Choosing the appropriate statistics

The contrast-based (CB) method for network Meta-analysis (NMA) on binary outcomes typically focuses on the relative treatment effects and is incapable to estimate the overall treatment-specific event rates. There have been a lot of papers reporting only odds ratios (ORs) based on the CB method. However, there is doubt about the appropriateness of merely reporting this summary statistic. In this chapter, we present a recently proposed arm-based (AB) method developed from a missing data perspective and illustrate how treatment-specific event proportions, risk differences (RDs), risk ratios (RRs), and ORs can be computed in NMAs. This more comprehensive and appropriate reporting of summary statistics facilitates patients and their caregivers to understand and trade-off efficacy and safety and thus the AB approach is recommended in practice.