Abstract
OBJECTIVES: During mass-casualty incidents (MCIs), prehospital triage is performed to identify which patients most urgently need medical care. Formal MCI triage tools exist, but their performance is variable. The Shock Index (SI; heart rate [HR] divided by systolic blood pressure [SBP]) has previously been shown to be an efficient screening tool for identifying critically ill patients in a variety of in-hospital contexts. The primary objective of this study was to assess the ability of the SI to identify trauma patients requiring urgent life-saving interventions in the prehospital setting. METHODS: Clinical data captured in the Alberta Trauma Registry (ATR) were used to determine the SI and the "true" triage category of each patient using previously published reference standard definitions. The ATR is a provincial trauma registry that captures clinical records of eligible patients in Alberta, Canada. The primary outcome was the sensitivity of SI to identify patients classified as "Priority 1 (Immediate)," meaning they received urgent life-saving interventions as defined by published consensus-based criteria. Specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated as secondary outcomes. These outcomes were compared to the performance of existing formal MCI triage tools referencing performance characteristics reported in a previously published study. RESULTS: Of the 9,448 records that were extracted from the ATR, a total of 8,650 were included in the analysis. The SI threshold maximizing Youden's index was 0.72. At this threshold, SI had a sensitivity of 0.53 for identifying "Priority 1" patients. At a threshold of 1.00, SI had a sensitivity of 0.19. CONCLUSIONS: The SI has a relatively low sensitivity and did not out-perform existing MCI triage tools at identifying trauma patients who met the definition of "Priority 1" patients.