Evidence-based practice in radiology: steps 3 and 4--appraise and apply systematic reviews and meta-analyses

A systematic review is performed in attempt to answer a specific research question by means of objective, unbiased evaluation of all pertinent available evidence. Component primary studies are selected on the basis of quality, and, if possible, their results are combined mathematically by using a process known as meta-analysis. While systematic review and meta-analysis are well-established methods to assess trials of therapeutic effects, they are increasingly more common in studies of diagnostic tests. In this article, the authors describe the benefits of a systematic approach over the traditional narrative review, illustrate the process, and examine some problems that are specific to systematic review and meta-analysis of diagnostic tests. They also explain how systematic review can help guide methodologic development for future research.