and meta-analysis by Lawler et al2 of 34 randomised controlled trials (RCT) (6111 patients) showed that CR attendees had a significantly lower risk of re-infarction (OR 0.53, 95% CI 0.38 to 0.76) than non-attendees. The evidence for CR has been extensively reviewed by the National Institute for Health and Clinical Excellence, the Scottish Intercollegiate Guidelines Network, the American Heart Association, the American College of Cardiology, health insurers, the World Health Organization and many other bodies, all of whom have recommended its numerous benefits. These include increased exercise capacity, improved quality of life and health behaviours and a reduction in cases of anxiety and depression.1 ,2 Moreover, economic evaluation shows CR to be more cost effective than most other cardiological or surgical treatments and rivalled only by aspirin and β-blockers.3 In addition to the above benefits in mortality, re-infarction risk, cost effectiveness, physical health and psychosocial wellbeing CR has the potential to reduce the burden on acute services and save money through a reduction in unplanned re-admissions.1 ,4 Despite this, until recently, in the UK there has been little systematic effort to ensure that every eligible or ‘in-scope’ patient received CR and, with some notable exceptions, few cardiologists in the UK have shown any interest in helping to change this situation.

The recent publication, 10 years after completion, of the Rehabilitation After Myocardial Infarction Trial (RAMIT)5 reporting no beneficial effect of CR on mortality (or any other outcome), has provided a rallying call for sceptics of CR. The trial has been deconstructed elsewhere,6 ,7 but in brief it was underpowered, closed early and the CR was poorly described. Adding the RAMIT results to the current Cochrane review data reduced the effect on relative all-cause mortality from 13% to 11%,7 the effect on cardiac mortality is unknown as this statistic was not reported in the RAMIT paper.

Critics of CR point out that many of the trials in the Cochrane review predate modern cardiological advances and it is true that RCT of CR have been reporting positive benefits for more than 40 years! The most recent Cochrane review1 and the systematic review of Lawler et al2 found no evidence that outcomes were influenced by publication date. A related criticism is that there is no single study with sufficient power to demonstrate an effect on mortality. In fact power can come from very large numbers or from moderate numbers with longer follow-up periods, and reduced mortality has been established in two individual studies in which long follow-up periods (>10 years) have been possible.8 ,9 It is clear that, using the methods universally accepted for the validation of all other medical treatments, the evidence that CR can reduce early mortality remains strong in spite of the addition of the negative results produced by the RAMIT study.

Although it is not regarded as essential, confidence in the efficacy of a treatment generally increases when there is a clear causal mechanism. What are the potential explanations for the effect of CR on survival? The mechanisms responsible for these benefits include a reduction in cardiovascular disease risk factors, improvement of endothelial function and inflammatory status, improved diastolic function, positive ventricular remodelling after heart attack and improved electrical stability of the myocardium.10–14

The British Heart Foundation-sponsored National Audit of Cardiac Rehabilitation (NACR) collects patient-level data from 60% of the CR programmes in England, Wales and Northern Ireland, and demonstrates changes over the 6–12-week period of a typical CR programmes that include a reduction in smoking of 5%, an increase in those achieving the guideline for physical activity of 22%, the number of cases with cholesterol scores below the clinical target increased by 21% and blood pressure under the target values (<90 diastolic blood pressure and <140 systolic blood pressure) by 3%.15 These changes are modest but are mean values suggesting that some programmes and individuals achieve more than others. We do not fully understand the exact mechanism by which CR prolongs life, but there are potential explanations involving individual and composite contributions from physiological to behavioural elements.10–14 In fact, looking for a single cause is to misunderstand the nature of CR because the ‘therapeutically active’ elements in a multi-element treatment will vary from patient to patient.

