Keywords: Bayesian, non-inferiority, confirmatory, innovative designs, simulations
Innovative Bayesian designs can be used for establishing non-inferiority to standard of care (SOC) without sacrificing the goal of evidence-based medicine. The Bayesian philosophy of combining direct data observed in the current study with evidence borrowed from historic clinical trials allows to design more cost effective and efficient clinical trials. In this presentation, the example of a Bayesian non-inferiority study will be discussed in the confirmatory setting. The augmentation of the SOC arm with historic efficacy information from publicly available clinical trials was proposed to more efficiently use the available data for SOC as well as to maximize information from treated patients in the study. Simulations were an integral part of the design development and justification. Great attention was given to the choice of prior distributions, the amount of efficacy data to borrow, and the selection of the decision rule to declare non-inferiority to show adequate Type 1 error rate control and other operating characteristics.