Attempt to supress legitimate scientific results
Case text (Anonymised)
The journal is operated by institute A, and the editor is an employee of institute A. A manuscript was submitted late in 2014 by authors from institute B, a similar type of organisation in the same country. The manuscript was reviewed by two referees who both recommended publication following minor revision. One of the reviewers noted that the abstract contained a vague statement related to the effectiveness of a treatment for a major type of export from the country but there was no further detail in the main text. In addition, the issue of worker safety was raised in the introduction but not discussed elsewhere. There was also a lack of context for the work and a lack of relevant conclusions. The authors were asked to add more detail on each of these points.
In May 2015, the authors submitted a revised version of their manuscript, which contained more relevant information and references to support their statements. The editor and associate editor considered that the authors had dealt with the issues raised by the referees and accepted the paper. The paper was published on 4 July 2015.
A senior manager at institute B rang the editor a few days after publication. The manger stated that the original manuscript had been approved for submission by institute B but the revised manuscript had not, and asked the editor to remove the paper from the journal’s website immediately. The editor said that this was not possible and further explained that if institute B wanted the paper retracted then they would need to provide a written justification.
Within an hour, the editor received three emails from the manager at institute B. The first one said that the paper “contains errors and speculation which were not able to be addressed at the time of finalising the text, as the paper was not re-submitted to the institute B editorial process following review by the journal” and that institute B believed “the paper is not in the interests of furthering an accurate and complete scientific record of the research in this field and therefore respectfully ask that it is retracted”.
The second email said that the manager at institute B had “since been in contact with the industry group who funded the project and they believe the inaccuracies are ‘significant’. As such, I’d like to emphasis the urgency in our request. I understand that in general journal referees remain anonymous unless they choose to identify themselves. In this instance I’ve been asked if we are able to learn the identity of the referees. Would you please comment on this so that I may respond appropriately?”
The third email said “Since I was in touch with you earlier today, the funding agency has had feedback from the relevant government department who have offered their support with responding to ensure the paper is corrected. I think this indicates the concern that these parties have in ensuring the information is accurate. The errors of fact and subsequent speculation to consequences related to worker safety and fumigation efficacy are problematic.”
The editor gave this matter urgent attention by reviewing the paper and checking the journal’s ethical guidelines (which are those of COPE). They concluded that a retraction was not warranted because the majority of this paper had not changed post revision so the potentially misleading revisions amounted to just a small portion of an otherwise reliable publication. The editor also concluded that, in accordance with COPE guidelines, the authors should submit an erratum detailing the specific passages of text that were incorrect and provide factually correct rewording. The editor sent these conclusions in an email to the manager at institute B the same day and waited for an erratum to be submitted promptly given the apparent seriousness of the situation. The editor did not reveal the names of the referees as the journal operates a closed review process. The editor also alerted the journal’s overseas publisher to this issue in order to fast track the erratum.
Ten days later, no erratum had been submitted to the journal but the editor was copied into an email from the funding agency to the manager at institute B. The agency thanked the manager for drafting an erratum but advised them that the funding agency and government department considered the risk associated with publishing an erratum to be too great so one should not be published. The funding agency also asked that both institutes A and B “manage their processes to ensure that any similar event does not occur in future”.
The manager at institute B forwarded that same email to the editor with a note saying that an erratum would not be submitted. The editor replied expressing surprise that an erratum had been drafted (but not submitted) and that it was not up to the funding agency/government department to decide whether or not an erratum should be published. The editor explained that the authors had an ethical obligation to correct the scientific record if errors existed. However, if there were no factual errors but simply statements that some people happened to disagree with, then no erratum was necessary. The Editor asked the manager at institute B to confirm which of these situations was the correct one and the manager replied stating that there were “no errors in the paper”.
Since then, the editor has been advised by colleagues that the funding agency has been alleging that publication of this paper could harm a key export industry for the country and cause substantial economic losses, and that institute A was at fault by allowing its journal to publish such sensitive work. These allegations are being strongly refuted by institute A. At no point have the authors of the paper communicated with the Editor.
The Editor has submitted this case to highlight concerns that:
• key stakeholders in a published work (but not the actual authors) have attempted to suppress legitimate scientific results because of possible economic and political damage to an export industry;
• this is a serious breach of scientific ethics;
• unfounded allegations have been levelled at the journal’s owner for allowing the paper to be published.
Question(s) for the COPE Forum
- What further steps or alternative actions does COPE recommend be taken?
This is essentially a conflict of interest issue, not with the authors, but with the employers/funders, emphasising the ubiquitous nature of conflicts of interest. The editor proposed that having clear guidelines and examples for similar situations from COPE would be helpful to resist pressure from funders or employers.
The Forum congratulated the editor for standing her ground, and agreed that the editor had done the right thing here in not bowing to pressure from the funders. As a way of avoiding a similar situation in the future, a suggestion was to ask authors to submit a statement on “the role of the funding source” as a way of outlining the role of the funder and clarifying issues such as: did the authors have control of the data at all times, and did the funder have a role in the analysis or submission of the paper? This would also help define the roles of the authors and funders. The authors are the researchers and should be in control of the data. There should not be any pressure from the funding source to try to manipulate the analysis or interpretation of the results or to influence the decision on where to submit the paper for publication. Both parties need to understand their roles.
Some journals ask each author to complete a separate conflict of interest form, and this may be something the editor might consider for the future.
A suggestion from the Forum was to write an editorial, highlighting the issue. This is an important issue and also raises the fact that government or other funders can be as conflicted as private companies, and this is especially true in small countries and for journals that are national journals.
The editor considers the case closed.