Causal Fairness Assessment of Treatment Allocation with Electronic Health Records

TL;DR

This work tackles fairness in clinical treatment allocation for coronary artery disease using causal fairness concepts applied to electronic health records. It defines a principal-fairness criterion based on principal strata $R_i=(Y_i(0),Y_i(1))$ and estimates $p(D|R,A)$, with stratum-specific violations measured by $\Delta(r)$. Through simulations and an EHR-based CAD cohort, it demonstrates gender and race disparities in CABG uptake within certain strata and shows that social determinants of health have limited impact when rich EHR covariates are accounted for. The approach enables retrospective, causally-informed assessment of treatment decisions in clinical practice and highlights the persistent equity challenges in cardiovascular care.

Abstract

Healthcare continues to grapple with the persistent issue of treatment disparities, sparking concerns regarding the equitable allocation of treatments in clinical practice. While various fairness metrics have emerged to assess fairness in decision-making processes, a growing focus has been on causality-based fairness concepts due to their capacity to mitigate confounding effects and reason about bias. However, the application of causal fairness notions in evaluating the fairness of clinical decision-making with electronic health record (EHR) data remains an understudied domain. This study aims to address the methodological gap in assessing causal fairness of treatment allocation with electronic health records data. We propose a causal fairness algorithm to assess fairness in clinical decision-making. Our algorithm accounts for the heterogeneity of patient populations and identifies potential unfairness in treatment allocation by conditioning on patients who have the same likelihood to benefit from the treatment. We apply this framework to a patient cohort with coronary artery disease derived from an EHR database to evaluate the fairness of treatment decisions. In addition, we investigate the impact of social determinants of health on the assessment of causal fairness of treatment allocation.

Figures (2)

• Figure 1: Principal fairness of decision in simulation. (a): Principal fairness decision probabilities and level of violation metric $\Delta(r)$. (b): Proportion of principal strata. The proposed algorithm can detect unfair decisions and accurately estimate levels of violation.
• Figure 2: Principal fairness of the allocation of CABG vs PCI. (a) Principal fairness with respect to gender. Negative $\Delta$ indicates that female were treated less frequently with CABG. (b) Principal fairness with respect to race. Negative $\Delta$ indicates that Black or African American patients were treated less frequently with CABG.

Theorems & Definitions (1)

• Definition 1: Principal fairness