Our journal was contacted by a representative of a company following acceptance of a manuscript that was based on a clinical study sponsored by that company. Upon acceptance, the senior author had forwarded a copy of the manuscript to the company, who had identified some discrepancies between the data presented in the article and an initial report that had been presented to them while the study was still underway. They stated that these discrepancies had been brought to the authors' attention but that a satisfactory resolution had not yet been reached and that they had requested a full independent audit.
The authors, who were copied on this correspondence, were asked by the journal to clarify the discrepancies. In a letter to the company and the journal, the authors responded to each of the concerns that had been raised. They stated that the discrepancies resulted to some degree from the use of a different analytic technique in the submitted article, citing the CONSORT statement as evidence that this technique has become the favoured form of analysis since the initial report, which was presented almost 15 years ago. They also acknowledged that one of the numbers presented in the initial report was inaccurate and that they had redone their analysis to ensure accuracy in their results. A revised manuscript was included with this letter.
The company responded that they were not satisfied with the authors' responses and repeated their request for an independent audit. The authors responded that they did not have the resources for an independent audit but would conduct an extensive internal one. The associate editor who had accepted the manuscript and a representative of the editorial board were asked to assess the company's claims and the authors' responses and they felt that the authors had satisfactorily answered the questions that had been raised.
The company contacted the ethics committee of the authors' institution and reported their concerns. The institution's CEO reported that the ethics committee, as well as the institution's board, had investigated the situation and found that while errors had occurred, they did not constitute research misconduct and that they did not see this as an ethics issue. They said that the decision regarding publication should be left to the discretion of the journal's editors.
Our publisher's legal team was consulted, and they felt that if the journal editors were satisfied with the authors' responses and the relevant changes to the manuscript, then they did not see any reason not to publish the paper. They also advised that our society's board of directors should review the situation and recommended that the case be submitted to COPE, a recommendation endorsed by the society's board.
The company continues to maintain that an independent audit is needed. The authors contend that the company is seeking to delay or stop publication of the article, perhaps because of its implications for a line of products produced by the company.
The Forum asked whether the company had any contract with the authors requiring them to have company approval before publication. The editor told the Forum that the only contract that was found was a 15 year old letter about the authors promising to produce results.
As the institutions found no evidence of misconduct, and the editor is satisfied with the paper, most agreed that the paper should be published but that the editor should perhaps consult the journal’s legal team and the society's board. All agreed the editor has done all the right things and ultimately it is his decision whether or not to publish. An editor does have a duty to publish good research.
The Forum thought it might be helpful to consider whether the editor is reacting differently because a commercial company is involved. Also, the editor should be prepared for any negative publicity or even legal ramifications if he publishes the paper. How will the editor handle letters from the company criticising the paper? Will he publish them?
Another suggestion was for the editor to write a commentary or editorial outlining the issues involved in this case.
Following review of the Forum discussion, the society's board of directors agreed to proceed with publication of the article. Rather than drafting a full editorial to explain the controversy, a footnote was added to the manuscript to explain the unusual delay between acceptance of the manuscript and publication. The authors were notified of this, and agreed to the proposed language for the footnote. The company that had raised the concerns was also notified that publication would proceed and provided with the relevant deadlines should they wish to submit an official correspondence. The correspondence arrived within days and it and a response from the authors were both published in the issue that followed the one with the article itself. Because the letters are now linked to the article and its entry in most citation databases, the company's concerns and the authors' response are now entered into the public record. The society and editors are grateful to the Forum for its advice; being able to say that an independent body had reviewed the situation was beneficial.