Introduction | The synthesis models | Unadjusted network meta-regression (NMR) | Using non-randomized studies (NRS) to construct priors for the treatment effects | Bias-adjusted model 1 | Bias-adjusted model 2 | Assumptions about the model parameters | Synthesis of studies comparing drugs for relapsing-remitting multiple sclerosis | Description of the data | Analysis | Unadjusted network meta-analysis | Unadjusted network meta-regression | References

Install R packages | Make R packages available | Default settings for R session | Import the dataset | Section 'Fixed effect and random effects meta-analysis' | Section 'Assessing the impact of missing outcome data' | Subgroup analysis of studies with and without missing data | Use imputation methods | Section 'Assessing and accounting for small-study effects' | Funnel plot | Harbord's score test for funnel plot asymmetry | Trim-and-fill method | Limit meta-analysis | References

Introduction | Data sets | Network meta-analysis in theory | Network meta-analysis in practice – Overview of netmeta | Three worked examples | Conclusions and outlook