Hospital resources do not predict accuracy of secondary trauma triage: A population-based analysis

BACKGROUND: The identification of patients who require transfer from non-trauma centers to trauma centers (secondary triage) is complicated by high rates of undertriage and overtriage. The objective of this study was to evaluate variations in secondary triage accuracy across non-trauma centers and identify factors associated with highly accurate secondary triage. METHODS: We performed a population-based study of injured patients who presented to non-trauma centers in a large regional trauma system. Patients were categorized as undertriaged, overtriaged, or appropriately triaged based on transfer status and presence of a severe injury (Injury Severity Score >15, death within 24 hours, or critical injury as defined by the American College of Surgeons). Mixed-effect models, adjusted for case mix and hospital resource, were used to compare triage accuracy across hospitals and identify factors associated with high-performing centers. RESULTS: Among 118,973 patients identified at 182 non-trauma centers, 37,528 (31.5%) had severe injuries. The majority (76.9%) of severely injured patients were not transferred to a trauma center (undertriaged), while 9.6% of nonseverely injured patients were transferred to a trauma center (overtriaged). Mixed-effect models demonstrated that at the average hospital severely injured patients were 3.76 times more likely to be transferred than nonseverely injured patients (diagnostic odds ratio, 3.76; 95% confidence interval, 3.20-4.31). Despite significant variation in triage accuracy across hospitals, adjusted analyses suggested that local resources bore no relationship to triage accuracy. CONCLUSION: Triage accuracy varies significantly across non-trauma centers, after adjusting for hospital resources. These findings suggest that other potentially modifiable factors play a key role in transfer decisions. LEVEL OF EVIDENCE: Therapeutic/care management, level IV.