Evaluation of pre-hospital trauma triage criteria: a prospective study at a Danish level I trauma centre

Background: The aim of the present study was to evaluate the precision of our trauma triage protocol [based on the American College of Surgeons, Committee on Trauma (ACS COT)] in identifying severely injured defined as an injury severity score (ISS) > 15. Our hypothesis was that isolated mechanism-of-injury criteria were responsible for a significant over-triage leading to over-use of our trauma team. Methods: Design: A prospective cohort study. Setting: A level I trauma centre, Aarhus, Denmark. Patients and participants: Among all injured patients admitted during a 6-month period in 2003 we identified severely injured. During the study period, trauma team activations were consecutively registered and triage criteria were prospectively collected. Sensitivity, specificity, positive predictive value, over-triage and under-triage were calculated. Results: Out of 15,162 patients in the emergency department, 848 injured patients were included and 59 (7%) were severely injured. We had 242 trauma team activations with 54 (22%) severely injured. Sensitivity was 92%, specificity 76%, giving an over-triage of 24% and an under-triage of 8%. The positive predictive value was 22%. Among 60 patients with mechanism-of-injury as the only criterion, five were severely injured in contrast to 12 out of 20 patients with mechanism-of-injury combined with physiological and/or anatomical criteria. Conclusion: The positive predictive value of our triage protocol was low, only 22%. This was mainly as a result of a significant over-triage from isolated mechanism-of-injury criteria. We recommend revision of the triage protocol and reallocation of our trauma team resources.