Double sampling for informatively missing data in electronic health record-based comparative effectiveness research

TL;DR

This paper proposes a double sampling design to address non-random missing data in electronic health records, enabling more reliable causal inference and estimation of treatment effects despite data missing not at random.

Contribution

It introduces a double sampling approach with identification assumptions and develops efficient estimators for causal effects under MNAR and MAR scenarios.

Findings

Double sampling improves causal effect estimation in MNAR settings.

Proposed estimators are robust and efficient under nonparametric models.

Simulation studies demonstrate advantages over traditional methods.

Abstract

Missing data arise in most applied settings and are ubiquitous in electronic health records (EHR). When data are missing not at random (MNAR) with respect to measured covariates, sensitivity analyses are often considered. These post-hoc solutions, however, are often unsatisfying in that they are not guaranteed to yield concrete conclusions. Motivated by an EHR-based study of long-term outcomes following bariatric surgery, we consider the use of double sampling as a means to mitigate MNAR outcome data when the statistical goals are estimation and inference regarding causal effects. We describe assumptions that are sufficient for the identification of the joint distribution of confounders, treatment, and outcome under this design. Additionally, we derive efficient and robust estimators of the average causal treatment effect under a nonparametric model and under a model assuming outcomes were, in fact, initially missing at random (MAR). We compare these in simulations to an approach that adaptively estimates based on evidence of violation of the MAR assumption. Finally, we also show that the proposed double sampling design can be extended to handle arbitrary coarsening mechanisms, and derive nonparametric efficient estimators of any smooth full data functional.