Maximally Flexible Bayesian Designs in Randomized Clinical Trials
View Presentation View Presentation
  *Frank Harrell, Jr., Vanderbilt University     


This presentation will cover the disadvantages of the traditional frequentist paradigm and will then turn to the development of a Bayesian flexible design for a biologic agent in a Phase II clinical trial. The design allows for infinitely many looks at the data and for possible study extension and conversion to adaptive allocation. Unlike frequentist sample size re-estimation procedures, the Bayesian procedure does not require penalizing the final analysis for having done earlier analyses nor does it require downweighting of data collected before the decision to extend the study. The study is easily extended if results obtained at the originally planned study termination are equivocal. The final analysis uses the same analysis procedure as used at the initial analysis, whereas there is no consensus in the frequentist world for how to analyze an extended study. Our primary analysis is based on a Bayesian Cox proportional hazards model using a skeptical prior distribution.