Secondary prevention of coronary heart disease

Editor—The review by McAlister et al of secondary prevention programmes in coronary heart disease does not adhere to some of the major principals of good practice when conducting systematic reviews of the medical literature.1–3 These include a clearly defined research question, strict inclusion criteria so that the review can be replicated, an exhaustive search of the medical literature to find all relevant studies, and findings that can be interpreted easily by the reader and relate to clinical practice. McAlister et al, at first glance, have chosen a huge area of the medical literature to review, encompassing both pharmacological and non-pharmacological interventions for the secondary prevention of coronary heart disease. On closer inspection, they say that it is their intention to review the literature concerned with disease management programmes for coronary heart disease. The definition of disease management programmes used is broad and is quoted as that proposed by Hunter et al as a combination of patient education, provider use of practice guidelines, appropriate consultation, and supplies of drugs and ancillary services; from the same source, Hunter et al also say that the spectrum of disease management extends from health promotion and disease prevention, through diagnosis, treatment and rehabilitation to long term care.4 This highlights the need to be absolutely explicit about the inclusion and exclusion criteria applied to a review in this area. The trials included in the review are a mix of nursing and multidisciplinary team interventions, and also comprehensive cardiac rehabilitation programmes. Single modality rehabilitation programmes were excluded. Many trials of comprehensive cardiac rehabilitation—which should be considered admissible under the umbrella of disease management programmes—are missing from studies included in the review. A recent Cochrane review of exercise based rehabilitation for coronary heart disease, which was not picked up with the search strategy used by McAlister et al, cites at least 17 trials concerned with comprehensive cardiac rehabilitation that could have also been included in the review by McAlister et al.5 Including these trials in the current review results in a pooled odds ratio for all cause mortality of 0.87 (95% confidence interval 0.76 to 1.0). A precise definition of disease management programmes is problematic, but using the authors' own description, it seems that deficiencies in searching and application of inclusion criteria have resulted in a review that that is difficult to interpret, is not replicable, and is potentially misleading.