RESUMO
Objective To evaluate the clinical effect and safety of early oral feeding versus delayed feeding in patients with acute pancreatitis using Meta-analysis.Methods An electronic search of PubMed,Cochrane Library,Embase,Web of Science,CNKI,Wanfang database and VIP database were retrieved from their inception to March 2019 to identify relevant literatures,by taking'acute pancreatitis,oral,food intake,refeeding,feeding,diet'as the keywords for retrieval.The evaluation indexes included hospitalization time,incidence of abdominal pain after feeding,incidence of nausea and vomiting after feeding,incidence of hyperglycemia and leucocyte level after feeding.To select literatures and extractdata strictly according to inclusion and exclusion criteria,then conducted data analysis by Revman 5.3 provided by Cochrane collaboration.Results A total of 9 articles,involving 856 patients,were included with 425 in the early feeding group and 431 in the delayed feeding group.The length of hospitalization in early feeding group was shorter than that in delayed feeding group (MD =-2.37,95％ CI:-3.45--1.28,P ＜ 0.000 1).The incidence of abdominal pain (P =0.67),nausea and vomiting (P =0.77),hyperglycemia (P =0.54),and leukocyte level (P =0.26) were not significantly different between two groups.Conclusion Early oral feeding is safe and effective for patients with acute pancreatitis,which can significantly shorten the hospital stay without increasing the incidence of adverse reactions.