CYBERSPACE PUBLISHING AND THE IRP:

'E' IS FOR EXCELLENCE—AS USUAL

Michael Gottesman

Earlier this year, I forwarded to the institute directors, scientific directors, and executive officers the report of a committee on electronic publishing. The committee was chaired by Henry Metzger (NIAMS) and included Robert Nussenblatt (NEI), Ed Liu (NCI), Annette Wysocki (NINR), Eugene Koonin (NLM), and Celia Hooper (OIR). This report has been discussed and has the concurrence of the Board of Scientific Directors. It is printed below.

The report includes the following recommendations, which I support:

Approval of manuscripts submitted for electronic publication will require the same process as is currently used for printed publications.

NIH intramural scientists who wish to establish electronic journals, as either official duty or outside activities, may do so following an approval process at the IC level.

The scientific directors will monitor and keep track of creation of electronic journals by intramural scientists and of submission of data to non-peer-reviewed electronic sites and report these events to the deputy director for intramural research (DDIR) on the Management Control Checklist.

The DDIR will, in turn, ascertain whether such electronic publications remain in keeping with the principle stated in the "Guidelines for the Conduct of Research in the Intramural Research Program at NIH" that "All research staff in the Intramural Research Programs should maintain exemplary standards of intellectual honesty in formulating, conducting, and presenting research, as befits the leadership role of the NIH."

—Michael Gottesman, Deputy Director for Intramural Research

REPORT OF THE COMMITTEE ON ELECTRONIC PUBLISHING

I. Introduction

The "Guidelines for the Conduct of Research in the Intramural Research Program at NIH" state, "All research staff in the Intramural Research Programs should maintain exemplary standards of intellectual honesty in formulating, conducting, and presenting research, as befits the leadership role of the NIH." (emphasis added)

The deputy director for intramural research, NIH, asked the committee to consider whether the increased use of electronic publishing in general, and the establishment of PubMed Central in particular, required additional regulatory or monitoring mechanisms to complement those already in place with respect to conventional publication activities of the intramural scientific staff ( see "Current Procedures for Overview of Publishing Activities by NIH Staff").

The committee feels that the electronic promotion of easy access to full-text material in conventionally published journals raises no new issues. However, the enhanced possibilities that electronic publishing permits for the easier establishment of new journals and the publication of materials that might not otherwise be published, given the constraints of conventional "hard copy" publishing, deserve consideration. The relative expense of conventional publishing usually necessitates the involvement of substantial organizations with a self-interest in maintaining high quality. The new opportunities for facilely disseminating information electronically could potentially lead to the flooding of the scientific literature with trivial and possibly even misleading material.

II. Is the NIH Intramural Research Program (IRP) a Special Case?

NIH is appropriately perceived by many as a source of credible and even official or "certified" information. It therefore has a self-interest in protecting its reputation in that regard. Another germane aspect is that the public and even many professionals often do not appreciate that the IRP is completely divorced from the decision-making process by which extramural research is funded. Thus any involvement by IRP scientists in the publication of research that can even give the appearance of promoting or suppressing the support of biomedical research in general should be avoided.

III. Who Should Act as Gatekeeper?

The decentralized organization of NIH into multiple ICs leads to a diversity of approaches to fostering biomedical research and is one of its great strengths. Nevertheless, much of the public and even many professionals think of NIH as a whole, and therefore lapses by individual members of an institute or center reflect not only on that particular IC but on NIH as a whole.

With respect to the new issues created by PubMed Central, the satisfactory solution that has been developed to resolve these nominally conflicting concerns in other contexts should be followed, that is, the Office of Intramural Research (OIR) should establish guidelines to which each IC will be expected to adhere, using whatever mechanisms it deems most appropriate. Periodically, the OIR should monitor these activities as part of the Management Control Checklist. For this purpose, the scientific directors are expected to track the participation of their staff in the organization of electronic publications and submission of non-peer-reviewed material for electronic distribution.

V. Establishing an Electronic Journal

The current PubMed Central system permits any group consisting of at least three senior scientists (independently funded investigators) to establish an electronic journal which will be automatically included in PubMed Central. This raises the question of whether IRP scientists should be permitted to use this new opportunity and, if so, whether any special requirements need be met before doing so.

With respect to publication activities, in general, IRP scientists enjoy the same academic freedoms as their extramural colleagues. Specifically, IRP scientists may serve as editors or members of editorial boards of existing journals, after the appropriate approval either as an outside activity or as part of their official duties. Involvement in electronic journals, including new ones, should be handled no differently, and IRP scientists should be permitted to establish electronic journals.

To ensure that such activities do not compromise the NIH or PubMed Central or the association between the two, the following are recommended:

An IRP group wishing to establish a new journal should include at least three tenure-track or tenured scientists.

Proposals for new journals, which are likely to include a policy statement, instructions for authors, and, perhaps, an inaugural editorial article, should be reviewed and cleared within the appropriate ICs in the same way as is currently done for manuscripts submitted for publication.

Establishing a new e-journal qualifies as an outside or official activity and should require approval by the same mechanisms that already exist for other professional activities.

To avoid the appearance of a conflict of interest stemming from the connection between NIH and PubMed Central, new e-journals edited by IRP scientists should include a prominent disclaimer stating that the views and opinions expressed in the journal do not necessarily conform to those of NIH or the U.S. government.

