Editors of the main journals in medical education have met informally to discuss issues of common interest and, since 2006, have made a particular point of holding regular meetings in conjunction with the RIME Conference. In 2008, we held an open symposium where we introduced a project to develop joint standards for medical education publication ethics.
One issue that has repeatedly surfaced in nearly all journals is unethical or questionable practices regarding submission. Although flagrant breaches such as plagiarism are uncommon (or undetected), some practices, such as copy of text from one manuscript to another, submission of nearly identical analyses to different journals (“salami slicing”), inadequate attribution of prior work, and inappropriate authorship reporting (honorary authors) appear all too frequently. We believe that most of these breaches are committed in good faith and not malevolence, and simply represent a lack of awareness on the part of authors of accepted standards, and perhaps some legitimate differences in interpreting ethical standards and guidelines.
Medical education editors are aware that general guidelines for scientific and biomedical publishing exist. Some examples are the International Committee of Medical Journal Editors Uniform Requirements (ICJME) Uniform Requirements for Manuscripts Submitted to Biomedical Journals (the Vancouver guidelines),1 the Council of Science Editors White Paper on Promoting Research Integrity in Scientific Publications,2 the Committee on Publication Ethics (COPE),3 and the guidelines published recently by Journal of the American Informatics Association.4 In an attempt to present a common standard for authors publishing in medical education journals, we have decided to create and disseminate common guidelines for publication in medical education journals, for a number of reasons: (1) While some of our journals may officially subscribe to the ICJME guidelines, others do not, and positions may change, (2) conversely, there are some issues, such as the very common practice of using data gathered routinely to monitor student progress, for example, for research purposes that are particularly relevant to our field, and (3) finally, our goal in initiating these discussions was to identify common strategies that the journals in the field may adopt to help protect and serve authors, in a way that may not be achievable for the broader array of journals covered by the existing guidelines.
We intend to disseminate, through the Instructions to Authors of our individual journals and other means, including presentations at international meetings, a set of common guidelines to which all subscribe.
To this end, in February 2008, the editors commissioned Julie Brice, formerly managing editor of Sociology and of Medical Education, and author of six articles5 related to publication ethics, to create a document of common guidelines. A survey was distributed to editors of 22 journals, and eight responded to this first round. The survey asked six questions:
- In what ways has the publication climate changed in recent years? What effect has this had on publication ethics?
- What key issues should be addressed by a consensus statement from medical education editors on standards in publication?
- Is medical education research different from other types of research?
- What are the key constraints on improving standards of publication ethics in medical education?
- What are the key constraints on improving standards of publication ethics in medical education?
- What actions should editors and journal owners take to reassure authors and readers concerning their own conduct?
To expand our survey to include researchers and authors, and to notify the research community of this initiative, a symposium was conducted at the RIME meeting in 2008. We reported on the editor survey and reiterated these questions for the symposium audience (although not all questions were covered). Below is a summary of the discussions:
Effect of Publication Climate Changes on Publication Ethics
Three topics were discussed, with some overlap: (1) plagiarism, (2) salami slicing, and (3) peer review.
There is a general view that editors need to be specific and consistent about what constitutes plagiarism: a sentence, a paragraph, a section from another publication? Although a sentence may get by, all three instances constitute copyright violations. The copyrights to a publication are usually signed off to the publisher (the owner) by the author. Thus, in doing so, authors no longer own the rights to their previously published work, and they cannot “recycle” it. An author copying his or her own work in a subsequent manuscript can be violating copyright.
Should pilot results be submitted for publication, before the story has been fully constructed? Is it too early for submission because the chances of duplication are high? Some journals do not publish pilot studies, but one editor did not see any problem with the publication of pilot results as long as the author is transparent about it and mostly if the publication contributes significantly to the literature and to knowledge in the field.
The method is one of “fair dealing”; provided authors are forthright in revealing previous publications and admitting the limitations of their work when it is part of an ongoing project, editors will be able to make reasonable judgments regarding publication, and the academic records will be protected from unoriginal and repetitive work.
The social sciences have a tradition of long papers in which authors present the theoretical foundation for their research along with results from a number of studies. By contrast, the tradition in medical education and medical sciences in general is to limit papers to 2,500 to 3,000 words. Although there are advantages to this approach, the “shorter” paper format creates a situation in which different bits of data, that are all part of the same study, are presented in multiple papers (salami slicing). This runs the risk of self-plagiarism, especially in the introduction and methods sections. It is the responsibility of the authors to be transparent about reporting the main story. The same applies when collaborative research is conducted and multiple publications ensue. It is the authors’ responsibility to refer to the common, underlying story.
Authors are encouraged to tell the whole story in one publication rather than slicing it into multiple publications. If an author cuts the information to be presented very finely to enable multiple submissions, he or she runs the risk of having subsequent papers, possibly with the most interesting results, rejected because of salami slicing.
Authors are encouraged to be clear about the process and to disclose any related submissions. They should be transparent about previous submissions rather than trying to disguise the multiple submissions. Most important, they should indicate how the current submission connects to what else was done and explain why the material merits reporting in separate papers.
Should editors share authors’ names and manuscript titles to avoid duplicate publication or salami slicing across journals, rather than having these situations identified fortuitously by reviewers? The answer was a resounding “no,” to protect privacy and guard against the risk of decreased chances of having a second (legitimate) submission to a second journal being sent for review.
However, if the journals do not communicate among themselves, determined authors can “divide and conquer” by, for example, making multiple submissions, confident in the knowledge that the individual journals will not know of the duplication. One alternative to a shared editors’ list is to have the submissions registered through a third party that does not communicate with editors except when there are signs of wrongdoing, like duplicate publication or salami slicing across journals. This alternative is akin to procedures used by some universities to monitor student misconduct.
