Prognostic performance of the mNUTRIC score for in-hospital mortality in critically Ill patients with severe and critical COVID-19: A retrospective cohort study

The prognostic value of the modified Nutrition Risk in the Critically Ill (mNUTRIC) score in patients with severe and critical Coronavirus disease 2019 (COVID-19) remains insufficiently defined. This study aimed to assess the predictive performance of the mNUTRIC score for in-hospital mortality in patients with severe and critical COVID-19 admitted to the intensive care unit. We conducted a retrospective cohort study including 355 critically ill patients with severe and critical COVID-19 admitted to a university-affiliated hospital in Vietnam. The mNUTRIC score was calculated using the first available clinical and laboratory values recorded within the first 24 hours of intensive care unit admission. The primary outcome was in-hospital mortality. Associations between the mNUTRIC score and mortality were assessed using multivariable logistic regression. Predictive performance was evaluated using the area under the receiver operating characteristic curve (AUC). Overall in-hospital mortality was 36.1%. High nutritional risk, defined as an mNUTRIC score ≥5, was identified in 29.6% of patients. In multivariable analysis, the mNUTRIC score remained independently associated with in-hospital mortality (adjusted odds ratio 1.58, 95% confidence interval 1.24-2.02; P < .001). The mNUTRIC score demonstrated moderate discriminative performance for predicting in-hospital mortality, with an AUC of 0.750 (95% confidence interval 0.698-0.803; P < .001). At the optimal cutoff value of ≥4.5, the sensitivity and specificity for predicting in-hospital mortality were 54.7% and 84.6%, respectively. In critically ill patients with severe and critical COVID-19, the mNUTRIC score was associated with in-hospital mortality and demonstrated moderate prognostic performance. Given the retrospective design, these findings should be interpreted with caution and warrant further prospective validation.