Basic concepts in meta-analysis: a primer for clinicians

ali khoshdel

With the expanding volume of medical literature, meta-analysis, a form of systematic review, has become indispensable for clinicians for evidence-based decision making. While the number of meta-analyses has substantially increased during recent years, there are still controversial issues regarding their methodology, interpretation and clinical application. In this review, the basic concepts of meta-analysis have been discussed from a clinician's perspective in order to facilitate its understanding, appraising and applicability in clinical practice. Although randomised controlled trials are the usual source for meta-analysis, observational studies are also being increasingly considered for meta-analysis. Like every other research design, meta-analysis starts with formulating a question, followed by searching for related data, based on predefined criteria and strategies. Inclusion of studies must be carried out with careful consideration of their quality and assessment of homogeneity using graphical means as well as statistical tools such as Q statistics, I2 statistics and meta-regression. The pooled effect size is commonly calculated using either a ‘fixed effect model’ or ‘random effect model’. Publication bias and other source of bias should be investigated and the impact of potential confounders should be eliminated as required. Given the above-mentioned considerations, meta-analysis can provide a more precise estimate of an effect size to be used in clinical decision making.