|
Activity Number:
|
335
- ASA Biometrics Section JSM Travel Awards (II)
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2018 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biometrics Section
|
|
Abstract #328305
|
|
|
Title:
|
Estimation and Optimization of Composite Outcomes
|
|
Author(s):
|
Daniel J Luckett* and Eric Laber and Michael Kosorok
|
|
Companies:
|
University of North Carolina at Chapel Hill and North Carlina State University and University of North Carolina at Chapel Hill
|
|
Keywords:
|
Individualized treatment rules;
Inverse reinforcement learning;
Precision medicine;
Utility functions
|
|
Abstract:
|
There is tremendous interest in precision medicine as a means to improve patient outcomes by tailoring treatment to individual characteristics. An individualized treatment rule formalizes precision medicine as a map from patient information to a recommended treatment. A rule is defined to be optimal if it maximizes the mean of a scalar outcome in a population. However, clinicians often must balance multiple and possibly competing outcomes. One approach to precision medicine in this setting is to elicit a composite outcome which balances all competing outcomes; unfortunately, eliciting a composite outcome from patients is difficult without a high-quality instrument. We consider estimation of composite outcomes using observational data under the assumption that clinicians are approximately making decisions to maximize individual patient utility. Estimated composite outcomes are used to construct an estimator of a treatment rule that maximizes the mean of patient-specific composite outcomes. We prove that the proposed estimators are consistent and demonstrate their finite sample performance through simulation experiments and an application to a study of bipolar depression.
|
Authors who are presenting talks have a * after their name.
