|
Abstract:
|
Phase II clinical trials typically are single-arm studies conducted to decide if an experimental treatment, E, is promising, relative to standard treatment, S, to warrant further investigation. Many methods exist for conducting phase II trials under the assumption of patient homogeneity. In the presence of patient heterogeneity, however, these designs are likely to draw incorrect conclusions. We propose a class of model-based Bayesian designs for single-arm phase II trials with a binary outcome and two or more prognostic subgroups. The designs' early stopping rules are subgroup-specific and allow the possibility of terminating some subgroups while continuing others. We require informative priors on S and subgroup main effects, and noninformative priors on E and treatment-subgroup interactions. An algorithm for computing prior hyperparameter values and a simulation study are presented.
|
