Policy in Development: Access to Outputs of Research

Survey Report - June 2006


Introduction

CIHR recognizes the importance of disseminating research results, data and materials to further advance scientific research and discovery. CIHR believes that the research process is not complete until the results are validated and openly transmitted to the appropriate audience. Access to research outputs, products and materials is imperative for reproducing results and maximizing the impact of these discoveries.

In February 2006, CIHR initiated a project to develop and implement a CIHR policy on access to research outputs. This work is being guided by an advisory committee whose members represent Canadian researchers across CIHR's four research theme areas (i.e., biomedical, clinical, health services, and population health). With the purpose of gathering information to support the committee's work, CIHR surveyed the health research community on a number of topics related to access, using a questionnaire that was posted on the CIHR Website in April 2006.

This report presents the results of survey, including key findings and some specific comments from respondents. In general, respondents supported the idea of CIHR establishing on access policy, and, in particular, a policy that would be broad and place minimal restrictions on access. Many respondents encouraged CIHR to look at the policies of other national and international research funding agencies in developing its policy.

Survey results

1) Respondents

We received 41 responses to the survey. Of the responses received, approximately half were from researchers with the rest coming from library affiliates, journals, research administrators, policymakers, health charities and others.


2) Are there any specific research resources, tools, and products that you think should be included in this policy (e.g., software or protocols)?

Key Finding: Generally there was agreement to include genomics and bioinformatics tools, software, protocols, algorithms, open-edit media, source code for software, survey instruments, measurement scales, and manuals for patient services.

Specific Comments:

  • CIHR should include everything, but with the caveat that intellectual property rights and proprietary issues are addressed.
  • Social science resources should be included like workshops, public reports, and pamphlets.
  • Researchers should store protocols in World Health Organization accredited clinical trial registries.
  • CIHR should enable researchers to provide sequences instead of primers, because of time and cost issues. Particularly if sequences can be readily ordered.
  • Concern over any intellectual property issues affecting publication.

3) How can we ensure that a policy increasing access to the physical products of research does not negatively impact the IP rights of inventors, and the commercialization of IP?

Key Finding: Most respondents felt that the protection of IP was important, but that a timeline for access should be stated.

Specific Comments:

  • CIHR's policy should not conflict with existing IP rights.
  • Accessibility, public interest, research integrity are paramount, IP rights of inventors is important but not as important.
  • Limit access to products that are currently under patent review or the equivalent.
  • Add a section to grant applications documenting how IP will be protected versus providing access.
  • A clear timeline should be specified by which researchers should file for patent protection from the time the public funds were given to them. After this timeline, researchers should provide materials regardless.
  • Allow for deferral of 6 months from date of disclosure to public access to allow for filing of IP protection which time may be extended upon request.
  • There should be an "opt-out" or deferral clause so that data contributing to commercialization of IP (e.g., patent or provisional patent applications) can be embargoed until such time as the patent or provisional patent is issued. It may be necessary to include maximum time limits for this deferral to prevent this policy from being abused.

4) When is it appropriate for researchers to put restrictions on products of research?
and
5) Can you suggest appropriate restrictions that will minimize harms (or maximize benefits) to further research?

Key Finding: Most respondents felt that restrictions should only be imposed in the case of innovations that are to be patented, but should do so under stated timelines.

Specific Comments:

  • Restrictions on access might be required if such access would affect commercialization.
  • Whenever confidentiality is at stake or when consent has not been given.
  • Restrictions are acceptable when there is a security risk.
  • Never, because publicly funded research is for public consumption and benefit.
  • Limitations on how the products are used and by whom they are used would be appropriate. Any limitation should have clearly-defined timelines.
  • It is tempting to require that Material Transfer Agreements be signed by both parties for each transaction. However, this may be onerous if many requests are received per year. Thus, if MTAs are required, there should be an option for the material provider to opt out should they wish (e.g., by signing a waiver). For some materials (e.g., plasmids), many researchers simply do not want to be bothered with MTAs because they do not perceive the material to have significant IP value.

6) Is there a specific type of data that should be covered with this policy statement?

Respondents identified the following types of data:

  • Genomic data;
  • Data specific to all other pillars (e.g., social science data);
  • Raw data;
  • All research data consistent with confidentiality agreements.