Why, in the face of so much convergent evidence, is some people's scepticism so resilient? Why has a few negative trials out of over 40 positive RCTs led to suggestions that the emperor has no clothes? Could it be that, as a crowd, we have all been persuaded, by the ‘emperor’, to believe that the only effective way to reduce premature cardiac mortality is through lifelong adherence to medicines, or dramatic (if crude) physical interventions such as stenting? Perhaps it is CR that is the brave boy in the crowd pointing out the obvious—that better results are obtained when modern medical treatments and improving the patient’s self-management of their illness by ‘empowering’, educating and ‘activating’ them are combined. The trials of CR have compared routine cardiological treatment plus CR versus routine care alone. The understanding that in long-term conditions healthcare has to be a partnership with the patient and that the patient is actually the main determinant of outcome is not new yet remains a strange concept to many clinicians; but at its best CR has always embodied these concepts. The definition of CR coined by the World Health Organization in 1993 deliberately included the words that the patient may ‘by their own effort’ regain as normal as possible a place in the community.

So the emperor is pretty well clothed. There is good evidence to believe that CR can reduce mortality and deliver other worthwhile benefits. However, proof of efficacy is not proof of effectiveness. A question that CR must still answer is, although RCT show that CR can reduce mortality, in practice, does it? Elsewhere in Europe CR programmes are generally run by cardiologists and the explanation for the RAMIT study results offered by some of our continental colleagues is that CR protocols and staff mix in the UK are so watered down from those used in the majority of RCT of CR as to render many UK programmes ineffectual.4 ,5 If we accept the findings of RAMIT at face value this could be what it is telling us: some programmes achieve very little.

CR is intended to be multifactorial and multidisciplinary, attending to health behaviour as well as to psychosocial factors, including return to work, anxiety and depression. Despite a long tradition of clinical guidelines setting out recommended staffing levels and principles such as, ‘CR should start as soon as possible after the acute event’, over the past 5 years the NACR has recorded a significant reduction in multidisciplinary input and a mean average ‘wait time’ of 7 weeks. There is little doubt that, despite recent advances, CR remains a Cinderella service. Currently, only approximately 40% of in-scope patients in the UK receive treatment.15

How can the effectiveness of CR in reducing early mortality be established? A multicentre pragmatic trial, such as RAMIT, which threatens to randomly assign patients out of a very popular treatment is likely to recruit an atypical sample of CR programmes and patients. The RAMIT investigators approached the majority of CR programmes and found only 14 prepared or able to take part. Current national policy is to increase the uptake of CR, which would prevent any multicentred large pragmatic RCT. However pragmatic and multicentered a trial is, it would never reveal what is happening in non-trial environments. The best answer to this question is to be found in patient-level, routinely collected, health-services data or audit data. Capitalising on the fact that less than half of the in-scope patients in England and Wales are being invited to take part in CR but can be found in the National Institute for Cardiovascular Outcomes Research (NICOR) datasets (surgery, Myocardial Ischaemia National Audit Project (MINAP) and the British Cardiovascular Intervention Society (BCIS)), we are hoping to create matched cohorts of those who did and did not attend CR to address the question of effectiveness in routine provision.


There are many things people living with stable coronary heart disease can do either to reduce or increase their risk of further acute events. At present, in most countries of the world, CR is the only systematic provision of such help on offer to cardiac patients. There is robust evidence that CR produces beneficial health behaviour change, improvements to cardiovascular status and can reduce morbidity and disability, symptom load, psychological distress, improve health-related quality of life and reduce relative all-cause and cardiac mortality. However, many CR programmes in the UK are dissimilar to those from which this evidence is derived, and few have the staffing mix, time or facilities to deliver such programmes. It is legitimate to ask if all programmes deliver all of the benefits expected of them. Linking national audit data to other routinely collected health datasets may provide the answer.


View Abstract


  • Contributors RL and PD contributed equally to the design, analysis and interpretation of subsequent drafts and the final version.

  • Competing interests None.

  • Provenance and peer review Commissioned; internally peer reviewed.

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.

Linked Articles