The Target Study: A Conceptual Model and Framework for Measuring Disparity

Area of Science:

• Epidemiology
• Biostatistics
• Health Equity Research

Background:

• Measuring social disparity is complex due to confounding and selection biases.
• Existing decomposition-based models may not adequately address these challenges in healthcare settings.

Purpose of the Study:

• To present a conceptual sampling-based model for measuring disparity.
• To extend the model for generalizing disparity under non-random selection and after interventions.
• To provide a framework for implementing the model using various estimators.

Main Methods:

• Developed a sampling design that does not alter group membership or covariates.
• Incorporated methods to address non-random sample selection and differential enrollment timing.
• Discussed study designs, data structures, and G-computation and weighting estimators.

Main Results:

• The proposed model offers a framework to measure disparity by balancing covariates and enrollment time.
• Methods are presented to transport or generalize disparity estimates.
• The model was applied to measure health system disparities in hypertension control.

Conclusions:

• The sampling-based model provides a robust approach to measuring social disparity in health.
• It offers solutions for bias mitigation due to sample selection and enrollment timing.
• The methods are applicable to real-world health equity research using electronic medical records.