Science and Technology CommitteeWritten evidence submitted by by BioMed Central and Current Controlled Trials

Background and Competing Interests

1. BioMed Central is an STM (Science, Technology and Medicine) publisher which has pioneered the open access publishing model. All peer-reviewed research articles published by BioMed Central are made immediately and freely accessible online, and are licensed to allow redistribution and reuse. BioMed Central is part of Springer Science+Business Media, a leading global publisher in the STM sector. All BioMed Central’s medical journals require prospective registration of clinical trials as a condition of publication.

2. Since its launch in 2000, BioMed Central has demonstrated that commercially viable business models exist which allow scientific publishers to make the peer-reviewed research articles they publish immediately and freely available online in their official form, with costs typically covered via a publication fee.

3. BioMed Central’s journal portfolio includes the journal Trials [http://www.trialsjournal.com], which is an open access, peer-reviewed journal that encompasses all aspects of the performance and findings of randomised controlled trials. Trials publishes articles on general trial methodology as well as protocols, commentaries and results and strongly supports publication of all trial results, regardless of the outcome of the trial.

4. Current Controlled Trials, part of BioMed Central Group, administers the ISRCTN register of clinical trials [http://www.controlled-trials.com/isrctn/]. The ISRCTN register allows users to search, register and share information about clinical trials. Access to all the information on this website is free; there are fees for the registration services offered by Current Controlled Trials.

5. The focus of this submission is on Terms of Reference numbers 3, 4 and 5. References are provided in square brackets [] throughout the text.

What evidence is there that pharmaceutical companies withhold clinical trial data and what impact does this have on public health?

6. There is much published evidence on the unavailability of clinical trial data leading to negative effects on public health, which we will not repeat detail. Reporting bias in medical research has been identified in around 50 different medical interventions [http://www.trialsjournal.com/content/11/1/37] and there are many examples of its negative impacts on public health. Reporting bias tends to favour positive results of trials, which support the effectiveness of the intervention (such as a drug or device) being tested, meaning treatment decisions are frequently being made by doctors who have incomplete information on the benefits and harms of medications. A particularly high profile case includes the widely prescribed antidepressant drug reboxetine [http://www.bmj.com/content/341/bmj.c4737] which was found to be ineffective or potentially harmful when unpublished data were analysed. There are many other serious examples, which are documented in the article by Peter C Gøtzsche published in the journal Trials [http://www.trialsjournal.com/content/12/1/249]. We encourage the Committee to consider the examples of harms to patients described in this article.

7. It is widely accepted in the scientific community that results which do not support the hypothesis or healthcare intervention being studied—negative results—are of vital importance. Many peer-reviewed journals in clinical medicine, from a number of publishers, strongly encourage publication of negative results including Trials, BMJ Open, BMC Research Notes, and PLOS One. And at least one journal makes publication of negative results its mission, Journal of Negative Results in Biomedicine, published by BioMed Central. However, despite many opportunities for investigators to publish all trial results in journals, and ethical (in the UK/EU) requirements to publicly register the existence of clinical trials at their inception, relying on ethical policies of journals, editorial organisations, and research funding agencies has so far been insufficient to address the problems of bias in clinical evidence. Despite prospective registration of clinical trials being required by major medical journals since 2004, evidence continues to emerge that adherence to policies for registration of trials [http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0025258] and reporting of results [http://www.bmj.com/content/344/bmj.d7292] is worryingly low.

How could the occurrence and results of clinical trials be made more open to scrutiny? Who should be responsible?

8. The initiative by the European Medicines Agency, to require sharing of raw data supporting all drug and device license applications by pharmaceutical companies in the EU from January 2014 [http://www.bmj.com/content/345/bmj.e8423], should be applauded. However, in the UK, there is no legislation requiring the public registration of trials or the public disclosure of results of trials.

9. Around early 2000, two major initiatives were launched in response to a general need for more transparency. (i) A group of UK based medical research organisations pushed for the creation of a public listing of clinical trials, which led to the creation of the UK based Current Controlled Trials website and its trials database the ISRCTN register [http://www.controlled-trials.com/mrct/meeting_ukcc_29jul98.asp]. (ii) US patient groups demanded better access to ongoing clinical trials in dangerous and life threatening disease areas such as cancer and HIV and this led to federal legislation and the creation of the ClinicalTrials.gov website [http://clinicaltrials.gov/ct2/about-site/background].

