Copyright and systematic reviews: do researchers have to break the rules to produce good quality research?

Jane Falconer is a medical librarian with over 20 years experience in medical charities, the NHS and Higher Education. She is currently the User Support Services Librarian at the London School of Hygiene & Tropical Medicine, responsible for all user-facing library services, including user training and support, membership, access and enquiry support, reading lists, interlibrary loans and liaison services. She also provides literature searching support for systematic reviews, she has created and run the searches on a number of projects including the Lancet Commission on Planetary Health and the WHO Guidelines on Heptatitis B and C Testing. ORCID: http://orcid.org/0000-0002-7329-0577. Jane got in touch via Twitter as she was frustrated by copyright laws that prevented her sharing medical articles with researchers around the world. Here’s what she told us…..

My name is Jane Falconer and I need advice from the copyright community.

In order to stop people dying I have been asked to upload a large number of academic articles to the cloud, so that six people situated in different institutions around the globe, can synthesise the information and recommend the best course of action. Do I do this, and break all sorts of copyright and licensing restrictions in the process?

This may sound melodramatic, but this situation happens all the time when medical researchers collaborate on a particular type of study called a systematic review.

A systematic review is a specific type of literature review, with a set methodology designed to provide an unbiased answer to a question, based on a summary of all the currently available studies. They were first developed in medicine in the 1970s, where doctors were being asked to make life-changing decisions based on scattered and contradictory evidence. The systematic review was intended to bring this evidence together to produce a more clinically and statistically significant conclusion. They are now used by health professionals around the world to ensure they implement best current practice. They are also used to produce documents such as medical guidelines which provide national or global medical practice advice.

Photo licensed under CC-0 from: https://www.pexels.com/photo/blue-skies-clouds-cut-fingers-335907/

Given the importance of reliable information to frontline health professionals, it is important that systematic reviews are conducted according to recognised and tested methodologies. Two of the best known are PRISMA and Cochrane. PRISMA is commonly used by authors to guide their methodology and reporting and they provide the evidence to back-up their methodology. Where it gets interesting from a copyright viewpoint, is when researchers are deciding which items to include in the systematic review, and which to exclude. They recommend that:

“Efforts to enhance objectivity and avoid mistakes in study selection are important. Thus authors should report whether each stage was carried out by one or several people, who these people were, and whether multiple independent investigators performed the selection, what the process was for resolving disagreements. The use of at least two investigators may reduce the possibility of rejecting relevant reports. The benefit may be greatest for topics where selection or rejection of an article requires difficult judgments.”[my emphasis][i]

The Cochrane Library (named after Archie Cochrane, the main advocate for systematic reviews) is generally regarded as producing the gold standard of systematic reviews. Their methodology has been adapted and is now used to produce systematic reviews in other subject areas including social sciences and ecology. When we look at their screening methodology we see that they are far more explicit than PRISMA:

“The methods section of both the protocol and the review should detail:

  • Whether more than one author examines each title and abstract to exclude obviously irrelevant reports:
  • Whether those who examine each full-text report to determine eligibility will do so independently (this should be done by at least two people) [my emphasis]”[ii]

Therefore, we need at least two people to have access to the same items. This is relatively easy when the researchers all belong to the same institution which has a well-funded library with subscriptions to all the relevant articles. In my experience, this has never happened.

Photo licensed under CC-0 from https://www.pexels.com/photo/black-and-white-blood-pressure-gauge-business-computer-263370/

