Randomised Blinding of a Phase I Trial Using Bayesian Hierarchical Optimisation

Abstract

This paper presents a brief overview of Bayesian design of clinical trials from a Bayesian perspective. Hierarchical Bayesian prior adaptive designs are adopted with sampling priors. The study explores how the variances of Bayesian Gaussian priors impact the optimal design parameters in phase I. This means that the uncertainty in the prior beliefs about model parameters will be minimized using our model set up and it will determine the optimal experimental setup.