10. Two other types of stakeholders were instrumental in making sure that clinical trials were more open to scrutiny:

(i)A number of medical journals editors declared that they would not consider the publication of the papers about clinical trials if specific details about those trials had not been publicly disclosed well before enrolment started [http://www.icmje.org/update.html]. (ii) The World Health Organization (WHO) described the public listing of clinical trials as a scientific, ethical and moral responsibility and set about defining standards and capacity building methodologies [http://www.who.int/ictrp/trial_reg/en/index.html]. This led to creating a public platform that brings together all vetted international, regional and national registers [http://apps.who.int/trialsearch/].

11. Increases in the numbers of registers worldwide divides opinion. For some, the multiplication of registers is seen as a waste of resources and efforts, leading to duplication of information with different levels of completeness and quality which makes global analyses on clinical evidence very difficult if not impossible. For others this should be seen as a positive step and a proof of increased awareness, regulations and protection of trial participants, and the acknowledgement of different geopolitical remits and language needs. A realistic view might well be that of the WHO which advocates a harmonised—rather than uniformed—approach.

12. ClinicalTrials.gov backed up by federal legislation and a substantial budget has grown to become the largest source of trial information in the world. Additional legislation in 2007 required that all trials be listed before enrolment starts and furthermore that basic results of all those trials should also be reported within well defined time frames [http://clinicaltrials.gov/ct2/about-site/results]. Germany is to follow suit [http://www.gesetze-im-internet.de/englisch_amg/englisch_amg.html#p0930].

13. An additional step towards increased transparency was the decision to make some sections of the confidential regulatory database EudraCT open to the public via the EU-CTR register [https://www.clinicaltrialsregister.eu/]. EudraCT will also add results and the way the information is required is very much along the lines developed by ClinicalTrials.gov [http://ec.europa.eu/health/files/eudralex/vol-10/2012_302–03/2012_302–03_en.pdf].

14. In the UK, a number of organisations including the Department of Health [http://www.crncc.nihr.ac.uk/about_us/processes/portfolio/isrctn_index/], the Medical Research [http://www.mrc.ac.uk/ResearchPortfolio/About/index.htm#P187_19342] Council and the Wellcome Trust [http://www.wellcome.ac.uk/Funding/Biomedical-science/Application-information/WTX022708.htm] support prospective registration of clinical trials. An effort has been made to simplify steps when seeking all required approvals for a trial by designing the Integrated Research Application System (IRAS) which provides researchers with a “one-stop shop” displaying all the relevant forms (including the clinical trial authorization to be submitted to the Medicines and Healthcare products Regulatory Authority (MHRA) then passed onto the EudraCT database and ultimately EU-CTR). But using IRAS is not compulsory [https://www.myresearchproject.org.uk/Help/UsingIRAS.aspx].

15. The Association of Medical Research Charities has advocated the need for a plain English summary for all scientific outputs in a recent report [http://www.amrc.org.uk/our-members_patients-participate].

16. As increased participation in trials remains very much supported by the UK government [http://cdn.hm-treasury.gov.uk/2011budget_growth.pdf], 2010 saw the launch of the UK Clinical Trials Gateway (UKCTG) which gives a overview of the trials that have enrolled or are enrolling UK participants, in a single environment [http://www.ukctg.nihr.ac.uk/default.aspx]. The UKCTG has two challenges: coverage (using other data sources apart from the ISRCTN database and ClinicalTrials.gov data might be required) and accessibility (clinical trial descriptions are often not in plain English).

17. As of 2013, there is still no legal requirement to publicly register clinical trials in the UK. Existing efforts to ensure trial registration have focused on ethical aspects and researchers’ motivations to publish in good journals but does not have the power to ensure all trials are registered and their results reported. Others have gone further and called for a global network to enable universal trial registration and data transparency [http://www.trialsjournal.com/content/11/1/64].

