|
Activity Number:
|
656
- Using Unique Associations to Address Health Policy Questions
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 1, 2019 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Health Policy Statistics Section
|
|
Abstract #304307
|
Presentation
|
|
Title:
|
A Unified Counterfactual Framework for Estimating Health Disparity
|
|
Author(s):
|
Chen-Pin Wang*
|
|
Companies:
|
University of Texas Health Science Center San Antonio
|
|
Keywords:
|
causal inference;
disparity;
balancing scores;
rank-and-replace;
Peters-Belson
|
|
Abstract:
|
This paper considered an integration of three Institute of Medicine concordant health disparity measures derived from the counterfactual framework: balancing scores weighting, rank-and-replace, and Peters-Belson methods. We began by comparing the three disparity estimates regarding (1) their underlying assumptions about the relationship between actionable and nonactionable factors; (2) their corresponding estimates when the predictivity and/or the distribution of actionable factors differ(s) between groups conditioned on both the default and constrained relationship between actionable and nonactionable factors; and (3) their implications for identifying intervention strategies to reduce health disparity. We then identified scenarios where a unification of these health disparity estimates can be formulated. We demonstrated the proposed unification method with simulated data and an application example that assessed racial/ethnic disparity in chronic liver diseases attributed to behavioral/clinical factors in a cohort study funded by the Cancer Prevention and Research Institute of Texas.
|
Authors who are presenting talks have a * after their name.
