Systematic Reviews: Critical Links in the Great Chain of Evidence

Successful clinical decisions, like most human decisions, are complex creatures [1]. In making them, we draw on information from many sources: primary data and patient preferences, our own clinical and personal experience, external rules and constraints, and scientific evidence (Figure 1). The mix of inputs to clinical decisions varies from moment to moment and from day to day, depending on the decision and the decision makers. In general, however, the proportion of scientific evidence in the mix has grown progressively over the past 150 years or so.

One major reason why the mix has changed is simply the explosive increase in the amount and quality of the scientific evidence that has come from both the laboratory bench and the bedside. The maelstrom of change wrought by the molecular biology revolution has been matched at the clinical level by a tidal wave of increasingly sophisticated clinical trials. It is estimated that since the results of the first randomized clinical trials in medicine were published in the 1940s [2], roughly 100 000 randomized and controlled clinical trials have appeared in print [3], and the results of many well-conducted, completed trials remain unpublished [4]. A second reason for the growing emphasis on scientific evidence is the increasing expectation, from both within and outside of the medical profession, that physicians will produce and use “the evidence” in delivering care.

The future holds the promise of continued expansion of the body of research information. However, it also holds the parallel threat of increasingly inadequate time and resources with which to find, evaluate, and incorporate new research knowledge into everyday clinical decision making. Fortunately, mechanisms are emerging that will help us acquire the best, …