Early ICU Mortality Prediction and Survival Analysis for Respiratory Failure

TL;DR

This paper presents a dynamic model for early mortality prediction in respiratory failure ICU patients, utilizing initial 24-hour physiological data to improve clinical decision-making and resource allocation.

Contribution

The study introduces a novel dynamic modeling approach validated on the eICU database, achieving higher AUROC and AUCPR than existing models for early ICU mortality prediction.

Findings

Achieved AUROC of 80-83% for mortality prediction.

Improved AUCPR by 4% on Day 5 post-ICU admission.

Demonstrated survival curve incorporating time-varying information.

Abstract

Respiratory failure is the one of major causes of death in critical care unit. During the outbreak of COVID-19, critical care units experienced an extreme shortage of mechanical ventilation because of respiratory failure related syndromes. To help this, the early mortality risk prediction in patients who suffer respiratory failure can provide timely support for clinical treatment and resource management. In the study, we propose a dynamic modeling approach for early mortality risk prediction of the respiratory failure patients based on the first 24 hours ICU physiological data. Our proposed model is validated on the eICU collaborate database. We achieved a high AUROC performance (80-83%) and significantly improved AUCPR 4% on Day 5 since ICU admission, compared to the state-of-art prediction models. In addition, we illustrated that the survival curve includes the time-varying information for the early ICU admission survival analysis.