Meta-analyses - important points to consider

Meta-analyses are very important instruments to evaluate the effect of therapeutical interventions, before all regarding evidence based medicine (EbM). This priority is problematic, especially if the inherent methodological problems of meta-analyses are not sufficiently taken into account. In contrast, the evaluation of single trials seems to be nowadays relatively unimportant. However, the correct interpretation of meta-analyses is strongly dependent on methodological issues, if those issues are not properly included into the interpretation, false conclusions might be drawn. The present paper focuses on two special points: selection criteria and multi-treatment comparisons. Taken together, results from meta-analyses must be weighted very carefully.