ABSTRACT
Network meta-analysis allows to jointly estimate the relative effectiveness of various treatments or interventions. To achieve this, network meta-analyses combine both direct (i.e. observed) and indirect evidence in a network.
This chapter provides a brief introduction into network meta-analysis, starting with a discussion of indirect and direct evidence, and how both can be combined to obtain more precise effect estimates. The concept of transitivity and consistency is also discussed.
The chapter covers both a frequentist and Bayesian network meta-analysis approach, and provides a hands-on tutorial on how these methods can be implemented using R. Furthermore, techniques to assess publication bias mechanisms in network meta-analyses are described, and how network meta-regression can be conducted using R.
