LAUSR

GOALS We aimed to assess outcomes of patients with liver cirrhosis who underwent therapeutic or diagnostic endoscopic retrograde cholangiopancreatography (ERCP) to determine whether these patients had different outcomes relative to patients without cirrhosis. BACKGROUND ERCP is an important procedure for treatment of biliary and pancreatic disease. However, ERCP is relatively technically difficult to perform when compared with procedures such as esophagogastroduodenoscopy or colonoscopy. Little is known about how ERCP use affects patients with liver cirrhosis. STUDY Using patient records from the National Inpatient Sample (NIS) database, we identified adult patients who underwent ERCP between 2009 and 2014 using International Classification of Disease, Ninth Revision coding and stratified data into 2 groups: patients with liver cirrhosis and those without liver cirrhosis. We compared baseline characteristics and multiple outcomes between groups and compared outcomes of diagnostic versus therapeutic ERCP in patients with cirrhosis. A multivariate regression model was used to estimate the association of cirrhosis with ERCP outcomes. RESULTS A total of 1,038,258 hospitalizations of patients who underwent ERCP between 2009 and 2014 were identified, of which 31,294 had cirrhosis and 994,681 did not have cirrhosis. Of the patients with cirrhosis, 21,835 (69.8%) received therapeutic ERCP and 9459 (30.2%) received diagnostic ERCP. Patients with cirrhosis had more ERCP-associated hemorrhages (2.5% vs. 1.2%; P<0.0001) compared with noncirrhosis patients but had lower incidence of perforations (0.1% vs. 0.2%; P<0.0001) and post-ERCP pancreatitis (8.6% vs. 7%; P<0.0001). Cholecystitis was the same between groups (2.3% vs. 2.3%; P<0.0001). In patients with cirrhosis, those who received therapeutic ERCP had higher post-ERCP pancreatitis (7.9% vs. 5.1%; P<0.0001) and ERCP-associated hemorrhage (2.7% vs. 2.1%; P<0.0001) but lower incidences of perforation and cholecystitis (0.1% vs. 0.3%; P<0.0001) and cholecystitis (1.9 vs. 3.1%; P<0.0001) compared with those who received diagnostic ERCP. CONCLUSIONS Use of therapeutic ERCP in patients with liver cirrhosis may lead to higher risk of complications such as pancreatitis and postprocedure hemorrhage, whereas diagnostic ERCP may increase the risk of pancreatitis and cholecystitis in patients with cirrhosis. Comorbidities in cirrhosis patients may increase the risk of post-ERCP complications and mortality; therefore, use of ERCP in cirrhosis patients should be carefully considered, and further studies on this patient population are needed.