Cochrane Collaboration - policy on commercial sponsorship

Introduction

The Steering Group of The Cochrane Collaboration has undertaken a process of consultation on commercial sponsorship. The current debate was stimulated by a letter from several members of The Cochrane Collaboration who felt that existing policy ought to be more restrictive - to provide still greater reassurance that the conclusions of Cochrane reviews were not biased through the influence of funding by commercial entities that stood to benefit financially from the results of reviews.

Commercial sponsorship of health-related research is, of course, not an issue of concern uniquely to The Cochrane Collaboration.

Many members of The Cochrane Collaboration have pointed out that external perception is also important. Any perception that for-profit commercial organisations, notably but not exclusively, the pharmaceutical industry and medical device manufacturers, were influencing the conclusions of Cochrane reviews would damage a carefully nourished reputation for impartiality and scientific rigour.

This issue was discussed at length at the 11^th annual Cochrane Colloquium in Barcelona in October 2003. A consultation document was disseminated during December 2003 with a request for views by 31 January 2004; 156 individuals or groups responded. Most were active members of The Cochrane Collaboration. The Steering Group met in Bergamo, Italy, from 29 February to 2 March 2004 and considered at length the very extensive and detailed documentation. As described below, for some questions, there was very clear consensus; for others, there was not.

Background

Since the decisions taken by The Cochrane Collaboration are also of interest to others it may be helpful to describe, briefly, the structure of The Cochrane Collaboration. It is a highly devolved organisation that involves more than 10,000 people, in different capacities, worldwide. Most do not receive any payment for the work they do within The Collaboration. They are drawn to The Collaboration through a wish to commit, either as a professional or as a consumer, to a movement to provide more sound evidence on which healthcare decisions can be made. The formal structure of The Collaboration comprises Collaborative Review Groups (which produce systematic reviews), Centres (with responsibilities that include support for Collaborative Review Groups within their area of geographical responsibility), Methods Groups, Fields, a Consumer Network, an elected Steering Group, and a small Secretariat. The Secretariat, Steering Group and Advisory Group meetings, and key generic developments (e.g. software for information management, production of the Cochrane Reviewers’ Handbook, and development of The Collaboration’s web site) are all funded, in part or in whole, through royalties on sales of The Cochrane Library. Everything else (including support of Collaborative Review Groups and Centres) is funded through applications to other sources (often government agencies), and these sources are almost all in the country in which the entity is located.

There is substantial variation internationally in the amount of funding for support of Cochrane activity and, in some parts of the world, it is extremely difficult to access government or charitable funds. In some areas, there has recently been an important decrease in financial support for Review Groups and Centres. Therefore, an alternative option, of seeking funding from commercial sources, could be attractive to, say, Co-ordinating Editors of Review Groups, or Centre Directors, who otherwise face the prospect of curtailing productivity and/or making skilled and experienced staff redundant. Setting policy on issues as sensitive and important as sources of funding in as complex an organisation as The Cochrane Collaboration is never an easy matter, and may be even more difficult at this time.

Definitions

By ‘commercial source’ we mean any for-profit manufacturer or provider of health care, or any other for-profit source with a real or potential vested interest in the findings of a specific review. Whilst government departments, not-for-profit medical insurance companies and health management organisations may find the conclusions of Cochrane reviews carry financial consequences for them, these are not included in this definition. Also not included are for-profit companies that do not have real or potential vested interests in Cochrane reviews (e.g. banks).
By ‘sponsorship’ of a review, we mean a sum of money given to a review author or group of review author to prepare, or update, a Cochrane review. Such sponsorship could include not only commissioning of specific systematic reviews, but also, for example, funding of a sabbatical period to work on a Cochrane review.
We used the term ‘firewall’ in the consultation document. By this, we mean, figuratively, a fireproof wall put in place to ensure that, if a fire occurs, it is confined to one area. We used the term to indicate a clear barrier or separation between a source of funding and the use to which that funding is put, so as to prevent any influence by the funding source on the outcome of, say, a Cochrane review.

Conclusions

There was overwhelming consensus that there should be a clear barrier between the production of Cochrane reviews and any funding from commercial sources with financial interests in the conclusions of Cochrane reviews.
Thus, sponsorship of a Cochrane review by any commercial source or sources (as defined above) is prohibited.
Other sponsorship is allowed, but:

• A sponsor should not be allowed to delay or prevent publication of a Cochrane review.

