A score recorded when adults arrived at hospital after traumatic brain injury was associated with unfavorable function at discharge, according to the study. In the main adjusted analysis, each one-point increase in TRIAGES was linked to 2.81 times the odds of an unfavorable outcome. The 95% confidence interval ran from 2.29 to 3.45, with p < 0.001. Unfavorable function meant a discharge score of 1 to 3 on the Glasgow Outcome Scale.
The result shows a statistical link, not that TRIAGES itself caused an unfavorable outcome. The research was a retrospective, single-center observational cohort, so it cannot establish that using the score would change what happens to patients.
What the score measured
TRIAGES combines age, the first Glasgow Coma Scale score, respiratory rate and systolic blood pressure from the initial emergency assessment. Those measurements were taken before in-hospital treatment. The primary outcome was functional status at discharge, with Glasgow Outcome Scale scores of 1 to 3 classed as unfavorable.
The final analysis included 753 adults: 572 had favorable discharge outcomes and 181 had unfavorable outcomes.
A small advantage over GCS
TRIAGES showed good separation between the two outcome groups in this cohort. Its area under the receiver operating characteristic curve, or AUC, was 0.905. At the cohort-derived threshold of 3 or higher, sensitivity was 95.6% and specificity was 70.1%, meaning the threshold identified most unfavorable outcomes while also correctly classifying about seven in ten favorable outcomes. The apparent Brier score was 0.101. Because the cutoff was derived and tested in the same cohort, it was not a clinically validated rule.
But the advantage over admission GCS alone was modest. TRIAGES had an AUC of 0.905, compared with 0.890 for GCS, a difference of 0.015. The 95% confidence interval for that difference was 0.001 to 0.029, and the paired DeLong test gave p = 0.034. The clinical importance of that small within-cohort difference remains uncertain.
The signal held up inside the same dataset
Researchers also tested whether the relationship changed shape at different TRIAGES scores. The pattern was approximately linear, with a strong overall association, P overall < 0.001, and no detected evidence of non-linearity, P for non-linearity = 0.366.
The association remained similar under different approaches to missing data. The main analysis used multiple imputation by chained equations under a missing-at-random assumption, pooling 20 completed datasets with Rubin's rules. In sensitivity analyses, the odds ratio was 2.76 in a complete-case analysis and 2.86 with single imputation.
An internal check using 1,000 bootstrap resamples produced an optimism-corrected AUC of 0.905, identical to the apparent AUC, and a corrected Brier score of 0.102. The calibration intercept was 0.008 and the calibration slope was 0.998. This validation was limited to TRIAGES alone; the comparative estimates were generated in the same dataset used to fit the models and were not optimism-corrected.
Why the result needs a wider test
The limitations narrow the claim. Fifty-seven adults were excluded because discharge GOS was unavailable or invalid after transfer or discharge against medical advice, so their later functional outcomes were unknown. The study also drew on retrospective records from one hospital, which can limit how readily the result transfers to other settings.
The comparison with GCS also has built-in overlap: admission GCS is one of the four measurements used to calculate TRIAGES. In other words, the small performance difference is not a comparison between two wholly independent scores.
Exploratory subgroup analyses found associations in both women and men, with an interaction by sex. Interactions involving age and admission GCS severity were not significant.
Taken together, the findings show a strong signal within this hospital's data, not a finished clinical rule. The score of 3 or higher was derived from this cohort and was not clinically validated, and the study did not show that TRIAGES improves treatment or patient outcomes. Its added value beyond admission GCS remains a question for testing in other settings.
Paper data and sources
Original title: Admission TRIAGES score for unfavorable functional outcome after traumatic brain injury: association, prognostic performance, and comparison with the Glasgow coma scale
Authors: Shi W, Xing L, Wu J et al.
Journal/Repository: Frontiers in neurology
Status: Peer-reviewed
First online: 2026-10-19
DOI: Not available
Original paper · Full text