FDA ISSUES GUIDANCE ON 'BAYESIAN' STATISTICS IN DEVICE TRIALS

The FDA's Center for Devices and Radiological Health (CDRH) May 23 issued draft guidance on the use of Bayesian statistical methods in device clinical trials that could significantly shorten those trials.

The guidance comes as part of the agency's device innovation initiative designed to bring innovative devices to market more quickly. The guidance outlines the use of adaptive clinical trial designs that may reduce the regulatory burden on devicemakers.

The document describes the Bayesian approach and its statistical design and analysis, pros and cons, how it compares with traditional methods, and how to use it in postmarket studies. The document does not describe the content of a device submission and is not all-inclusive, the agency notes.

Bayesian statistics provide "a coherent method for learning from evidence as it accumulates," the guidance states. "When correctly employed, [it] may be a less burdensome approach." ()a href="http://www.fdanews.com/ddl/33_22/" target=_blank>