Temporal Causal Mediation through a Point Process: Direct and Indirect Effects of Healthcare Interventions

TL;DR

This paper introduces a non-parametric causal mediation analysis method for complex, irregular healthcare time-series data, modeling mediators as temporal point processes to distinguish direct and indirect effects of interventions.

Contribution

It presents a novel non-parametric model that captures dynamic mediator-outcome interactions in irregular time-series, enabling accurate estimation of causal effects in healthcare.

Findings

Accurately estimates direct and indirect effects on semi-synthetic data.

Infers meaningful effect trajectories for blood glucose after surgery.

Addresses limitations of existing methods in irregular, long-range interactions.

Abstract

Deciding on an appropriate intervention requires a causal model of a treatment, the outcome, and potential mediators. Causal mediation analysis lets us distinguish between direct and indirect effects of the intervention, but has mostly been studied in a static setting. In healthcare, data come in the form of complex, irregularly sampled time-series, with dynamic interdependencies between a treatment, outcomes, and mediators across time. Existing approaches to dynamic causal mediation analysis are limited to regular measurement intervals, simple parametric models, and disregard long-range mediator--outcome interactions. To address these limitations, we propose a non-parametric mediator--outcome model where the mediator is assumed to be a temporal point process that interacts with the outcome process. With this model, we estimate the direct and indirect effects of an external intervention on the outcome, showing how each of these affects the whole future trajectory. We demonstrate on semi-synthetic data that our method can accurately estimate direct and indirect effects. On real-world healthcare data, our model infers clinically meaningful direct and indirect effect trajectories for blood glucose after a surgery.