The baseline nutritional status assessed by MUST score has a low accuracy in predicting the risk of hospitalization during follow-up in patients with chronic pancreatitis: A cohort study.

BACKGROUND: Hospitalization and death in patients with chronic pancreatitis (CP) are often due to extra-pancreatic events. Recent guidelines recommend the use of the MUST score to assess CP patients' nutritional status, but its association with clinical outcomes has been poorly investigated. The aims of this study are to evaluate the incidence of extra-pancreatic events in patients with CP during follow-up and their association with the nutritional status. METHODS: Retrospective analysis of single-centre cohort of CP patients prospectively enrolled and followed-up. Exocrine pancreatic insufficiency (EPI) was assessed by fecal elastase, MUST score calculated at diagnosis. The occurrence of hospitalizations or death were recorded. Differences between subgroups were analysed by Fisher's and T-test and hospitalization-free survival with Kaplan-Meier curves and Cox regression analysis. RESULTS: Of 111 enrolled patients (64% male; mean age 57); 52% had alcoholic aetiology, 53% EPI, 10% severe CP and 26% a MUST score≥2 at diagnosis. During a median follow-up of 37 months, 3.6% of patients died and 34.2% needed hospitalization, in 50% of cases for extrapancreatic events (2% cardiovascular events, 8% infections and 3% cancer). There was no significant association between EPI, BMI<20 kg/m2, MUST score≥2, alcoholic aetiology and extra pancreatic events or need of hospitalization. A baseline MUST score≥2 had an accuracy of only 64.8% in predicting subsequent hospitalization. CONCLUSIONS: A sizeable portion of CP patients are at high risk of malnutrition and are hospitalized during the follow-up, often for extra-pancreatic events. The nutritional status evaluated with the MUST score lacks accuracy in predicting the risk of these events.