Introduction Needleknife assisted cannulation has been shown to be effective in ERCPs where standard techniques have failed. Concerns regarding risk of complication, particularly pancreatitis and perforation, have led to it being used only as a last resort. We evaluated the current practice and safety profile in a regional tertiary referral centre.

Methods We performed a prospective observational study of ERCP outcomes in patients with intact ampullae. Experienced endoscopists with HPB expertise performed all ERCPs, and were asked to follow their standard practice. We used three ampulla classifications; non-prominent, prominent and distorted by tumour. The number of attempts at cannulation was recorded, as were the techniques used. Primary outcome measures were cannulation success and complication rates.

Results Over a period of 8 months, 222 procedures were performed on patients who had not had previous ERCP. Successful cannulation in this group was achieved in 91.7%. Needleknife assisted cannulation was performed in 37 cases (17%). All needleknife cuts were started at the ampullary orifice.

Needleknife use varied between different ampulla types (p = 0.44). Needleknife cannulation was most frequently attempted in ampullae involving tumour (33% attempted) but often unsuccessful (60% failure), compared to non-distorted (16.7% attempted, 22% failure rate).

There was a wide range in the number of cannulation attempts made in both the needleknife and non needleknife groups (range 1–25) but there was a significant difference between the number of cannulation attempts in the standard cannulation and needleknife groups (p < 0.001). Despite this, there was no difference in the complication rate between standard cannulation and needleknife groups 5.6% v 7.3% (p = 0.522).

Conclusion Needleknife assisted cannulation is more likely to be used where the ampulla is involved with tumour and where ERCP is indicated for malignant disease. However, in this context, needleknife assisted cannulation is more likely to fail.

Reassuringly despite being used after failed attempts at cannulation using standard techniques, the complication rate for needleknife-assisted cannulation is not statistically different.

The likelihood of progression to needleknife use may be predicted by ERCP indication and ampullary characteristics. This may facilitate consideration of an early conversion to needleknife-assisted cannulation, but also early abandonment of procedure for alternative methods (percutaneous or surgical) in these groups.

Disclosure of Interest None Declared