Bayesian designs and the control of frequentist characteristics: A practical solution

摘要: Frequentist concepts, such as the control of type I error or false discovery rate, are well established in medical literature and often required by regulators. Most Bayesian designs defined without explicit considerations frequentist characteristics. Once design is structured, statisticians use simulations adjust tuning parameters to comply with a set targeted operating These adjustments affect prior information utility functions. Here we consider decision theoretic approach for experimental requisites. We define optimal under constraints regulator. Our combines interpretable functions criteria, selects an that satisfies illustrate using group-sequential multi-arm Phase II trial bridging trial.