Meta-analysis of the effectiveness of early endoscopic treatment of Acute biliary pancreatitis based on lightweight deep learning model

Abstract

Abstract Background Acute biliary pancreatitis (ABP) is a clinical common acute abdomen. After the first pancreatitis, relapse rate is high, which seriously affects human life and health and causes great economic burdens to family and society. According to a great many research findings, endoscopic retrograde cholangiopancreatography (ERCP) is an effective treatment method. However, whether ERCP should be performed in early stage of ABP is still controversial in clinical practice. Methods Related articles were retrieved from Pubmed, Web of Science core library, Nature, Science Direct, and other databases published from January 2000 until now. The keywords included early ERCP, delayed ERCP, ABP, laparoscopy, and cholecystectomy, all which were connected by “or” and “and”. The language of articles was not restricted during the retrieval and Review Manager5.3 was employed to perform meta-analysis of experimental data. Finally, a total of 8 eligible articles were selected, including 8,801 patients. Results The results of the meta-analysis demonstrated that no remarkable differences were detected in the incidence of complications, mortality, and operation time between patients undergoing ERCP in early stage and those receiving delayed ERCP. However, the hospitalization time of patients in experimental group was notably shorter than that among patients in control group. Conclusins Early ERCP treatment is as safe as late ERCP treatment for biliary pancreatitis, and can significantly shorten the hospital stay. Hence, the therapy was worthy of clinical promotion. The research findings provided reference and basis for clinical treatment of relevant diseases.

Keywords

• Acute biliary pancreatitis
• Early endoscopic retrograde cholangiopancreatography
• Light-weight deep learning
• Meta-analysis