• A sponsor should not be able to interfere with the independence of the authors of reviews in regard to the conduct of their reviews.

• The protocol for a Cochrane review should specifically mention that a sponsor cannot prevent certain outcome measures being assessed in the review.

4. These rules also apply to ‘derivative products’ (containing Cochrane reviews) so that commercial sponsors could not prevent or influence what would be included in such products.

5. To ensure the integrity (real and perceived) of the ‘firewall’, it is also prohibited for a commercial source or sources (as defined above) to sponsor Cochrane entities that produce Cochrane reviews, that is, Collaborative Review Groups.

6. It was agreed that these same restrictions should apply to Fields and to the Consumer Network because of the close proximity of these entities to review production.

7. It was agreed that commercial sources of funding to Methods Groups should not be prohibited. However, the Screening and Diagnostic Tests Methods Group needs to be considered as a special case because of its likely close involvement in the preparation and maintenance of Cochrane reviews of diagnostic test accuracy. The Funding Arbiter (see below) should be asked to advise on those situations that are not clear-cut.

8. The situation with regard to Cochrane Centres is more complex than for other Cochrane entities. For example, Centres can be both close to review production (like Fields and the Consumer Network) but can also engage in methodological work (like Methods Groups). It was agreed, therefore, that a further, short, period of consultation should take place specifically in relation to the sponsorship of Cochrane Centres by commercial sources.

9. Some entities may find themselves in financial difficulty because of the need to shed current commercial funding. Therefore, although this policy is mandatory now in relation to any new funding, it will become mandatory in relation to existing sources of funding two years after the date of adoption, to allow time for entities to seek alternative sources of funding. If any entity has contractual obligations that mean that they cannot shed current commercial funding within the next two years, they should discuss this urgently with the Funding Arbiter.

10. The position of ‘Funding Arbiter’ will be established, analogous to the Publication Arbiter. The Funding Arbiter will be a Steering Group member and will convene a standing panel of three to give guidance on difficult cases.

11. The responsible Collaborative Review Group should refer any existing Cochrane reviews that have been produced by a process that would no longer be permissible to the Funding Arbiter. A decision will be taken within the first twelve months of the implementation of this policy to consider what should happen to these Cochrane reviews (e.g. whether they should be withdrawn from The Cochrane Library).

12. Authors of reviews should declare financial support for the review, private clinical practice (if relevant), stocks, legal advice, consultancies, involvement in primary research in the subject area of their review, and any other ‘competing interests’ that they judge relevant.

13. Such declarations will be described in the review. The declarations will not be published outside of the review itself, for example with the abstract or synopsis.

14. If an author has been actively involved in a study/studies that was/were eligible for their review, they should have, as a co-author, someone who was not involved in the study/studies). The co-author would not necessarily be the contact author for the review, but could act as a ‘guarantor’.

15. If a review has been done, or is proposed, by people who are employed by a pharmaceutical or medical devices company that relates to the products of that company, it will be referred to the Funding Arbiter. In such circumstances, The Cochrane Collaboration will insist on a multi-disciplinary review team with a majority of the team of review author not being employed by the relevant company.

16. People with a direct financial interest in a particular intervention should not be involved in a review of that intervention, either as review authors, editors or peer review authors.

17. It was agreed to establish a central fund or Foundation into which unrestricted donations could be made. It was further agreed that there should not be a prohibition on donations from any single company or type of industry but that all funding of activity in The Cochrane Collaboration should be in keeping with the principles of The Cochrane Collaboration.

18. There is an existing Collaboration policy on sponsorship of Colloquia. The Colloquium Policy Advisory Group will be asked to reconsider this in light of changes to the policy on commercial sponsorship, so that any recommendations can be brought to the next Steering Group meeting in Ottawa in October 2004.

19. Reviewers and Collaborative Review Groups should not receive royalties on sales of reprints of their reviews, since these sales are likely to have been made to commercial sources and might, therefore, be assumed to be equivalent to direct sponsorship of the review or Group. Therefore, the current policy that royalties on reprint sales go to The Cochrane Collaboration centrally, via the Collaboration Trading Company, will continue. When a Foundation is established, the possibility that such income should go into it will be discussed.

20. John Wiley and Sons Limited should continue to be encouraged to make bulk sales of The Cochrane Library and derivative products to commercial sources.

21. All Cochrane Collaboration policies are kept under continual review, but these decisions will be formally reviewed after three years.

6 April 2004