Estudio descriptivo de dos herramientas de cribado de riesgo nutricional al ingreso hospitalario en el sanatorio Allende de Argentina

: The aim of this study was to compare two nutritional screening tools, Nutritional Risk Screening 2002 (NRS-2002) and Nutritional Analytical Phase Filter (FILNUT) on a sample of patients to hospital admission, analyze the prevalence of nutritional risk, assess the agreement between the two and their validity for identifying patients at risk. Material and methods : Descriptive, cross-sectional study, where 271 admitted patients were evaluated within the first 72 hours, using the NRS-2002 method used as the gold standard and the FILNUT tool. Chi2 Test was used for the statistical association between the different methods and the agreement was analyzed with the kappa index. Accuracy was assessed by sensitivity, specificity, positive predictive value, negative predictive value, and likelihood ratio. The validity of the nutritional screening test to identify patients at risk was analyzed by comparing ROC curves with determination of the area under the curve (AUC). Results : The prevalence of patients at nutritional risk was 61% with FILNUT and 31%t with NRS-2002. The FILNUT method with high sensitivity (92.8%), throws a considerable number of false positives by low specificity (53.3%). The agreement between both tools was fair (Kappa = 0.37). Using ROC curve, both tests were valid for identify patients at risk: AUC = 0.999 FILNUT (95% IC 0.963 to 0.100); NRS-2002 AUC = 0.708 (95% IC 0.643-0.767). Conclusions : The FILNUT method is a valid screening tool that reveals higher prevalence of nutritional risk and with high sensitivity to dismiss properly those patients without risk. Comparing both tools, FILNUT incorporates the use of prealbumin, whose low levels should be considered as potential nutritional risk. On the other hand, its low time factor and direct costs in its implementation ensure it as an efficient tool. The simplicity and facility required for its performance contrast with the NRS-2002 that requires trained personnel.