Fig. 5From: Borrowing information across patient subgroups in clinical trials, with application to a paediatric trialTreatment difference estimates and 95% intervals in Example 2, obtained from a pooled analysis of all children and a Bayesian analysis for children weighing < 14 kg (incorporating evidence from children weighing ≥ 14 kg as a prior distribution), together with results from separate subgroupsBack to article page