Reply to the letter of Park et al.

We would like to thank Dr. Park and colleagues for their interest in our study [1]. The authors acknowledge the limitations of measuring post-ERCP pancreatitis (PEP) in a crossover design, especially for difficult cannulation where crossover is more likely. We believe however, that our study design more closely replicates the realities of everyday ERCP practice by virtue of the involvement of multiple operators and the crossover protocol. Operator dependence is well recognized as an important factor for success and complication rates. Caution should be exercised before generalizing studies involving a single operator [2] [3] to wider practice. Such studies, by their nature, introduce an immediate bias in favor of a particular technique; usually the preferred technique of the single operator. Multicenter randomized controlled studies remain the gold standard and single-center studies with multiple operators are second best.

In our study there were 90 patients who proceeded to another technique: 72 with crossover to the alternative technique and 18 who proceeded directly to needle-knife sphincterotomy. The incidence of pancreatitis for the 323 patients in whom cannulation was successful with the randomized technique without crossover was 10/167 (6.0 %) in the guide-wire arm and 7/156 (4.5 %) in the contrast arm. This type of analysis excludes those patients with difficult cannulation and is thus less informative.

In the absence of successful cannulation the indication for ERCP cannot be addressed as a risk factor for PEP. Our crossover study design allowed for the use of both techniques if necessary to achieve this goal. We would suggest that there are some patients who are better suited to contrast-assisted cannulation, and others where a guide wire is the best method to maximise success and minimize complications. We have demonstrated that the risk of PEP increases with the number of attempts at the papilla. Thus persistence with a technique that is not successful appears to increase the risk of PEP. In our study [1], the finding that seven patients from the guide wire arm developed PEP without any contrast injection demonstrates that use of a guide wire does not prevent PEP. Clearly there are other factors involved. Perhaps the challenge for those who work in this area is to identify which patients are best suited to a particular cannulation technique, and select those at high risk of PEP for pancreatic stent placement.

Competing interests: None

References

• 1 Bailey A A, Bourke M J, Williams S J. et al . A prospective randomized trial of cannulation technique in ERCP: effects on technical success and post-ERCP pancreatitis.  Endoscopy. 2008;  40 296-301
• 2 Lella F, Bagnolo F, Colombo E. et al . A simple way of avoiding post-ERCP pancreatitis.  Gastrointest Endosc. 2004;  59 830-834
• 3 Artifon E, Sakai P, Cunha J. et al . Guidewire cannulation reduces risk of post-ERCP pancreatitis and facilitates bile duct cannulation.  Am J Gastroenterol. 2007;  102 2147-2153