|
Activity Number:
|
251
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 1, 2016 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
|
Abstract #321058
|
|
|
Title:
|
Bayesian Statistical Models Assessing Disease Modification Drug Effects Using Doubly Randomized Delayed-Start and Matched Control Designs
|
|
Author(s):
|
Ibrahim Turkoz* and Marc Sobel
|
|
Companies:
|
Janssen R&D and Temple University
|
|
Keywords:
|
Disease Modification ;
Model Averaging ;
Spline Models ;
Bayesian Hierarchical Models
|
|
Abstract:
|
Randomized delayed-start designs for assessing disease modification have significant analytic limitations.Because following initial randomization, high percentage of subjects in the control arm are likely to drop out before entering into the delayed-start period.Proposed study design addresses limitations associated with delayed-start designs that might be used to support a disease modification indication.Proposed innovations represent hybrids of randomized and epidemiologic designs.We describe a design with a run-in period followed by second randomization just prior to initiation of the delayed start.The run-in period is used to provide a basis for a matched control re-randomization.The second randomization generates groups that are comparable with respect to disease characteristics.A number of models have been developed to identify "disease modification."The "failure-to-catch-up" concept is evaluated using Bayesian hierarchical models.Bayesian methodology characterizes the truth of the failure-to-catch-up concept using posterior probabilities.This is a natural fit for establishing disease modification compared to traditional noninferiority margins or parallel-line approaches.
|
Authors who are presenting talks have a * after their name.
