Calibration in the Continual Reassessment Method for Phase I Clinical Trials

• Activity Number: 226
• Type: Topic Contributed
• Date/Time: Monday, August 5, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Xiaoyu Jia*+ and Shing M. Lee
• Companies: Columbia University and Columbia University
• Keywords: dose finding ; calibration ; two-stage continual reassessment method

Abstract

The continual reassessment method is an adaptive model-based design used to estimate the maximum tolerated dose in phase I clinical trials. In finite sample settings, it is important to specify a reasonable model. We propose a systematic approach to calibrate the two-stage continual reassessment method whereby model based dose escalation is activated after an initial sequence of patients are tested on lower dose levels. The calibration approach is based on the concepts of indifference interval and coherence. We compare the approaches to the traditional trial and error approach in the context of examples. The models selected by the systematic approach are competitive to the models obtained through a time consuming trial and error process. The semi-automatic calibration approaches simplify the model calibration process for the continual reassessment method. We also share our experience using the calibration technique in real life applications using the dfcrm package in R.