BACKGROUND Trauma center activation criteria aim to triage patients and ensure appropriate activation level and allocation of resources. We hypothesized that a triage system based on prehospital physiologic variables could predict children at risk for injuries requiring acute interventions and a mechanistic rule would capture any remaining injured patients and those requiring prolonged intensive care unit (ICU) admission or need for urgent surgery. METHODS Data sets from two multicenter prospective studies of pediatric blunt trauma were analyzed. A physiologic rule was developed using prehospital vitals to predict “serious injury,” defined as the need for acute intervention on the head, chest, or abdomen; blood transfusion; or mortality. Next, a mechanistic rule was determined using logistic regression to identify mechanisms of injury associated with serious injury, need for prolonged ICU admission, or surgical intervention. We then evaluated the diagnostic accuracy as well as under- and overtriage rates. RESULTS The combined data sets included 14,232 patients (mean age, 9.5 years; ICU admission, 12.1%; transfusion, 1%; and mortality, 1%). Interventions included 170 abdominal surgeries, 18 angio-interventions, 265 head procedures, and 227 chest procedures. A physiologic rule (full activation) was developed with 97.8% negative predictive value for serious injury; the four variables included prehospital intubation (odds ratio OR, 17.1), Glasgow Coma Scale score of 10 ft (OR, 1.7), and pedestrian struck by automobile (OR, 1.2). Combined, the rules identified 82% of children requiring intervention, transfusion, and ICU care or who died. CONCLUSION Pediatric blunt trauma triage benefits from a two-tier approach. Our physiologic rule effectively identifies children needing acute intervention, while the mechanistic rule detects others requiring trauma center resources. LEVEL OF EVIDENCE Therapeutic/Care Management or Prognostic and Epidemiological; Level III.
Frederick et al. (Fri,) studied this question.