However, if authors are frank in revealing a prior rejection and include the initial reviews and explanations as to how they addressed the initial shortcomings, this may actually increase their chances of having work accepted, since some editors will accept the submission as is without further review, or may even dismiss the initial objections.
Some reviewers need training in providing constructive feedback. Highly negative (or highly positive) reviews are useless if they do not provide constructive comments and specific suggestions for improvement. Editors should screen out the “nasty” or useless reviews. One can be forthright without being nasty.
One view is that reviewers should sign their reviews on the principle that comments that cannot be said to the author’s face should not be written. However, this was countered by others in the audience who expressed the desire from many authors for anonymous submissions and reviews, especially to protect younger researchers from the wrath of senior researchers and interpersonal rivalries.6 There is no uniformity in editorial policy; Medical Education has a policy of blinding reviewers to the identity of authors; AHSE does not use anonymized manuscripts; both leave disclosure of the reviewer’s identity to the reviewer’s discretion. Other journals, such as BMJ, require reviewers to sign reports.
Is Medical Education Research Similar to or Different from Other Research?
The question is important because it underlies the more basic question: Do medical education’s particular characteristics mean that it needs different publication standards than other disciplines?
The general view is that differences are not so much across disciplines but across countries and cultures. For example, the Netherlands does not have rule-based ethics review for education projects. However, medical education journals could have a philosophy-based approach to ethical conduct and require that authors show how they addressed the spirit of protecting subjects and how they articulated any issues of risk. It is not the role of the journal to define or police ethical standards but, rather, to be clear on what statement is required from the authors about ethical requirements in their country of origin. In other words, authors should be encouraged to be transparent about issues of research ethics.7,8
One difficulty that was raised is the issue of accommodating authors for whom English is a second language, or authors from countries with different standards than those generally adhered to by the medical education journals. Ideally, reviewers and/or editors might show some understanding, even help, in dealing with such reports’ need to follow the rules and conditions for this type of reporting.
Because the subjects (participants) in medical education studies are more likely to be students and faculty than patients, review boards and ethics committees frequently find it difficult to handle requests for review from the medical education community. Workshops or sensitivity training would be useful at the medical center or university levels to help everyone understand that students and faculty can be at risk of harm in education projects, like privacy breaches, poor study design, coercion, and stress.9
What constitutes authorship? This issue is often contentious, and it is a particular problem in medical education because of the strongly collaborative nature of much of the work. Unless resolved early in the publication process, it can cause real conflict and affect publication. “Guest” and “honorary” authorship is frowned on, but difficult to prevent. One solution adopted by Medical Education is to explicitly describe conditions for authorship and have each author sign a statement indicating his or her level of involvement related to each criterion for authorship before the manuscript is accepted for review. The Vancouver guidelines1 also make specific reference to criteria for authorship, and Academic Medicine asks the corresponding author to certify that all authors meet these guidelines.
Raising Standards and Using Common Standards: What Else Should Editors Do to Raise Standards?
Editor misconduct can sometimes be a problem. What happens when there is a change of editorship and the new editor does not honor prior commitments made by the outgoing editor? How do authors complain about slow or capricious decisions? Where can the author report this type of conduct? How can authors get support? In cases in which authors feel they have been treated unfairly, they may appeal to the editorial board or to the journal’s society governing body (if there is one). Two other mechanisms are available: (1) If the journal is a member of COPE, the author could send a complaint to the forum for investigation (see www.publicationethics.org), and (2) if the journal has a mechanism independent of the editor, the author may address the complaint at that level (e.g., Medical Education has a Quality and Standards Advisory Group to whom authors can directly address such complaints).10
Many, and possibly most, researchers in medical education are unaware of the type and frequency of ethical misconduct reported by editors. It was suggested that journals should make public the list of possible types of misconduct along with the consequences related to such misconduct (e.g., in the case of plagiarism, the author’s institutional official may be informed of the situation, as required by the Vancouver agreement; being blacklisted by a journal). This should serve as a deterrent to potential offenders. As for possible gray-zone misconduct, use of journals or meetings is one avenue for discussing possible violations to better sensitize authors. Authors should also be encouraged to diffuse and discuss case studies of publication misconduct within their institutions (e.g., as part of grand rounds, research seminars, or collaborative project meetings).
It is heartening to see that the meeting was well attended, and much debate has been stimulated at all levels in the community. The general view of the editors is that the most appropriate strategy is to provide preventive measures that should help decrease the need for policing.
Julie Brice is continuing to take leadership on this initiative. Discussions and consultations continue around this important issue. When this data-gathering step is complete, a report will be produced, which we anticipate will constitute a “joint statement” on publishing ethics in medical education. This will take the form of a set of principles for medical education publication that will support the journals’ guidelines to authors. The joint statement will be accessible from all journal Web sites. Journals will be encouraged to refer to the statement in their author guidelines.
It is uniformly believed among the authors of this paper and seemingly among the participants of this symposium that issues of publication ethics breaches are more often attributable to ignorance than malfeasance. As such, the key message for everyone is one of transparency, and a key task for all of us is to educate ourselves and the medical education community about the latest thinking in publication ethics. Each editor takes it as his or her responsibility to work with authors to find the most appropriate solutions to problems that reside in the aforementioned “gray zone.” Such issues are much more easily resolved with upfront discussion than by trying to dissect after the fact what needs to be done in response to critical incidents. The journals desire a fair and just process for all authors, just as we hope authors will engage in their publication activities in a fair and just manner rather than falling prey to the temptation to “game the system.”