The Future of Medical Journal Publishing
The Journal Editor’s Perspective: Looking Back, Looking Forward
After the birth of printing, books became widespread…men quickly learned [about] so many topics….[E]specially since 1563, the number of [publications] in every field is greater than all those produced in the past thousand years….I really believe that at last the world is alive, indeed [intellectually] stimulating.
—Johannes Kepler, De Stella Nova in Johannes Kepler Gesammelte Werke, vol 1. 1937, pp 330 to 332
Scientific information is of no value without dissemination. Before the advent of the printing press, new scientific knowledge was shared by word-of-mouth, by hand-written correspondence, or by the assiduous efforts of scribes who copied information on semipermanent media in libraries of papyrus scrolls. These methods of communication all suffered from limited reader access. With the invention of the printing press in the 15th century, mass production of printed materials became possible. The practical limitations of earlier methods of information sharing were immediately eliminated, with the resulting explosive growth in the publication of books of all sorts.
Historical Role of Biomedical Journals and Journal Editors
The sciences, in particular, benefitted greatly from mechanical printing. The scientific method and the progress of science require that new observations are shared with the scientific community, both to encourage their replication and to advance the discipline. Each new scientific observation adds to the corpus of scientific knowledge, and the more rapidly the observation can be shared, the more rapidly the field progresses. To this end, the first scientific journals were established in 1665: the French Le Journal des Sçavans (January 1665) and the English Philosophical Transactions of the Royal Society of London (March 1665). These journals presented reports on topics across the spectrum of science (natural philosophy), without emphasis on any particular branch of the discipline. The first biomedical journal was Acta Medica et Philosophica Hafniensia (1673). This journal was edited by Thomas Bartholin, physician and professor of anatomy at the University of Copenhagen, and published in Latin. The first French medical journal, Nouvelles Descouvertes sur Toutes les Parties de la Medecine (1679), was published 4 years later and lasted only 5 years. The first English medical journal appeared in 1684, Medicina Curiosa (full title: Medicine Curiosa, or a Variety of New Communications in Physick, Chirurgery, and Anatomy, from the Ingenious of Many Parts of Europe, and from Other Parts of the World), and, whereas Philosophical Transactions contained some medical articles (≈15% of content), Medicina was devoted exclusively to medicine.1 Medicina Curiosa was published and edited by Thomas Basset, a bookseller who viewed this publication as a means to summarize (abstract) existing medical publications (books, other journals not published in English) for the benefit of practicing physicians. His view of this role of the medical editor presaged one of the responsibilities of the contemporary editor: to guide the reader to useful information, culling that information from a much larger (and growing) universe of medical information of variable quality and reliability. In keeping with this notion, then-President of the American Heart Association, H. M. Marvin, pointed out that Circulation was established in 1950 to permit investigators and clinicians “to keep abreast of advances” in cardiovascular biology and medicine through “the creation of a scientific journal that shall be acknowledged as foremost in the world among those devoted to a specific field of medicine.”2
In addition to directing the attention of the reader to areas of routine progress relevant to a scientific discipline, journal editors are also responsible for identifying those articles likely to have a significant impact on the field. Although most scientific observations are of modest consequence, a (very) few transform a field and have long-lasting effects. How can a reader of the scientific literature, even one well versed in the specific discipline, appreciate the difference between articles of incremental impact and those of great impact? Some journal readers feel more than competent to distinguish between run-of-the-mill reports and transformative reports; however, many readers rely on the editor to help them discriminate. This editorial function is particularly relevant to medical journals whose clinician readers often lead very busy professional lives and, therefore, depend on the editor’s and reviewers’ breadth of experience and knowledge to guide their focused reading. Admittedly, the long-term transformative impact of any scientific observation is often not predictable at the time of publication, with many of the most important, groundbreaking observations requiring years to be appreciated as a discipline matures. Yet, the editor can and does shape opinion, and may presage true impact as a result of thoughtful editorials or commentaries on an original report at the time of its publication. Because of these important editorial roles, the biomedical journal has thrived for more than 300 years as a repository of new scientific knowledge filtered or vetted by an editorial process that, in the ideal, involves thoughtful and committed expert reviewers and unbiased, knowledgeable, high-minded editors of exquisite scientific taste.
