Summer 2006 Workshop

Held on June 22, 2006 at the Renaissance Chicago North Shore Hotel.

The 22st Annual Summer Workshop of the Northeastern Illinois Chapter of the American Statistical Association was given on the topic of:

Bayesian Clinical Trials
Dr. Scott M. Berry, Berry Consultants, College Station, TX

Summary

The course will describe recent Bayesian innovations in the design and analysis of clinical trials. The goals are (i) more efficient clinical trials and clinical development programs, and (ii) treating patients more effectively, both those in and outside of trials. The presenter will provide some background on Bayesian designs for clinical trials and give several case studies of the adaptive approach used in actual designs and analyses presented to the FDA. These examples include the possibility of early stopping, seamless phase II and III trials with sequential sampling, using early endpoints to guide sample size and adaptive dose finding. The cost savings of such an approach are usually substantial. The only prerequisite knowledge necessary is a basic understanding of statistics and of clinical trials.  

Learning Outcomes

The course is designed with two purposes in mind. The first purpose is to inform attendees of the Bayesian approach to the design of clinical trials. They will learn new methods for conducting “better” clinical trials. The hope is that a general knowledge of these new methods and related issues will be understood. The second purpose is to provide enough detail that someone involved in the design of clinical trials could conduct a Bayesian clinical trial.

The instructional method will be hands-on. The presenter will provide real examples, with the real issues involved. The best way to understand the benefits of the Bayesian design is to see how it works in real trials being conducted. By teaching through these real problems the attendees will see the potential advantages of the Bayesian design as well as gain the ability to carry out similar trials.

Biographical Background

Scott M. Berry received his Ph.D. from Carnegie Mellon University (1994) and his B.S. from the University of Minnesota (1990). He spent 5 years at Texas A&M University in the Statistics Department. The last five years he has been consulting for Berry Consultants, where he has designed more than 20 purely Bayesian clinical trials for pharmaceutical and medical device companies. He leads the statistical aspects of these trials as well as writing the software and doing the modeling needed.