RESUMO
BACKGROUND: Malnutrition in hospitalized patients is associated with an increased risk of death, in both the short and the long term. AIMS: The purpose of this study was to determine which nutrition-related risk index predicts long-term mortality better (three years) in patients who receive total parenteral nutrition (TPN). METHODS: This prospective, multicenter study involved noncritically ill patients who were prescribed TPN during hospitalization. Data were collected on Subjective Global Assessment (SGA), Nutritional Risk Index (NRI), Geriatric Nutritional Risk Index (GNRI), body mass index, albumin and prealbumin, as well as long-term mortality. RESULTS: Over the 1- and 3-year follow-up periods, 174 and 244 study subjects (28.8% and 40.3%) respectively, died. Based on the Cox proportional hazards survival model, the nutrition-related risk indexes most strongly associated with mortality were SGA and albumin (<2.5 g/dL) (after adjustment for age, gender, C-reactive protein levels, prior comorbidity, mean capillary blood glucose during TPN infusion, diabetes status prior to TPN, diagnosis, and infectious complications during hospitalization). CONCLUSIONS: The SGA and very low albumin levels are simple tools that predict the risk of long-term mortality better than other tools in noncritically ill patients who receive TPN during hospitalization.