AN UNVARNISHED TRUTH

Every 5 years, the Editorial Board of the JCI changes hands. To some degree, so does the JCI’s philosophical bent. We have just passed the midway point in our term. As such, we look back with some degree of pride at developments ranging from the move to self-publication and a new journal format to our continued ability to provide free access to the Journal’s content on the Internet. However, significant as these changes may be, they impact little on the scientific content of the JCI. Now, with over 10,000 decisions having passed through our hands, we have steadily begun to find, and exercise, our own editorial “voice.” Since the start of this year, the entire Board of 14 Editors here at the University of Michigan has steadily increased the stringency of the selection process in an effort to improve the quality of the Journal’s content. While we have always prized novelty and scientific rigor, we are focusing more attention on identifying those papers with the broadest appeal to our general readership. In Ajit Varki’s inaugural address as Editor-in-Chief of the JCI in 1992, he emphasized that “excellence and originality are the major criteria for acceptance rather than trends and fashions” (1). Our commitment to this important tradition continues. However, new developments dictate that the JCI can only retain its leadership if we stay true to the original mission of disseminating the best of the biomedical sciences to as broad an audience as possible (2).

Why change?

Through the 1970s and 1980s, the JCI enjoyed a commanding position as a general-interest publication with a focus on the cellular or molecular basis of key physiologic and pathophysiologic processes. However, coincident with the application of increasingly powerful methodologies and the attendant accelerating speed of scientific discovery, new journals appeared and began to draw from the same pool of work traditionally directed to the JCI. As the pace of science quickened even further in the mid-1990s, subspecialty journals reaped similar benefits, and the quality of their published works improved as well. This trend continues, and the JCI finds itself as one among many attractive venues available to authors in the biomedical sciences.

Since assuming the editorship, we have noted — with increasing concern — that many of the manuscripts submitted to the JCI have appeared to be virtually indistinguishable from those found in quality subspecialty journals. If the JCI were to become only a depository for articles identical to those found in more specialized venues, we believe that the value of the “JCI stamp of approval” eventually would be diminished. Clearly — at least to the Editorial Board — our most important mission is to ensure that the Journal maintains, or even advances, its ability to attract those works that capture the interest of readers across multiple disciplines.

What are Editors for?

In a manner different from other journals, each new Editor-in-Chief selects an Editorial Board comprising active scientists at the same (or at a nearby) institution, who meet weekly to review all submitted manuscripts. To many, the rationale underlying this requirement is unclear. If submitted works are evaluated by external referees, aren’t the editors merely “rubber-stamping” the collected recommendations? The answer to this query is a simple “no.” A recent editorial in Nature Genetics discussing the merits of publishing certain types of manuscripts concluded that the Journal “will continue to publish selected… studies of broad interest and high caliber. It is an editor’s job to ascertain the former, but s/he depends on expert advice for the latter [italics added]” (3). We could not agree more. However, in the case of the JCI, this editorial responsibility does not rest on a single set of shoulders but instead is shared by a group of 14 peer scientists. Thus, the Board convenes each week to evaluate general appeal as well as scientific rigor. We view this dual mandate as the most unique characteristic of the Journal. Our intent is to apply our skills as “editor-scientists” to the best advantage of the JCI and the biomedical sciences.

How do we make decisions?

Every paper submitted to the JCI is initially read by one or more of the Board members to judge its overall suitability for external review. Until recently, the Editors would render a “screening decision” only on those papers that contained obvious design flaws or fell directly within their area of expertise (about 10% of all submissions). In an effort to reduce the overall workload of the external reviewers (and to accelerate the authors’ ability to submit declined work elsewhere), the former Board at University of California–San Diego (UCSD) instituted a policy wherein referees were given 72 hours in which to prescreen a work for “suitability.” If a submission were judged to fall into the bottom 50% of related works in the field, the manuscript could be declined without a full review. While this process might deprive authors of a full review, the UCSD Editors noted — and we concur — that “the primary purpose of the Journal is not to provide authors with detailed feedback regarding the flaws in their work and advice for future studies. Rather, it is to attempt to publish the best possible collection of suitable original articles following fair and expedient peer review” (4).

