|
Abstract:
|
Let's consider a clinical trial with a binary outcome. The parameter of interest, say, ?, is the success (or failure) rate. For example, a critical safety parameter in the evaluation of a factor VIII product is the inhibitor rate, that is, the proportion of subjects who develop inhibitors. One key concept of Bayesian analysis in clinical trials is the exchangeability of the studies as discussed in FDA/CDRH (2006). To better understand Bayesian analyses, a basketball free throw (BFT) example will be discussed in this paper. The objective is to describe three key concepts in Bayesian approaches: (1) that ? is fixed for a given clinical trial, yet it can be considered random, (2) how historical data may be utilized through the assumption that studies are of exchangeable and (3) how the (Bayesian) credible interval may be interpreted from a Frequentist perspective.
|