18. Requiring registration of trials and reporting of results is less complicated to implement than the sharing of raw data but is equally important. Raw data is important for researchers, such as systematic reviewers, wishing to build on or validate previous research. However, other important stakeholders such as patients, research funding agencies, ethics committees and journal editors, would greatly benefit from human readable (understandable) summary information about all trials. Furthermore, there are fewer considerations for patient privacy—a barrier to full public disclosure of clinical data—when sharing summary information about trials, compared to sharing raw data. The public registration and reporting of results of all trials is the aim of the AllTrials initiative [http://www.alltrials.net/], which is supported by BioMed Central and Current Controlled Trials.

19. Furthermore, registration of all trials and reporting of all results, if made law, could be achieved through incremental developments to existing tools for public disclosure of information about trials. The ISRCTN register already accepts trial registrations globally of all trial designs and at any stage of the trial. And at Current Controlled Trials we are investigating the feasibility of providing a results reporting service within the ISRCTN database. As a commercial organisation which operates a trial registration service with a strong UK focus we clearly have an interest in stronger requirements to register trials. However, the ISRCTN database is just one of many trial registers operating in the EU and we regularly collaborate with other registers in the WHO network for mutual benefit.

20. Services for trial registration and reporting of results are already widely available, and so these activities are supported by the publishing industry and other organisations which provide trial registration services. Responsibility, however, for registration of trials and reporting of results is ultimately that of the investigators and their sponsors or employers. Trial sponsors and investigators have a responsibility to patients recruited to trials—and been exposed to unknown benefits and harms of treatments—to disclose results, so research is not repeated unnecessarily. Trial registration also helps reduce wasteful duplication of research as it creates a public record of a trial, reducing the potential for patients to be recruited to redundant trials and put at unnecessary risk. Evidence of adherence to legislation could be provided by provision of a unique trial identifier from an approved trial register, such as an ISRCTN number, and an equivalent or updated identifier for the reporting of results. These publishing services largely exist already, and could partner with Government bodies to ensure implementation is effective and adherence simple to ascertain.

21. Increased transparency in clinical trials could also be achieved by: (i) Leveraging increased support for open access to publicly funded scientific research in the UK, and extend this to all clinical trials, regardless of the source of funding; (ii) Health research funding bodies encouraging researchers to provide high quality information publicly about trials they fund, in plain English, with funding retained in the future if this criterion is not met; (iii) Encourage publishers and journals to demand proof of prospective registration (this may even go as far as a commitment to data sharing); (iv) Engage with patient groups to understand what participants need to read and understand about trials before contacting their doctor and/or a researcher directly; (v) Develop consensus guidelines regarding basic results posting that do not, critically, jeopardize peer-reviewed journal publication in the future; (vi) Legislate in a way that takes into account the views of all parties: industry (intellectual property), research communities (help them design better studies and meet their recruitment targets), prospective participants (easy to understand information).

Can lessons about transparency and disclosure of clinical data be learned from other countries?

22. There has been growing awareness of the need for transparency under the aegis of the WHO. But transparency and access comes at a cost and funding is a constant concern. Trial registration services are often funded through research or governmental grants, the availability of which may change over time. The ISRCTN register is ensured to be sustainable through fees which are levied for each trial which is accepted for public registration in the database. This model of charging “authors” is increasingly common with the growth of open access journals and publishers which operate a model of charging authors of accepted papers. It is employed by BioMed Central, parent company of Current Controlled Trials, and is also known as “gold” open access which already has the support of the UK Government for financially sustainable approaches to open access scientific publishing.

23. The UK was a pioneer regarding transparency for clinical trials but over the years we may have lost some momentum. Legislation has been key to increasing trial registration in the USA and Germany and initiatives such as the UK Clinical Trials Gateway (UKCTG) are an opportunity to build up on past efforts.

24. ClinicalTrials.gov is very open about the challenges represented by results reporting. Although the focus is on reporting numeric data, data quality is highly variable and the uptake is slow [http://www.nejm.org/doi/full/10.1056/NEJMsa1012065]. Complimentary initiatives to legislation could improve matters. These include reporting standards agreed by relevant stakeholders for the reporting of summary trial results. Reporting guidelines for complete reports of clinical trials are already widely adopted by journals—in the CONSORT statement and checklist [http://www.trialsjournal.com/content/11/1/32].

25. Although the volume and quality of information that is publicly available has evolved dramatically over the past decade, expectations about coverage, completeness and usability still need to be managed. More efforts will be needed in defining, applying and enforcing standards.

February 2013

Prepared 16th September 2013