For the same reasons, the inclusion of non-NIH scientists among those administering the journal, and sensitivity to the appearance of conflicts of interest in the selection of referees for IRP-authored manuscripts, is strongly advised.

V. Participation in an E-Journal

As an Author: Clearance and Submission of Manuscripts. The choice of where to publish NIH research has traditionally been the prerogative of the authors, in consultation with colleagues and supervisors. The advent of electronic publishing is no reason to change this policy or in any way restrict the choices of NIH scientists in their selection of the best forum for airing their findings. In general, research reports should be published in peer-reviewed journals.

Because there should be no difference in the quality or rigor of electronically compared to conventionally published research, authors should follow the same procedures for NIH review and clearance of electronic publications as are followed for traditional publications. Clearance procedures were recently reviewed by a subcommittee of the scientific directors and simplified to a check sheet that can be amended or used as is by the institutes. The manuscript clearance can be found at this web address:

<http://www1.od.nih.gov/oir/sourcebook/oversight/pub-clear-form.htm>.

Some institutes have amended the check sheet procedures, but all should have in place a system for timely review of publications prior to submission. Where the in-house review process is backlogged, lengthy, or burdensome, institutes should streamline and delegate the review workload more efficiently. The potentially shorter time-course and increased volume of electronic publications—and the potential increase in submissions of material that will not be subject to further peer review (see Section V)—intensify the need for prompt, careful clearance.

As an Editor or Reviewer. Editing or reviewing articles for an electronic journal raises no special issues relative to editing a conventional journal and may be carried out either as part of one’s official duties or as an outside activity. A good guide to what is and is not allowed for official duty activities and outside activities may be found at

<http://www1.od.nih.gov/oir/source book/ethic-conduct/outside-act.htm>.

Given the ease of mass distribution of documents via e-mail, reviewers and editors should be extremely careful when they address and electronically send manuscripts and reviews to avoid accidental release of pre-publication data.

VI. Posting of Non-Peer-Reviewed Material

Background. Currently, PubMed Central is not accepting non-peer-reviewed material. However, the committee is aware of at least three journals (one conventionally published, two electronic) that are now accepting non-peer-reviewed material, or will soon. Though there are concerns about publishing non-peer reviewed results in the new online formats, it should be noted that IRP scientists, like their extramural colleagues, already regularly publish or otherwise publicize non-peer-reviewed material in reviews, some abstracts, and articles in conference proceedings, and deposit non-reviewed data such as nucleotide sequences or x-ray coordinates in various data banks. Attempts to limit these contributions especially to repositories would be detrimental to science. It is unclear whether, in the field of biomedical research, the publicizing of non-reviewed material in online journals will mushroom or if it will have harmful effects. The committee recommends that the individual institutes track such contributions by their own staff and that the OIR review the matter one to two years from now.

Non-Clinical Data. NIH authors should weigh carefully the ramifications of publishing their results in non-reviewed electronic publications. For very theoretical reports submitted to a discerning online audience of peers and colleagues, errors may be quickly spotted, discussed, and corrected. For very specialized research with limited response, it may take much longer for errors to be uncovered in a non-peer-reviewed e-journal. For research of high interest to the lay public, unreviewed findings may be disseminated widely before errors are caught and may be misinterpreted and taken out of context.

In any case, there should be no lowering of the standards of quality for NIH research that goes into journals, whether these are rigorously reviewed or minimally screened. It is particularly important that the Institute’s internal procedures for clearance of manuscripts be strictly adhered to for material that will not be otherwise reviewed.

Clinical Research. Manuscripts submitted by author(s) from the NIH that will receive no further peer review before posting and that report results of research studies involving the use of human subjects or materials should receive special attention because the potential cost of error in the scientific record is very high. They should be rigorously peer reviewed by individuals with expertise in the area of the study, either from another branch or laboratory within the same IC or from other ICs or the extramural community.

All manuscripts involving clinical research should include a statement that the study was approved by an Institutional Review Board (IRB) and should provide the approved NIH protocol number. Manuscripts should seek to protect patient confidentiality in any text, figures, or images. In addition, for all clinical studies, investigators should include a statement that all applicable regulations and institutional rules regarding protection of human subjects have been followed.

If the study was conducted with the support of nonfederal funds, then this information should be stated in an appropriate place consistent with the journal or PubMed Central guidelines. All manuscripts should identify the author(s), their section, branch, laboratory, institute or center, academic affiliation (for non-NIH collaborators), and contact information.

It is recommended that all clinical research studies contain the following information:

background, including whether the trial is phase I, II, or III

objective or statement of the problem

information about subject recruitment, selection, and, when applicable, randomization

statistical test(s) used

results (both positive and negative)

conclusions

references

VII. Other Comments

Although outside professional activities are currently subject to approval, it is uncertain whether publications stemming from such activities are regularly subjected to the institutes’ procedures for clearance (see "Current Procedures for Overview of Publishing Activities by NIH Staff"). The Committee recommends that all such material either be submitted for clearance or include a prominent disclaimer stating that the views and opinions expressed do not necessarily conform to those of NIH or the U.S. government.

Return to Table of Contents