Activity Number:
|
579
- Tests of Noninferiority in the Clinical Trial Setting
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, August 2, 2017 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Biopharmaceutical Section
|
Abstract #323615
|
|
Title:
|
Bayesian Approach of Assessing Non-Inferiority of a New Treatment with Binary Outcomes in a Three-Arm Trial
|
Author(s):
|
Shrabanti Chowdhury* and Samiran Ghosh and Ram Tiwari
|
Companies:
|
and Wayne State University - Detroit, MI and FDA/CDER/OT/OB
|
Keywords:
|
Non inferiority ;
Bayesian ;
Placebo ;
Historical ;
Prior ;
Sample size
|
Abstract:
|
Non inferiority (NI) trials are of great importance for comparative effectiveness research when one cannot guarantee superiority of one treatment over the other. Such trials aim to demonstrate that an experimental treatment is non-inferior to an existing clinically proven active comparator by not more than a pre-specified margin. The inclusion of a placebo arm in NI trial giving rise to three-arm trial is prudent, if ethically reasonable, since it requires less stringent assumptions compared to the two-arm placebo-free trial. We consider fraction margin approach where NI margin is formulated as a pre-determined negative fraction of the unknown effect size of the reference drug in current trial. We propose a Bayesian procedure for testing NI in three-arm trial with binary outcomes. Bayesian paradigm provides a natural path to integrate historical and current trial as well as uses patients'/clinicians' opinions as prior information. Inclusion of prior information may greatly reduce the cost burden on current trial in terms of effective sample size and manpower. The developed statistical methods are assessed via extensive simulation studies and illustrated with a clinical trial data.
|
Authors who are presenting talks have a * after their name.