Diagnostic accuracy of shock index, revised trauma score, and injury severity score for mortality prediction in trauma: A hierarchical meta-analysis.
Injury · 2026-10-08 · Meta-analysis
Abstract
BACKGROUND: Shock Index (SI) and Revised Trauma Score (RTS) are physiological tools available at the point of trauma triage. Injury Severity Score (ISS) is an anatomical score that requires completed injury characterization and is typically available only after initial resuscitation. Their comparative value for predicting mortality and whether ISS meaningfully adds to the acute tools remains unclear. AIMS: The aim of this systematic review and hierarchical meta-analysis is to pool the diagnostic accuracy of SI and RTS as acute mortality-prediction tools in adult trauma patients; to benchmark these against ISS; and to formally test, in studies reporting both, whether RTS and ISS differ in diagnostic accuracy. METHODS: Electronic databases were searched from inception to May 2026. In-hospital mortality was prespecified as the primary endpoint. Pooled sensitivity, specificity and hierarchical summary receiver operating characteristic (HSROC) analyses were performed with a bivariate random-effects model for each tool separately. For studies reporting both RTS and ISS in the same cohort, a paired comparison was performed using a bivariate model with test type as a covariate and a pooled within-study difference analysis. RESULTS: Twelve studies (234,129 patients) were included. Pooled sensitivity/specificity were 0.679/0.763 for SI (5 studies, AUC 0.747), 0.788/0.801 for RTS (6 studies, AUC 0.852), and 0.788/0.796 for ISS (9 studies, AUC 0.855). In the six studies reporting both RTS and ISS, a paired bivariate comparison found no significant difference in diagnostic accuracy (likelihood-ratio test, p = 0.91; within-study sensitivity OR 1.59, 95% CI 0.82-3.07; specificity OR 0.89, 95% CI 0.55-1.45). CONCLUSION: RTS and ISS did not differ significantly in a formal paired comparison, and neither tool, alone, reliably rules in or out death at realistic trauma prevalence. SI, RTS and ISS should be understood as serving different, complementary roles across the trauma timeline - SI and RTS at the point of acute triage, ISS as a later anatomical benchmark - rather than as competing predictors to be ranked against one another.
Abstract © the authors or publisher, via PubMed. Reuse and removal
Tags
Early warning scores · Mortality & survival · Triage instruments
Related articles
- On-scene predictors of injury severity, deterioration, and death following road injury: a systematic scoping review of information available to responders and its use in triage tools. (Scand J Trauma Resusc Emerg Med, 2025)
- The accuracy of acuity scoring tools to predict 24-h mortality in traumatic brain injury patients: A guide to triage criteria. (Int Emerg Nurs, 2018)
- Prehospital national early warning score outperforms shock index in predicting mortality and resource utilisation in adult trauma: A systematic review. (Injury, 2026)
- Shock index and shock index, pediatric age-adjusted as predictors of mortality in pediatric patients with trauma: A systematic review and meta-analysis. (PLoS One, 2024)
- Predictive Utility of the Total Glasgow Coma Scale Versus the Motor Component of the Glasgow Coma Scale for Identification of Patients With Serious Traumatic Injuries. (Ann Emerg Med, 2017)
- Evaluating performance and potential clinical benefit of the Swedish On Scene Injury Severity Prediction (OSISP) model for prehospital field triage on Norwegian trauma data. (Scand J Trauma Resusc Emerg Med, 2026)