A proxy-based approach for unmeasured confounding in electronic health records research

TL;DR

This paper presents a novel proxy-based method to adjust for unmeasured confounding in EHR research, using proxies like vitals to improve causal effect estimation in observational health data.

Contribution

It introduces a new approach combining proxies with factor analysis and regression, demonstrating its effectiveness and robustness compared to existing methods in realistic settings.

Findings

The method effectively adjusts for unmeasured confounding in EHR data.

It remains robust under model misspecification and incorrect variable classification.

Application to real EHR data illustrates practical utility in clinical decision studies.

Abstract

Electronic health records (EHR) are widely used to study clinical decisions, yet unmeasured confounding remains a persistent challenge. Proxy variables offer a potential solution. In EHR data, clinicians already record many such measurements (e.g., vitals), each revealing something about a patient's underlying health. Despite this, proxy-based methods are rarely used in practice. We introduce a new way to use proxies to adjust for unmeasured confounding. Our approach uses a vector of proxies to construct covariates that capture aspects of the unmeasured confounder, which are then included in a regression model. As one implementation, we use factor analysis followed by regression. We compare this approach with existing methods, including proximal causal inference, across a range of realistic settings. In practice, assumptions rarely hold exactly, so we study what happens when models are misspecified and variables are used incorrectly: e.g., a confounder or instrument is treated as a proxy. Finally, we apply the method to EHR data to estimate the effect of hospital admission for older adults presenting to the emergency department with chest pain, a setting where unmeasured confounding is a substantial concern. This work provides a practical way to use proxies and may help bring proxy-based methods into broader use.