A Hybrid Frequentist-Bayesian Approach to a Pilot Study

• Activity Number: 503
• Type: Topic Contributed
• Date/Time: Wednesday, August 4, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Nicole Blackman*+
• Companies: GlaxoSmithKline
• Keywords: Bayesian designs ; interim monitoring ; early stopping ; clinical trials

Abstract

Clinical trials in humans generally progress from dose-finding ("Phase I") to initial investigations of efficacy ("Phase II"). However, it is often prudent to address safety and efficacy simultaneously, with formulation of statistical guidelines for early termination for harm. Pilot studies in hematology/oncology settings that combine the elements of dose-finding with initial investigations of efficacy are presented. Traditional two- or three-stage designs are used to evaluate efficacy. While efficacy is being assessed, a Bayesian approach is adopted to monitor safety, formally incorporating prior information about the incidence of important adverse events. Specific examples together with considerations for application in hematology/oncology settings are discussed. Design characteristics under various scenarios are presented.