The evaluation of consumed food portions as a screening test for malnutrition in the older people living in a nursing home: A cross-sectional pilot study

IntroductionThe Self Evaluation of Food Intake (SEFI®) is a simple tool to assess food intake.AimsTo evaluate the validity of SEFI® to screen malnutrition compared to the Mini Nutritional Assessment (MNA)® as the reference method (primary) and other screening tools (secondary); to assess the feasibility of SEFI® (secondary) in older people living in a nursing home.MethodsQuantitative, non-interventional, cross-sectional, monocentric pilot study. Patients were included in a 143-beds nursing home. The SEFI® visual analogue scale (VAS) from 0 to 10, and the evaluation of consumed portions of served food by the general practioner (SEFI® consumed portions) were performed after the MNA®. Malnutrition was defined as MNA®<17.StatisticsYouden method and receiver operating characteristic (ROC) curves.ResultsOut of 128 patients, 93 were included: 72% women, mean age (±SD), 88 ± 7 yr, body mass index, 25.9 ± 6.1, MNA®<17, 26% (24/93), SEFI® consumed portions ≤50%, 29% (27/93) of patients. The feasibility of SEFI®  VAS and consumed portions were respectively 0 and 100% (93/93). The predictive performance of SEFI® consumed portions for the diagnosis of malnutrition was good: area under the curve = 0.81 [95% confidence interval (CI), 0.69–0.93]; sensitivity 75.0% (n = 18/24, 95%CI, 57.7–92.3), specificity 87.0% (n = 60/69, 77.0–94.9), positive predictive value 66.7% (n = 18/27, 48.9–84.4), and negative predictive value 90.9% (n = 60/66, 80.0–93.9).ConclusionsSEFI® is feasible and useful for detecting malnutrition in nursing home residents.