[How to interpret systematic reviews with multiple comparisons or network meta-analysis].

Systematic reviews evaluating multiple interventions can be useful in different clinical situations. However, some concerns arise when more than two interventions are compared and there is a paucity of good quality randomized clinical trials. A novel statistical method based on indirect comparisons, called network meta-analysis (NMA), can be a useful approach to find a clinical answer when multiple interventions are evaluated for the same outcome or comparator. The aim of this review is to describe the main characteristics and provide a user guide for a critical analysis of NMA focusing on its three main domains, namely homogeneity, transitivity and consistency.