Making predictive inference using Bayesian methods – two case studies
*Xuefeng Li, FDA 

Keywords: Bayesian, Hierarchical mode, Adaptive design

Bayesian method is the most intuitive and coherent way to synthesize observed data from different studies or different stage and make predictive inferences. In this talk, I will present two Bayesian case studies in medical device trials. The first one explores the between-trial predictions by borrowing information from a historical study via a hierarchical model. The second one demonstrates the within-trial prediction by implementing a Bayesian adaptive design. Some practice issues were also discussed from regulatory perspectives.