Pooling of Data Across Multiple Subgroups in FDA Submissions

• Activity Number: 70
• Type: Topic Contributed
• Date/Time: Sunday, July 29, 2012 : 4:00 PM to 5:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Tyson Rogers*+ and Jeff Cerkvenik
• Companies: Medtronic and Medtronic
• Keywords: Pooling, FDA, medical devices

Abstract

Every year, medical device use increases in countries outside the United States (OUS) causing medical device manufacturers to want to test their products in many different countries simultaneously. In recent international studies, the U.S. Food and Drug Administration (FDA) Center for Devices and Radiological Health (CDRH) division has advocated an alpha level of 0.15 on a test of pooling data between US and OUS subgroups of patients within the same study. A case study and simulations will show that if this informal policy is adopted as standard practice, it will greatly reduces the chance (power) of a positive study outcome for device manufacturers, increasing costs and risks. A 15% chance of a false finding that US and OUS patients are not poolable would deny the timely availability of some safe and effective medical devices to US consumers.