Heterogeneous Causal Discovery of Repeated Undesirable Health Outcomes

TL;DR

The paper addresses discovering causes and effect modifiers of repeated undesirable health outcomes from observational data. It introduces an end-to-end heterogeneous causal discovery framework that ensembles multiple CSL algorithms and jointly estimates heterogeneous effects, enabling identification of subpopulation-specific interventions. Through real-world (repeat ER visits for diabetics) and synthetic/semi-synthetic experiments, the approach shows robustness, improved recall for true causal factors, and actionable heterogeneity insights aligned with existing literature. This framework has practical implications for targeted health policies and resource allocation by surfacing high-impact subpopulations and context-specific interventions.

Abstract

Understanding factors triggering or preventing undesirable health outcomes across patient subpopulations is essential for designing targeted interventions. While randomized controlled trials and expert-led patient interviews are standard methods for identifying these factors, they can be time-consuming and infeasible. Causal discovery offers an alternative to conventional approaches by generating cause-and-effect hypotheses from observational data. However, it often relies on strong or untestable assumptions, which can limit its practical application. This work aims to make causal discovery more practical by considering multiple assumptions and identifying heterogeneous effects. We formulate the problem of discovering causes and effect modifiers of an outcome, where effect modifiers are contexts (e.g., age groups) with heterogeneous causal effects. Then, we present a novel, end-to-end framework that incorporates an ensemble of causal discovery algorithms and estimation of heterogeneous effects to discover causes and effect modifiers that trigger or inhibit the outcome. We demonstrate that the ensemble approach improves robustness by enhancing recall of causal factors while maintaining precision. Our study examines the causes of repeat emergency room visits for diabetic patients and hospital readmissions for ICU patients. Our framework generates causal hypotheses consistent with existing literature and can help practitioners identify potential interventions and patient subpopulations to focus on.

Figures (9)

• Figure 1: Illustration of data model with timeline $E \in \{0,1\}^{4 \times T}$ for $N$ units with three endogenous events ($e^1, e^2,$ and $e^3$) and an undesirable outcome ($o$).
• Figure 2: Characterization of effect modifiers (EM) with a structural causal model (SCM).
• Figure 3: Heterogeneous causal discovery framework.
• Figure 4: Multi-set average total causal effect of top diagnosis codes on the repeat ER visits.
• Figure 5: Effect modifiers (EM) of two preventive diagnosis codes on repeat ER visits.