Introduction to Network Meta-Analysis

Many treatment decisions in hospital medicine are not divalent choices between 2 options but, rather, decisions among multiple possible interventions. While traditional pairwise meta-analysis can provide information on the effectiveness of 1 or 2 interventions, a different strategy is required to evaluate the relative merits of multiple possible interventions. Network meta-analysis is a robust statistical tool to analyze 3 or more interventions for a single problem, allowing for comparisons of interventions that may not have been directly compared in head-to-head studies. This guide provides an introductory overview of network meta-analysis, focusing on how this analysis differs from traditional pairwise meta-analysis. We focus on key assumptions in network meta-analysis, including transitivity and consistency as well as considerations of evidence strength and bias. This guide can serve as a starting point for a researcher seeking to perform their own network meta-analysis or a guide for a clinician looking to interpret published studies.