Endoscopic retrograde cholangiopancreatography (ERCP) is one of the most difficult endoscopic procedures to learn. It has a significant complication rate even when undertaken by experienced practitioners. During the procedure, patients have a right to expect a competent practitioner to be in charge as complications may result in cholangitis, bleeding, pancreatitis, duodenal and biliary perforation, and consequent death.1 However, in the United Kingdom, there are no guidelines either for the training required by these doctors or the assessment of proficiency.

This issue has become increasingly important. The use of ERCP has increased with the advent of laparoscopic cholecystectomy2 ,3 and the minimal access management of gallstones.4 Developments such as magnetic resonance imaging (MRI) are likely to reduce the number of purely diagnostic ERCPs needed, leaving only the more difficult therapeutic procedures to be done using this method. It has become increasingly obvious that competence in ERCP reduces complications,5 maximises therapeutic potential and reduces costs incurred in unsuccessful and repeated procedures. Finally, time available for ERCP training within the confines of specialist registrar training schemes has decreased.

In the past, recommendations (mainly from the USA) have emphasised the number of procedures that a trainee needs to master in order to achieve competence. Often, these recommendations represent a pragmatic, consensus view of the number of procedures a trainee could realistically expect to perform under supervision, rather than the number actually needed to achieve a definitive level of success—for example, an 80% cannulation rate of the desired duct.6 A survey of the American College of Physicians suggests that 50 ERCPs are sufficient7 and guidelines from American Society of Gastrointestinal Endoscopy (ASGE) propose that 35 ERCPs are enough.8 In practice, it seems that as the number of endoscopic procedures performed by trainees increase,9 ,10there …