Clinical Trial Design - Bayesian and Frequentist Adaptive Methods

There has been enormous interest and development in Bayesian adaptive designs, especially for early phases of clinical trials. nevertheless, for phase III trials, frequentist methods still play a dominant role through controlling type I and type II errors in the hypothesis testing framework. This book provides an overview of the fundamentals of clinical trials, the key terminologies and concepts, and a brief review and comparison on Bayesian and frequentist estimation and inference procedures. From the practical point of view, this book introduces various statistical methods that are commonly used for designing clinical trials and interim monitoring and analysis. Adaptation has a broad meaning in both Bayesian and frequentist perspectives, such as dose finding, trial early stopping for futility or superiority, dropping or adding an arm, seamless transitions between consecutive phases, group sequential methods, sample size re-estimation, adaptive randomization, and subpopulation enrichment, etc.§Comprehensive discussions on a variety of statistical designs, their properties, and operating characteristics for phase I, II, and III clinical trials are provided, as well as an introduction on phase IV trials. Many practical issues and challenges arising in clinical trials are addressed, and while the book mainly focuses on the Bayesian approaches for phase I and II trial designs, many important frequentist methods for phase III clinical trails are included. In addition, advanced and up-to-date topics such as jointly modeling toxicity and efficacy, seamless phase I/II trial designs, multiple testing, causal inference and noncompliance, adaptive randomization, issues associated with delayed outcomes, dose finding with combined drugs, and targeted therapy designs in personalized medicine development are discussed. Chapter coverage includes Fundamentals of Clinical Trials; Frequentist versus Bayesian Statistics; Phase I, II, and III Trial Designs, Adaptive Randomization, Late-onset Toxicity, Drug-combination Trials, and Targeted Therapy Design.