Background Identification of nutritional risk in critically ill patients is a challenge because each nutritional screening tool has its specific characteristics. The objective was to evaluate the performance of the modified Nutrition Risk in Critically ill (mNUTRIC) score, used alone or in combination with the Nutritional Risk Screening 2002 (NRS‐2002) score, to predict hospital mortality in critically ill patients. Methods A prospective study was performed with patients admitted to the intensive care unit (ICU) from October 2017 to April 2018. Multiple logistic regression analysis was used to test for complementarity between the mNUTRIC and NRS‐2002. A receiver operating characteristic (ROC) curve was used to identify the performance of the instruments to predict mortality. This study was conducted in accordance with the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) statement. Results 384 patients were evaluated (51.8% female mean age 59.6 ± 16.7 years). High nutritional risk was detected in 54.4% by the NRS‐2002 and 48.4% by mNUTRIC. The overall mortality rate was 36.5% (n = 140). Patients in whom nutritional risk was identified both by mNUTRIC and by NRS‐2002 (score ≥5) had a twofold greater risk of in‐hospital mortality (RR = 2.29; 95%CI: 1.42–3.68; p = 0.001). The area under the ROC curve to predict mortality was 0.693 for mNUTRIC; 0.645 for NRS‐2002; and 0.666 for mNUTRIC and NRS‐2002 combined. Conclusions The mNUTRIC and NRS‐2002 scores had similar performance in predicting hospital mortality. The mNUTRIC has better discriminant ability to quantify the risk of mortality in critically ill patients.

No abstract