Open access journals: Why are we not there yet? (!)

Although I am not aged by current criteria, in my formative years as a junior investigator, I remember long hours in the library searching for references. There was no PubMed; instead, a book called IndexMedicus was searched for topics relevant to a field of research, or an information expert in the library was asked to perform a computerized search for a fee. Once articles of interest were identified, they were copied in the library (often at a cost of $0.05/page). Copy machines were crucial for obtaining the articles. Book binding often precluded laying the book flat and interfered with accurate copying near the seam of the binding. Articles were then read, further references were identified, and this prompted additional trips to the library and repetitive use of the copying machine. Obtaining and managing references for grants and articles was costly, tiresome, and exasperating. Relevant articles were so difficult to identify and retrieve that senior investigators could often play one-upmanship by quoting pertinent publications of which others were simply unaware.

Currently, we can receive an electronic table of contents, download articles as PDF files, and print or read them on our computer monitor; alternatively, we can run computer searches on PubMed, locate the relevant article, and download the PDF file. Considerable information is now readily available in a real-time, efficient manner. Because of the increasing number of journals and articles, the current challenge is focusing and managing the search process. However, for access to many recent articles, an individual or institutional subscription to the journal is needed. This fact still poses a barrier to obtaining critical information. Indeed, I am currently writing a grant (still a painful, time-intensive experience), and several articles that I needed to review for emerging and evolving concepts were not available because neither I nor the institution had a subscription. This experience was frustrating and emphasized that barriers between the consumer and scientific information still exist. This problem was meant to be solved by the development of a process for open access to scientific information (the availability of articles online without fees or subscriptions), but obviously universal open access has yet to be obtained.

The driving force for open access has been the World Wide Web, which has facilitated a shift from print-only journals to parallel print and electronic formats. Two types of open access now exist: an article can be published in a truly open access journal such as a Public Library of Science (PLoS) journal1, 2 or in a closed access journal with subsequent deposition by the author in an open access forum such as PubMed Central. Often, this second scenario results in delayed open access. For example, the publisher of Hepatology, Wiley-Blackwell, deposits articles funded by the National Institutes of Health into PubMed Central 6 months after the publication of the print version; all original research articles are ultimately electronically unlocked 12 months after publication. Although articles are available to Hepatology subscribers before print publication via Early View, this is not an open access proposition, nor are Early View articles searchable on PubMed. Bjork et al.1 recently analyzed the status of open access publication in multiple disciplines. They reported that in the broad category of medicine, approximately 14% of the publications are available online free of charge from the onset of publication, and another 8% are available on a delayed path to open access. Thus, only a disappointing 22% of articles in medicine are available in an open access format. Why has open access not become the predominant publication format?

The answer to this question lies in the economics of publishing. Open access saves the direct costs of print publication and dissemination, although the costs related to copyediting, typesetting, and image treatment are not obviated. Open access also results in more article citations.3 All these features of open access reduce costs and enhance the impact factor and prestige of the journal. However, direct open access reverses the business plan of publishing. The costs of publishing are transferred to the authors rather than the subscribers, and the authors, rather than the readers, become the clients of the publication process. Some fear that if the authors end up paying the piper, they will also end up calling the tune and subverting the financial independence of the journal from its authors; this is perhaps a risk, but it is an unlikely one in scientific publishing with peer review. The cost to the authors ranges from $500 to $3000 per article. The cost of publishing an article in PLoSOne is $1350. These costs compare favorably with charges for publishing color figures in print journals (this cost is avoided with online-only publications) and, therefore, may not be too exuberant for well-funded investigators. (PLoSOne has a process for subsidizing authors who cannot afford this fee, such as underfunded researchers from the developing world, and thereby averts the fear that only the wealthy may publish.) Thus, open access shifts charges to the investigators and research institutions producing information and away from those readers and institutions not producing research and no longer paying a subscription (many may view this shift as unfair). This open access business model, however, is not as lucrative as the current business model, in which individual and institutional subscriptions and advertising revenue provide the economic incentives for publication. In an access control or subscription model, journals are profitable for the publisher and the societies, which often own the journals. Authors do not have the resources to replace all these revenues in the immediate open access paradigm. This reality is a financial disincentive for either publishers or societies to further encourage open access. This financial disincentive explains the failure of most publishers and societies to openly embrace open access to their journals. However, the overall goal should be to obtain the right balance between revenue and costs so that the publication provides the resources to conduct a peer-review system but makes information as widely, easily, and freely accessible as possible.

The financial argument by the publishers is that institutional and perhaps individual subscriptions will decrease if all articles are published with immediate open access. In support of this perspective, institutional subscriptions for the Journal of Clinical Investigation decreased by 40% from 1996 to 2003 when the American Society for Clinical Investigators elected to provide open access to the journal.4 In 2009, the Journal of Clinical Investigation re-instituted access control (a subscription structure) for nonresearch articles, but it still retains open access for research articles; this is an option for other journals as well. Other sources of print media such as newspapers have a declining subscription and advertising base.5 The conversion from print plus electronic formats to only electronic media will continue rapidly and is inevitable. Marketing departments in industry will likely respond to this change by decreasing advertising in print journals, and this will result in a loss of revenue. Marketing efforts will likely rely more on direct-to-consumer and direct-to-physician advertising and social media networks. The business model for print journals will be less lucrative, and this will further propel an immediate open access format into reality for all scientific publications. However, this does not indicate that online journals cannot be financially viable. PLoSOne has become financially viable, albeit with a business model different from that used by print journals.6 It achieves financial stability by high-volume publishing: approximately 7500 articles in 2010. In comparison, Hepatology will publish approximately 350 original manuscripts in 2010. PLoSOne has an acceptance rate of 69%, whereas Hepatology has an acceptance rate of approximately 20%.6 Yet, PLoSOne is still able to achieve an impact factor greater than 4, which places it in the top 25% of biology journals (the impact factor for Hepatology is 10). These data suggest that the business model of open access journals relies more heavily on quantity versus quality. However, the philosophy of PLoSOne is that the science must be robust, but ultimately the scientific accuracy and value of an article can be truly assessed only over time after its publication and should not be prejudged. This is a valid perspective.

In my opinion, scientific and clinically relevant information needs to be available to scientists, clinicians, and the public at large to accelerate the pace of scientific progress and thus to further the care of our patients. PubMed Central attracts 420,000 visitors per day, of which only a quarter originate from a university log-on; this indicates a high demand for scientific information outside academia.6 Practicing physicians, patients, and others seeking medical information likely account for a large number of these visits. Access to scientific and clinical information will be even more important as we emphasize global advances for patients with liver and other diseases because subscriptions and fees for articles remain an impediment to information in the developing world. If we are to help advance patient care and science, open access for these physicians and their patients is critical. As consumers, we should do all we can to facilitate this progress to immediate open access publications.