Monitoring Clinical Trials Based on the Bayesian Predictive Probability Using Data from Both Completers and Non-Completers

• Activity Number: 307
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Qi Tang*+ and Weining Zhao Robieson and Yili Lu Pritchett
• Companies: AbbVie Inc. and AbbVie and Astellas Pharma Global Development, Inc.
• Keywords: Probability of Success ; Adaptive Design ; Bayesian Predictive Probability ; Interim Analysis ; Repeated Measure ; Non-completer

Abstract

We have successfully applied a multiple-look futility criterion to monitoring clinical trials using the Bayesian predictive probability that predicts the chance of achieving a positive result at the end of study. In this algorithm, response data from subjects who have completed the study at interim evaluations were utilized to calculate Bayesian predictive probabilities because of the availability of a closed form expression, while partially observed data from ongoing subjects were not utilized to calculate the Bayesian predictive probability. To overcome this drawback while retain the benefit of having a closed form expression, in this presentation, we propose a new method for calculating Bayesian predictive probabilities where information from partially observed data are included using imputation methods. Comparing with the method based on completers, we conclude through simulations that the proposed method yields better operating characteristics under conditions such as fast enrollment speed, presence of data heterogeneity and frequent intermediate measurement.