Keywords: Bayesian method, pediatric trial, hierarchical model
There are many challenges facing the conduct of pediatric trials. And many pediatric trials are small and lack statistical power. Innovative and unconventional statistical methods are imperative in assisting drug development in this area. Bayesian statistical methods have been used more and more often to integrate historical information via using a prior. The prior can be built applying Bayesian hierarchical model, power prior, or a meta-analytic predictive prior. In certain situations, the empirical Bayes approach may also be considered. In this presentation, we discuss the application of the Bayesian meta-analytic approach to integrate information from multiple existing trials into a new pediatric trial. We use simulations to quantify the gains.