American Statistical Association
The impact of one's prior on adaptive decisions in a Bayesian clinical trial design can be profound. For example, the prior may have a substantive effect on treatment assignments involving life-and-death outcomes in trials of treatments for rapidly fatal diseases. This issue is at the heart of what may be considered the strongest frequentist criticism of Bayesian methods. Despite this, little attention has been paid to the problem of eliciting and calibrating priors in such settings. In this talk, I will discuss how priors were established in a variety of different Bayesian clinical trials in which I have been involved. As time permits, topics will include Dirichlet priors for discrete 2x2 (response, toxicity) outcomes, the notion of prior effective sample size, logistic regression and generalized logistic regression models for toxicity as a function of either the dose of a single agent or the doses of two agents in phase I trials, using penalized least squares to establish a prior for bivariate (efficacy, toxicity) probabilities as a function of dose, eliciting a hyperprior for a hierarchical model in a phase II activity trial, and eliciting priors for event-free survival time and toxicity as functions of patient covariates in a pediatric brain tumor trial.
Peter F. Thall received his PhD from The Florida State University Department of Statistics in 1975. He was a faculty member in the Program in Mathematical Sciences at the University of Texas at Dallas from 1975-1980, and in the Department of Statistics at George Washington University from 1980-1990. During 1986-87, he spent a sabbatical year collaborating with Richard Simon at the Biometric Research Branch of the Cancer Therapy Evaluation Program at NCI. Dr. Thall has worked at MD Anderson Cancer Center since 1990, where he holds the Anise J. Sorrell Endowed Professorship. He has pioneered the application of Bayesian statistical methods in medical research, and has designed over 300 clinical trials. He has published over 170 research papers and book chapters and has presented 23 short courses and over 150 invited talks. He has been the head biostatistician on program project grants in acute myelogenous leukemia, chronic myelogenous leukemia, neuro-oncology, phase II/III clinical trials, immunology, and soft tissue sarcoma. He has served as an Associate Editor for Statistics in Medicine, J. National Cancer Institute, and Biometrics, and regularly serves on clinical trial data safety monitoring boards, and NIH steering committees, working groups, and advisory boards. Currently, he is an Associate Editor of Clinical Trials and Statistics in Biosciences, serves as an American Statistical Association Media Expert, is the 2012 Program Chair-elect for the American Statistical Association Section on Bayesian Statistical Sciences, and is the Principal Investigator of the National Cancer Institute RO1 grant for Statistical Methods for Complex Cancer Trials.
|Date:||Thursday, June 30, 2011|
|Time:||2:00 - 3:00 P.M.|
Mailman School of Public Health
Department of Biostatistics
722 West 168th Street
Biostatistics Conference Room
6th Floor - Room 627
New York, New York