Predicting on-going hemorrhage and transfusion requirement after severe trauma: a validation of six scoring systems and algorithms on the TraumaRegister DGU®
BACKGROUND An early predictive model for massive transfusion (MT) is critical for management of combat casualties because of limited blood product availability, component preparation, and the time necessary to mobilize fresh whole blood donors. The purpose of this study was to determine which variables, available early after injury, are associated with MT. We hypothesized that International Normalized Ratio and penetrating mechanism would be predictive. STUDY DESIGN We performed a retrospective cohort analysis in two combat support hospitals in Iraq. Patients who required MT were compared with patients who did not. Eight potentially predictive variables were subjected to univariate analysis. Variables associated with need for MT were then subjected to stepwise logistic regression. RESULTS Two hundred forty-seven patients required MT and 311 did not. Mean Injury Severity Score was 22 in the MT group and 5 in the non-MT group (p < 0.001). Patients in the MT group received 17.9 U stored RBCs and 2.0 U fresh whole blood, versus 1.1 U RBCs and 0.2 U whole blood in the non-MT group (p < 0.001). Mortality was 39% in the MT group and 1% in the non-MT group (p < 0.001). Variables that independently predicted the need for MT were: hemoglobin <or= 11 g/dL, International Normalized Ratio > 1.5, and a penetrating mechanism. The area under the receiver operator characteristic curve was 0.804 and Hosmer-Lemeshow goodness-of-fit test was 0.98. CONCLUSION MT after combat injury is associated with high mortality. Simple variables available early after admission allow accurate prediction of MT.