Bayesian Adaptive D-Optimal Design with Delayed Responses

Abstract

In some disease areas such as diabetes and obesity, patients may need to be studied for several weeks or months for a drug effect to emerge, where longitudinal data are normally collected for each patient. These delayed responses provide challenges to current adaptive design methods. Fu and Manner (2010) provided an integrated two-component prediction (ITP) model for delayed response adaptive design. In this paper, we extend their ITP model to incorporate a dose-response model in it and propose an ITP Emax model. Furthermore, we propose a utility function for decision making and this utility function is based on D-optimal design theory. We demonstrate our method by a simulation study, and potential sample size reduction is also discussed.

Key Words:

• Adaptive design
• Bayesian method
• Delayed responses
• D-optimal
• ITP model
• Pseudo-Bayesian D-optimal design