RESUMO
Understanding of the association between nutritional risk and clinical outcomes in hospitalised patients with overweight is still at an early stage. Given the lack of specific tools for the patient with overweight, the objective of this study was to compare two of the main general screening instruments for assessing nutritional risk in predicting clinical outcomes in a population of hospitalised patients with overweight. A retrospective study was carried out in a medium-sized hospital in Brazil, with patients ≥20 years old admitted between July 2017 and December 2019. Patients who were overweight and had records of Nutritional Risk Screening-2002 (NRS-2002) and Mini-Nutritional Assessment-Short Form (MNA-SF) in their medical files were included in the study. Clinical outcomes data (longer length of stay, readmission during the study period and mortality before the end of study or during hospitalisation) were obtained. The Kappa coefficient assessed agreement between both tools, and their performance for predicting outcomes was analysed using characteristic receiver operating curves (ROC). Data were collected from 643 patients. The prevalence of nutritional risk was 17.7% and 36.1% according to the NRS-2002 and MNA-SF (k = 0.390; p < 0.001), respectively. According to both tools, all clinical outcomes were significantly more common among individuals at nutritional risk (p < 0.05). Only the MNA-SF showed a significant percentage of predictions for readmission (57.2%) and death during hospitalisation (65.7%). For mortality until the end of the study, the area under the ROC curve was similar for MNA-SF (60.5%) and NRS-2002 (60.7%; p = 0.057). The MNA-SF detected a greater proportion of nutritional risk among hospitalised patients with overweight and better predicted all clinical outcomes compared to the NRS-2002 and should be used to screen patients with overweight for nutritional risk.