Reply to Kadayifci and Yildirim

We thank Drs. Kadayifci and Yildirim for their interest in our study [1] and for the opportunity to respond to their observations and insightful comments. In their Letters to the editor, Kadayifci and Yildirim questioned the role of “late” precut in increasing the risk of post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis (PEP), while at the same time acknowledging the protective effect of early precut on PEP shown in our study, an effect already instantiating the main goal of the study.

Kadayifci and Yildirim suggested to focus on two additional subgroups of patients, named the “delayed precut” and “without precut” subgroups. These patients all belong to the group of patients who, in the original study, underwent repeated papillary cannulation attempts followed, in case of failure, by late precut (Group B). Most importantly, however, Kadayifci and Yildirim affirmed that these two (sub-)groups of patients would ostensibly and importantly differ from each other in the risk of developing PEP, a difference they objectified with an odds ratio (OR) of 2.09. However, whereas the original primary outcome reported in our study was supported, of course, by statistically significant results, the OR of 2.09 reported by Kadayifci and Yildirim, is, in our reiterated control calculations, very far from statistical significance. To be sure, the “new” pattern of data suggested by Kadayifci and Yildirim for subgroups leads to an (unreliable) OR of 2.09, which must be paired with a (nonsignificant) P value of 0.20, and a (too-large) 95 % confidence interval ranging from 0.67 (which would qualify the variable under scrutiny as a protective factor for PEP) to 6.45 (which, on the contrary, would qualify the same variable as a risk factor; Z test = 1.28, P = 0.20).

Additional follow-up analyses performed in our laboratory shed further light on the issue, by showing how the suggested pattern of results proves fallacious also from different and equally important analytical angles. A Fisher’s exact test, for instance, leads to P = 0.23, whereas a chi-squared analysis (with Yates’s continuity correction) yields a statistic of 1.12 which, again, signals that there are no reliable differences of any sort between “observed” and “expected” values in the “new” pattern of results for the subgroups suggested by Kadayifci and Yildirim. We are, again, very far from statistical significance (P = 0.29).

On the basis of the above considerations, all firmly grounded in experimental data, sound design and methodology, and proper statistical analyses, we must thus reject in toto the claims advanced by Kadayifci and Yildirim that delayed precut has a more significant effect on PEP risk than repeated attempt itself. On this basis, we must of course also confirm to the readers the conclusion of our study that “repeated biliary cannulation attempts are a real risk factor” (emphasis added) for PEP. At present, data from our study and other published series do not support the suggestion that a second ERCP session might be safer if compared with only one ERCP session with a delayed precut procedure. In the same vein, we must reassert that we did not ignore any of the possible effects of a late precut procedure, as in the original study there were no additional (reliable) effects to be detected and, consequently, to be systematically reported to the reader with any degree of trustworthiness. However, as pointed out in recent European guidelines on papillary cannulation and sphincterotomy techniques at ERCP [2], in patients with difficult biliary cannulation and unintended pancreatic guidewire insertion, as was the case for some of the patients enrolled in our study, transpancreatic biliary sphincterotomy can be a preferable technique to late precut for reducing the papillary trauma.

Finally, we must rebut Kadayifci and Yildirim’s assertion that their newly suggested pattern of results would “contradict” with the discussion and the conclusions drawn by our study because, again, published outcomes cannot be simply questioned and dismissed on the basis of nonsignificant results and unproven effects, however suggestive and appealing they may appear.

• References

• 1 Mariani A, Di Leo M, Giardullo N et al. Early precut sphincterotomy for difficult biliary access to reduce post-ERCP pancreatitis: a randomized trial. Endoscopy 2016; 48: 530-535
• 2 Testoni PA, Mariani A, Aabakken L et al. Papillary cannulation and sphincterotomy techniques at ERCP: European Society of Gastrointestinal Endoscopy (ESGE) Clinical Guideline. Endoscopy 2016; 48: 657-683