A comprehensive comparison of phase I dose escalation methods

Abstract

Phase I clinical trials are fundamental in drug development because they bring proposed designs to initial clinical testing. Recently, several dose finding methods have been developed. However, the comparison of those designs and traditional designs are not intensive. This study compares the most commonly used phase I dose finding methods and determines which one performs better. To do so, two different real life stories are analyzed through simulation studies. It was found that the 3+3 design, the most popular method employed by scientists, produced the worst results. More reliable and applicable results for phase I dose escalation trials can be produced by BMA-CRM, CRM, and BCRM designs.