It was brought to our attention that there is considerable overlap and duplication of data in two papers that a group of authors submitted and that were subsequently published in two different journals.
The control groups are identical in the two papers although it is claimed that they were matched controls.
The data in several columns in the tables are identical; one figure has been reproduced.
The response from the authors was as follows.
Regarding the considerable overlap and duplication of data, I'd like to explain this:
- In both papers the control group is the same.
- One figure has been reproduced in both papers. The aim of including this figure in each paper is different, being in one case to show the difference against one option and in the other case to compare against a different technique, with their respective different discussion, implications and discussion.
- Having this in mind, it is our opinion that this cannot be considered redundant publication as the rationale behind each of the papers aims to address a different hypothesis and the discussion dealing with the explanation of the results is largely different. Moreover, in the results section, representing the bulk of the achievements of our research, only one figure out of six is present in both papers.
- Both papers are clearly different in their scope, bulk of results, discussion and clinical implications.
The Editor of the journal is not happy with this response, feeling that the author has explained why they did it but does not seem to be concerned that they did it. The editor and publisher of the other journal involved have not responded to our emails.
We would like some advice from COPE on where to take this case and whether retraction is fair.
The Forum suggested there are two issues here: (1) possible fraudulent data as it is unlikely that both studies would have the same control group (if they were matched controls); (2) duplicate publication. It is the responsibility of the other journal to act, as they published second. The suggestion was for the editor to contact the second journal and ask the editor what he understands to be the case. It can also be more influential if two editors act together if they need to contact the author’s institution, for example.
It may be that the author needs to be educated regarding the fact this is unacceptable behaviour. The editor should write to the author and ask him to explain why the control groups were the same. The editor should give the author a specific deadline in which to respond and the author should be informed that if a response is not received, the editor will contact the institution. The Forum advised against retraction until the full facts of the case are known.