Drug Evaluation Committee How to use simulation in Bayesian study design
October 2025
In 2020, the FDA issued “Interacting with the FDA on Complex Innovative Trial Designs for Drugs and Biological Products,” which discusses the use of Bayesian approaches in specialized clinical trial designs such as adaptive designs, external controls, and master protocols.In Japan as well, the “Basic Guidelines on the Application of Bayesian Approaches in Clinical Trials for Rare Diseases” were compiled as part of AMED’s Research Project on Harmonization and Evaluation of Regulations for Pharmaceuticals and Related Products, andand the “Points to Note Regarding the Use of Master Protocol Trials in Drug Development,” issued in June 2024, also refers to Bayesian analytical methods, indicating that this framework is becoming increasingly established.
In trials using Bayesian analysis, while decisions based on interim trial results can allow for flexibility in the trial, and incorporating external information can enable efficient drug development,depending on the prior distribution and decision criteria, there is an increased risk of drawing erroneous conclusions; therefore, their performance must be carefully evaluated before conducting the trial. Consequently, simulation plays a crucial role in evaluating the performance of Bayesian trial designs.
This document examines simulation methods for measuring the behavioral characteristics of Bayesian trial designs, aimed at those responsible for the statistical analysis of clinical trials, with the goal of enabling them to plan appropriate Bayesian trials when necessary.
Japan Pharmaceutical Manufacturers Association (JPMA) Drug Evaluation Committee
Data Science Subcommittee FY2024 Task Force 3-2