Determination of relationship between NRS-2002, M-NUTRIC scores, and GLIM criteria used for malnutrition screening and mortality in the intensive care

There is still no ideal malnutrition risk screening tool for intensive care patients. This study, it was aimed to screen and detect malnutrition in critically ill patients with NRS-2002 score, m-NUTRIC score, and GLIM criteria, to determine the relationship between malnutrition and mortality in high-risk patients. The study included one hundred six patients hospitalized in Hacettepe University Hospital, Internal Medicine ICU between 15.05.2022-31.08.2022. 59.4% of the study population was male and the median age was 64. According to the NRS-2002 score, 71 (67%) patients and 54 (50.9%) patients according to the m-NUTRIC score were found to be at high risk for malnutrition. According to GLIM criteria, malnutrition was detected in 72 (67.9%) patients. According to three screening scores, intensive care mortality, hospital mortality, 28-day mortality, and 90-day mortality were evaluated in patients with and without malnutrition. In Kaplan-Meier survival analyses, 28-day mortality and 90-day mortality were found to be significant in malnourished patients according to all three screening scores. NRS-2002 and mNUTRIC were found to be significant in ROC analyses to predict intensive care mortality (p=0.002 and <0.001). All three scores were significant in the ROC curve to predict hospital mortality, 28-day mortality, and 90-day mortality. It was determined that the m-NUTRIC score predicted intensive care, hospital, 28-day and 90-day mortality with the best sensitivity and specificity at a cut-off value of <5 points in our population. At the same time, it has been shown that the risk of 28-day mortality decreases with reaching the target calorie in patients who are malnourished according to all three scores. The use of NRS 2002 and m-NUTRIC scores are appropriate for screening for malnutrition in critically ill patients. Although GLIM criteria were also found to be significant for predicting mortality in our study, more studies are needed in intensive care patients to recommend their use.