Calculating trauma undertriage rates: Are we using the wrong formula—and the wrong definition?

BACKGROUND: Undertriage (UT) is central to trauma center quality improvement, yet UT definitions vary, and no data-driven consensus target exists for interfacility benchmarking. We compared facility-level UT rates across four methods and assessed the impact on facility benchmarking. METHODS: This retrospective study included patients aged 18 to 89 from Level I & II trauma centers in the ACS TQP Files (2021-2023). Facility-level UT was calculated using the Cribari Matrix Method (CMM), modified Cribari Matrix Method (mCMM), Standardized Triage Assessment Tool (STAT), and High-Intensity Time-Sensitive (HITS) intervention-based definition representing trauma team actions. Paired facility-level UT rates were compared across methods. To estimate benchmarking impact, we calculated within-facility rank change (higher rank = higher UT) relative to CMM. RESULTS: 2,643,881 patients and 515 facilities were included. Median facility-level UT significantly differed by method: CMM 11% [interquartile range: 9-14], mCMM 60% [50-67], STAT 6% [5-8], and HITS 30% [22-40] (all P<0.001). At individual hospitals, the median range of maximum UT values between methods was 53% [44-60]. The median absolute facility rank change between the CMM method and other tools was 129 [54.5-236.5] for mCMM, 48 [17-90] for STAT, and 142 [62.5-252.5] for HITS. CONCLUSIONS: These UT data highlight a concerning level of variation and uncertainty in the calculation and interpretation of a key trauma center performance metric. Until a UT definition is standardized and data-driven UT target goals established, cross-center benchmarking using UT should be interpreted cautiously and may be better suited for within-center trending. LEVEL OF EVIDENCE: Level III, therapeutic/care management.