A Causal Framework for Evaluating ICU Discharge Strategies

2026-03-26

AI Summary

This study addresses the challenge of ICU discharge decision-making, which requires balancing intervention duration against patient outcomes and contends with missing post-intervention variables after treatment cessation. The authors propose a causal inference framework based on an extended g-formula to evaluate multi-objective policies incorporating stopping rules. Under standard positivity and coverage conditions, the method effectively handles the non-observability of post-intervention data. Applying this approach to the MIMIC-IV dataset with an open-source Python toolkit, the study identifies discharge strategies that outperform current clinical practice. The work establishes a verifiable and reproducible causal evaluation paradigm for complex medical decision-making under dynamic treatment regimes.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

In this applied paper, we address the difficult open problem of when to discharge patients from the Intensive Care Unit. This can be conceived as an optimal stopping scenario with three added challenges: 1) the evaluation of a stopping strategy from observational data is itself a complex causal inference problem, 2) the composite objective is to minimize the length of intervention and maximize the outcome, but the two cannot be collapsed to a single dimension, and 3) the recording of variables stops when the intervention is discontinued. Our contributions are two-fold. First, we generalize the implementation of the g-formula Python package, providing a framework to evaluate stopping strategies for problems with the aforementioned structure, including positivity and coverage checks. Second, with a fully open-source pipeline, we apply this approach to MIMIC-IV, a public ICU dataset, demonstrating the potential for strategies that improve upon current care.

Problem

• ICU discharge
• optimal stopping
• causal inference
• observational data
• treatment duration

Innovation

• causal inference
• optimal stopping
• g-formula
• ICU discharge
• observational data