7) Would you support a policy statement that involved sharing research data obtained with the help of CIHR funding? Please elaborate on your answer in the comments section.
and
8) If you answered No to question number 7, please explain

Key Finding: Respondents supported the inclusion of research data in the policy, but suggested CIHR move forward with caution in this area.

Specific Comments:

  • Sharing is useful, but simply spreading around unreviewed, unsummarized and uncontextualized data would not be particularly useful. Data should have some form of validation such as peer review for papers to ensure quality control.
  • Concern that Open Access to data would negatively impact commercialization. There would be an ability to withhold data.
  • CIHR should reward 'good' behaviour. For example, on grant renewal applications, could ask people what they'd done to share their research in substantive ways (e.g., providing tissue samples to collaborators, or preparing a plain language summary and presenting it to a health care managers group or policy makers).
  • CIHR should not make this a sweeping requirement.
  • Data from CIHR-funded clinical trials should be shared and made public via a registry.
  • It is imperative that control over the materials/data, especially if commercializable, remain largely with the producer of the material.
  • Suggest researchers deposit data into international databases to which Canadians have access.
  • Will this policy address the sharing of "primary data" (e.g., meta-analysis)?

9) Do you think this policy should cover products other than peer-reviewed publications, such as book chapters, editorials, reviews, or conference proceedings?

Key Finding: Most respondents would like CIHR to include other types of publications in this policy.

Specific Comments:

  • Everything CIHR sponsors regardless of the medium should be openly accessible.
  • CIHR should mandate open access to peer-reviewed articles (and peer-reviewed conference papers). Open access to the other kinds of publications should be strongly encouraged.
  • There is concern that this may be too much work for grant holders.
  • Mandating open access may be difficult for certain types of publications like book chapters, which consist of variable amounts of opinion, and generate revenue.
  • Open access to these publications would provide a better return on investment, would reduce duplicative research and focus energies towards new research.
  • CIHR could also consider this as a mechanism to police fraudulent researchers.
  • All "grey literature" should be included. Other materials could include audio, video and image files.

10) Do you support self-archiving of peer-reviewed research publications an Institutional Repository (IR) at a Canadian university?

Key Finding: A majority of respondents supported self-archiving of peer-reviewed publications in appropriate Institutional Respositories.

Specific Comments:

  • CIHR should consider available substitutes: 33 institutional repositories.
  • Little experience with using institutional repositories and/or unaware of self-archiving.
  • CIHR should setup a Canadian central repository for archiving publications.
  • Researchers expressed concern that there was no time for curation.
  • Who will pay for archiving?
  • IRs or a Canadian central IR would be preferable, as long as it followed the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH) standards.
  • CIHR should mandate self-archiving of funded research into institutional repositories.
  • CIHR should ensure manuscript preservation in a digital IR, that is free and interoperable.

11) Would you prefer self-archiving of research publications using your personal web site?

Key Finding: Most respondents would prefer institutional repositories because they are interoperable and better preserved.

Specific Comments:

  • CIHR should mandate archiving in institutional repositories.
  • The lifetime of a personal web site is limited due to retirement, lack of resources, or other reasons.
  • Both repositories and web sites should be options.
  • National resource needs national presence.
  • Little or no experience archiving on a personal web site.
  • Personal Websites are not recommended because they are not interoperable. Most are not included on well-known search engines such as OAIster, Google Scholar and Microsoft Live Academic Search.

12) Would you endorse archiving of peer-reviewed results in the National Library of Medicine's PubMed Central? Or, would you prefer that Canadian peer-reviewed results be archived in a Canadian repository?

Key Finding: Respondents rank order of preference: PubMed Central > both PubMed Central and a Canadian repository > Canadian repository > Institutional repositories. Primary reasons for preferring PubMed Central over a Canadian repository were the fact that PubMed is well established and that CIHR should reserve its funds for supporting research.