Notwithstanding the desirability of this ideal, it is important to point out a more negative view of editorial choice. If the bright side of editorial function is to serve as wise sage encouraging the progress of a scientific discipline, the dark side is to serve as biased judge impeding the progress of a discipline. In the latter case, the editor prevents the publication of what the authors view as scientifically sound articles of potential importance because he/she does not find them sufficiently interesting for the readership, or because of the adverse effect publishing such an article may have on various use metrics by which journal quality is assessed, or because of (undisclosed or unrecognized) bias or conflict between a reviewer and the author about which the editor is unaware.
The Rise of Open Access Publishing
This long-standing natural tension between authors and editors reached a tipping point at the turn of this century with the advent of open access biomedical publishing. This initiative, which is becoming increasingly commonplace, was designed to eliminate the restrictions on dissemination of journal articles imposed by publishers who held copyright on the articles and whose conventional business model was a pay-to-read model in which the cost of publishing is borne by the users (subscribers). Despite an ever-increasing number of new journal titles, this model has been increasingly challenged by many interested parties who feel that free access to information—the great promise of the Internet—was impeded by publishers’ business strategies, which were designed for success in the printing era. Willingly, in some cases, but by regulation in others, publishers responded to this groundswell of pressure by offering free access to all (or most) published articles after a period of time (6–12 months). Under these circumstances, subscription fees give immediate access to new information as soon as it is published, with nonsubscribers given delayed access (by which time the competitive advantage may have waned considerably). Although an improvement, delayed free access in a subscription model failed to meet the immediacy of open access, for which reason the latter blossomed. Recent data show that, in the past 7 years alone, open access journals have increased in number by >300% to 10 475 journal titles.3 This incredible proliferation of titles argues for a socialization of the publishing process, basically providing a venue for publishing virtually any article, provided the author does not have concerns about price or quality (obviously unrealistic, as discussed further below).
True open access publishing serves the purpose of shifting the publishing model from the conventional pay-to-read format to a pay-to-publish format in which the authors pay for their work to be published, retain copyright, and offer access to anyone with an Internet connection free of charge as soon as the article is published. Notwithstanding efforts to contain it, open access publishing was ultimately established as a viable publishing model as a consequence of the confluence of several forces: the continued restriction on acceptable papers (page limitations) imposed by editors and publishers; the ever-increasing cost of journal subscriptions; the decline in advertising revenue used to offset publication costs (especially in clinical journals); the concern by funding agencies that the fruits of scientific research supported by federal (or philanthropic) dollars were being transferred from the investigator to the publisher with the former’s relinquishment of copyright; the increasingly competitive research grant environment necessitating supportive publications of the highest quality by the applicant before grant submission; and pressure from the federal research community and the public to release research results funded by the taxpayer freely for public scrutiny and benefit as soon as possible after completion of the work.
Many professional society journals are published by the society itself (as the American Heart Association journals were before 1998), although most now are published through contractual relationships with commercial publishers (as the American Heart Association journals are currently). Although the commercial publisher in these cases seeks to make a profit, the sponsoring society also benefits from that profit, funneling the proceeds into the operating funds of the society. In the case of the American Heart Association, proceeds from its publications are used largely to support its research programs, a fact that is not commonly appreciated. Thus, any change in the flow of funds or the business model can have significant consequences for the society’s missions. With the diminishing pool of subscribers for print journals and the decrease in advertising revenue, many society journals are facing a challenging future without a serious reconsideration of their working model. With the advent of open access publishing, the American Heart Association recognized the opportunity in the publishing landscape and established the Journal of the American Heart Association as its fully open access journal. The success of the journal has exceeded expectations, with submissions and publications growing faster than anticipated. As a result, the journal promises to offer a new, successful strategy for continued publishing growth and success for the American Heart Association. The American Heart Association has also supported the development of hybrid publishing models—as exist for Circulation–in which the journal maintains a subscription-based format, but also offers authors the option to publish their articles as true open access articles for a fee.
Although addressing many of the issues that led to its creation, both globally and locally, open access publishing has created other problems that have yet to be solved. Shifting the cost burden from subscriber to author stabilizes the financial infrastructure of many journals; yet, with expense inflation and the need for commercial publishers to profit, these added costs will be (and are being) shifted directly to the authors because there are no or only limited other revenue sources for most contemporary journals. These authors often feel a need to publish their results quickly and, therefore, may be compelled to pay whatever is charged, especially if the publication venue is a desirable one under these urgent circumstances. Free-market forces do not work in this system because costs appear to be second-order concerns when an investigator chooses a journal venue. Expense and publication cost inflation is, therefore, likely to become a growing problem for many authors, especially as extramural funding sources continue to decrease.
