Topic Review - Newest First (only newest 5 are displayed)

rqayyum

Bayesian Statistics

It is interesting to note that, yesterday, FDA published draft guidelines for Bayesian methods in medical device trials. Bayesian methods differ from the commonly used statistical methods (generally known as frequentist) in some interesting ways.
The way FDA draft points out the difference is interesting. It says:

“Bayesian statistics is a statistical theory and approach to data analysis that provides a coherent method for learning from evidence as it accumulates. Traditional (frequentist) statistical methods formally use prior information only in the design of a clinical trial. In the data analysis stage, prior information is considered only informally, as a complement to, but not part of the analysis. In contrast, the Bayesian approach uses a consistent, mathematically formal method called Bayes’ Theorem for combining prior information with current information on a quantity of interest. This is done throughout both the design and analysis stages of a trial.”

Bayesian methods are intuitively attractive and extremely flexible yet still capable of performing very complex tasks that are either very difficult or impossible with standard (frequentist) methods. Unfortunately, Bayesian methods are rare in Medical (basic and clinical) research. I hope that this proactive approach by FDA will be helpful in wider dissemination of Bayesian methods.

Here is the link:
http://www.fda.gov/cdrh/osb/guidance/1601.html