RESUMO
BACKGROUND & AIMS: No universally accepted diagnostic criteria for malnutrition are available to date. The aim was to assess the concordance for the diagnosis of hospital malnutrition between the bioelectrical impedance vector analysis (BIVA) and the new definition of malnutrition proposed by the European Society for Clinical Nutrition and Metabolism (ESPEN). METHODS: A total of 140 adult hospitalized patients were included. The malnutrition risk was assessed by the Nutritional Risk Screening. The ESPEN malnutrition criteria (body mass index (BMI) <18.5 kg/m2, weight loss (WL) + low BMI, and WL + low fat free mass index (FFMI) were applied to each patient. The bioelectrical impedance of each patient was measured, and malnutrition was diagnosed using the BIVA method. Diagnostic concordance between the BIVA and the new ESPEN definition was assessed with the Kappa coefficient. RESULTS: The malnutrition prevalence was higher with the BIVA vs ESPEN definition (22.1% vs 13.6%) in the global population and was similar (12.8% vs 12.1%) in the population at risk of malnutrition. The diagnostic performance of the BIVA was acceptable, with higher sensitivity in patients with fluid overload (FO) and more specificity in euvolemic patients. Diagnostic concordance between the BIVA and the ESPEN definition was poor for the global population (kappa = 0.56) and the population at risk of malnutrition (kappa = 0.67) but was acceptable in patients with FO (kappa = 0.78). However, the discordant BIVA+/ESPEN- patients classified as false positives for BIVA showed clinical and body composition data (low FFMI, low phase angle) consistent with malnutrition. CONCLUSIONS: According to the clinical and bioelectrical characteristics of the discordant BIVA+/ESPEN- patients, the BIVA could perform better that the new ESPEN definition for the diagnosis of hospital malnutrition, which should be confirmed with other studies.
