An observational cohort study investigating risk of malnutrition using the Malnutrition Universal Screening Tool in patients with stroke.

Background: Malnutrition in patients hospitalised with a stroke have been assessed using different nutritional screening methods but there is a paucity of data linking risk of malnutrition to clinical outcomes using a validated tool. Aims: To identify the prevalence of malnutrition risk in patients after a stroke and assess the predictive value of the Malnutrition Universal Screening Tool on clinical outcomes. Patients and methods: Using data from electronic records and the Sentinel Stroke National Audit Programme (January 2013 and March 2016), patients aged > 18 years with confirmed stroke admitted to a tertiary care stroke unit were assessed for risk of malnutrition. The association between malnutrition risk and clinical outcomes was investigated and adjusted for confounding variables. Results: Of 1101 patients, 66% were screened at admission. Most patients (n= 571, 78.5%) were identified as being at low risk, 4.1% (n=30) at medium risk and 17.4% (n=126) at high risk of malnutrition. Compared with low risk, patients with medium or high risk of malnutrition were more likely to have a longer hospital stay (IRR 1.30, 95% CI 1.07, 1.58), and had greater risk of mortality (10.9% versus 3.5%, 95% CI 0.03, 0.13). Conclusions: Prevalence of malnutrition assessed by Malnutrition Universal Screening Tool in patients after a stroke was relatively low, but nearly a third of patients were not screened. Patients classified as being at medium or high risk of malnutrition were more likely to experience negative outcomes. Early identification of this population may improve outcome if appropriate care is provided.