External Validation of the Traumatic Bleeding Severity Score for the Predictive Accuracy of the Need for Massive Transfusion.

Background: The Traumatic Bleeding Severity Score (TBSS), developed in Japan, has shown high predictive accuracy in internal validation; however, external validation has still not been performed. We conducted the first external validation of the TBSS in Japan and compared it with other massive transfusion (MT) prediction scores. Methods: This multicenter retrospective study included severe trauma patients (aged ≥ 18 years, Injury Severity Score [ISS] ≥ 16) from 25 tertiary critical care centers (April 2018-March 2019), excluding those with isolated head injury or cardiac arrest on arrival. TBSS's accuracy for predicting MT was analyzed using the area under the curve (AUC). A complete case was the primary approach, with sensitivity analysis using multiple imputed datasets. The Trauma-Associated Severe Hemorrhage (TASH) score and Assessment of Blood Consumption (ABC) score were also evaluated for comparison. Results: Of 1193 patients, 873 were eligible, and 365 patients were included in the complete case analysis. The median age was 63 (interquartile range [IQR]: 45-75) years, the median ISS was 25 (IQR: 20-34), 69.3% were male individuals, 98.1% experienced blunt trauma, and 24.7% received MT. TBSS' AUC was 0.86 (sensitivity: 70.1%; specificity: 88.4%). The TASH score showed a comparable AUC of 0.88 (p = 0.850), whereas the ABC score had a significantly lower AUC of 0.80 (p = 0.024). Sensitivity analysis confirmed these findings. Conclusion: TBSS' accuracy for predicting MT was comparable to that of the TASH score and superior to that of the ABC score. Further research is needed to determine its clinical utility and broader applicability in trauma care.