A Unified Decision Framework for Phase I Dose-Finding Designs

The purpose of a phase I dose-finding clinical trial is to investigate the toxicity profiles of various doses for a new drug and identify the maximum tolerate dose. Over the past three decades, various dose-finding designs have been proposed and discussed, including conventional model-based designs, new model-based designs using toxicity probability intervals, and rule-based designs. We present a simple decision framework that can generate several popular designs as special cases. We show that these designs share common elements under the framework, such as the same likelihood function, the use of the loss functions, and the nature of the optimal decisions as Bayes rules. They differ mostly in the choice of the prior distributions. We present theoretical results on the decision framework and its link to specific and popular designs like mTPI, BOIN, and CRM. These results provide useful insights into the similar theoretical foundations of these designs. We also show that the designs exhibit similar operating characteristics. Therefore, the choice of a design for a practical trial among the ones we reviewed may be up to the statistician's and clinician's own preference, such as preference of more model-based approach or more simple and transparent decisions.