Background & aims: There is no method to be used as a reference standard for nutritional assessment. This study aims to develop and verify the performance of a new tool, based on the Item Response Theory (IRT), from the Subjective Global Assessment (SGA) questionnaire, in hospitalized adult patients.
Methods: Retrospective cohort study, composed by secondary database, formed by patients included from October 2005 to June 2006. The new tool presented was developed through the usage of cumulative models from the IRT. Out of 1503 evaluated patients, 2/3 were randomly selected to the development sample of the new tool and 1/3 to the performance verification sample.
Results: After item adjustments, the "Nutritional Assessment Score" (NAS) was proposed, with reduced number of questions, and, in comparison to SGA, less polytomic items. NAS demonstrates association to variables that are clinically relevant (hospital mortality, long hospital stay, serum albumin and body mass index) and has shown itself to be more precise to patients with the worst degrees of nutritional status.
Conclusions: Results point to the validation of the NAS in detecting, accurately, the nutritional status of hospitalized patients.
Keywords: Inpatients; Malnutrition; Nutritional Assessment Score; Nutritional assessment; Nutritional status; Subjective Global Assessment.
Copyright © 2017 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.