Research integrity: clinical trials transparency Contents

2The current state of clinical trials transparency and related legislation

Definitions

6.The World Health Organization (WHO) defines a clinical trial in broad terms as “any research study that prospectively assigns human participants or groups of humans to one or more health-related interventions to evaluate the effects on health outcomes”. Specifically, the WHO definition includes “drugs, cells and other biological products, surgical procedures, radiological procedures, devices, behavioural treatments, process-of-care changes, preventive care, etc”.14

7.‘Clinical trials transparency’ is typically discussed in terms of three interconnected issues:

Existing legislation, rules and current compliance

8.Clinical trials are heavily regulated, and a wide range of relevant rules and initiatives exists to address non-registration, non-reporting and mis-reporting. However, despite the existence of these “innumerable regulations, edicts, reports, guidelines and strategy documents” relating to clinical trials transparency, Dr Goldacre told us that “none have been enforced or implemented, and breaches are not documented”.15 Examples to illustrate this are explored below.

Non-registration

9.In 2013, the Health Research Authority (HRA) made it a condition of a trial receiving a ‘favourable opinion’ from a research ethics committee that the trial16 must be registered—or a deferral for specific reasons requested17—before participants are recruited.18 The HRA subsequently conducted several audits of registration, with follow-up contact with lead investigators where non-compliance was found. These demonstrated that registration was still not universal, even among trials that had received a favourable opinion, and that some who fell short on their compliance ignored contact from the HRA. The HRA’s 2017 audit revealed that 32% of 599 studies that received a ‘favourable opinion’ (and no agreed deferral) could not be found on a publicly accessible registry.19 Moreover, of the 194 lead researchers contacted regarding non-registration, 73 failed to respond to the HRA within a week to provide an explanation or take steps to register.20 The HRA concluded that “awareness that the requirement to register a clinical trial as a condition of favourable opinion was variable. Many responders did not know how or where to register their study and what was an acceptable register for their study type”.21 Four of the non-respondents related to phase I drug trials;22 we were told that the current status of these trials had been explored further after the audit: one was already on the US public register, clinicaltrials.gov, two were taking action to register publicly, prompted by the HRA’s follow-up, and the other trial did not start so it did not need public registration.23

Non-reporting

10.Since July 2014 the European Commission has required all trials on the EU Clinical Trials Register (i.e. trials of medicinal products) to post results to the registry within 12 months of completion, with a final compliance date of 21 December 2016.24 The first analysis of compliance with this requirement was undertaken by Dr Goldacre and his team at the Evidence-Based Medicine DataLab this year. That study identified 7,274 trials where it could be verified that results were now due. Of these, just 49.5% had reported results.25 This further reinforces the assessment made by the AllTrials campaign made that “around half” of trials go unreported, drawing on a range of research on this topic.26 The Medicines and Healthcare products Regulatory Agency (MHRA) is responsible for ensuring that sponsors provide results for drug trials (i.e. a subset of all clinical trials) in the UK that are registered in the European Clinical Trials Register. The HRA told us that it would “seek to work with the MHRA to understand better the situation regarding the trials they regulate”.27

11.Reporting requirements are also specified by UK funders. In October 2016 the Medical Research Council (MRC) clarified its policy on clinical trials reporting, including a requirement for award holders to ensure that findings are publicly available within 12 months of completion of the trial.28 This appears to have been successful as compliance rates are relatively high for MRC-funded trials. In 2017, the MRC examined reporting rates for trials funded between 2011 and 2016, finding that 33 out of 40 trials (~82%) that had been completed for at least a year had reported in at least one publication.29 The Government referred several times during our inquiry to the relatively high compliance rate for MRC-funded trials.30 However, the MRC’s role in clinical trials is relatively small; it funds around 120 clinical trials each year,31 representing less than 3% of the 4,568 studies reviewed by the HRA in 2017/18.32

12.The current EU Clinical Trials Directive33 (incorporated into UK law via the 2004 Medicines for Human Use (Clinical Trials) Regulations34) will soon be superseded by a new EU Clinical Trials Regulation.35 This Regulation includes new transparency requirements regarding publication of trial results and requires Member States to lay down rules on penalties for non-compliance.36 Although the new Regulation entered into force on 16 June 2014, the point at which it becomes applicable to the Member States is contingent on the completion of a new EU clinical trials portal and database; the Regulation becomes applicable six months after the European Commission publishes confirmation that the portal and database are fully functional. This is expected to happen in 2019, after the UK leaves the EU, and so the Regulation will not be incorporated into UK law by the EU (Withdrawal) Act. The Government has stated that it will “align where possible with the [new Regulation] without delay when it does come into force in the EU, subject to usual parliamentary approvals”.37 What this commitment to “align” with the Regulation will mean in the context of forthcoming clinical trials transparency requirements (and any corresponding penalties for non-reporting) is currently unclear.

Mis-reporting

13.A multi-journal initiative exists to ensure that the full results from trials are reported. The Consolidated Standards of Reporting Trials (CONSORT) statement, first published in 1996 and updated in 2010, sets out “a standard way for authors to prepare reports of trial findings, facilitating their complete and transparent reporting, and aiding their critical appraisal and interpretation”.38 Its checklist requires all outcomes to be defined and identified, and results to be reported for each outcome.

