Obligation to tell journal of flaws in paper previously reviewed?

Question

The blog Retraction Watch posted an interesting question the other day about reviewing the same paper twice, rejecting it both times, and then later finding out it was published in a different journal. The reason for rejection was that the author(s) failed to mention previous work using the same dataset; seemingly a case of very similar publications (or piecemeal publication, or borderline plagiarism, however you want to describe it).

So what is an appropriate course of action for the reviewer in this circumstance?

Answer 1

Now that the paper has been published, your role is not that of a reviewer anymore, but is the same as any reader. Anyone might read the paper and draw the same conclusions as you.

As such, you can act just as any reader would act, without having to disclose your earlier role as a reviewer. It seems to me that you then have a choice of three options:

1. Write to the authors
2. Write a comment to their paper
3. Write to the editor, mentioning your doubts about the adequacy of references to the existing literature (including previous studies by the authors)

In your particular case, the only additional information that you (as a former reviewer) have is that the authors have been journal-shopping for their paper, and do not want to make the necessary modifications to it. This means that solution #1 is probably not going to be productive, because the authors acted in bad faith. So, you're left with #2 and #3.

If you choose to write a public comment, you definitely cannot say that you were a reviewer for a previous version of the journal. On the other hand, if you write to the editor, I think you could reasonably give him that information (“I was a reviewer of this particular paper for another journal”), because (i) it may change his point of view of the author's honesty, (ii) he will treat it as confidential information.

Answer 2

I would suggest contacting the editors of the journal where it was published and pointing out the problem. However, the problem of salami slicing (to chose one expression) publications are not necessarily grounds for retraction so not much may happen. What will at least be achieved by such a contact is to make editors aware of the behaviour, which may be a very small victory in the battle against such publications.

In cases like this it also pays to look at cases brought up by the Committee on Publication Ethics (COPE). I found one case on An attempt to publish data already published elsewhere which may be interesting to read. Each individual case must be weighed against these COPE cases and evaluated separately. In the linked case, retraction of both papers was in fact the resolution. It may be useful to point the editors towards the COPE site and any COPE case that might be applicable.

I handled a case where someone published a paper in what they thought was in-house "publication" ("" indicating the didn't see it as published). The submitted paper was more or less a copy of the other with another years worth of data added. The problem was that the in-house publication had an ISSN number and had to be counted as a real publication. We rejected the paper much to the dismay of the first author. The point is that some people try to publish and from pressure or ignorance end up in these bad situations.