The Responsibilities of Authorship

Occasional events observed by the editor during JAMIA's manuscript review process indicate that potential authors should become more familiar with their responsibilities to ensure the integrity of peer-reviewed publications. This responsibility is critical in communicating new findings and developments to the scientific community.

Authorship is a responsibility, not an honor or acknowledgment. Huth has outlined five principles for determining authorship.¹ The key idea is that each author should be in a position to take public responsibility for the content of the publication. To this end, Huth requires each author to participate in each of three steps: design or performance of the work; drafting or revision of the manuscript; and approval of the final version prior to publication. Contribution short of involvement in those three steps must be acknowledged but does not constitute authorship.

Duplicative, or redundant, publication is both wasteful and misleading.² The instructions for authors included in the January issue of the journal state that “authors should not submit material that substantively duplicates content previously published or in press.” Direct re-publication is a clear-cut violation of policy. The appropriateness of dividing reports from one project into multiple manuscripts is more subjective. A good rule is that a project should be reported in a single manuscript whenever “a single paper would be more cohesive and more informative than two, without being excessively long.”³

Regardless of how authors decide to best present their work, they must disclose any potentially redundant or related fragmented publications to the editors at the time the work is submitted for consideration. The journal's instructions for authors require that “any possibly duplicative published material” be included with a manuscript to expedite determination of the degree to which overlap exists. The Annals of the Rheumatic Diseases has taken a hard-line position against “the misconduct of redundant publication.”⁴ Its editors have outlined a process for investigating suspected duplicate or redundant publication. Upon confirmation of abuse of the policy, they notify the home institution of each author, and they refuse to review a paper by any of the authors for 5 years.

The purpose of a manuscript that reports an original investigation is to communicate to other interested parties. Readers must understand the context of the work before they can apply it to their situations. References to prior work and parallel efforts, together with a discussion of the way in which the reported work builds on or extends other work, is a critical piece of that context. I would suggest avoiding claims of precedence. A variation on claiming precedence is the practice of citing references predominantly from the authors' own laboratory when similar work has been done elsewhere. Let historians judge whose work is important. In large part they will make that judgement based upon the rate at which work is built upon or re-used by others. Work that is presented in the correct context is more likely to be re-used than work that is presented with a bias.

The same principle applies to the investigators' statements about the importance of a finding. Equal attention should be given to identifying each factor that may limit application or generalization of a finding. A finding with limited application can be re-used effectively if those limits are understood. Without that understanding, attempts at application of the finding will fail, and the finding will fall from sight.

The health professions should be able to look to the field of medical informatics to set the standard with regard to practices that increase the integrity and efficiency of information communication. As a minimum, we should adopt Huth's principles of author-ship and the policy developed by the Annals of Internal Medicine regarding redundant publication. I would recommend that we extend the concept of author accountability to the areas of reference to prior work and discussion of limitations. I would also suggest that we require full disclosure of the history of prior submissions of a manuscript and the reports from any prior reviews. Such practices would emphasize the author's responsibility to make the message non-ambiguous through clarification and revision.

Issues surrounding the integrity of our scientific communication are paramount. At the same time, we do not want to hold back ideas or have ideas blocked by differences of opinion. I would like to see these questions and trade-offs discussed by the Association's Publications and Ethics Committees. In the meantime, each of us should make an extra effort to set an example by acting responsibly.