[Bayesian methodology for innovative clinical trial designs]

With the rapid growth of precision medicine and the increasing demand for novel drug development, traditional development strategies and trial designs are no longer adequate for contemporary clinical research, prompting regulatory agencies, industry, and academia to explore more flexible and efficient approaches. Bayesian methods are emerging as a vital statistical tool for innovative clinical trial design due to their flexibility and strong interpretability. In critical fields of clinical trials such as dose finding and optimization, adaptive design, external information borrowing, and master protocol design, Bayesian methods have demonstrated unique value distinct from traditional frequentist approaches. This paper systematically reviews common Bayesian methods in innovative clinical trial design and discusses crucial considerations for their application, aiming to guide the standardized, rational use of Bayesian methods.