Department of Biostatistics Seminar/Workshop Series

Bayesian trial monitoring: Separating the baby from the bathwater

John Kittelson, PhD

Associate Professor, Department of Biostatistics and Informatics, University of Colorado at Denver

Wednesday, December 8, 1:30-2:30pm, MRBIII Conference Room 1220

As others have observed, clinical trials function as diagnostic tests for identifying beneficial (or non-beneficial) therapies. Within clinical trials, the positive predictive value (PPV) is the prevalence of truly beneficial therapies among all trials that show significant benefit. In this talk I evaluate common approaches to trial design and monitoring in terms of their effect on PPV. In particular I focus on Bayesian approaches to trial monitoring. In order to solve some shortcomings, I propose and evaluate a general Bayesian design and monitoring framework that is parameterized to focus on the scientifically important aspects of the trial. I illustrate the methods in the context of a clinical trial to evaluate a new approach for treating pre-hypertension in young adults.

Topic revision: r5 - 26 Apr 2013, JohnBock

This site is powered by FoswikiCopyright © 2013-2017 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Vanderbilt Biostatistics Wiki? Send feedback