Multicenter validation of a simplified score to predict massive transfusion in trauma

Several studies have described predictive models to identify trauma patients who require massive transfusion (MT). Early identification of lethal exsanguination may improve survival in this patient population. The purpose of the current study was to validate a simplified score to predict MT at multiple Level I trauma centers.

All adult trauma patients treated at three Level I trauma centers from July 2006 to June 2007 who (1) were transported directly from the scene, (2) were trauma activations, and (3) received any blood transfusions during admission were included. Assessment of Blood Consumption (ABC) score developed using the same inclusion criteria for patients admitted to a single trauma center (Vanderbilt University Medical Center [VUMC]-1) between July 2005 and June 2006. ABC score calculated by assigning a value (0 or 1) to each of the four parameters: penetrating mechanism, positive focused assessment with sonography for trauma for fluid, arrival blood pressure <90 mm Hg, and arrival pulse >120 bpm. A score of 2 was used as "positive" to predict MT. Area under receiver-operating characteristic curve was calculated to compare the predictive ability of the score at each institution.

There were 586 patients in the developmental (VUMC-1), 513 patients at trauma center 1 (VUMC-2), 372 at trauma center 2 (PMH), and 133 at trauma center 3 (Johns Hopkins Hospital). MT rate was similar between centers: 14% to 15%. Sensitivity and specificity for the ABC score predicting MT ranged from 75% to 90% and 67% to 88%, respectively. Correctly classified patients and area under receiver-operating characteristic curve, however, were 84% to 87% and 0.83 to 0.90, respectively.

The ABC score is a valid instrument to predict MT early in the patient's care and across various demographically diverse trauma centers. Future research should focus on this score's ability to prospectively identify patients who will receive MT.