Restarting anticoagulation at discharge after GIB was associated with fewer thromboembolic events without a significantly increased risk of recurrent GIB at 90 days. The benefits of continuing anticoagulation at discharge may outweigh the risks of recurrent GIB.

No abstract

Acute pancreatitis remains the most common complication of endoscopic retrograde cholangiopancreatography (ERCP). It is reported to occur in 2–10% of unselected patient samples and up to 40% of high-risk patients. The purpose of this article is to review the evidence behind the known risk factors for post-ERCP pancreatitis, as well as the technical and medical approaches developed to prevent it. There have been many advances in identifying the causes of this condition. Based on this knowledge, a variety of preventive strategies have been developed and studied. The approach to prevention begins with careful patient selection and performing ERCP for specific indications, while considering alternative diagnostic modalities when appropriate. Patients should also be classified by high-risk factors such as young age, female sex, suspected sphincter of Oddi dysfunction, a history of post-ERCP pancreatitis, and normal serum bilirubin, all of which have been identified in numerous research studies. The pathways of injury that are believed to cause post-ERCP pancreatitis eventually lead to the common endpoint of inflammation, and these individual steps can be targeted for preventive therapies through procedural techniques and medical management. This includes the use of a guide wire for cannulation, minimizing the number of cannulation attempts, avoiding contrast injections or trauma to the pancreatic duct, and placement of a temporary pancreatic duct stent in high-risk patients. Administration of rectal non-steroidal anti-inflammatory agents (NSAIDs) in high-risk patients is the proven pharmacological measure for prevention of post-ERCP pancreatitis. The evidence for or against numerous other attempted therapies is still unclear, and ongoing investigation is required.

No abstract