Bedside Web-Based Model Predicts Mortality in Patients with Traumatic Bleeding

Rapid identification of patients at risk of dying from traumatic bleeding is essential to guide initial management and avoid preventable death. To develop a prognostic model for such patients, investigators met with prehospital, battlefield, and emergency department personnel to identify variables they considered important and convenient for assessing risk from traumatic bleeding. To identify which factors predicted in-hospital death within 4 weeks of injury, the researchers performed multivariable logistic regression analyses using data from the CRASH-2 trial cohort of 20,127 trauma patients (age >15) with estimated blood loss of at least 20%. The authors externally validated the model in 14,220 patients (primarily from the U.K.) from the …