Related Experiment Videos
Predicting Prehospital Blood Transfusion After Motor Vehicle Trauma
Amanda Pope1, Jonathan Zadra1, Kent Page1
1Utah Data Coordinating Center, University of Utah, 303 Chipeta Way, Rm H-111, Mailbox #11, Salt Lake City, UT 84108.
Prehospital Emergency Care
|August 10, 2026
Summary
Predictive models using National Emergency Medical Services Information System data accurately identify motor vehicle crash patients needing prehospital blood transfusion (PHBT). This aids in estimating demand and expanding PHBT programs for better trauma care.
Area of Science:
- Emergency Medicine
- Trauma Surgery
- Public Health
Background:
- Prehospital blood transfusion (PHBT) improves outcomes in traumatic hemorrhagic shock.
- The epidemiology and geographic distribution of patients requiring PHBT are not well understood.
Purpose of the Study:
- Develop predictive models for PHBT in motor vehicle crash (MVC) patients.
- Estimate the national distribution of patients with a high probability of needing PHBT.
Main Methods:
- Retrospective cohort study using 2024-2025 National Emergency Medical Services Information System (NEMSIS) data.
- Developed predictive models using random forest and penalized logistic regression.
- Assessed model performance using AUC, sensitivity, and specificity.
Main Results:
- Models demonstrated strong discrimination (AUC > 0.90).
- Identified key predictors including composite physiologic measures and shock designation.
- Estimated 3.6% and 2.4% of evaluated MVC patients as predicted PHBT recipients.
Conclusions:
- NEMSIS data and predictive modeling accurately identify patients likely to receive PHBT.
- Findings support estimating geographic demand for prehospital blood products.
- Informs data-driven expansion of PHBT programs within trauma systems.