Exclusion of 'nonRCT evidence' in guidelines for chronic diseases - is it always appropriate? The Look AHEAD study.

Evidence-based medicine (EBM) is the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients. The introduction of EBM was a conceptual and practical milestone in the history of medicine, with far-reaching impact yet to be fully realized. EBM has limitations, including inapplicability to populations dissimilar to those in studies, and may not reflect duration of exposure to risk factors, details of lifestyle, incubation period, latency, or environmental changes during chronic diseases. Routine exclusion of evidence other than randomized controlled trials (RCTs) or meta-analyses from consideration in treatment may not always be wise. This review is not a result of a search, but rather a conceptual unification of (a) the increasing restrictions in guideline-writing favoring more RCTs, and rejecting observational studies when chronic diseases with a long incubation period may sometimes be best probed by the latter; (b) the possibility RCTs may be inconclusive, nonapplicable, or result in 'negative' results which may misdirect future therapy by physicians and undermine adherence by patients; (c) the potential improvement in patient care from having all available information evaluated (especially epidemiological studies of chronic diseases) and synthesized in guidelines. The example of the Look AHEAD study is chosen - a 'negative' RCT with significant information overlooked by reviewers, who initially declared that weight loss and physical activity were ineffective in treating diabetes, or in preventing cardiovascular complications. In this review, placing this study in perspective, among others, suggests the opposite - exercise and weight loss are effective if done early and sufficiently. Synthesizing worthy data from many sources, including prospective and pathophysiological studies, particularly when RCTs are unavailable, has the potential to add depth and expand the understanding of disease. In addition, integrated data may generate useful, rich material for use during shared decision making discussions with patients, and clarify future hypotheses.