Prognostic Accuracy of Scoring Indices Assessment for Massive Transfusion in Trauma (SIAM-T): A Five-Year Retrospective Cohort Study
TL;DR
Both NEWS and rSIG predicted MT and 24-hour mortality more accurately than ABC and RABT, and may support early pre-arrival notification and blood bank preparation without delaying rapid transport.
Abstract
Abstract Introduction: Hemorrhagic shock represents a major cause of preventable mortality following trauma. Identifying shock early on the scene is critical for survival, particularly within resource-limited settings like Thailand, where prehospital blood products are unavailable. Although several scoring tools exist, uncertainty remains regarding which tool offers the optimal diagnostic accuracy for predicting massive transfusion (MT) and early death in this specific setting. Study Objective: This research aimed to compare the predictive performance of five prehospital scoring tools – Shock Index (SI), Reverse Shock Index multiplied by Glasgow Coma Scale (rSIG), National Early Warning Score (NEWS), Assessment of Blood Consumption (ABC), and Revised Assessment of Bleeding and Transfusion (RABT) – for predicting MT and 24-hour mortality. Methods: This retrospective prognostic accuracy study included adult trauma patients transported by Emergency Medical Services (EMS) to a university-affiliated Level 1 Trauma Center in Bangkok from January 2019 through December 2024. The primary endpoint was MT; the secondary endpoint was 24-hour mortality. Area Under the Receiver Operating Characteristics (AUROCs) assessed discrimination and the DeLong test compared performance among scores. Results: A total of 584 eligible patients were included, of whom 63 (10.8%) required MT. The NEWS demonstrated the highest discriminatory power for MT (AUROC 0.84; 95% CI, 0.79–0.90), followed by rSIG (AUROC 0.83; 95% CI, 0.76–0.90) and SI (AUROC 0.78; 95% CI, 0.72–0.86). In contrast, trauma-specific scores showed lower performance, with AUROCs of 0.69 for ABC and 0.74 for RABT. For 24-hour mortality, NEWS and rSIG demonstrated excellent discrimination (AUROCs of 0.93 and 0.91, respectively). At a cut-off of four or higher, NEWS showed a sensitivity of 81.0% for MT, while rSIG ≤ 16 provided a balanced sensitivity of 74.6% and specificity of 88.5%. Conclusion: Both NEWS and rSIG predicted MT and 24-hour mortality more accurately than ABC and RABT. These tools may support early pre-arrival notification and blood bank preparation without delaying rapid transport.