Added Value of Quadriceps Muscle Layer Thickness to modified Nutrition Risk in the Critically Ill (mNUTRIC) Score in Predicting Clinical Outcomes in Critically Ill Patients: A Prospective Observational Study

Résumé fourni par la source

Background and aims: This study evaluated whether adding quadriceps muscle layer thickness (QMLT) to the modified nutrition risk in the critically ill (mNUTRIC) score improves prediction of 28 days mortality in critically ill patients. Patients and methods:We conducted a prospective observational study in a single-center intensive care unit (ICU), enrolling 201 adult patients.Nutritional risk was assessed using the mNUTRIC score, and QMLT was measured using bedside ultrasound at ICU admission.The primary objective was to compare the predictive performance of mNUTRIC, QMLT, and their combination (mNUTRIC + QMLT) for 28 days mortality using receiver operating characteristic (ROC) curves for discrimination, Brier scores for accuracy, and decision curve analysis (DCA) for clinical utility.Results: Our cohort had a median age of 60 years (IQR, 50-69.5),with a 28 days mortality rate of 35.8%.Median mNUTRIC and QMLT were 4 (3-5) and 1.71 cm (1.31-2.13),respectively.Non-survivors had significantly lower QMLT (1.54 vs 1.84 cm, p = 0.010) and higher mNUTRIC scores (5 vs 3, p < 0.001).Mortality was highest (67.56%) in patients with both high mNUTRIC (≥5) and below-median QMLT.The combined model showed superior discrimination [area under the curve (AUC) = 0.783] compared to mNUTRIC (0.708) and QMLT (0.614).Calibration was acceptable for the combined and mNUTRIC models.The Brier score was lowest for the combined model (0.1769), indicating highest accuracy.Decision curve analysis showed greater clinical utility for the combined model, especially at mortality risk thresholds of 20-55% and 75-90%.Conclusions: Combining QMLT with mNUTRIC improved prediction of 28 days mortality, offering a more comprehensive tool for risk stratification and potential guidance for targeted ICU interventions.Clinical trial number: CTRI/2023/09/057364.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.