|
Activity Number:
|
74
- Invited E-Poster Session I
|
|
Type:
|
Invited
|
|
Date/Time:
|
Sunday, August 7, 2022 : 8:30 PM to 9:25 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
|
Abstract #323256
|
|
|
Title:
|
A Bayesian Approach for Integrating External Data in Clinical Trials Using Overlap Prior
|
|
Author(s):
|
Yanxun Xu*
|
|
Companies:
|
Johns Hopkins University
|
|
Keywords:
|
Clinical trials;
Power priors;
Overlap weights;
Real world data
|
|
Abstract:
|
Integrating external data from past clinical trials or observational studies becomes increasingly important in oncology clinical trials. It has been extensively studied using propensity score-based and Bayesian dynamic borrowing-based approaches. While the former the cannot adjust for discrepancies due to unmeasured confounding, the latter often fails to reward the similarity in covariates. We develop a new class of power priors using overlap weights that can integrate individual patient data and address the shortcomings of the traditionally-used propensity score and Bayesian dynamic borrowing approaches.
|
Authors who are presenting talks have a * after their name.
