Prognostic value of the NUTRIC score for 28-day mortality in patients with sepsis-associated acute respiratory distress syndrome and its comparison with other nutritional indices: a retrospective cohort study

Abstract

Background Nutritional and metabolic derangements are common in patients with sepsis-associated acute respiratory distress syndrome (ARDS) and may contribute to adverse outcomes. This study evaluated the prognostic value of the Nutrition Risk in Critically Ill (NUTRIC) score for 28-day mortality in this population. Methods A retrospective analysis was conducted on clinical data from 152 sepsis patients with ARDS admitted to the intensive care unit (ICU) at Yiwu Central Hospital between December 2023 and January 2025. Associations between NUTRIC score and 28-day mortality were examined using multivariable logistic and Cox regression. Restricted cubic spline (RCS) analysis assessed dose–response relationship. Discrimination, calibration, and clinical utility were evaluated using receiver operating characteristic (ROC) curves, calibration analysis, Brier score, and decision curve analysis (DCA). Internal validation used 1,000 bootstrap resamples. Subgroup, interaction, and Kaplan–Meier analyses were performed. Results The 28-day mortality rate was 34.2% (52/152). After multivariable adjustment, each 1-point increase in NUTRIC score was associated with higher odds of mortality (odds ratio, 1.65; 95% CI, 1.34–2.02; p < 0.001) and a higher mortality hazard (hazard ratio, 1.41; 95% CI, 1.24–1.61; p < 0.001). Quartile analysis demonstrated a significant trend (p for trend < 0.001). RCS analysis indicated an approximately linear positive association (p for nonlinearity = 0.764), robust to alternative knot placements. The NUTRIC score showed moderate discriminative (AUC = 0.743, bootstrap 95% CI, 0.661–0.827) and performed better than mNUTRIC (AUC = 0.712, DeLong p = 0.003). Compared with mNUTRIC, NUTRIC improved reclassification and discrimination (net reclassification improvement, 0.219; integrated discrimination improvement, 0.044; both p < 0.001). Calibration was satisfactory (Brier score, 0.194; slope, 1.018; intercept, 0.020). Decision curve analysis showed net benefit at threshold probabilities of 11–60%. A significant interaction was found for septic shock (p for interaction = 0.038). Patients with NUTRIC scores ≥6 had lower cumulative survival (log-rank p < 0.001). Conclusion In patients with sepsis-associated ARDS, higher NUTRIC scores were independently associated with increased 28-day mortality. The NUTRIC score may serve as a practical tool for early risk stratification and integrated assessment of nutritional and disease burden in critically ill patients with sepsis-associated ARDS.