TY - JOUR T1 - Interpretation of Metaanalyses: Pitfalls Should Be More Widely Recognized JF - The Journal of Rheumatology JO - J Rheumatol SP - 1107 LP - 1109 DO - 10.3899/jrheum.111420 VL - 39 IS - 6 AU - YVES E. HENROTIN Y1 - 2012/06/01 UR - http://www.jrheum.org/content/39/6/1107.abstract N2 - On the basis of metaanalysis of randomized studies assessing the symptoms and radiological progression of patients with osteoarthritis (OA), investigators have concluded that health authorities and health insurers should not cover the costs of glucosamine and chondroitin, and new prescriptions to patients who have not received treatment should be discouraged. This conclusion has the potential to change current management of OA mainly in Europe, where these compounds are prescribed drugs. But should metaanalysis be considered the ultimate level of evidence and sole support for these conclusions? This editorial addresses this question.A metaanalysis is defined as a panel of statistical methods of combining data coming from a set of comparable studies addressing a particular question. A metaanalysis may or may not be a part of a systematic review yielding a quantitative summary of the pooled results1. In general, metaanalyses are used to support evidence-based recommendations. The general aim of a metaanalysis is to more powerfully estimate the true “effect size” as opposed to a smaller effect size derived in a single study under a given single set of assumptions and conditions. Reasons for considering a metaanalysis in a review include increasing the power (higher chance to detect an effect) and precision to answer questions not posed by individual studies, and to settle controversies arising from apparently conflicting studies or to generate new hypotheses. Of course, the use of statistical methods does not guarantee that the results of a review are valid, any more than it does for a primary study.It has been largely recognized that good research practices in conducting … Address correspondence to Dr. Henrotin; E-mail: yhenrotin{at}ulg.ac.be ER -