Identification of patients with stable coronary artery disease who benefit from ACE inhibitors using Cox mixture model for heterogeneous treatment effects
Event Type
Research Presentation
Academic Department
Mathematics and Statistics
Location
Dana Science Building, 2nd floor
Start Date
14-4-2023 1:30 PM
End Date
14-4-2023 3:00 PM
Description
Under the direction of Chirag Nagpal (Ph.D. candidate) and Artur Dubrawski (Ph.D. and faculty), Auton Lab, School of Computer Science, Carnegie Mellon University
The presence of heterogeneity in response to treatment effect is often excluded in the evaluation process of Randomized Clinical Trials. In this talk, we utilize the existing Deep Cox Mixtures with Heterogeneous Effects (CMHE) model to study the heterogeneous treatment effects in The Prevention of Events with Angiotensin Converting Enzyme Inhibition (PEACE) study. The PEACE primary publication found no benefit of using ACE-Inhibitors in patients with stable coronary disease and reduced left-ventricular function. With the CMHE to subgroup the PEACE trial population based on response to ACE-inhibitors, we identified a patient subgroup whose outcomes could improve with treatment, even though population level on-average analysis shows no desirable effects.
Identification of patients with stable coronary artery disease who benefit from ACE inhibitors using Cox mixture model for heterogeneous treatment effects
Dana Science Building, 2nd floor
Under the direction of Chirag Nagpal (Ph.D. candidate) and Artur Dubrawski (Ph.D. and faculty), Auton Lab, School of Computer Science, Carnegie Mellon University
The presence of heterogeneity in response to treatment effect is often excluded in the evaluation process of Randomized Clinical Trials. In this talk, we utilize the existing Deep Cox Mixtures with Heterogeneous Effects (CMHE) model to study the heterogeneous treatment effects in The Prevention of Events with Angiotensin Converting Enzyme Inhibition (PEACE) study. The PEACE primary publication found no benefit of using ACE-Inhibitors in patients with stable coronary disease and reduced left-ventricular function. With the CMHE to subgroup the PEACE trial population based on response to ACE-inhibitors, we identified a patient subgroup whose outcomes could improve with treatment, even though population level on-average analysis shows no desirable effects.