LAUSR

BACKGROUND The Cribari matrix method (CMM) is the standard to identify potential overtriage and undertriage but requires case reviews to correct for the fact that Injury Severity Score does not account for physiology or comorbidities, nor is it well correlated with resource consumption. Further, the secondary reviews introduce undesirable subjectivity. This study assessed if the Standardized Triage Assessment Tool (STAT)—a combination of the CMM and the Need For Trauma Intervention—could more accurately determine overtriage and undertriage than the CMM alone. METHODS The registry of an American College of Surgeons verified Level I adult trauma center in Texas was queried for all new emergency department traumas 2013 to 2016 (n = 11,110). Binary logistic regressions were used to test the associations between the triage determinations of each metric against indicators of injury severity (risk factors, complications, and mortality) and resource consumption (number of procedures in 3 days and total length of stay). RESULTS Both metrics were associated with the indicators of injury severity and resource consumption in the expected directions, but STAT had stronger or equivalent associations with all variables tested. Using the CMM, there was 50.4% overtriage and 9.1% undertriage. Using STAT, overtriage was reduced to 30.8% (relative reduction = 38.9%) and undertriage was reduced to 3.3% (relative reduction = 63.7%). CONCLUSION Using the CMM with secondary case reviews makes valid multi-institutional triage rate comparisons impossible because of the subjective and unstandardized nature of these reviews. STAT’s out-of-box triage determinations (i.e., without manual case review) outperformed CMM in almost every tested variable for both over- and undertriage. STAT, an automatic, standardized method offers significant improvements compared to the current subjective system. Further, by accounting for both anatomic injury severity and resource consumption, STAT may allow trauma centers to better allocate resources and predict patient needs with fewer cases requiring manual review. LEVEL OF EVIDENCE Prognostic, level III.