Salami slicing and heart attacks don’t mix: Duplication, lack of transparency lead to retraction

A group of French cardiology researchers have retracted a study of a potential way to rule out heart attacks, after it became clear they had used data from another study without alerting the journal.

In an unusually forthright letter accompanying the retraction of “Concomitant measurement of copeptin and high-sensitivity troponin for fast and reliable rule out of acute myocardial infarction,” originally published in Intensive Care Medicine, Bruno Riou and colleagues note:

You raised the question of self-plagiarism because we had previously published a paper in the International Journal of Cardiology which tested copeptin/conventional (not high-sensitivity) troponin [2]. In fact, there are two different problems with this manuscript: self-plagiarism and respect of the rules of the peer-review and editorial processes.

Massimo Antonelli, editor in chief of Intensive Care Medicine, tells Retraction Watch:

This was reported by one of our readers who could compare our on line publication ( the article stayed on line for two weeks only and never assigned to an issue) with the companion article published in the International J of Cardiology.

It turns out that the authors had disclosed they were using the same patients they had used in another study, but they were referring to a paper in Critical Care, not the one in the International Journal of Cardiology.

We made an error in that we did not cite the study published in the International Journal of Cardiology.

We believe that the new results obtained with high-sensitivity troponin are important to guide future trials and that this information has not been previously provided.

The authors note, however, that the then-editors of the journal had made it very clear, in a 2007 editorial, when they’d consider a study to be slicing the salami a bit too thinly:

Deciding on when a report is redundant, there is no (and there will be no) definitive cut-off for these boundaries, but the response is called transparency. Transparency is needed with editors and reviewers firstly in the cover letter. They will decide, with the author, whether the novelty and magnitude of new information justifies a new publication. They may also decide with the author what format is the most appropriate: a full article; a brief report; or a simple letter. The question indeed is certainly not to stop publication at any price, but to decide openly. Transparency is then due to readers, and simply quoting a previous publication from the same group is not sufficient.

The authors continue:

Rather than self-plagiarism, the main problem of our manuscript lies here. We must recognize that we failed to provide the reviewers and the Editor-in-Chief with all of the relevant information to enable them to take a decision with transparency. Although this was not intentional, we broke this rule and thus we must ask for retraction of this manuscript. As the senior author of these publications, Bruno Riou take the entire responsibility for this neglect and apologize for the consequences.

Riou is vice president for medicine at the Sorbonne’s Université Pierre et Marie Curie and a senior editor of the journal Anesthesiology.

The authors say they will submit a new version of the manuscript

to Intensive Care Medicine because we are convinced that it contains new and important information. This is the only possibility for us to show that we made an error but in good faith. Whatever the future decision, we will accept it with all humility.

Hat tip: Hervé Maisonneuve

One thought on “Salami slicing and heart attacks don’t mix: Duplication, lack of transparency lead to retraction”

  1. The authors seem to be honest and forthright about their “mistake.” Unlike some others we could mention, such as Anil Potti. I hope they don’t get in too much trouble for this salami slicing…

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.