[Meta-analyses-Explained my means of examples from anesthesia and pain medicine].

Abstract

Meta-analyses are a central part of systematic reviews. The term meta-analysis describes the statistical methods to summarize the results of the available scientific studies providing the highest possible evidence. In medicine, meta-analyses aim to guide clinical decisions. This article provides an overview of the necessary work steps.The classical meta-analysis summarizes the results of randomized controlled trials that compare an intervention against a control intervention. This is illustrated by means of an example from a Cochrane Review on videolaryngoscopy in comparison to direct laryngoscopy. Crucial methodological aspects such as the weighting of individual studies when pooling their results as well as the evaluation of study heterogeneity and potential publication bias are explained.The second part of the article focusses on two extensions of meta-analyses: the individual patient data meta-analysis and the network meta-analysis. The individual patient data meta-analysis makes use of the information that is available from the patient-level data of the included studies. As an example, the work accomplished by an international collaboration on the efficacy of acupuncture in chronic pain is presented. A network meta-analysis enables the comparison of more than two interventions by making use not only of the available direct but also of the respective indirect evidence. This is illustrated by means of a Cochrane Review on drugs for the prophylaxis of postoperative nausea and vomiting.