Abstract
Introduction Journal publishers increasingly offer governmental and charitable research funders the option to pay for open access with a Creative Commons Attribution (CC BY) licence, which allows sharing and adaptation of published materials for commercial as well as non-commercial use. The Open Access Scholarly Publishers Association recommends this licence as the least restrictive Creative Commons licence available. We set out to investigate whether pharmaceutical companies are offered the same options.
Methods Using Journal Selector (Sylogent, Newtown, PA, USA), we identified journals with a 2015 impact factor of at least 15 on 24 May 2017, and excluded journals that only publish review articles from the analysis. Between 29 June 2017 and 26 July 2017, we collected information about the journals’ open access policies from their websites and/or by email contact. We contacted the journals by email again between 6 December 2017 and 2 January 2018 to confirm our findings.
Results Thirty-seven non-review journals listed in the Journal Selector database, from 14 publishers, had a 2015 impact factor of at least 15. All 37 journals offered some form of access with varying embargo periods of up to 12 months. Of these journals, 23 (62%) offered immediate open access with a CC BY licence under certain circumstances (e.g. to specific research funders). Of these 23, only one journal confirmed that it offered a CC BY licence to commercial funders/pharmaceutical companies.
Conclusion The open access policies of most medical journals with high impact factors restrict the dissemination of medical research funded by the pharmaceutical industry. To give the scientific community freedom to read, reuse and adapt medical publications, publishers and academic journal editors would ideally allow pharmaceutical companies to fund unrestricted and immediate open access with a CC BY licence.
Introduction
Freely accessible research publications support innovation in the public and private sectors by allowing previous research to be built upon, encouraging collaboration, speeding up research dissemination and improving the transparency of research across the world.1-3 It also benefits public health by supporting informed decision-making.4,5
Open access means that readers do not have to pay to access peer-reviewed scientific research articles; however, open access is more ambitious than just making content free to read.2 To realize the full potential of open access in supporting innovation, research publications should, on the condition that the source is appropriately attributed:
be freely available to the end user at the time of publication
allow for unrestricted reuse of content.3
Creative Commons licences attached to publications allow authors to retain copyright of their work and to stipulate the restrictions that apply to its reuse.6 A Creative Commons Attribution (CC BY) licence allows sharing and adaptation of published materials for commercial as well as non-commercial use.6 It is recommended by the Open Access Scholarly Publishers Association as the least restrictive type of Creative Commons licence available and by the Budapest Open Access Initiative as the optimal licence for open access publishing.7,8
There is no harmonization between journals in their options for publishing open access and a key area of difference is the types of Creative Commons licence on offer: common derivatives of the CC BY licence include CC BY-NC, which restricts commercial reuse; CC BY-ND, which restricts adaptation; and CC BY-NC-ND, which restricts both (Table S1).9
Governmental and charitable research funders increasingly mandate publishing open access with a CC BY licence.10-13 The Medical Research Council, for example, requires grantees to publish their work in journals that offer immediate access with a CC BY licence.10,14 In 2014, the Charity Open Access Fund was established by six partner UK charities (Arthritis Research UK, Bloodwise, the British Heart Foundation, Cancer Research UK, Parkinson’s UK and the Wellcome Trust) to ensure that the researchers and organizations they support can access the funds required to pay article processing charges and make research papers open access with a CC BY licence.15
Although the pharmaceutical industry funds approximately half of all medical research,16,17 no pharmaceutical company currently mandates that the research they fund is published open access with a CC BY licence.
We set out to clarify the open access options offered by journals with high impact factors and whether pharmaceutical companies are given the same options as non-commercial funders.
Methods
Using Journal Selector (Sylogent, Newtow n, PA, USA), we identified journals with a 2015 impact factor of at least 15 (accurate on 24 May 2017); we excluded journals that only publish review articles. We collected information on the open access policies of the selected journals from their websites and by email contact, making up to three attempts between 29 June 2017 and 26 July 2017.
We categorized the least restrictive open access variant offered by each journal, using our own classification, according to:
whether a CC BY licence was offered
whether there was an embargo period
which version of the article would be available (published, accepted, submitted).
For journals that offered a CC BY licence, we also assessed:
article processing charges
funding requirements for obtaining a CC BY licence.
Between 6 December 2017 and 2 January 2018 we sent emails to the journals requesting confirmation of our findings (Table S2).
