Applications of meta-analysis in pathology practice.

Abstract

Meta-analysis is the structured and systematic integration of information from different studies of a given problem. It has been widely used to integrate findings of randomized controlled trials (RCTs) investigating the efficacy of therapeutic interventions, but its use in pathology has lagged behind its use in clinical medicine. The purpose of this review is to present pathologists with the necessary background for reading and evaluating published reports of metaanalyses. The concepts and methods of meta-analyses of RCTs are presented, followed by a discussion of how these concepts and methods potentially can be applied to studies of diagnostic test accuracy. Three differences between RCTs and studies of diagnostic test accuracy are identified, and 4 possible obstacles to the use of meta-analysis in pathology are discussed. Despite these specific difficulties in the meta-analysis of diagnostic test data, meta-analysis can (and should) be used to produce valid summary estimates of the diagnostic accuracy of laboratory tests across all available studies.