Predictive validity of the Mini Nutritional Assessment Short‐Form for rehabilitation patients: A retrospective analysis of the Japan Rehabilitation Nutrition Database

<jats:title>Abstract</jats:title><jats:sec><jats:title>Background</jats:title><jats:p>Malnutrition is associated with worse outcome in rehabilitation patients; however, appropriate malnutrition screening tools for this population have not been investigated. We examined the predictive validity of specific cut‐off values of the Mini Nutritional Assessment Short‐Form version 2 (MNA‐SFv2) for Japanese rehabilitation patients.</jats:p></jats:sec><jats:sec><jats:title>Methods</jats:title><jats:p>This retrospective cohort study analyzed adult patients (≥ 20 years) in the Japan Rehabilitation Nutrition Database who were in convalescent rehabilitation wards after stroke or hip fracture. Patients were classified into three categories based on MNA‐SFv2 original (0–7, 8–11 and 12–14 points, respectively) or modified (0–5, 6–7 and 8–14 points, respectively) cut‐off values: malnutrition, at risk of malnutrition or well‐nourished. Functional independence measure (FIM) and home discharge were compared between the categories.</jats:p></jats:sec><jats:sec><jats:title>Results</jats:title><jats:p>Overall, 489 patients were analyzed. Based on the MNA‐SFv2 original and modified cut‐off values, 64.4% and 36.0% were malnourished, 32.3% and 28.4% were at risk of malnutrition, and 3.3% and 35.6% were well‐nourished, respectively. Malnutrition defined by both cut‐off values was significantly associated with the FIM at admission, whereas only those defined by modified cut‐off values predicted the FIM at discharge (B, −7.1; 95% confidence interval = −12.3 to −1.9). Neither original, nor modified cut‐off values predicted discharge to home and long‐term care facilities.</jats:p></jats:sec><jats:sec><jats:title>Conclusions</jats:title><jats:p>An MNA‐SFv2 score of 0–5 points may be useful to identify Japanese patients with poor outcomes in a rehabilitation setting.</jats:p></jats:sec>