14.Despite “over 500 individually named academic journals”39 having formally endorsed the CONSORT guidelines for trial reporting, mis-reporting remains an issue. A study of the work of Hampshire A research ethics committee by Dr Kolstoe found that 57% of publications associated with trials approved by the committee showed “inconsistencies with the outcomes originally declared in the ethics application”.40 Similarly the Medical Research Council told us that while reporting rates for the trials it funds were high, in its most recent audit of corresponding publications “only half of these appeared to include the main trial results”.41

15.A range of UK and EU rules and guidelines are in force to improve clinical trials transparency, in terms of tackling non-registration, non-reporting and mis-reporting. Despite these rules, around half of clinical trials are left unreported, clinical trial registration is not yet universal in the UK and reported outcomes do not always align with the original study proposal. Further action is needed to improve reporting and registration of clinical trials, as we set out in this Report. The Government should not rely on the higher reporting rates for trials funded by the Medical Research Council to suggest that the problem is being addressed, as these represent a small proportion of the total number of clinical trials undertaken in the UK.

16.The Government should explicitly commit to introducing the clinical trials transparency requirements in the EU Clinical Trials Regulation that are expected to be applied in the EU shortly after Brexit.

Compliance rates by sponsor

17.Dr Goldacre’s recent work has also revealed that compliance with reporting requirements varied by sponsor—i.e. the “individual, company, institution, organisation or group of organisations that takes on responsibility for initiation, management and financing (or arranging the financing) of the research”.42 Trials with a commercial sponsor were substantially more likely to post results than those with a non-commercial sponsor (68.1% vs 11.0%). In particular, “compliance among pharmaceutical companies has been good; while universities have performed poorly”.43 However, the study found considerable variation between universities. The University of Dundee, for instance, has a compliance rate of 82%, whereas the equivalent figure for the University of Nottingham is 5.9%.44

18.Dr Patrick Vallance, the Government’s Chief Scientific Adviser, told us that his message to universities with low trial reporting rates was to “sort it out”.45 The Science Minister, Sam Gyimah MP, agreed with this message.46 The work being undertaken by Universities UK to review the Concordat to Support Research Integrity in response to our earlier Report on this subject represents an important opportunity to press for progress on this issue.47

19.Compliance with transparency rules varies by sponsor—while pharmaceutical companies have good rates of reporting within a reasonable timeframe, the picture is much more mixed for universities. We welcome the Science Minister and the Government Chief Scientist’s call for universities to deal with this problem and expect universities to take heed. Every university should aim for 100% compliance. We recommend that the updated and strengthened Concordat to Support Research Integrity being developed by Universities UK should include requirements on universities to ensure that all trials are reported, and that efforts are made to share best practice in achieving compliance with reporting rules within the university sector.

20.The ‘EU Trials Tracker’ website set up by Dr Goldacre and colleagues at the Evidence-Based Medicine Data Lab also reveals that Public Health England has three overdue trials dating from 2010–2016 relating to meningitis vaccines.48 Meanwhile, many NHS Trusts have high numbers of unreported clinical trials according to the site: the Manchester University NHS Foundation Trust has 13 overdue trials, NHS Greater Glasgow and Clyde has 12 that are due to have reported, and both Newcastle upon Tyne Hospitals NHS Foundation Trust and Hull and East Yorkshire Hospitals NHS Trust have 11 outstanding trials.49

21.It is particularly disappointing that trusted bodies such as Public Health England and a range of NHS Foundation Trusts are also failing to report results from clinical trials. Public trust in medicine could easily be eroded by failures in clinical trials transparency from such important parts of the health system. Public Health England should write to us with an explanation and the steps it will take to correct this.


14 World Health Organization, ‘Health topics: clinical trials’, accessed 31 August 2018

15 Dr Ben Goldacre (RIN0073) para 3

16 The requirement encompasses: clinical trials of an investigational medicinal product; clinical investigations or other study of a medical device; combined trial of an investigational medicinal product and an investigational medical device; and other clinical trials to study a novel intervention or randomised clinical trial to compare interventions in clinical practice.

17 Reasons for deferral could include commercial sensitivity. There is an expectation that the trial will be registered when the reason for the deferral is no longer valid or immediately should the trial be terminated early for safety reasons.

18 Health Research Authority, Clinical Trial Registration: Audit of compliance for period 1 January 2016 to 30 June 2016 (August 2017), p1. The registration condition applies throughout the UK as a result of the HRA’s coordination with the Devolved Administrations, as per s111(4) of the Care Act 2014.

20 Health Research Authority, Clinical Trial Registration: Audit of compliance for period 1 January 2016 to 30 June 2016 (August 2017), p8. The audit report notes that “The response rate from investigators / sponsors when contacted regarding study registration was reasonably good, given the short response window (one week from when the email was sent and no reminder emails were issued)”.

22 ‘Phase I’ refers to a therapy being given to a group of people for the first time.

23 Health Research Authority (RES0047)

26 AllTrials campaign (RIN0067)

27 Health Research Authority (RES0047)

28 Medical Research Council (RES0041) para 1.1

30 Q598 [Sir Mark Walport], Department for Business, Energy and Industrial Strategy (RES0057),

33 Council Directive 2001/20/EC

34 The Medicines for Human Use (Clinical Trials) Regulations 2004 (SI 2004/1031)

35 Council Regulation (EC) No. 536/2014

36 For further details see ‘Clinical trial regulation’, European Medicines Agency, accessed 24 August 2018, and Article 94 of Council Regulation (EC) No. 536/2014.

39 Dr Ben Goldacre (RIN0073), para 1.3

40 Dr Simon Kolstoe (RIN0022)

41 Medical Research Council (RES0041) para 3.3

42 Health Research Authority, ‘Roles and responsibilities’, accessed 30 August 2018

44 According to data published at https://eu.trialstracker.net as at 22 October 2018

47 See Science and Technology Committee, Sixth Report of Session 2017–19, Research integrity, HC 350, para 43

49 Data taken from https://eu.trialstracker.net accessed 22 October 2018.




Published: 30 October 2018