In May 2011 a letter from the Vice-Rector for Personnel of a reputable university was sent to the editor mentioning that two articles published in the journal contained two statements not supported by documented evidence. The two statements related to: (1) approval of the local ethics committee and (2) representation of the experimental evidence.
With regard to point (1), the authors stated in the article that they had approval in 1995 for their research protocols but the authorities state that there is no written documentation of this agreement and that this cannot substitute for formal approval of the research. The journal and the university rules indicate that formal approval of an ethics committee is required.
With regard to point (2), a statement in both articles cannot be sustained for one of three patients in one of the articles and for one case in the second article. Laboratory analysis revealed contradictory evidence from the authors’ statements in the articles. The authors gave three reasons why they ignored this information. Unfortunately, the samples kept in the authors’ laboratory were destroyed in a fire.
According to the letter from the academic authorities, “the authors have been kept informed of these facts, which are in breach of the rules of good scientific conduct”.
One of the articles is coauthored by three colleagues from another university. They have asked the journal that their names be removed.
The author of the articles, who received a copy of the letter from the Vice-Rector, asked to have some time to send in his rebuttal of the accusations. For both issues the answers provided by the authors were submitted to the university and were judged as unsatisfactory.
Long discussions within the publications committee of the journal with representatives of the publisher and the scientific society led to the decision that an “expression of concern” should be published. Prior to publication, the expression of concern was sent to the authors and the university for their comments. Just before the deadline, a letter arrived from the university (signed by the Vice-Rector and the Rector). The conclusion of the letter was that the university believed an expression of concern was not needed. The university believed that the authors recognized that they made mistakes in relation to both issues but since they acted “in good faith” the university had closed the case and did not consider an expression of concern appropriate.
So the journal was faced with an author admitting two serious “mistakes” in two articles. The institution that originally raised the concerns backed off in the end. After consulting with the editorial team I wrote to the author asking him to send a letter to the editor signed by all authors correcting the serious mistakes in the literature. A confidential draft letter was received from the author, and edited and completed by the editor so that both issues were mentioned. This letter was signed by all authors at this author’s institution. Three authors at a different institution refused to sign the letter as they believed that the letter to the editor did not clarify the situation. These three authors confirm their initial position and encourage the other authors to retract both articles. A copy of the email correspondence between the author and a spokesman for the other institution indicates that the author does not want to do this.
A possible conclusion would be to publish the letter to the editor signed by the authors from the institution of the first author as well as a letter to the editor from the three authors from the other institution. This would be accompanied by an expression of concern or an editorial by the editor, highlighting the necessity of proper ethics approval and reporting all experimental data.
An additional question to COPE: should other editors be informed of this? In a sister journal, an article was submitted mentioning the same very outdated ethics approval.
The Forum agreed that this was a very interesting but complicated case. There are two issues here: (1) approval of the local ethics committee and (2) representation of the experimental evidence.
Regarding the first point, the Forum suggested that if the validity of the ethics approval is not in question, then this may not be an issue. The editor does have a right to expect a higher standard from the authors, but they do not seem to have broken the rules as at the time of submission (in 1995), formal approval was not considered mandatory.
In terms of the data, the Forum agreed that the editor needs to decide whether the basic findings of the study are sound. By leaving out some of the data, were the readers mislead? If the data that have been omitted are incidental and do not change the findings of the study, then the advice was for the editor to issue a correction. If however, the editor feels that the study is flawed and the findings were presented in a misleading way, then the article should be retracted.
Some felt that as the authors at the second institute want the article retracted, then the editor should consider retracting it (since these authors no longer stand by the findings). As these three authors believe there are grounds for removing their names because they were unaware of the lack of ethics approval and the omitted data, then the editor should consider this option too.
In the end, it is up to the editor to decide. If s/he decides to issue a correction, s/he could detail in the correction notice which authors were aware of (or responsible for) the errors that occurred. But if the editor has doubts about the underlying science, then s/he should retract the paper.
After the COPE Forum discussion, a decision was taken to correct the literature by publishing two letters to the editor. The first letter from all of the authors recognises the errors made and explains the reason why the omission of the experimental evidence did not put into question the validity of the work. The letter was signed by all of the authors except the three authors from the other institution; they had asked that their names be removed as authors and they explain in a letter to the editor why they requested this. An editorial was written in relation to these two letters, reiterating the facts and insisting that proper ethical approval is required and that all experimental evidence needs to be reported. The conclusion of the editorial is that the journal decided not to retract the paper but to publish a correction. Both letters to the editor and the editorial will be linked to the two articles in the literature.