The medical review article revisited: Has the science improved?

Background: The validity of a review depends on its methodologic quality. Objective: To determine the methodologic quality of recently published review articles. Design: Critical appraisal. Setting: All reviews of clinical topics published in six general medical journals in 1996. Measurements: Explicit criteria that have been published and validated were used. Results: Of 158 review articles, only 2 satisfied all 10 methodologic criteria (median number of criteria satisfied, 1), Less than a quarter of the articles described how evidence was identified, evaluated, or integrated; 34% addressed a focused clinical question; and 39% identified gaps in existing knowledge. Of the 111 reviews that made treatment recommendations, 48% provided an estimate of the magnitude of potential benefits (and 34%, the potential adverse effects) of the treatment options, 45% cited randomized clinical trials to support their recommendations, and only 6% made any reference to costs. Conclusions: The methodologic quality of clinical review articles is highly variable, and many of these articles do not specify systematic methods.