Cochrane’s sinking ship and conflicts of interest
We believe Cochrane is a key source of scientific evidence on diagnostic and therapeutic medical issues. The very recent expulsion of Peter Gøtzsche from the Cochrane Board and subsequent retirement of four Board members have great impact on the existence of this organisation. Beside issues mentioned by other authors1-3 we believe this crisis is a good opportunity to fix a big problem raised by Peter Gøtzsche and others about Cochrane’s conflict of interest policy.
Currently, Cochrane allows some authors of its reviews to have conflicts of interest with drug companies, a policy that is widely criticized by insiders, and largely unknown to the public.
It is well known that researchers with conflicts of interest judge more positively about drug therapies than researchers without such ties. Gøtzsche said that Cochrane’s policy regarding conflicts of interest of authors of reviews was inadequate. But Cochrane did not solve this problem. The International Society of Drug Bulletins (ISDB) has criticized this policy already in 2013.4
Because organisations like Cochrane play a key role in assessing clinical trials and other evidence regarding medicines, it is essential that they have robust policies with regard to conflicts of interest.
The International Society of Drug Bulletins (ISDB) is a worldwide network of bulletins and journals on drugs and therapeutics that are financially and intellectually independent of the pharmaceutical industry.5 Cochrane reviews have been considered important scientific and trustworthy references for ISDB members.
In 2016 ISDB adopted a policy that will be totally implemented in 2019, in which members are not allowed to have conflicts of interest with the healthcare industry. Those who have not fulfilled the criteria will be removed from the full membership list.
A Conflict of interest is defined as any financial or advisory relationship (paid or unpaid) with the pharmaceutical industry or related healthcare industry (e.g. medical devices or diagnostics), including the conduct of industry funded clinical trials. Members of the editorial team must be free from conflicts of interest with these industries. All authors who write articles which could influence therapeutic choices (e.g. drug and treatment reviews or guidelines) must be free from conflicts of interest.
What is at stake is not the transparency of conflicts of interest or whether or not it is feasible to get rid of conflicts of interest it is definitely about trust, credibility and scientific integrity.6
Cochrane is damaging the trust and credibility doctors, pharmacists, scientists and patients have put in them. Cochrane’s credibility and trust are largely at stake if they do not adequately deal with this issue immediately.
ISDB therefore supports Gøtzsche’s claim that a recovery from this dire situation would call for the dissolution of the present board, new elections and a broad-based participatory debate about the future strategy and governance of the organisation.
- Demasi M. Cochrane – A sinking ship? https://blogs.bmj.com/bmjebmspotlight/2018/09/16/cochrane-a-sinking-ship/
- Greenhalgh T. The Cochrane Collaboration – what crisis? https://blogs.bmj.com/bmj/2018/09/17/trish-greenhalgh-the-cochrane-collaboration-what-crisis/
- Moynihan R. Let’s stop the burning and the bleeding at Cochrane — there’s too much at stake. https://blogs.bmj.com/bmj/2018/09/17/ray-moynihan-lets-stop-the-burning-and-the-bleeding-at-cochrane-theres-too-much-at-stake/
- Schaaber J. Cochrane under influence: assessment of the HPV vaccines and conflict of interest. ISDB Newsletter 2013;2:8.
- Menkes D, Bijl D. Credibility and trust are required to judge the benefits and harms of medicines. BMJ 2017;358:j4204.
Dick Bijl, president International Society of Drug Bulletins (ISDB),
on behalf of the ISDB Committee: Luis Carlos Saiz Fernandez, Maria Font, Ciprian Jauca, Christophe Kopp, Benito Marchand, Joerg Schaaber.
Click here to read this ‘Rapid response’ online at bmj.com: https://www.bmj.com/content/362/bmj.k3945/rr-0