Introduction Endoscopic retrograde cholangiopancreatography (ERCP) has become a specialised service and over the last few years training has been limtied to tertiary centres. NCEPOD “Scoping out Practice” (2004) highlighted the poor outcomes for therapeutic endoscopy and the modernisation of Endoscopy service has required regular audits of all endoscopy procedures. Locally ERCP has been performed by a single endoscopist for the last 5 years.

Methods A local clinical information system is used to record all ERCP data and six monthly audits are carried out on all cases performed and presented to endoscopy users group and audit meetings.

Results A summary of the 5 years from 2005 to 2009 are summarised below:

Total No=578.

Primary selective cannulation=528 (91.3%).

Indications: Jaundice 47.1%, Pancreatitis 5.2%, Abnormal imaging 18%, Redo (stones and stent change) 27.3%, Other 2.4%.

Findings: Stones 56.2%, Strictures 23.2%, Other 1.9% and Failed 8.7%.

Procedures: Sphintertomy and stone extraction 45% or stent insertion 9.9%, Stent insertion or exchange 21.6%, Other 4.2%.

Complications: None 88.1%, pancreatitis 1.2%, Cholangitis 0.9%, Bleeding 0.9%, Perforation 0.3%, Hypotension 2.2%, and Failed 8.7%.

Mortality: Deaths in 30 days 6.1% but majority due to cancers.

Delays: Decompression for obstructive jaundice in <5 days was 15% Average 2005=11 days, 2006=13 days, 2007=18 days 2008=17 days.

Conclusion ERCP perfomed by a single operator within a hospital site can meet the quality assurance standards except for the number of trained ERCPists. The majority of procedures (>90%) are therapeutic and overall complications are low (5.5%). Mortality at 30 days is over 5% but the majority are related inoperable cancers presenting with obstructive jaundice. Delays due to access of x-ray screening could be improved with imaging facilities in the endoscopy department to provide a more flexible service.