Meta-analysis: When and how

Systematic reviews have a central role in evidence-based medicine. The quan titative systematic review, also known as meta-analysis, provides a logical structure for quantifying evidence and for exploring bias and diversity in research systematically. It is essential that clinicians, educators, and r esearchers understand the methods that comprise this research tool, particu larly the basic step-by-step process, and know when numerical pooling of da ta is appropriate. The essay describes how systematic reviews are best cond ucted and when statistical pooling of data is appropriate. Systematic revie ws are scientific investigations with planned methods that use original stu dies as subjects and synthesize the results of multiple studies using strat egies to limit bias and random error. This process requires judgments to be made explicit, and should be question driven, protocol based, reproducible , and comprehensive in scope. Meta-analysis provides a framework for resear ch synthesis, increases power and precision, provides an overall estimate a nd range of effect, and identifies greater-than-expected variability among study results (heterogeneity). Metaanalysis does not remove subjectivity fr om the process of synthesis, identify sources of variability among studies, or obviate the need for sound, compassionate clinical reasoning. Statistic al heterogeneity should be anticipated and welcomed. It forces a considerat ion of clinical heterogeneity as well as variation in study protocol and qu ality. Statistical tests for homogeneity are insensitive and do not indicat e sources of heterogeneity, making such consideration imperative, The most common and popular measures of efficacy for a meta-analysis are the standar dized difference between two means, the relative risk, and the odds ratio. An additional measure, the number needed to treat, with its 95% confidence interval is the most clinically useful measure of the effects of an interve ntion and is useful for comparing the relative effectiveness of different i nterventions for the same condition. Important parts of metaanalysis and se nsitivity and subgroup analyses are best considered a priori and should be used to explore heterogeneity and to test for publication bias and variatio n in study quality.