Drug Evaluation Committee Study on Statistical Inference for Adaptive Design
February 2023
The FDA finalized guidance on clinical trials using adaptive designs in 2019, and discussions are underway to establish harmonized guidelines under ICH-E20.
Because adaptive designs allow for modifications to the clinical trial protocol midway through the study, they are expected to offer benefits such as increased opportunities for trial participants to receive better treatment, more efficient drug development, and more effective utilization of available resources.On the other hand, if statistical methods are not used appropriately, undesirable phenomena from the perspective of statistical validity may occur, such as an increase in the Type I error rate, the introduction of bias into point estimates, and a coverage probability of the confidence interval falling below the nominal confidence level.
The Data Science Subcommittee of the Drug Evaluation Committee’s FY2022 Continuing Task Force 1 focused on adaptive designs for sample size re-estimation and treatment group selection, investigated methods of statistical inference, and compiled a report.
While this report is primarily intended for statistical analysts involved in the planning and conduct of clinical trials, the case studies of clinical trials using adaptive designs and the challenges regarding statistical validity will also be useful to members other than statistical analysts. We hope this report will contribute to a better understanding of adaptive designs and the proper conduct of clinical trials.
Published February 2023, Revised April 2024
Japan Pharmaceutical Manufacturers Association (JPMA) Drug Evaluation Committee
DS Subcommittee FY2022 Continuing Task Force 1