Effect of Laparoscopic Cholecystectomy After Idiopathic Acute Pancreatitis May Be Overestimated
Being a part of the original study group of the recent article by Räty et al,1 I am very concerned about the validity of the results as there are a few serious methodological flaws. First of all, the study was interrupted after unplanned interim analysis. Such policy may lead to large overestimation of treatment benefit.2 Secondly, according to the study protocol, all randomized patients should have had 2 years follow-up for recurrences. However, after discontinuation of randomization, this 2-year follow-up was not required for the analysis resulting in 13 (15%) patients with incomplete follow-up in the article. Statistical analysis presented in the article (and used in the interim analysis) was based on the wrong statistical testing (Fisher exact test), which does not take into account differences in follow-up time. As an example, during the preparation of the manuscript in December 2013, I noticed that 1 patient randomized to laparoscopic cholecystectomy group in May 2012 in Helsinki University Hospital had developed recurrent pancreatitis within 2 years from randomization, but after interim analysis was done. This patient is included in the article, but the recurrence is not reported, although I suggested that it should be. Due to the small number of events, even 1 recurrent case in the laparoscopic group could have changed the results of statistical tests significantly. In order to avoid these kinds of problems in reporting randomized trials, authors and journals should follow Consolidated Standards of Reporting Trials (CONSORT) checklist and guidelines more strictly.