The reality more closely matches the situation described at the beginning of this blog. Let me put some more detail on that, loosely based on a real study. An international group of 6 researchers have been commissioned by the World Health Organization to conduct a systematic review on a topic, let’s say the efficacy of low-cost diagnostic tests for sexually transmitted diseases. The production of these tests is important in resource poor countries where current tests requiring samples be sent to a lab for analysis are too expensive to roll out to the entire population. If the review can show that these new tests are sufficiently sensitive and specific, the review will form a key part in the production of new guidelines which will update global recommendations. This will lead to more people being successfully diagnosed and treated, with a subsequent reduction in infection rates. The researchers are based in institutions in China, Tanzania, Brazil, USA and UK, in order to ensure regional differences are reflected in the study. To facilitate communication, the lead author creates a Dropbox folder, to share project-specific documents. Given the complexity of the topic, I am commissioned to conduct the literature search and help source the required information. I conduct a search of the literature, searching 12 bibliographic databases and several sources of grey literature, being deliberately broad and comprehensive to limit the possibility that a study has been missed. Once the results are de-duplicated, 6348 items are identified for screening. I export the titles and abstracts to Excel and upload to the Dropbox folder where the team exclude any obviously irrelevant items. We are left with 301 items where the abstract does not provide sufficient information to include or exclude. Therefore, the team need to read the full-text of all 301 papers. The researcher in the USA is based at an Ivy league institution, the researcher in Tanzania does not have access to a library at all, the researchers in China and Brazil only have access to a limited number of western journals and some local titles. Therefore, the lead researcher asks me to source the full-text of these items and upload them to Dropbox.

How can I do this without breaking copyright and licensing laws?

  • 25 items are available open access. These are the easy ones to deal with and everyone has access to these.
  • 150 items are available through my institution’s e-journal subscriptions. Can I upload these to the cloud for others in the research group to access, even if they are not affiliated with my institution?
  • 50 items are available through the subscriptions of the American institution. Are US licences and copyright restrictions different to those in the UK?
  • 2 items are from the 1970s and are available in print in my library. Can I scan these and upload them to the cloud for others to view?
  • 99 items have to be ordered through our inter-library loans system. This is where it gets really difficult. If I order these from the British Library, they are linked to my BL account. I either have to share my account details with the group, or print them, re-scan without DRM and upload the papers. The National Library of Medicine in the USA sends PDFs of items without DRM but is clear that the documents should not be shared, copied, or stored in electronic format.

It seems to me that copyright and licensing restrictions are at loggerheads with study methodology. I cannot see any way the study can continue without breaking the rules. However, not completing the research to the highest methodological standard is not an option. Compromising the quality of the study risks drawing the wrong conclusion, causing potential harm to a large number of people, as well as damaging the reputation of the researchers and the institution.

I was commissioned to write this blog as the result of a twitter conversation I had with UK Copyright Literacy over a blog post on the ACRLog. It details a discussion around why academics consistently break copyright and licensing rules. It’s a good post, which brings up many important points. However, nowhere does it discuss the realities of trying to do research in an interconnected, digital world where funding grants are awarded for groups who collaborate between institutions, and money is specifically available for collaborative projects in resource poor countries (particularly where they have a capacity building component). I believe that copyright and licensing rules have not kept pace with technology and the reason many academics break the rules is because it is impossible to complete their research without doing so.

As a librarian I am caught in the middle. Can we tell academics that in order to conduct a systematic review, the balance of risk is such that they can break the rules to complete their study? If so, how can we then tell the same academic that they can’t email a PDF to someone, or upload a chapter to their course site? Like many others, we are working with academics to encourage them to publish open access. However, this is a solution for newly published information, it doesn’t help people like myself and my research group who are trying our best to source already published information in order to improve health across the world.

I’d be really interested to read everyone’s advice and comments.

[i] Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, et al. (2009) The PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses of Studies That Evaluate Health Care Interventions: Explanation and Elaboration. PLOS Medicine 6(7): e1000100. https://doi.org/10.1371/journal.pmed.1000100. Pg 9.

[ii] Higgins JPT, Green S. (2011) Cochrane Handbook for Systematic Reviews of Interventions version 5.1.0. [online] http://handbook-5-1.cochrane.org/ [accessed 31 January 2018]. Section 7.2.4.

