Bayesian Statistics at FDA, Especially for Clinical Trials of Medical Devices - SOLD OUT

• Activity Number: 236
• Type: Luncheons
• Date/Time: Tuesday, August 10, 2004 : 12:30 PM to 1:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Gene A. Pennello*+
• Companies: U.S. Food and Drug Administration
• Keywords: therapeutic devices and implants ; diagnostic test ; microarray ; hierarchical model ; interim analysis ; post-market adverse event data-mining

Abstract

Bayesian design and analysis is part of an increasing number of pre-market submissions to the Center for Devices and Radiological Health (CDRH) of the Food and Drug Administration (FDA). This initiative, which began approximately five years ago, takes advantage of good prior information on safety and effectiveness that is often available from studies on the same or recent generation devices. To date, about fourteen PreMarket Approval (PMA) applications have been approved by CDRH based on Bayesian analyses, including coronary stents (both drug-eluting and not), spinal fixation devices, a diagnostic breast imager, an artificial urinary sphincter for urinary incontinence, and a female permanent contraception implant. Four were presented to an FDA Advisory Committee Panel of outside experts for a recommendation of approval or disapproval by FDA. Bayesian methodology is also being used at the Center for Drug Evaluation and Research (CDER) and the Center for Biologics Evaluation and Research (CBER) to datamine post-market adverse event reports for associations of events with drugs and vaccines, and CDRH is looking to do the same. Bayesian methodology may also be helpful.