A high-volume ERCP service led by surgeons is associated with good outcomes and meets national key performance indicators: results from a British district general hospital.

Background: Endoscopic retrograde cholangiopancreatography (ERCP) is a common, but technically challenging procedure used in the management of hepatopancreaticobiliary (HPB) disease. It is traditionally performed by medical gastroenterologists. In 2014, the British Society of Gastroenterology (BSG) proposed key performance indicators to evaluate and set standards of ERCP practice. This study aimed to compare our ERCP outcomes against these targets, in a centre where ERCP is exclusively performed by surgeons.

Methods: A retrospective analysis of all ERCPs undertaken over a 38 months in a District General Hospital in the United Kingdom (UK), by three Upper Gastrointestinal Surgeons. Study outcomes were based upon, and compared against, BSG key performance indicators, including number of ERCPs per annum, proportion of successful cannulations of bile duct and stone clearance, ERCP-specific complications and mortality.

Results: The unit's caseload over this period was 1324, equating to approximately 418 per annum (BSG minimum 200 per unit). Management of bile duct stones was the commonest indication for ERCP. Overall, 95% (1253/1324) of bile ducts were cannulated and 92% (645/698) for those undergoing their first ERCP. Bile duct clearance was achieved in 80% of patients (BSG recommend > 75%) and the successful stenting of extra-hepatic strictures in 94% (BSG recommend > 80%). The overall complication rate was 4.3% (BSG standard < 6%). Procedure-specific mortality was 0.3% (4/1324) where death was either caused by pancreatitis or sepsis.

Conclusion: A high-volume ERCP service led and performed exclusively by surgeons meets all BSG performance indicators, with good procedural and patient outcomes. Formal training pathways should be developed to encourage more surgical centres to provide an ERCP service and deal with what are common surgical pathologies.