Meta Analysis – An Alternative to Large Trials

A large series of studies on the same treatment and for the same indication often include trials that lean one way while others go in the opposite direction and a few are just neutral. A relatively new statistical technique, the meta analysis, is seen as a hopeful method that can take a large body of studies and provide an overall assessment of a treatment effect. The technique, admired because of its ability to take many small trials and blend them into a consensus finding has its detractors as well. There are methodological issues such as accounting for unpublished studies, or the inclusion of published results that are of poor quality. The absence of individual patient data can also be a deterrent to a sound analysis. Examples of meta analyses cases illustrate where the method has provided useful revelations as well as questionable conclusions. In the end, however, because there is an incomplete understanding of the technique, caution is called for when interpreting the results from a meta analysis. There is clearly a need for meta analysts to educate the health care professions about the strengths and weaknesses of this promising technique.