Specific Comments:

  • PubMed Central already exists and has high visibility on the national and international scene.
  • No need to duplicate effort (and costs). Those funds are better used for the research itself.
  • If interoperable, it doesn't matter. PMC or a Canadian equivalent could be the fall-back for researchers without an IR at their institution.
  • The archive should not be as medically oriented as NLM. Not everyone publishes in journals covered by PubMed.
  • Support something Canadian, if it can be done inexpensively.
  • A similar Canadian counterpart that functioned in the same way. Ideally, links from PubMed would link to the full text in the Canadian repository, just like those that currently link to PubMed Central.
  • Peer reviewed results should be archived in the author's own institutional repository that can then be harvested by central repositories such as PMC.
  • CIHR should carefully consider federal and other policy constraints that may mitigate against using an American-based repository.
  • Issues at the level of the data should also be addressed, for example, making provisions for languages other than English and providing tags within the metadata that indicates country of origin.
  • The advantage of a Canadian repository are that it could: provide bilingual abstracts (not full-text); ensure that Canadian copyright is applied appropriately; contain Canadian-specific metadata allowing for searching and reporting by Canadian criteria - e.g., province, organization, granter; ensure long-term preservation of and access to Canadian content; and ensure that any publications submitted to an open-access archive remain open access to the Canadian public.

13) As a journal editor, or representative of a professional scientific society, what are potential positive or negative impacts that you see with the implementation of a policy requiring CIHR funded researchers to follow one of the aforementioned mechanisms?

Key Finding: Most respondents felt that the benefits outweighed any negative aspects.

Specific Comments:

  • There are only positive benefits, for research impact and impact factor.
  • The policy may negatively impact on the viability of journals and should not detract from the journals' ability to raise revenue (from subscribers or from advertising).
  • Lost revenues and loss of "prestige" associated with open access nature of archives.
  • Greater visibility for all research; access for health care professionals not affiliated to institutions with subscriptions; access for the public.
  • A potential negative impact may be the time commitment, which in reality is minimal.
  • CIHR should encourage access to drug company funded clinical trials.
  • As an editor of a free journal, I can only foresee a positive effect of CIHR requiring archiving on PubMed as it is likely to increase the number of high quality submissions.
  • The concerns of editors and societies running subscription-based journals are well known and predictable - they obviously fear to lose subscribers. The real question is do we really want to continue to fund over 40.000 STM journals with library budgets and public money in the age of the Internet? To put it bluntly: Any journal which can be replaced by an institutional repository deserves to be.
  • Speaking as journal editors, we would be cognisant of the fact that it is generally accepted that open access increases the impact of the research, including the citation rate. Open access offers a better return on investment on publicly-funded research. Publicly-funded research can be accessible within public institutions, without those institutions having to spend public monies to private parties for access to that research. Editors see an internationalization of contributors with open access which is an advantage for regionally-based journals such as publicly funded journals in Canada.
  • Concern that if existing journals went OA this would mean charging very large publishing fees.

14) While considering IP rights and the commercialization of IP, what should be the minimum time required for release of these forms of data into the public domain?

Total number of responses: 34

Responses:
Immediately (10)
Three months (6)
Six months (8)
Other (10)

Other Responses:

  • Immediate deposit; 3-6 months for OA release.
  • After protection of IP.
  • As soon as ready.
  • As most appropriate for commercialization.
  • As soon as first patents are filed.
  • One year.
  • 12 months may be acceptable to journals.

15) If CIHR were to mandate self-archiving of peer-reviewed publications, how long after publication should this occur?

Total number of responses: 34

Responses:
Immediately (19)
Three months (6)
Six months (2)
Other (7)

Other Responses:

  • Immediate deposit and release of metadata.
  • Consultation with journals required.
  • One year.
  • Terrible idea.
  • Don't think it is necessary.
  • Researchers shouldn't be curators.
  • As the national science library, we recommend immediate deposit and availability in an IR.

16) Can you think of any responsibilities or obligations for those requesting CIHR materials? (e.g., acknowledgements)

Key Finding: Generally it was suggested that there should be appropriate citation and acknowledgements.

Specific Comments:

  • Requested materials should not be used for commercialization unless otherwise granted permission.
  • Requestors should agree to use materials in accordance with best scientific practices.
  • Privacy rights must be respected.
  • Limitations on acceptable use - i.e., not for profit; acknowledgements of both the original researchers and the archive required.
  • The costs of providing the materials (e.g., shipping, perhaps production - depending on the material) should be borne by the requestor.

