Author disagreement regarding article corrections

Case number:
15-08

Case text (Anonymised)

We received an original article which was accepted and published. The article was written by multiple authors from several centres, and the corresponding author undertook the task of standardising the content, making several corrections to the original text. The author proofs were sent to the corresponding author, who reviewed them.

However, once published, one of the co-authors indicated his disagreement with the changes that had been made by the corresponding author on his part of the article. This author has asked us to publish an erratum to include, under “Conflicts of interest”, his disagreement with the final version, because he wants to respect the original text (the author already expressed their disagreement before the article was published, but we did not know this until now).

The corresponding author disagrees with the addition to the conflict of interest statement because there is no conflict of interest, only a difference of opinion. Moreover, he argues that, as scientists, we must be aware that all of our work has limitations, and recognize that these limitations are part of what drives scientific progress.

Question(s) for the COPE Forum

  • What is the procedure we should follow in this case?
Advice: 

The Forum agreed with the editor that these are not conflicts of interest, but differences in opinion between the authors. A conflict of interest statement would not be appropriate in this instance. The Forum suggested that the journal could ask the disgruntled co-author to submit a letter to the editor for publication in the journal, outlining his concerns. Raising these concerns in public could be a way of resolving this issue. It is important that the letter is linked to the original article, so that the two items are permanently linked. If the journal does not have a mechanism for publishing letters to the editor, the journal could use PubMed Commons which enables authors to share opinions about publications in PubMed. In extreme cases, it has been known to have two different discussion sections in the same paper.

A way to avoid a similar situation in the future would be to ask the authors to provide a contributorship statement, which outlines the contribution of each author to the study. The statement could also include agreement on the final version, so that all authors confirm that they agree with the final version for publication.

Follow up: 

The journal decided to invite the disgruntled co-author to submit a letter to the editor and to ask the other co-authors to reply. They have yet to receive the letter.

Follow up (February 2016):
The disgruntled co-author submitted a letter to the editor. The corresponding author of the original article was given the opportunity to reply. Both letters were published and permanently linked to the original article.

Resolution: 
Case Closed
Year: 

Comments

  • Posted by Mike Curtis, 2/9/2015 9.25am

When a paper is submitted or resubmitted it should be part of the process that all authors are required to approve the version as part of the submission process. The Circulation journals do this. If this is not done then it should be stated as part of the submission process that when a individual submits a paper he/she has obtained the written agreement of co-authors. If this is later found not to be the case and a co-author objects to the content, then the authors should be offered the option of publication of an erratum or corrigendum (typically one page in length) with all correspondence countersigned by all authors. If agreement cannot be reached among authors the paper should be retracted by the publisher ('toombstoned' is what Elsevier calls this). I appreciate that it can be argued that letting the submitting author 'speak' on behalf of all authors should not be allowed. But if so it should not be allowed at any stage in the peer review process, not just on initial submission. I have had a case where an author objected after publication - because their name was not on the paper! So getting all authors to sign off every part of the process would bot have prevented this problem, since the missing author was unknown until after publication. I resolved this issue by consulting with the participants.