What do you think? Are you a publisher or rightsholder and do you have any thoughts about the issues that Jane has raised? We’d love to hear from you and publish a response to this blog post, but also to ask other librarians how they deal with these sorts of tricky issues related to systematic reviews. 

19 Comments

  1. Hi Jane (and Jane!)
    I read your blog with interest. I think if you worked for a large pharma company with a CCC or CLA Multinational Licence, and you were sharing articles with colleagues globally, you’d pretty much be able to do everything you describe – with a few limitations admittedly. Which means it’s likely that collective licensing could solve a lot of your problems – there just isn’t an off-the-shelf licence that suits the Cochrane Library at the moment. Perhaps you could get in touch with me at james.bennett@cla.co.uk and we can have a detailed conversation about whether CLA could help?

    James Bennett
    Head of Rights and Licensing
    Copyright Licensing Agency

    Like

  2. Thanks for highlighting the real realities of doing research in an interconnected, digital world. Help us fix the problem! Join EIFL, IFLA and other groups advocating at WIPO for an international copyright treaty to enable cross-border access to knowledge. What we say at WIPO, for example: http://bit.ly/2BQZScX

    Liked by 1 person

  3. Thanks for clearly summarising one massive challenge around collaboratively produced systematic reviews.

    I’m wondering about CLA licensing and potential development – my understanding is limited because it’s a new area to me so please forgive me if I’ve misunderstood, but – if organisations both hold CLA licenses then they can share articles with each other, but this doesn’t extend to international collaboration. However, CLA has just extended its overseas campus based students pilot – and as far as I know this isn’t for academics but just for students. Perhaps this is an avenue to explore for expansion?

    Like

  4. Load the papers to Dropbox and leave it to the CLA copyright police to work out if you’ve infringed their rules and they want to seek redress. And if they do let the world know.

    Like

    1. Thanks for the comment Ben, that would clearly be a bold act of civil disobedience that not everyone would be comfortable taking. Although the CLA has responsibility for licensing the copying of published works in certain contexts they aren’t actually “the copyright police”. Not to say that they don’t have an enforcement role, but they represent certain rights for certain publications. The issue that Jane has identified here is with the patchwork of different licences that underpin access to the literature, of which the CLA’s is one. I would say this makes a pretty convincing case for open access.

      Liked by 1 person

      1. Hi Jane, thanks for flagging up the particular copyright difficulties faced when carrying out a systematic review – you have my sympathy! While CLA does offer licences to multinational organisations that permit project-based cloud storage of journal articles, these are aimed primarily at pharmaceutical companies at the moment. I’d be very happy to have an offline chat with you about your particular situation and how a collective licence may be able to help you and other organisations in the same position. Do please get in touch with me at james.bennett@cla.co.uk

        Like

  5. Hi Jane, thanks for flagging up the particular copyright difficulties faced when carrying out a systematic review – you have my sympathy! While CLA does offer licences to multinational organisations that permit project-based cloud storage of journal articles, these are aimed primarily at pharmaceutical companies at the moment. I’d be very happy to have an offline chat with you about your particular situation and how a collective licence may be able to help you and other organisations in the same position. Do please get in touch with me at james.bennett@cla.co.uk

    Like

  6. Excellent, but very sad, article. I empathize with your phrase “In order to stop people dying … [I have to break the law]”. This is the language we should use. Some years ago I coined and promoted the phrase “Closed Access Means People Die”. I got some support, some hostility (“you haven’t done a controlled study” – no it’s a political truth, not a scientific one ). But deafening apathy.
    The sad and simple fact is that academia does not care. I have been banging this drum for years. The Liberian government was appalled that the paper in 1983 predicting the outbreak (“Liberia should be regarded as an Ebola endemic zone”) was still paywalled. The NY times cared. But academia doesn’t.
    Should you break the civil law? If I say you should, I am technically breaking the criminal law (“Incitement to commit a felony”) and could go to jail. Universities and their Libraries take the view that Copyright is supreme and act as police for the publishers against the researchers. Publishers base their models on restricting access to material.
    I and colleagues work with systematic reviewers – we know one who had to sift through 50,000 papers for one review. We have developed Open software to help remove the burden of human review (please contact if you are interested!) . But we know of researchers who have been cut off and vilified for “downloading too many papers” and for using automated methods.

