A comparison of five approaches to decision-making for a first clinical trial of efficacy

The first trial of clinical efficacy is an important step in the development of a compound. Such a trial gives the first indication of whether a compound is likely to have the efficacy needed to be successful. Good decisions dictate that good compounds have a large probability of being progressed and poor compounds have a large probability of being stopped. In this paper, we consider and contrast five approaches to decision-making that have been used. To illustrate the use of the five approaches, we conduct a comparison for two plausible scenarios with associated assumptions for sample sizing. The comparison shows some large differences in performance characteristics of the different procedures. Which decision-making procedures and associated performance characteristics are preferred will depend on the focus of interest and the decision maker's attitude to risk. Copyright © 2016 John Wiley & Sons, Ltd.