Dealing with editorial misconduct

“I am that wicked editor,” announced the email from Richard Smith, then editor of the BMJ, to members of the World Association of Medical Editors (WAME) on their listserve two years ago. An aggrieved author had, at Smith's suggestion, complained to WAME's ethics committee after the BMJ went back on its promise to publish a paper. When the anonymised case was posted on the listserve, opinion from WAME's members was unanimous—the editor in question had behaved wrongly and the journal should honour its commitment to publish. The BMJ did.

This case is important. Although of a different order of magnitude than serious cases of editorial misconduct that have been uncovered over the past 10 years,1 it may be the first example of self regulation by journal editors. An author complained, a body of editors responded, and right—as perceived by those editors—was done.

Editors have traditionally enjoyed power without well codified responsibilities. They decide what gets published and they control an author's right of reply. Much effort has been focused, largely by editors, on what constitutes good and bad behaviour on the part of authors and peer reviewers.2 Far less attention has been paid to the behaviour of editors themselves. Over the years, the International Committee of Medical Journal Editors has included guidance to editors in its Uniform Requirements for BioMedical Publication,2 and in 2001 WAME produced a statement on the responsibilities of editors.3 This week, building on these initiatives, the Committee on Publication Ethics (COPE) launches a new code of conduct for editors (http://www.publicationethics.org.uk/).

The code, initiated by Richard Smith and Michael Farthing, former editor of Gut, comprises a set of standards for good editorial conduct (box).

But the code takes editorial responsibility beyond these—what may be considered basic hygiene measures. It places a new and potentially onerous duty on editors to pursue allegations of research misconduct. Editors are in a unique and difficult position. They are often the first to hear concerns about a paper's integrity. Researchers may contact the journal to say that they should have been, or should not have been, included as authors. A peer reviewer may spot that a paper, or something very similar, has been published by the same authors elsewhere, or that it plagiarises their own or someone else's work, or that the authors have failed to declare important conflicts of interest, or that the data are just too good to be true.

Editors have no mandate, and usually inadequate resources, to investigate such cases themselves. In view of this, they may be tempted to take the route of least resistance, especially when allegations are made in relation to a paper that has landed on their desk but has not yet been published in their journal. A published paper is clearly within the editor's jurisdiction and may be corrected or retracted. But what of allegations about an unpublished paper? Rather than embarking on a potentially troublesome and protracted investigation, an editor may be tempted simply to reject the paper on other grounds. COPE takes the view that this is not acceptable. The code specifically calls upon editors to take seriously their role as guardians of biomedical science and to make all reasonable efforts to ensure that allegations, whether relating to published or unpublished papers, are properly investigated.

Supporting editors in this difficult but crucial aspect of their work is one of COPE's main objectives. It does this by publishing guidance and providing a forum in which editors can discuss individual, anonymised cases. COPE now has 245 member journals and has heard over 200 cases since it was established in 1997.4 The summaries of these cases, along with COPE's recommendations, now form a unique publicly accessible archive on COPE's website (http://www.publicationethics.org.uk/).

COPE's other main objective is to promote the setting up in the United Kingdom of an agency to which allegations of serious research misconduct can be referred. Such agencies already exist in Australia, Denmark, Finland, Norway, and the United States reflecting the seriousness with which those countries take the threat of research misconduct.5 Efforts to establish a similar agency in the United Kingdom are at long last beginning to bear fruit: Universities UK, the Department of Health, and the NHS are now working together on a framework to establish a panel for research integrity. COPE hopes that cases referred to such a body will receive serious attention and prompt action.

The COPE code stops short of compulsion. No sanctions exist for editors who breach the code. We felt that this would be counterproductive and impossible to enforce, given that COPE has no legal jurisdiction. At most, editors who seriously breach the code will be removed from COPE's list of members. But the code does provide a mechanism for people to refer complaints against editors who are members of COPE. Complaints can be referred by readers, authors, peer reviewers, other editors, or publishers, but COPE will only consider complaints that have been through the journal's own complaints procedure—in effect providing its members with a mechanism for dealing with unresolved and difficult cases.

The two professions most closely allied to medical journal editing, medicine and the press, have well established systems for self regulation. But self regulation is a privilege not a right. It brings with it responsibilities to establish and enforce standards of good practice. The COPE code of conduct is a first step.