First published November 3, 2008 - More info
Unfortunately, we seem to run article amendments (corrections, errata, retractions, addenda) in every issue these days. In the current issue, we have a correction and a retraction — both coming after intensive investigations and peculiar situations we hadn’t encountered before.
We are saddened to run another retraction in this issue, especially as the article (1) has only been in print for one month. Soon after the paper was posted online, one of the authors listed on the paper e-mailed our office, stating that he was surprised to see his name listed as a coauthor. He said he had been unaware of the preparation, submission, and acceptance of the article. He had provided an unpublished (at the time of submission) mouse with a floxed GSK-3β gene to the senior author’s colleague for an unrelated set of experiments and had not specifically authorized the current set of experiments. He also stated that he had not signed an authorship agreement form, one of our prerequisites for publication.
The JCI authorship agreement form is clear about the need for each author to sign it, as it stipulates specific criteria and responsibility for authorship and is meant to ensure that the authors are not infringing on anyone else’s proprietary rights. We queried the senior author of the study, Andrew Leask, after verifying that the signature provided on our form was inauthentic. Leask replied that in order to meet our production deadline, he signed the form for his coauthor. As the coauthor denies knowledge of the manuscript or consent to either its submission to or publication in the JCI, Leask has agreed to retract the article.
We issue this retraction with regret, knowing that the other coauthors, most likely postdoctoral fellows or students, did years of work and were rewarded with a high-profile publication that now ceases to exist. This is particularly a pity as the data themselves are not under question, but we cannot continue to endorse them. There are also the members of the lab who created the mice to consider — what if they wanted to do the same crosses and investigate the outcomes? Will a retracted paper reflect on their future work?
Our reason for retraction of the manuscript is based solely on the unauthorized signature on the authorship agreement form, but there are other issues at play here. Some journals only require the signature of the senior/corresponding author on their copyright and agreement forms. Are they adequately protecting themselves and the other authors? We suspect that many senior authors, with coauthors’ verbal assent, have signed for their colleagues when those individuals were not readily available. Is verbal agreement enough? Should e-mails or an online verification system be employed? Most vexingly, how should authors deal with a theoretical situation in which a coauthor either is unavailable for an extended period of time (e.g., due to illness) or withholds approval? Removing an author who has contributed key data does not seem to be the answer, but what is?
This retraction isn’t the only authorship issue we’ve encountered lately. We received a letter, after publication of an article, from a researcher who claims he was not listed as a coauthor, despite being substantially involved in the design and discussion of the study and even in the execution of some of the key preliminary experiments. We cannot get involved in the claims of authorship and have referred this to the senior author’s institution; perhaps a correction may be forthcoming, but this case underscores our point from a different perspective — the need for open, clear communication among collaborators.
On the topic of authors, yet another recent matter led to one of the most spirited Editorial Board meetings we have yet had. We received a controversial manuscript for review, with the source of the debate being that some of the authors’ previous publications had been proven wrong — or, at least, others in the field had not been able to replicate their findings. The Editors had a long discussion about this issue — should we hold these authors to a higher burden of proof? Should we be suspicious of their data? Is it fair to ask more than three (our default number of referees) experts to evaluate it? Should those who claim to have been unable to replicate the data be particularly sought or instead specifically excluded as reviewers? In the end, we chose from a list of referees who were agreed by all to be above the fray, and we will abide by their recommendations.
In addition to the retraction this month, we are issuing a correction of an article from 2002 (2) about which there was originally some concern (3). After a thorough investigation into whether duplication of some panels in the published article was a deliberate falsification, the authors’ institution has determined that the authors made an honest error — a panel that was duplicated in our version was correct in the first author’s PhD thesis and is now being replaced; some omitted wording in a figure legend has now been added to clarify further duplication of panels later in the article. Another paper, however, from the same set of authors is likely to be retracted from Blood after the investigating committee found evidence that indicated falsification of figures (4). But in the same vein as in the previous case, we felt that the current data in question are the only set that is relevant — a cloud of suspicion over other or previous works shouldn’t unduly influence our judgment.
The JCI may not always have the perfect solution to these problems, nor may we always be able to prevent them, but our policy is to have open discussion to promote fairness, both in our actions and those of our authors.