INITIATION OF QUALITY ASSURANCE PROGRAM REDUCED ERCP COMPLICATION RATES IN HIGH VOLUME CENTER

Aims:

We initiated a quality assurance program for ERCP to evaluate its feasibility and utility.

Methods:

We prospectively collected data between April 2016-April 2017 about cannulation success and complication rates as widely accepted quality indicators of ERCP. As majority of ERCP complications occur in patients with naive papilla and cannulation rate is most reliable in this settings as well, we considered only patients with naive papilla. After evaluation of these quality parameters an action plan was carried out. After implementation of determined changes the same data were collected during the next 6 months.

Results:

During baseline 12 months we performed 1102 ERCPs, 445 of them with naive papilla. The primary and overall cannulation rates were 88.5% and 96.6%, respectively. The rate of overall complications was 5.7% (3.4% postERCP pancreatitis/PEP/(none severe), 1.6% bleeding requiring transfusion, 0.7% perforation). We had PEP in 1.4% in cases when we accessed biliary duct by conventional method and in 5.5% when we had to use one of the advanced techniques. In latter cases we had PEP in 5.7% and 4.9% without and with the placement of prophylactic pancreatic stent (PPS), respectively. After evaluation these results we replaced our old electrosurgical unit with a newer one, and placed more frequently PPS in the second period after advanced cannulation techniques than in baseline year (27% vs. 18.8%). In the next 6 months we had 246 patients with naive papilla. The primary and overall cannulation success rates were 87% (214/246) and 96.8% (238/246), respectively. The rate of overall complications decreased from 5.6% to 2.4% (2.0% post PEP (none severe), 0.4% bleeding that needed transfusion, no perforation).

Conclusions:

Rates of severe bleeding and perforation were too high at beginning but by newer electrosurgical unit we could decrease them. The PEP rate could also be decreased by placing PPS more frequently.