Results
Least restrictive open access variant offered by journals
Fifty-three journals listed in the Journal Selector database had a 2015 impact factor of at least 15. Sixteen review journals were excluded and 37 journals from 14 publishers were included in this analysis. Table 1 shows the categories for the least restrictive open access variant offered by each journal.
Immediate open access with a CC licence was offered by:
3 (30%) of the 10 journals with the highest impact factor (impact factor 30.4–137.6)
The types of CC licence available from these 23 journals were:
CC BY from 23 journals (100%)
CC BY-NC from 5 journals (22%)
CC BY-NC-ND from 20 journals (87%).
The 14 journals, from six publishers, that did not offer open access with a CC licence offered access to different versions of the article either immediately, after a 6-month embargo period or after a 12-month embargo period (Table 2).
The cost of open access with a CC BY licence
Of the 23 journals that offered a CC BY licence, 21 (91%) disclosed article processing charges on their websites. Across these journals, charges ranged from US$3000 to US$5000; the most common article processing charge (65% of journals) was US$5000 (Figure 3). Details of the fees charged by the remaining 2 journals (9%) are not available from their websites because the details are only provided when the article is accepted.
The article processing charge for obtaining a CC BY licence from Acc Chem Res (US$5000) is lower if authors are paid members of the American Chemical Society.18
aDetails on processing fees are provided at acceptance.19
Relationship between authors’ funding source and the availability of open access variants
Table 3 shows the open access policies of journals with high impact factors that offer open access with a CC BY licence. Of the 23 journals listed, only one – Accounts of Chemical Research – had an open access policy that allowed this option for commercial funders/pharmaceutical companies.
Discussion
We found that the majority of journals with a high impact factor that were included in this study offer open access with a CC BY licence; however, with the exception of one journal that does not publish evidence from clinical trials, this option appeared to be available only to authors funded by non-commercial organizations. Some journals allowed authors to obtain a CC BY licence when the work was supported by funders who mandated its use; however, at present, pharmaceutical companies do not mandate CC BY licences.
Here, we present a thorough analysis of open access policies of journals with a high impact factor, including society-owned journals, from multiple publishers. We met our objective to clarify the open access policies of all journals analysed, and received confirmation of our findings by email from 97% of the journals. Although our study covers only a small number of journals, extending such a manual analysis to a greater number of journals without loss of detail and still being able to confirm our findings efficiently would be difficult. When more extensive mining of journal (meta-)data becomes available, however, this study could be easily repeated for a bigger cohort of journals. To our knowledge, it has not been reported previously that the availability of open access options depends on authors’ source of funding.
Limitations of this study are that we only investigated journals listed in the Journal Selector database with an impact factor of at least 15, and that, because impact factors and the open access policies of journals and publishers are subject to change, the information may not be current. Furthermore, by selecting journals with a high impact factor, our analysis does not include prestigious journals from certain therapy areas and regional journals.
The constraints imposed by journals and publishers that prohibit commercial funders from publishing open access with a CC BY licence are incompatible with key political declarations such as the Budapest Declaration,8 the Berlin Declaration20 and the Bethesda statement,21 which aim not only to give end users immediate access to research articles but also to give them the opportunity to reuse material without restrictions. Moreover, pharmaceutical companies fund medical research with the aim of helping to develop effective treatments that address unmet patient needs; placing restrictions on access to this research is not in line with key ethical principles such as those laid out in the Declaration of Helsinki.22
Good Publication Practice version 3 (GPP3) guidelines state that authors should take responsibility for the way research findings are published.23 In line with these recommendations, pharmaceutical companies can and should advise authors to reach a consensus on which journal to publish with, to avoid predatory journals, and to adhere to sponsor guidelines and regulations. There may be scope for pharmaceutical companies to introduce internal guidelines recommending open access publishing.
