SurgCalc Home
Add to your home screen for offline access
SurgCalc
Trauma

Trauma Mortality Predictor

Enter injuries by body region — AIS scores and ISS are calculated automatically. Add ED physiology to generate an individualized mortality risk estimate.

Train N3,817,361
Test AUC0.909
TRISS AUC0.879
cNRI vs TRISS+0.517
Model lockApr 2026
Clinical Decision Support Tool — Not a Substitute for Physician Judgment For educational and research use. Not FDA cleared. No data leaves your browser. Full disclosures, limitations & terms of use
🦴
Step 1 — Injuries by Body Region
Select the most severe injury in each region. AIS scores and ISS update automatically.
🧠 Head / Brain
AIS 0
😶 Face
AIS 0
🫁 Chest / Thorax
AIS 0
Abdomen / Pelvis
AIS 0
🦵 Extremity / Pelvis
AIS 0
🩹 External / Burns
AIS 0
ISS
0
Top 3 regions²
Head
0
Face
0
Chest
0
Abdomen
0
Extremity
0
External
0
ISS (auto)
0
Sum of top 3 AIS²
AIS Head (auto)
0
Model predictor
AIS Chest (auto)
0
Model predictor
AIS Abdomen (auto)
0
Model predictor
Mechanism
From Step 2
🩺
Step 2 — Demographics, Mechanism & ED Physiology
All values from ED presentation at time of trauma activation.
Glasgow Coma Scale
GCS Total (E+V+M) 15
Severe TBI: ≤8
Moderate: 9–12
Mild: 13–15
📊
Step 3 — Predicted Mortality Risk
TQIP logistic model — temporally validated prospective test set 2023–2024
In-Hospital Mortality
0%10%25%50%75+%
Predicted probability
Linear predictor (b)
Risk tier
ISS used
GCS used
Model AUC (test set 2023–2024)0.909
Brier score (test set)0.0189
Test set N1,581,953 patients

Model Comparison — This Calculator vs. TRISS

This model — predicted risk
TRISS test-set AUC0.879 (historical benchmark)
Improvement (ΔAUC)+0.030 absolute
Net reclassification (cNRI)+0.517 (51.7% correctly reclassified)