The FDA's Center for Devices and Radiological Health (CDRH) is making a guidance available that provides agency recommendations on the use of Bayesian statistical methods in the design and analysis of device clinical trials, the agency has announced. The guidance is not final, and it not in effect at this time, the agency said.
Bayesian statistical methods are named after English theologian and mathematician Thomas Bayes, and involve calculating the probability of a new event based on earlier probability estimates that come out of empirical, or observation-based, data. They are currently used in a variety of medical device applications to FDA.
The guidance includes a general description of Bayesian methods, discussions on design and analysis of Bayesian medical device clinical trials, the benefits and difficulties with the Bayesian approach, and comparisons with standard, or "frequentist," statistical methods. The document also presents ideas on how to use Bayesian methods in postmarket studies.
To view the pre-publication announcement on the ''Guidance for the Use of Bayesian Statistics in Medical Device Clinical Trials," go to http://www.fda.gov/OHRMS/DOCKETS/98fr/ch0626.pdf (http://www.fda.gov/OHRMS/DOCKETS/98fr/ch0626.pdf).