Prospective validation of a hospital triage predictive model to decrease undertriage: an EAST multicenter study.

Area of Science:

• Emergency Medicine
• Trauma Surgery
• Health Services Research

Background:

• Tiered trauma team activation (TTA) aims to optimize resource allocation for injured patients.
• Achieving target undertriage (<5%) and overtriage (<35%) rates is challenging, with significant performance variability across centers.
• Existing trauma triage models require optimization and external validation to improve accuracy and consistency.

Purpose of the Study:

• To optimize and externally validate a previously developed hospital trauma triage prediction model.
• To predict the need for emergent intervention in 6 hours (NEI-6), a key indicator for full TTA.
• To reduce variability in TTA and improve undertriage/overtriate rates.

Main Methods:

• A weighted multiple logistic regression model was used to retrain and optimize the NEI-6 model.
• Prospective data from five trauma centers were used, with a portion for retraining and the remainder for external validation.
• Model performance was assessed using Area Under the Receiver Operating Characteristic Curve (AUROC) and Area Under the Precision-Recall Curve (AUPRC).

Main Results:

• The external validation cohort included 2476 patients.
• The optimized model achieved an AUROC of 0.80 and an AUPRC of 0.63.
• Undertriage rates were 9.1% overall, 8.8% for blunt trauma, 8.4% for patients ≥65, and 7.7% for Black or African American patients.

Conclusions:

• The externally validated NEI-6 model approaches recommended undertriage and overtriage rates.
• The model significantly reduces TTA variability, particularly for blunt trauma patients.
• The model demonstrates effective performance in populations historically experiencing high undertriage rates.