European Journal of Trauma and Emergency Surgery · Nov 16, 2026
Critically ill trauma in Portugal: a multicenter intensive care cohort study of frailty, organ dysfunction, and outcomes
By Gonçalves Pereira M, Pinho A, Guimarães T et al.
Purpose Multicenter data on critically ill trauma patients admitted to Portuguese intensive care units (ICUs) are lacking. We aimed to characterize their epidemiology, injury profile, organ dysfunction trajectories and outcomes, and to evaluate the prognostic performance of severity and frailty measures. Methods Multicenter retrospective cohort study within the National Intensive Care Registry (RNMI) of adults (≥ 18 years) admitted to an ICU as a direct consequence of trauma (1 October–31 December 2024) in 32 of 50 voluntarily participating Portuguese ICUs; not nationally representative. SAPS II, ISS, TRISS, the Clinical Frailty Scale, daily SOFA scores, organ support, quality of life and mortality were collected. Prognostic performance was assessed by ROC analysis, multivariable logistic regression, calibration, and mixed-effects modelling of serial SOFA. Results 457 patients were included; median age 61 years (IQR 45–74), 72.0% male. Falls were the leading injury mechanism (46.9%). Traumatic brain injury (32.9%) and thoracic trauma (21.4%) predominated. Observed ICU and hospital mortality were 11.0% and 15.3%, respectively. SOFA scores were consistently higher in non-survivors from day one. SAPS II showed the highest discrimination for hospital mortality (AUC 0.874, 95% CI 0.825–0.924), followed by TRISS (0.758) and the Clinical Frailty Scale (0.702). SAPS II, day-1 SOFA, and frailty were independently associated with hospital mortality. SAPS II over-predicted mortality (observed/expected 0.64), whereas TRISS showed a non-significant tendency to under-predict it (1.32, 95% CI 0.99–1.72). Conclusions In this multicenter cohort from 32 voluntarily participating ICUs, patients were older, with frequent falls, neurotrauma and thoracic injury, and moderate mortality. Organ dysfunction trajectories and frailty were associated with hospital mortality, but incremental predictive value was not assessed; both merit inclusion in future trauma registry frameworks.