RESUMO
BACKGROUND: Identification of intensive care unit (ICU) patients who require nutrition intervention is crucial to initiating nutrition therapy. This prospective quality improvement study evaluated the Nutritional Risk Screening (NRS) 2002, Malnutrition Universal Screening Tool (MUST), and Nutrition Risk in Critically Ill (NUTRIC) score in comparison with the Veterans Administration Nutrition Status Classification (VANSC) tool to determine which best identified the need for nutrition intervention. METHODS: A convenience sample of 150 ICU patients was evaluated using the VANSC, NRS 2002, MUST, and the NUTRIC score. The resultant score, need for nutrition intervention, and presence of malnutrition were recorded for patients. Interventions were defined as need for enteral or parenteral nutrition, nutritional supplements, or diet change. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), Matthews Correlation Coefficient (MCC), F1 score, and accuracy to predict need for nutrition intervention were calculated for each screening tool. RESULTS: Of the 150 patients, 49 (33%) required 1 or more nutrition interventions. The NRS 2002 (0.878) and VANSC (0.816) had the highest sensitivity. The NUTRIC (0.921) and VANSC (0.911) had the highest specificity. The VANSC (0.816) and MUST (0.687) had the highest PPV. The VANSC (0.911) and NRS 2002 (0.872) had the highest NPV. The VANSC (0.727) and MUST (0.528) had the highest MCC. The VANSC (0.816) and MUST (0.680) had the highest F1 score. CONCLUSIONS: Trialing several tools to identify their efficacy and reliability individual setting may help determine the most appropriate tool to utilize for your patient population and specific goals.