Reproducible Bayesian hierarchical evidence synthesis of mortality outcomes across beta-blocker trials, with heterogeneity assessment, prior sensitivity, posterior diagnostics, and partial pooling.