Meta-Analysis

Meta-analysis, a tool developed to summarize the findings from randomized clinical trials (RCTs), can be used by many scientific fields, including health services research, to statistically combine data from many individual studies. A meta-analysis adds up the results for each participant in the experimental group and in the control group of all the relevant studies and presents an easily understood summary; it also provides a visual depiction of the outcome, a forest diagram, in which the results of each study are shown, making it obvious if all the studies agree or not. For example, if some studies find that an intervention or experimental group is worse than the control group, and other studies find it better, the disagreement can be seen at a glance.

The term meta-analysis was coined by the American statistician Gene V. Glass while he was a faculty member at the University of Colorado at Boulder in 1976. However, the practice actually originated before 1976 as many meta-analyses were published earlier. The use of meta-analysis in clinical medicine was systematically developed in the United Kingdom by the Cochrane Collaboration, an international group of thousands of volunteers founded in 1993 and named after the British epidemiologist Archibald “Archie” L. Cochrane (1909–1988). The Cochrane Collaboration is an international, not-for-profit organization that produces and maintains systematic reviews of healthcare interventions, doing their meta-analysis in a standard way. These meta-analyses are published electronically in the Cochrane Database of Systematic Reviews, which are published many times a year and can be easily updated.

Meta-analysis consists of (a) a systematic search of the literature, identifying studies by predefined criteria; (b) extracting numerical results from each study for the experimental and control subjects, on various outcomes and their difference; plus (c) the calculation of parameters reflecting their statistical confidence (e.g., standard deviation and sample size).