RESUMEN
BACKGROUND: The emergency department (ED) is the most frequent access route to the hospital. Nutrition risk (NR) screening allows the early identification of patients at risk of malnutrition. This study aimed to evaluate the feasibility and predictive validity of five different tools in EDs: Nutritional Risk Screening 2002 (NRS-2002), Nutritional Risk Emergency 2017 (NRE-2017), Royal Free Hospital-Nutritional Prioritizing Tool (RFH-NPT), Malnutrition Universal Screening (MUST), and Malnutrition Screening Tool (MST). METHODS: Patients with scores ≥3 according to the NRS-2002, ≥1.5 according to the NRE-2017, and ≥2 according to the MUST, RFH-NPT, or MST were classified with NR. Prolonged length of stay (LOS) and 1-year mortality were evaluated. RESULTS: 431 patients were evaluated (57.31 ± 15.6 years of age; 54.4% women) in a public hospital in southern Brazil. The prevalence of NR was: 35% according to the NRS-2002, 43% according to the MST, 45% according to the NRE-2017 and MUST, and 49% according to the RFH-NPT. Patients with NR, had a greater risk of prolonged LOS (P < 0.001). The presence of NR was associated with an increased risk of 1-year mortality according to the NRS-2002 (hazard ratio [HR]: 4.04; 95% CI, 2.513-6.503), MST (HR: 2.60; 95% CI, 1.701-3.996), NRE-2017 (HR: 4.82; 95% CI, 2.753-8.443), MUST (HR: 4.00; 95% CI, 2.385-6.710), and RFH-NPT (HR: 5.43; 95% CI, 2.984-9.907). CONCLUSIONS: NRE-2017 does not require objective data and presented predictive validity for all outcomes assessed, regardless of the severity of the disease, and thus appears to be the most appropriate tool for carrying out NR screening in the ED.