Another problem is that of newer open access publishers (predatory publishers) who seek to publish articles simply to acquire the fee an author is willing to pay, and do so with limited regard for quality or readership.4 Peer review is perfunctory, at best, and the likelihood that any of these journals has achieved or will ever achieve any semblance of quality or authority is low. In recent years, these open access journals have expanded considerably, proving a different aspect of the free market, namely, low barrier to entry of new commercial entities, no matter their perceived value. How to flag these unscrupulous publishers remains a matter of concern and debate.
Yet a third problem concerns the true cost of publishing an article in electronic format and its implications for authors and sponsoring institutions. Van Noorden5 has given a thorough analysis of this problem, and points out that the costs depend on many factors, including subsidies from sponsoring organizations, the services provided by the journal (copyediting, rigorous peer review, professional editorial staff, etc), and the perceived quality of the publication venue (those of lesser perceived quality charging less than those of higher perceived quality). Given these wide-ranging contributors to the publication fee, it should come as no surprise that the average per-article fees range widely from $0 (for fully subsidized journals) to ≈$5000. This range of prices suggests that the true cost is likely to be far less than the mean or median cost from a comprehensive survey, given the financial incentive that drives the for-profit arm of open access publishing.
Defining Journal Quality and Article Impact
This cost and fee analysis raises the important issue of what the author gets for his fee. Why publish in open access journal X rather than Y, if X charges 3 times as much per article as Y? The primary motivation for open access journal choice is no different from that for print journal choice: perceived quality and reader impact. Authors choose to publish in journals that are viewed as the best arbiters of scientific quality with the widest audience. This ability to discriminate among scientifically valid papers of varying quality remains the primary reason for the continued existence of journals in formats that have remained largely unchanged for the past 3 centuries. Editorial guidance through the morass of published articles is appreciated by readers whose limited time and, often, inexperience are greatly benefitted by the wisdom of the editorial process. We shall, for the moment, put aside the self-reinforcing pretense implicit in defining journal hierarchy—that is, one journal is viewed as superior to another because those who choose to publish in it believe that it is. Provided that those who choose to publish in it are widely admired in their scientific community as important contributors, the journals in which they publish are widely admired as discriminating arbiters of scientific taste. Rather than relying on this tautological definition of quality and impact, information specialists have developed a variety of bibliometric parameters linked largely to citation frequency in some parameterized way that are viewed as objective measures of quality, including citation frequency itself, impact factor, eigenfactor, etc. These and many other metrics suffer from the same problems: none is perfect, each can be subject to manipulation, and, in the end, none is a truly objective measure of journal quality (ie, none can serve as an objective comparator across all journals). Factors ranging from the size of the scientific community likely to be interested in the topic to lack of distinction between reviews and original articles to excessive autocitation (directly or through subsidiary family journal intermediaries) all influence the citation frequency and its derivatives in ways that do not reflect intrinsic value, discounting their ability to serve as objective measures of impact and importance.
Recognizing the intrinsic weakness of these conventional factors that purport to define journal or article impact on the basis of citation frequency, Wang and colleagues6 explored the determinants of long-term predictablility of article impact. Using a rigorous mathematical analysis, they were able to define the intrinsic, ultimate impact of an article without regard for its publication venue. They did so by recognizing that the true effect of journals believed to be more authoritative in a discipline is simply to increase the citation frequency soon after publication in comparison with comparable articles published in other less authoritative venues. In the end, each article will achieve its maximal level of recognition as perceived by the scientific community, albeit with kinetics that differ as a function of the perceived authoritative value of the journal in which it is published. This study represents a key contribution in the growing field of the science of science7 in which rigorous scientific approaches are being increasingly applied to this central question of assessing the importance of scientific reports.
The Future of Biomedical Publishing
In view of the extraordinary changes in scientific publishing currently upon us, how will a clinician or investigator seeking information from the literature go about his or her work in the future? Will there be journals as we currently recognize them, open access or not? What will be the role, if any, of the person or persons we currently recognize as journal editors? These are important questions for which there are no simple answers, but, in the spirit of the purpose of this article, I will give my personal view of this future. I should point out first that there are publishing futurists who feel that the journal as we know it will quickly become an anachronism, to be superseded by smart information algorithms that instantaneously “filter, rate, and disseminate scholarship as it happens,” as cogently argued by Jason Priem.8 Traditional journal functions will be offered by a variety of service providers in a universal system that incorporates certification of peer review, aggregated commentary from knowledgeable readers, Web-based marketing, conversion to an appropriate e-based format with a unique electronic identifier, and dissemination broadly and openly to the interested reader, from scientist to clinician to patient to policy maker. According to Priem’s insightful review, many of these services are either currently available or can readily be developed. F1000Research is one such publishing platform that claims to offer “immediate publication of posters, slides, and articles with no editorial bias.”9 These publications incorporate all primary source data and do undergo signed peer review after posting with the reviewers’ comments published directly linked to the article.