    P

    Like

  7. Hi Jane,

    a very serious issue indeed. Open Access may help, but its limited still and not retrospective.

    But there is an alternative route of action.

    I would certainly support an initiative where we draft a charter asking publishers to commit to supporting the practice of sharing full texts for systematic review purposes. It should be an exemption on their copyright statements.

    Start with a list of all publishers in red and after a publisher has filed its support and publicly states on a webpage (to be archived) on their website that they permit this by creating an exemption and that they will not press charges, make them green (or more colour blind friendly colours).

    Ask all publishers listed to support it. After that publicly state that the red ones are knowingly obstructing the development of health care. Make sure that you got a few green ones to start with.

    Let’s ask Médecins Sans Frontières, WHO and other organisations to support such a campaign.

    Like

    1. Hi Jeroen. Thanks for the suggestion. I think it sounds like a really interesting idea and we had comments from STM on twitter saying their sharing principles were not really designed to support this. But perhaps they could be? I was thinking of asking IFLA and EIFL for their thoughts.

      Like

  8. I would just tell the authors to go get the papers from sci-hub themselves. What they do on Tor in their own time does not implicate you in their copyright infringement.

    You could try and do this the ‘right’ way as other commenters here have argued, but is it really worth your time?

    Like

    1. Thanks for your comment Joe. Most information professionals would avoid directing people to infringing sites because this could be seen as authorisation of infringement. Not only is it likely to conflict with institutional policies, it is also likely to conflict with a strong ethical code to ‘do the right thing’.

      Of course there is a debate to be had over what ‘the right thing’ is – is it about complying with the systems, or challenging them? Jane and I explored some of this in our research, particularly “Understanding librarians’ experiences of copyright: findings from a phenomenographic study of UK information professionals”, Library Management , https://doi.org/10.1108/LM-01-2017-0011

      Have a look at Jane S’s follow up post at https://copyrightliteracy.org/2018/02/12/copyright-and-systematic-reviews-a-response/ to see where we’d like to go next with this.

      Like

  9. I don’t know about copyright laws in the UK, but in the US it would almost certainly be fair use for SR collaborators from multiple institutions to share full-text articles. Section 107 (*https://www.copyright.gov/title17/92chap1.html#107) of the Copyright Act explicitly states that “reproduction in copies” of copyrighted words “for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright.”

    If writing a systematic review is not an example of “scholarship” or “research” then I don’t know what is.

    It would simply be a major hindrance to scientific progress and especially to writing SRs if project collaborators could not use the same set of PDFs or freely share them.

    Ben Harnke

    Education/Reference Librarian
    University of Colorado, Health Sciences Library

    Like

    1. This is an interesting point. Amongst other things, it indicates the extent of confusion about copyright law, in this case the difference between ‘fair use’ and ‘fair dealing’. As a result, in the discussion we are having here, something suggested as acceptable in the US has been interpreted as a bold act of civil disobedience in the UK.

      The wish among librarians not to transgress perceived restrictions is understandable. But it is also unfortunate. It wastes the time of users and librarians, and it involves the library community devoting precious resources to indemnifying itself against largely imaginary self-imposed restrictions.

      To spell it out. It is in the interests of agencies who manage copyright for uncertainty to persist. Librarians have to operate with this uncertainty. But, because there is no legal definition of fair dealing, it is not principally by seeking advice from copyright agencies or publishers or colleagues that librarians will get clarity. It is through reasoned and reasonable action – in this case uploading articles to a shared server as part of the research process – that progress will be made and clarification gained. The public funds librarians to deal effectively on a day to day basis with copyright on its behalf. At the moment the evidence suggests this isn’t happening.

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.