Abstract 63: Risk of Malnutrition on Admission Predicts Mortality, Length of Hospital Stay and Hospitalisation Costs at 6 Months Post Stroke

Abstract

INTRODUCTION Malnutrition is related with poor outcomes after stroke which justifies the need for screening to identify and treat nutritionally vulnerable patients; however, no nutrition screening tool has been validated for this population. This study aimed to establish the predictive validity of the Malnutrition Universal Screening Tool (MUST) in stroke patients, using several outcomes: mortality, cumulative length of hospital stay (LOS) and hospitalisation costs at 6 months post stroke. METHODS Patients were recruited from consecutive admissions at 2 hyper-acute stroke units in London and were screened for their risk of malnutrition (low, medium and high) according to MUST. Outcomes were obtained for each patient through a national database that contains details of all hospital admissions. RESULTS Of 543 recruited patients, 537 were screened within 72h of admission with MUST, 51% were males and 87% had an ischaemic stroke, with a mean age of 74.7years (range 22-99). Results (see table) showed a strong positive association between risk of malnutrition and mortality rate (Chi square test, p<0.001 ), which remained significant after adjustment for possible confounders (Multivariable Cox Proportional Hazards Model, p<0.001 ). For patients who survived, there was a strong positive association between the risk of malnutrition and both LOS and hospitalisation costs (Kruskal-Wallis test, p<0.001 and p=0.049 , respectively), which again remained significant after adjustment for possible confounders (univariate analysis of variance, p<0.001 and p=0.001 , respectively). CONCLUSION Risk of malnutrition (as assessed by MUST) is an independent strong predictor of mortality, LOS and hospitalisation costs at 6 months post stroke, supporting the routine screening of stroke patients for risk of malnutrition on admission.