|
Activity Number:
|
71
- Statistical Methods for Personalized Medicine
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 29, 2018 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Mental Health Statistics Section
|
|
Abstract #328404
|
Presentation
|
|
Title:
|
Q-Learning for Dynamic Treatment Regimes on CODIACS Vanguard Randomized Controlled Trial
|
|
Author(s):
|
Eun Jeong Oh* and Min Qian and Ying Kuen Ken Cheung
|
|
Companies:
|
Columbia and Columbia University and Columbia University
|
|
Keywords:
|
Personalized medicine;
Q-learning;
Dynamic treatment regimes;
Adaptive Interventions;
Two-stage procedure
|
|
Abstract:
|
There has been increasing development in personalized interventions that are adaptive to the uniquely evolving health status of each patient over time. Deriving dynamic treatment regimes (DTRs) from a single training set of finite horizon trajectories is one of the key goals in medical applications to accommodate heterogeneity among patients' responses to drugs. In this paper, we introduce two-stage Q-learning procedures to estimate optimal DTRs for post-acute coronary syndrome (ACS) patients with depressive symptoms. While controlling for the effects of both past and subsequent adaptive interventions, we utilize a series of linear regressions to provide an individually tailored sequence of decision rules that would maximize the expected outcome if implemented. The performance of the method is demonstrated through the example of comparison of depression interventions after acute coronary syndrome (CODIACS) vanguard trial.
|
Authors who are presenting talks have a * after their name.