Until recently, we continued using the UCSD system and provided our external reviewers with the option of exercising a screening decision. Similarly to the UCSD experience, another 10–15% of the submitted articles were declined via this route (for a combined total of approximately 25%). However, as reviewers have become increasingly taxed, we have found that the number of reviews performed within the 72-hour guideline has fallen dramatically. Uncomfortably, we have found ourselves apologizing to authors whose papers were declined after 4 or more weeks in the review process without detailed comments from the referees. In response to this trend, the Board has assumed greater responsibility for screening manuscripts. Currently, we only submit works for external review if we judge the manuscript to have potential merit (or if it falls entirely outside the collective expertise of our Editorial Board). As in the past, all papers not submitted for external review are examined each week by the full Board before the work is formally declined. A dissenting vote from even a single member is generally sufficient to warrant external review.

Except for those works declined in the screening process, all other manuscripts are reviewed by 2 or more referees selected from our 9,000-member database. Following receipt of the reviewers’ comments, the manuscript is re-evaluated by the responsible Associate Editor and an initial recommendation is presented to the entire Board for evaluation. In a small number of instances, papers fall outside our collective expertise and prove difficult for us to evaluate. However, in such cases, we rely on our Board of Consulting Editors (their names can be found on the front page of the Journal or at our website, http://www.jci.org) and defer our decision to their recommendation.

Within the past year, we have also instituted a new policy for submission of Rapid publications. In this process, papers of exceptional importance are reviewed and a decision rendered within 2 weeks of receipt. These manuscripts are first screened by the Editors for suitability and then refereed externally by either experts in the field or members of the Board of Consulting Editors, who agree to review the work within 5 days. Because “Rapids” must be accepted either directly or after only minor modification, detailed comments are seldom provided to the authors. However, the Editors strive to provide authors with an explanation for our decision and an indication as to whether, in the case of a rejected manuscript, the work may be suitable for resubmission as a Regular publication.

Is this process fair?

The screening review process. “Fairness” is, of course, in the eye of the beholder. However, a review of recent statistics is instructive. First, despite the 40% decrease in the number of articles published in the Journal in 1999 (relative to 1998) the number of new submissions has remained steady over the last 4 years at 3000–3200 annually. Second, the percentage of manuscripts declined at the screening stage in 1999 stands at approximately 25%, a number almost identical to that reported by the UCSD Board in 1994 (4). The bulk of these submissions are judged to be technically sound but fail to convince the Editors that they represent a significant advance in the field.

Obviously, absolute rules for selection criteria at the screening level are difficult to establish, as our standards — like those of other top-tier journals — evolve as fields develop. For example, whereas genetic linkage analyses, association studies, the mapping of disease genes, and genome scans to map QTLs for complex traits all attracted considerable attention even a few years ago, we now consider these types of studies more appropriate for specialized journals unless new biological insights are forthcoming. While we apply similar selection criteria to all fields of the biomedical sciences, an examination of the rebuttal rates for screening decisions in 1999 suggests that major “miscarriages of justice” are few. Indeed, of the 802 screening decisions rendered in the last year, only 9 rebuttals were generated. In each case, we carefully detailed our rationale to the authors and offered them the opportunity to have the work reviewed by external referees. When authors took advantage of this option, the external reviewers concurred with our original evaluation and no decisions were overturned throughout 1999. This record is not to be construed as proof that we are infallible. Errors will always occur and, as a consequence, all rebuttals are treated seriously and given every consideration.