17) How do you see co-funding influencing access to physical products of research?

Key Finding: Most respondents thought that co-funding should not influence access.

Specific Comments:

  • CIHR's policy should apply to all research funded in whole or in part by CIHR funds similar to NIH.
  • The co-funder needs to know the consequences of partnering before they commit resources.
  • May have an adverse impact on students who need the data/results for their theses.
  • Concern that conflict of interest could significantly restrict access.
  • If the research is sponsored by an organization that requires development of IP, then one should not impose on them.
  • If industry is co-funding, they should have first right of commercialization refusal to partner with the researcher on the physical products of the research.
  • Restriction has to fall within particular categories currently accepted in courts of law (e.g., national security, corporate confidentiality), and for which a reasonable time limit is set.

18) Please comment on any experiences with other organizations, both nationally and internationally, regarding sharing or access to resources, data, and publications? Do you have suggestions or comments that CIHR should consider during policy development?

Key Finding: The Wellcome Trust position statement in support of open and unrestricted access to published research was most often cited as the gold standard. Many respondents felt that the NIH Public Access policy was unsuccessful because of voluntary compliance.

Specific Comments:

  • The NIH ask for voluntary compliance (with self-archiving of publications), and its compliance rate is a dismal 3.8%. The NIH now recognize that a mandate is necessary to achieve compliance.
  • Recommend using The Wellcome Trust policy as a model.
  • NIH did it wrong; Wellcome Trust did it better; RCUK might do it best, if it mandates OA self-archiving in author's own IR immediately upon acceptance for publication.
  • The NIH experiment has not been a success - voluntary deposition may not work.
  • The experience with other funding bodies is that encouragement is not helpful - open access should be mandated, and a short time frame given.
  • Some organizations have a mandate to place what might be considered important information into the public domain without restriction on use. It has neither impeded science, the subsequent development of IP or our ability to publish.
  • CIHR should refer to the International Council for Science: Committee on Data for Science and Technology (CODATA) with regard to issues surrounding access to data. The NIH "delayed access" policy delays the speed of important scientific discovery, particularly in medical research.

Other Comments and Suggestions:

  • CIHR should become a signatory to the Berlin Declaration on Open Access to Knowledge.
  • CIHR should consider the recommendations of National Consultation on Access to Scientific Research Data and work with others to implement a national infrastructure for research data in Canada.
  • Canada needs a made in Canada research data accessibility/maintenance policy.
  • CIHR needs to do a better job of working together with other jurisdictions rather than developing incompatible policies that make research and its dissemination even more difficult.
  • Concern over repositories/databases containing anonymous patient information collected to study natural history of disease or genotype-phenotype correlations.
  • Would like to see greater support and help in publicizing journals such as the CMA's Journal of Psychiatry and Neuroscience which support CIHR policies about open access and are Canadian.
  • CIHR should consult with Canadian universities heavily during policy development.
  • A "moving wall" model with openness after 6 or 12 months has been implemented by many journals/publishers, but it should be realized that such "delayed" open access slows down knowledge dissemination and uptake within the scientific community.
  • CIHR could maximize research benefits by including research users on review panels (e.g., health care decision makers).
  • Material transfer agreements cause long delays between the transfer of research materials.
  • CIHR should provide incentives for grant holders to publish in open access journals.
  • CIHR's grants should include a non-exclusive reservation of rights to disseminate research papers.
  • CIHR should provide free online access to each CIHR-funded manuscript as soon as possible, and no later than six months after the article has been published in a peer-reviewed journal.

Conclusion and next steps

The responses to this survey have provided valuable feedback and many useful suggestions that will undoubtedly encourage internal and advisory committee discussions. This information will be of great assistance in the development of a draft policy on access to research products. CIHR anticipates posting a draft policy on its Website for consultation during the latter part of this summer.

CIHR continues to welcome comments and suggestions. Please direct them to:

Geoff Hynes
Research Officer
President's Office/Corporate Planning and Policy
Canadian Institutes of Health Research
Tel.: (613) 952-8965
Fax: (613) 941-1040
E-mail: geoff.hynes@cihr-irsc.gc.ca

Appendix

Invitation to provide advice and suggestions for CIHR's Policy in Development - Access to Products of Research.