Repeated Measures vs. Crossover: Selecting an Optimal Design for a Pivotal Trial

• Activity Number: 345
• Type: Topic Contributed
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Theodore Lystig*
• Companies: Medtronic, Inc.
• Keywords: crossover ; medical devices ; pivotal trial ; repeated measures ; sham control

Abstract

Sponsors have multiple reasons for desiring smaller trials. Cost containment is a popular reason, as is a desire to minimize exposure of patients to a potentially inferior treatment. In this talk, I work in a setting where an important concern is that the pool of available patients will be small. The experimental intervention is surgical, and the control treatment will be a sham surgery. I will discuss options to a traditional two parallel arm randomized controlled trial, primarily in terms of alternative treatment sequences. This will be done be considering the alternatives as special cases nested within more general models, using two different parent models to reflect different assumptions. We can think of these as repeated measures designs, with crossover examples as special cases. The mathematical framework for comparing the models will be derived, and power curves presented.