Is effective pharmacological prevention of post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis impossible, uncertain or easy? Just 2 years ago I wrote an editorial on this topic and concluded at the time that the use of drugs to reduce the risk of post-ERCP pancreatitis was depressingly uncertain, and pancreatic stenting appeared the best available way to reduce—though not abolish—this complication.1 This conclusion was based on a literature overview evaluating 39 reports published between 2000 and 2005, including the only two available meta-analyses.

About 40% of clinical trials proved some efficacy of the drug tested; meta-analysis showed that octreotide did not prevent post-ERCP pancreatitis,2 whereas both somatostatin and gabexate appeared significantly associated with a reduction.3 However, the same group carried out a subsequent evidence-based medicine study 2 years later, starting from an updated meta-analysis they themselves had done on somatostatin and gabexate in patients at standard and high risk for post-ERCP pancreatitis, and concluded there was no evidence of effective prophylactic use of either drug.4 5

The main issue concerning pharmacological prophylaxis was that a drug that had been proved effective in a few trials on a limited number of patients gave largely disappointing results over the long term when it was adopted in routine clinical practice. This is very probably the consequence of the most important bias of most published trials: the sample size. Because of the low incidence of post-ERCP pancreatitis (overall mean rate 5%; low-risk subjects 2%), large numbers of patients are required …