linical practice should follow evidence-based medicine, which is derived from clinical trials. The outcomes of clinical practice, however, may not equal those of trials if there are differences in the patient outcomes or the quality of treatment they receive. We report the example of cardiac rehabilitation to illustrate this point, comparing the characteristics of patients and treatments offered in randomized controlled trials (RCTs) in this area with those included in 2 large surveys of cardiac rehabilitation in the United Kingdom. We found that cardiac rehabilitation as currently practiced in the United Kingdom is unlikely to be as effective as clinical trials may suggest.

Editor's Comment. The article compares data from cardiac rehabilitation programs in England to those published in a Cochrane review with respect to patient demographics, program offering, and outcomes. This is a brief, easy-to-read article that makes 2 relevant points. The first salient point is that RCTs often represent the "best" of a given treatment, not everyday, "real-world" clinical intervention. We all know this to be true from many of the landmark RCTs that involved not only lifestyle intervention and secondary prevention in individuals with cardiovascular disease but also delivery of those interventions in ways that are difficult for real-world, hospital-based programs to provide. The second point is that the differences in outcomes are sometimes due to differences in demographic or diagnostic characteristics of the groups, which is also the case in this study. However, at times the comparison can shed light to one set of programs. In this case, the authors suggest that an increase in the number of exercise sessions offered by English programs might improve outcomes.- JLR