OBJECTIVE Decompensated liver cirrhosis (LC) can negatively affect the outcome of endoscopic retrograde cholangiopancreatography (ERCP). We aimed to compare the efficacy and safety of ERCP in patients with clinically defined compensated and decompensated LC.

METHODS In a single tertiary hospital, 146 endoscopic sphincterotomy-naive patients with LC who underwent ERCP between 2005 and 2016 were reviewed. Patients with LC who had experienced variceal bleeding, ascites or hepatic encephalopathy were included in the decompensated LC group. Cannulation, technical and clinical successes, and major post-ERCP adverse events including bleeding, pancreatitis, cholangitis and perforation were compared between the two groups.

RESULTS Patients were divided into compensated and decompensated LC groups. Their baseline characteristics were not different, except for comorbid malignancy (22.3 % vs 38.5%, P = 0.038) and preprocedural transfusion (7.4% vs 36.5%, P < 0.001). The cannulation (97,9% vs 94.2%, P = 0.348) and technical (95.7% vs 88.5%, P = 0.167) success rates were not different. The clinical success rate was lower in the decompensated LC group (95.7% and 78.8%, P = 0.003), mainly due to comorbid hepatobiliary malignancy. Post-ERCP pancreatitis (6.4% vs 30.8%, P = 0.008) and cholangitis (18.1% vs 32.7%, P = 0.046) rates were higher in the decompensated LC group.

CONCLUSIONS Despite lower clinical success rates due to comorbid hepatobiliary malignancy, ERCP in patients with decompensated LC is technically feasible. Because postprocedural cholangitis and pancreatitis are more frequent in patients with decompensated LC, greater procedural precautions are needed in these patients.