Clearly defining pediatric massive transfusion: cutting through the fog and friction with combat data.

BACKGROUND: Massive transfusion (MT) in pediatric patients remains poorly defined. Using the largest existing registry of transfused pediatric trauma patients, we sought a data-driven MT threshold.

METHODS: The Department of Defense Trauma Registry was queried from 2001 to 2013 for pediatric trauma patients (<18 years). Burns, drowning, isolated head injury, and missing Injury Severity Score (ISS) were excluded. MT was evaluated as a weight-based volume of all blood products transfused in the first 24 hours. Mortality at 24 hours and in the hospital was calculated for increasing transfusion volumes. Sensitivity and specificity curves for predicting mortality were used to identify an optimal MT threshold. Patients above and below this threshold (MT+ and MT-, respectively) were compared.

RESULTS: The Department of Defense Trauma Registry yielded 4,990 combat-injured pediatric trauma patients, of whom 1,113 were transfused and constituted the study cohort. Sensitivity and specificity for 24-hour and in-hospital mortality were optimal at 40.1-mL/kg and 38.6-mL/kg total blood products in the first 24 hours, respectively. With the use of a pragmatic threshold of 40 mL/kg, patients were divided into MT+ (n = 443) and MT- (n = 670). MT+ patients were more often in shock (68.1% vs. 47.0%, p < 0.001), hypothermic (13.0% vs. 3.4%, p < 0.001), coagulopathic (45.0% vs. 29.6%, p < 0.001), and thrombocytopenic (10.6% vs. 5.0%, p = 0.002) on presentation. MT+ patients had a higher ISS, more mechanical ventilator days, and longer intensive care unit and hospital stay. MT+ was independently associated with an increased 24-hour mortality (odds ratio, 2.50; 95% confidence interval, 1.28-4.88; p = 0.007) and in-hospital mortality (odds ratio, 2.58; 95% confidence interval, 1.70-3.92; p < 0.001).

CONCLUSION: Based on this large cohort of transfused combat-injured pediatric patients, a threshold of 40 mL/kg of all blood products given at any time in the first 24 hours reliably identifies critically injured children at high risk for early and in-hospital death. This evidence-based definition will provide a consistent framework for future research and protocol development in pediatric resuscitation.

LEVEL OF EVIDENCE: Diagnostic study, level II. Prognostic/epidemiologic study, level III.