Stepwise Fine and Gray: Subject-Specific Variable Selection Shows When Hemodynamic Data Improves Prognostication of Comatose Post-Cardiac Arrest Patients

2025-08-08

AI Summary

Integrating time-invariant baseline features with dynamic hemodynamic data remains challenging for prognostic prediction in comatose post-cardiac arrest patients. Method: We propose a stage-wise dynamic competing-risks model that extends the Fine-Gray framework to identify—individually—*when* and *for which patients* time-varying hemodynamic data significantly improve prediction performance, while leveraging neural networks to capture nonlinear temporal dependencies. Contribution/Results: Evaluated on a retrospective cohort of 2,278 patients, the model demonstrates robust discrimination (AUC significantly superior to conventional models) across competing outcomes—including awakening, withdrawal of life-sustaining therapy, and death under continued treatment. It enables dynamic, patient-specific selection of clinical features and multi-stage integration of heterogeneous data, establishing a novel paradigm for individualized prognostic assessment in post-cardiac arrest care.

Technology Category

• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: Adaptive Behavior
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Prognostication for comatose post-cardiac arrest patients is a critical challenge that directly impacts clinical decision-making in the ICU. Clinical information that informs prognostication is collected serially over time. Shortly after cardiac arrest, various time-invariant baseline features are collected (e.g., demographics, cardiac arrest characteristics). After ICU admission, additional features are gathered, including time-varying hemodynamic data (e.g., blood pressure, doses of vasopressor medications). We view these as two phases in which we collect new features. In this study, we propose a novel stepwise dynamic competing risks model that improves the prediction of neurological outcomes by automatically determining when to take advantage of time-invariant features (first phase) and time-varying features (second phase). Notably, our model finds patients for whom this second phase (time-varying hemodynamic) information is beneficial for prognostication and also when this information is beneficial (as we collect more hemodynamic data for a patient over time, how important these data are for prognostication varies). Our approach extends the standard Fine and Gray model to explicitly model the two phases and to incorporate neural networks to flexibly capture complex nonlinear feature relationships. Evaluated on a retrospective cohort of 2,278 comatose post-arrest patients, our model demonstrates robust discriminative performance for the competing outcomes of awakening, withdrawal of life-sustaining therapy, and death despite maximal support. Our approach generalizes to more than two phases in which new features are collected and could be used in other dynamic prediction tasks, where it may be helpful to know when and for whom newly collected features significantly improve prediction.

Problem

• Improving prognostication for comatose post-cardiac arrest patients
• Determining when time-varying hemodynamic data enhances prediction
• Modeling dynamic feature phases for better outcome prediction

Innovation

• Stepwise dynamic competing risks model
• Neural networks for nonlinear relationships
• Two-phase feature selection timing