Review of refereed submissions. For those papers that are judged to be suitable for external review, any submission that receives a priority ranking in the top 25% from at least one of the referees is discussed at the Board meetings, even if the other reviewer recommends rejection. Regardless of whether the responsible Associate Editor recommends declination, an invitation to revise, or acceptance, the paper is presented to the entire Board for discussion. We hope that our readers appreciate that this is a very time-consuming affair, but that it helps ensure equitable treatment for all works. Special attention is, of course, given to those studies that are judged by referees and editors alike to be carefully performed and presented. However, “broad appeal” remains a key watchword. This having been said, we would be remiss if we did not clarify the criteria applied for determining “appeal” — especially since it hardly requires a unanimous or even a majority vote. Simply stated, our experience to date shows that if 3 of the 14 members of the Board feel strongly that the work should be accepted, the manuscript is almost inevitably published.

Invariably, we are asked about the actual percentage rate for accepted papers, i.e., “the unvarnished truth.” Given the noticeable decrease in size of the Journal over the last year, “varnishing” is simply not an option. For 1999, 16% of the papers submitted for external review were accepted (or 12% of the total, if we include those declined in the screening review process). Many have asked if we have a limit, an arbitrarily drawn line separating the “accepts” (i.e., 16th percentile and lower) from the “declines” (i.e., 17th percentile and higher). Nothing could be further from the truth. We simply accept every well-executed paper that we believe will be of interest to a cross-section of our readership. Given the opportunity, we could, and would like to, publish more. However, our standards for acceptance will continue to be applied independent of the absolute number or percentage of papers accepted. Are our criteria too stringent? Are we out of touch with our expert referees? Significantly, a diverse mix of external reviewers culled from virtually every biomedical discipline appear to rank papers in a manner very similar to our own. In an analysis of 700 manuscripts that were reviewed externally in 1999, 17% received priority rankings of 25% or better by both referees. Given our own accept rate of 16%, we believe that the editors and referees are in step with their commitment to excellence.

Are even more changes coming?

During the remaining period of our tenure, we would like to reinforce the changes instituted to date. By way of providing our readership with a bimonthly periodical that is readable and combines rigor with novelty as well as broad appeal we hope to stabilize the JCI at 18–20 original articles per issue. We would also like to encourage submissions in areas that have not been traditionally well-represented in the Journal’s pages, e.g., cancer, the neurosciences, and developmental biology. To further develop interest in the JCI’s content, we plan to expand further the Commentary and Perspective series. As some of you may have noted, Fintan Steele, the former Science Editor of the JCI, assumed a new position as Editor of Molecular Therapy earlier this year. In his place, we have recently recruited John Ashkenas, the former Science Editor of the American Journal of Human Genetics. We are delighted to have him aboard and look forward to his efforts in helping us to advance the Journal’s mission.

Finally, in the near future, we hope to implement additional systems that will allow for electronic manuscript submission and review. We have made a successful transition to self-publishing under the careful direction of our Managing Editor, John Hawley. Average turnaround times for review and publication continue to improve (Table 1). We anticipate that additional gains will be realized with our transition to a fully “electronic office.” Likewise, we have recently instituted a telecommuting network that allows off-site editors to participate actively in the weekly Board meetings. This process has worked so well that we should soon be able to expand the Editorial Board to increase further our range of expertise.

Table 1

Average turnaround times (in weeks)

As in the past, we hope to enjoy the continued confidence and support of our readers in our efforts to make the Journal more valuable to you. In the rapidly changing world of the biomedical sciences, we are striving to maintain the standards that have long been associated with the JCI. As always, we welcome your suggestions and constructive criticisms.

References

1. Varki, AJ. The times they are a’changin’: changing with the times. J Clin Invest 1992. 89:721-722.
   View this article via: JCI PubMed Google Scholar
2. Weiss, SJ. Speed, competition, rigor and creativity: striking a delicate balance. J Clin Invest 1997. 99:817-818.
   View this article via: JCI CrossRef Google Scholar
3. Freely associating. Nat Genet 1999. 22:1-2.
   View this article via: PubMed CrossRef Google Scholar
4. Varki, AJ. The screening review system: fair or foul? J Clin Invest 1994. 93:1871-1874.
   View this article via: JCI PubMed Google Scholar