As intriguing as this idealized format may appear, we are a long way from meeting it, particularly in clinical medicine. Remember that a primary driver for the development of open access publishing is to make new information rapidly available for interested readers. Yet, there is an important limitation of this idealized goal in clinical medicine: if treatment strategies are published without proper review and vetting a priori, the information can be applied indiscriminately or prematurely to clinical care with potentially dire consequences. In addition and importantly, in contrast to the physical sciences, the biomedical sciences yield published results that are often far less predictable (if not irreproducible) given the natural variability of biological systems, including human biology. Thus, clinicians, investigators, and the public will (or should) expect rigorous, fair peer review coupled with authoritative evaluation by an experienced, objective, and thoughtful editorial team to ensure the safe and effective application of the published results to patient care.
If one accepts the need for an editorial process that at some level mirrors contemporary editorial review, does this necessarily mean that the review must occur under the aegis of an authoritative journal structure? My own view is that it does not. I believe the future will offer a global open access publishing format whose authority will not be linked to a journal title, but to a generic review and editorial process in which a panel of (anonymous) experts will assess each publication for influence and impact by using a consistent and optimally objective set of criteria that are not subject to manipulation. These criteria may include citation frequency after some predefined period of time, a measure of the intrinsic impact of each article as derived by Wang and colleagues,6 the impact of derivative articles (second-order impact, not simply citation frequency), and accompanying editorial commentary. Each article will invariably be accompanied by some published editorial assessment that may take different forms, including formal expert editorials, briefer editorial commentary, less formal reader comments, or comparative comments in the context of other related articles. Derivative publication formats will likely expand, such as virtual compendia of topic-related articles with accompanying expert assessments and commentary. With these and other approaches or metrics, articles will be tiered by a panel of editorial experts in the discipline in this unfettered open access world to guide the reader effectively. Articles will, therefore, be judged on their own intrinsic merits without regard for a specific publication venue. This evaluative process will, in turn, offer a standard (if not become the standard) for quality and importance that guides the interested reader. Readers will, of course, make their own judgments regarding the importance of a published article, but that judgment will be informed by a rigorous, overarching editorial process helping guide the field. In addition and importantly, the published article will serve as a portal to an array of related resources that inform, amplify, and enrich the readers’ experience. These related resources may include links to in-depth reviews of related topics, to detailed analytic approaches relevant to the article’s methods, or to elements found in the Internet of Things10 that provide real-world data on parameters that relate to the theme of the article (eg, an article on heart rate changes with exercise in women might be linked to mean heart rates with exercise in 50-year-old women recorded from FitBits reported to the Internet of Things). Identifying these resources now becomes the purview of the editor who serves not only as gatekeeper of the published material, but also as purveyor of the universe of relevant resources for the reader.
Although this is but one view of the future of scientific, biomedical publishing, there are others, no doubt, that are likely to appear and compete for the ultimate format. Whatever that future format, it is very likely to be quite different from contemporary publishing, now viewed as an anachronism perpetuating an outdated 17th century model. The future of publishing is thus uncertain, but very likely to be exciting and transformative for those of us who care for patients and conduct biomedical research. Thus, although the transformative technology to which they might be applied is quite different, Kepler’s words are as relevant today as they were 450 years ago.
I thank Elliott Antman, Karen Barry, and Anita Loscalzo for helpful comments.
About the Author
Joseph Loscalzo, MD, PhD, is the Editor-in-Chief of Circulation.
The opinions expressed in this article are not necessarily those of the editors or of the American Heart Association.
- © 2016 American Heart Association, Inc.
- Marvin HM
- 3.↵Directory of Open Access Journals. https://doaj.org/. Accessed June 25, 2015.
- Wang D,
- Song C,
- Barabási AL
- 9.↵F1000Research. http://f1000research.com/. Accessed July 29, 2015.
- 10.↵Internet of Things. https://en.wikipedia.org/wiki/Internet_of_Things/. Accessed July 29, 2015.