301 – Synthetic Data Approaches to Disclosure Limitation
Bayesian Inference for Meta-analysis of 2X2 Contingency Tables
Qi Tang
AbbVie Inc.
Natalia Kan-Dobrosky
AbbVie Inc.
Shihua Wen
Abbvie
Yuzhen Wang
AbbVie Inc.
Meta-analysis has been widely used for synthesizing safety and efficacy treatment effect information from multiple clinical trials to support landmark decision-making during drug development. Bayesian meta-analysis approach offers a very flexible modeling strategy. In this presentation, several fixed and random effect meta-analysis approaches under Bayesian framework are critically reviewed and applied to binary data (2X2 binary response contingency table). Simulations were used to evaluate the performance of different Bayesian Meta analysis approaches under the challenges of rare events, data heterogeneity, and unbalanced randomization by design.