Timing of enteral nutrition in acute pancreatitis: Meta-analysis of individuals using a single-arm of randomised trials

• Published in: Pancreatology : official journal of the International Association of Pancreatology (IAP) ... [et al.] 2014-09, Vol.14 (5), p.340-346
• Main Authors: Bakker, Olaf J, van Brunschot, Sandra, Farre, Antoni, Johnson, Colin D, Kalfarentzos, Fotis, Louie, Brian E, Oláh, Attila, O'Keefe, Stephen J, Petrov, Maxim S, Powell, James J, Besselink, Marc G, van Santvoort, Hjalmar C, Rovers, Maroeska M, Gooszen, Hein G
• Format: Article
• Language: English
• Subjects: Acute Disease; Acute pancreatitis; Confidence intervals; Endocrinology & Metabolism; Enteral nutrition; Enteral Nutrition - methods; Gastroenterology and Hepatology; Hospitalization; Humans; Infected necrosis; Logistic Models; Meta-analysis; Mortality; Motility; Multiple Organ Failure - etiology; Multiple Organ Failure - prevention & control; Multivariate Analysis; Necrosis; Nutrition; Organ failure; Pancreas; Pancreatitis - complications; Pancreatitis - mortality; Pancreatitis - therapy; Pancreatitis, Acute Necrotizing - prevention & control; Parenteral nutrition; Physiology; Randomized Controlled Trials as Topic; Regression analysis; Studies; Time Factors; Treatment Outcome; Variables
• ISSN: 1424-3903; 1424-3911
• DOI: 10.1016/j.pan.2014.07.008

Abstract Introduction In acute pancreatitis, enteral nutrition (EN) reduces the rate of complications, such as infected pancreatic necrosis, organ failure, and mortality, as compared to parenteral nutrition (PN). Starting EN within 24 h of admission might further reduce complications. Methods A literature search for trials of EN in acute pancreatitis was performed. Authors of eligible trials were requested to provide the data of all patients in the EN-arm of their trials. A meta-analysis of individual patient data was performed. The cohort of patients with EN was divided into patients receiving EN within 24 h or after 24 h of admission. Multivariable logistic regression, adjusting for predicted disease severity and trial, was used to study the effect of timing of EN on a composite endpoint of infected pancreatic necrosis, organ failure, or mortality. Results Observational data from 165 individuals from 8 randomised trials were obtained; 100 patients with EN within 24 h and 65 patients with EN after 24 h of admission. In the multivariable model, EN started within 24 h of admission compared to EN started after 24 h of admission, reduced the composite endpoint from 45% to 19% (adjusted odds ratio [OR] of 0.44; 95% confidence interval [CI] 0.20–0.96). Within the composite endpoint, organ failure was reduced from 42% to 16% (adjusted OR 0.42; 95% CI 0.19–0.94). Conclusions In this meta-analysis of observational data from individuals with acute pancreatitis, starting EN within 24 h after hospital admission, compared with after 24 h, was associated with a reduction in complications.