Our research shows that roughly one-third of the journals with a high impact factor that we assessed do not offer immediate access to the published version of a manuscript upon publication, even though the open access policies of many funders with respect to embargo periods echo the recommendations set out by open access declarations.8,20,21 The World Health Organization expects trial data to be published in a journal within 24 months of study completion.24 The Research Councils UK policy supports open access, with a preference for immediate access with the maximum opportunity for reuse, which would be granted by a CC BY licence.10 Horizon 2020 requires its beneficiaries to make publications open access as soon as possible upon publication and to make every effort to allow for maximum reuse of the materials, whether that be copying, distributing, searching, linking, crawling, mining or some other use.2 This position is supported by the European Research Council.25
A potential conflict of interest for publishers in relation to their willingness to offer CC BY licences to the pharmaceutical industry is the revenue generated from reproducing article content and reprints. Fees for the reproduction of copyrighted materials can, for example, range above several thousand dollars for using a single chart.26 Recent research by Lundh et al.27 aimed to quantify reprint revenues as a proportion of journal income. Of the six journals investigated, only the two European journals – the BMJ and The Lancet – disclosed the information requested. The editors of the US journals – Archives of Internal Medicine, Annals of Internal Medicine, JAMA and the New England Journal of Medicine – did not provide the data. For the BMJ, reprint revenues constituted 3% of its overall income; The Lancet obtained 41% of its revenue from reprints.27 In The Lancet, industry-funded publications constituted a large proportion of highly reprinted articles (63/88) compared with a sample of control articles from the same journal (23/88).28
Another factor that may be causing publishers to place restrictions on the reuse of materials funded by the pharmaceutical industry could be concerns over the possible misuse or misrepresentation of pharmaceutical data by any end user; however, such activities would be offences under existing compliance regulations and would attract fines.29 In support of credible evidence-based medicine, and owing to the competitive nature of pharmaceutical companies, it is likely that they scrutinize their competitors’ evidence, and identify and report any breaches of good practice to regulators.
Two of the journals included in our analysis, Science and Science Translational Medicine, both published by the American Association for the Advancement of Science, do not disclose article processing charges on their websites; instead, they provide this information upon their acceptance of an article.19 This practice is common among predatory journals and does not comply with the Directory of Open Access Journals guidelines,30 which state that processing fees must be stated clearly on journal websites in a place that is easy to find for potential authors prior to submitting their manuscript.
Future research could focus on a larger selection of journals than the current study, or on journals from a specific therapy area, to further clarify the use of open access variants in the medical publication landscape.
Conclusion
To maximize societal benefit from medical research and to align journal policies with open access declarations, funders of research, including the pharmaceutical industry, should be allowed to have their research read and reused immediately after its publication. This is best achieved by publishing research open access with a CC BY licence, which is recommended as the least restrictive licence available. At present, journals with a high impact factor restrict commercially funded research from being published with this licence, and pharmaceutical companies do not mandate its use.
To align publishing policies with the open access declarations,8,20,21 the Horizon 2020 program supported by the European Research Council,2,25 and the International Committee of Medical Journal Editors31 and GPP323 guidelines, the publishing industry and the pharmaceutical industry must work together. By offering authors the option of publishing their research with a CC BY licence, irrespective of funding source, publishers and journals can support immediate and unrestricted open access to publications to give the global scientific community, patients and the public the opportunity to learn from and build on medical research.
Funding statement
This research was funded by Oxford PharmaGenesis.
Competing interests
Tim Ellison, Tim Koder, Laura Schmidt and Amy Williams are employees of Oxford PharmaGenesis, Oxford, UK. Chris Winchester is a Director and a shareholder of Oxford PharmaGenesis Holdings Ltd.
Author contributions
Conceptualization, project administration, T.E. (https://orcid.org/0000-0003-0307-725X), T.K. (https://orcid.org/0000-0001-6152-7365), L.S. (https://orcid.org/0000-0001-6117-781X), A.W. (https://orcid.org/0000-0002-9354-6402); methodology, resources, investigation, formal analysis, T.E.; writing – original draft, T.E. and L.S.; visualization, T.E; w r it in g – review and editing, T.E., T.K., L.S., A.W., C.W. (https://orcid.org/0000-0003-3267-3990); supervision, T.K., L.S.
Supplemental information
Acknowledgements
Robert Kiley (https://orcid.org/0000-0003-4733-2558) is Head of Open Research at the Wellcome Trust, London, UK, and contributed to the review of this manuscript. Paul Farrow (https://orcid.org/0000-0002-0569-9688) is an employee of Oxford PharmaGenesis, Oxford, UK, and contributed significantly to the review of this manuscript. Sarah Stokes (https://orcid.org/0000-0002-8761-8588) is an employee of Oxford PharmaGenesis, Oxford, UK, and contributed significantly to the review and editing of this manuscript. This work will be presented as a poster at the European Meeting of the International Society for Medical Publication Professionals on 23 January 2018.