Real-Time Risk Prediction on the Wards: A Feasibility Study.

Area of Science:

• Clinical Informatics
• Patient Safety
• Health Systems Engineering

Background:

• Clinical deterioration in hospitals is often missed, leading to adverse patient outcomes and increased healthcare costs.
• Early detection of patient decline is crucial for timely intervention and improved care quality.

Purpose of the Study:

• To assess the feasibility and accuracy of real-time risk stratification using the electronic Cardiac Arrest Risk Triage (eCART) score.
• To compare the eCART score's performance against the standard Rapid Response Team (RRT) activation for predicting critical events.

Main Methods:

• A prospective, black-box validation study was conducted on three general care wards.
• Patient data were transmitted in real-time via HL7 feed for eCART score calculation.
• The eCART score's sensitivity and timing were compared with RRT activations for cardiac arrests and ICU transfers.

Main Results:

• The eCART score demonstrated high accuracy with an area under the receiver operating characteristic curve of 0.88 for cardiac arrest and 0.80 for ICU transfer.
• eCART identified 80% of cardiac arrests and 52% of ICU transfers, significantly outperforming the RRT (20% and 34%, respectively).
• High-risk eCART alerts occurred a median of 30 hours before events, compared to 1.7 hours for RRT activation.

Conclusions:

• The eCART score is a feasible and accurate tool for real-time risk stratification in hospitalized patients.
• eCART significantly improves the detection of impending cardiac arrests and ICU transfers compared to the standard RRT.
• Early identification via eCART allows for substantially earlier intervention, potentially improving patient outcomes.