PURPOSE: While Bayesian methods have grown in popularity in the last few decades, applications in clinical development have only recently gained momentum. Bayesian methods may increase trial efficiency by leveraging prior information such as historical or real-world data to reduce the number of enrolled study participants. This workshop will provide a gentle introduction to the principles of Bayesian statistics including Bayes’ theorem, prior distributions, likelihoods, and posterior distributions. Attendees will learn about current opportunities for use in clinical development with a focus on practical considerations when implementing these methods. Common methods, their advantages and limitations, and use cases will be highlighted.

DESCRIPTION: Workshop attendees will receive a gentle introduction to Bayesian principles and understand opportunities for application in clinical development. Mr. Parzynski will introduce Bayesian thinking and how it differs from traditional statistical methods. Focus will be placed on the core components of Bayesian methodology: Bayes’ theorem, prior distributions, likelihoods, and posterior distributions (~15 minutes). Dr. Clark will then introduce clinical development settings in which Bayesian applications have been considered including early drug/biological product development, non-inferiority designs, adaptive designs, rare diseases, and pediatric studies. These will be presented alongside current challenges to their use. (~15 minutes). Next, Dr. Zeldow will introduce commonly used methods for increasing power for clinical trials using historical or real-world data (~15 minutes). The session will end with Dr. Saranadasa walking through primary considerations when pursuing these approaches in clinical development, highlighting limitations and decision hurdles with a particular emphasis on power borrowing methods (~15 minutes).