Supplementary file 5_Admission TRIAGES score for unfavorable functional outcome after traumatic brain injury: association, prognostic performance, and comparison with the Glasgow coma scale.docx

Multivariable logistic regression, restricted cubic spline analysis, receiver operating characteristic (ROC) analysis, calibration, Brier score, decision curve analysis, bootstrap internal validation, sensitivity analyses, and subgroup analyses were performed.</p>Results<p>Among 753 patients, 181 (24.0%) had an unfavorable functional outcome. Each 1-point increase in TRIAGES was independently associated with higher odds of unfavorable functional status after adjustment for sex, injury mechanism, number of non-reactive pupils, and Rotterdam CT score (adjusted odds ratio 2.81, 95% confidence interval 2.29–3.45; p < 0.001).

The association was approximately linear (P for non-linearity = 0.366). TRIAGES alone achieved an area under the ROC curve (AUC) of 0.905 compared with 0.890 for admission GCS alone (ΔAUC = 0.015, 95% CI 0.001–0.029; paired DeLong p = 0.034). Within this cohort, a Youden index-derived threshold of TRIAGES ≥3 yielded a sensitivity of 0.956 and specificity of 0.701.

Bootstrap internal validation showed an optimism-corrected AUC of 0.905, and the findings remained consistent across multiple sensitivity analyses.</p>Conclusion<p>Higher TRIAGES scores at initial emergency presentation were independently associated with unfavorable functional status at hospital discharge in adults with TBI. TRIAGES showed good prognostic performance and a small within-cohort increase in discrimination relative to admission GCS alone while retaining the practical advantage of using only four routinely available variables.

Prospective multicenter studies with external validation and standardized long-term functional outcomes are warranted.</p>