Improving Clinical Study Design and Understanding via Modeling and Simulation

• Activity Number: 13
• Type: Topic Contributed
• Date/Time: Sunday, July 29, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Chyi-Hung Hsu*+ and Jose Carlos Pinheiro and Fei Chen
• Companies: Janssen Pharmaceuticals, Inc. and Johnson & Johnson and Janssen Pharmaceuticals, Inc.
• Keywords: clinical trial design ; modeling and simultion

Abstract

Sample size calculations for and overall design of clinical trials have since long relied on the statistical power for hypothesis test paradigm. This is true also for early development and dose-ranging studies, even though the key questions to be addressed in such studies are more of an estimation nature than related to testing a pre-determined hypothesis. Many reasons have contributed to this "Phase 3 view" of clinical drug development, including software availability and methodological training/knowledge. Modeling and simulation (M&S) approaches have freed statisticians and modelers from the "power calculation" dogma for trial design, allowing better understanding of the true underlying goes of the trial and appropriately tailoring the design to them. This talk will discuss the M&S view of study design, using examples from real trials to motivate and illustrate the key ideas.