|
Activity Number:
|
85
- Contributed Poster Presentations: Statistics and Pharmacometrics Interest Group
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2020 : 10:00 AM to 2:00 PM
|
|
Sponsor:
|
Statistics and Pharmacometrics Interest Group
|
|
Abstract #311111
|
|
|
Title:
|
Estimation of Treatment Effects in a Tumor-Agnostic Trial Based on Non-Randomized Control Group
|
|
Author(s):
|
Wei Wang* and Kan Li and Razi Ghori and Sanatan Saraf and Lei Xu
|
|
Companies:
|
Merck and Merck and Merck and Merck and Merck
|
|
Keywords:
|
Generalized linear mixed model;
Non-randomized control group;
Propensity score ;
Treatment effect;
Tumor-agnostic trial;
Baseline characteristics
|
|
Abstract:
|
Tumor-agnostic drug approvals in recent years marked a paradigm shift for cancer treatment and cancer drug development. A drug can be used to treat a tumor based on a specific molecular signature, regardless of where in the body the tumor started. In cancer clinical trials, estimation of treatment effects based on non-randomized control group occurs frequently. Propensity score methods are widely used to account for systematic differences in baseline characteristics between treatment and control groups. The advent of tumor-agnostic trials adds another layer of complexity to the data analysis. Our research aimed to investigate statistical analysis methods based on propensity score for the estimation of treatment effects in a tumor-agnostic setting using non-randomized control group. A generalized linear mixed model (GLMM) in conjunction with propensity score methods were used as the framework. Simulation studies were conducted to evaluate the performance of different methods using propensity scores, in particular, the inverse probability weighting using propensity score and stratification based on the propensity score. The simulation results were discussed.
|
Authors who are presenting talks have a * after their name.
