Call for sanctions as data shows hundreds of UK clinical trials (and over 39,000 medical studies wor
A retrospective audit of clinical trials conducted across the UK has found that over 30% of non-regulated clinical trials still remain unregistered, in breach of long-standing national requirements and global ethics rules.
The team conducting the audit concluded that achieving universal trial registration in the UK will require comprehensive audits to be combined with sanctions.
Trial registration is a central pillar of medical research transparency and integrity. For example, unregistered trials can result in the needless duplication of costly research efforts when multiple scientists unwittingly initiate similar studies, needlessly putting patients at risk.
The audit team reviewed ethics approvals issued by Research Ethics Committees nationwide during the first half of 2016 to generate “the first ‘true’ registration rate (certainly at a national level)” data. Their audit covered 599 weakly regulated trials that had received ethics approval in the UK during a six month period in 2016. In this cohort, which includes trials of medical devices, surgery and other non-drug treatments like physiotherapy, 184 trials (30.7%) had remained unregistered.
This represents progress against even lower rates uncovered by an earlier audit, with the exception of trials of medical devices, for which registration rates had declined even further.
The figures suggest that UK trial sponsors still fail to register over 300 clinical trials per year, and that globally, at least 39,000 medical studies still go unregistered every year.
(The calculation of the global estimate is explained further below.)
Subsequent attempts by the audit team to contact researchers and institutions that had run the unregistered trials were only partially successful. While some respondents pledged to retrospectively register their trials, other responses betrayed confusion about registration rules, even though these have been in place since 2013. Parties responsible for 73 trials did not respond to emails, and a further 20 parties were not contactable because the email addresses held on file were invalid.
The audit team, which included two staff members of the Health Research Authority (HRA), noted that a recent parliamentary enquiry had recommended that measures be put in place to ensure that 100% of clinical trials are registered. It concluded that:
“It is difficult to see how this [100% trial registration] target can be achieved without a more complete audit… followed by organisations such as the HRA considering the use of sanctions with sponsors or investigators who are found not to have registered their studies.”
In October 2018, the parliamentary committee had explicitly recommended that the HRA “introduce a system of sanctions to drive improvements in clinical trials transparency”.
Numerous independent groups, including Cochrane, Transparency International, TranspariMED and the AllTrials campaign have also called for sanctions to be imposed.
The nationwide audit of all clinical trials was possible due to the structure of the UK’s ethics approval system. Ethics approvals are made by more than 80 individual Research Ethics Committees spread across the country, but the Health Research Authority (which oversees their work) centrally archives all approval records.
The Health Research Authority is currently developing a blueprint for annual nation-wide audits of clinical trial registration and reporting based on ethics approval records. TranspariMED recently set out eight transparency benchmarks for the planned national monitoring system. One of the benchmarks is the adoption of effective sanctions, including debarment from conducting future trials and the imposition of financial penalties.
Legal and regulatory context
Global ethical requirement. Prospective trial registration has been a global ethical requirement since the 2008 revision of the Declaration of Helsinki, irrespective of national rules:
“Every research study involving human subjects must be registered in a publicly accessible database before recruitment of the first subject.”
UK regulations. Committing to register a trial has been a condition of ethics approval in the UK since 2013. According to the HRA website:
“Registration should occur before the first participant is recruited and no later than six weeks after recruitment of the first participant.”
The Health Research Authority was set up in January 2015 with a mandate to promote transparency in research. However, for many years, compliance remained unmonitored, and no sanctions have ever been imposed by the HRA.
Other frameworks. The European Union (2004) and the United States (2007) already require some, but not all, clinical trials to be registered.
Note on methodology
UK audit methodology. The UK audit team identified 1,014 clinical trials in total. Of these, 397 trials were so-called CTIMPs involving investigative medicines. As CTIMPs are covered by European Union registration requirements and their oversight is tight in the UK, the team assumed all these had been registered. A further 18 trials had received permission to defer registration. This left 599 applicable weakly regulated trials, of which 184 had failed to register. The ethics approval data held by the Health Research Authority is considered confidential and is thus not publicly available.
Estimate of non-registered medical studies worldwide. In 2018, a total of 95,494 medical studies were registered on trial registries worldwide (60,690 across the network of WHO primary registries plus 34,804 on Clinicaltrials.gov). Of these, only 2,679 studies were so-called CTIMP drug trials that are subject to European Union registration requirements, while a further 4,409 trials were subject to US legal registration requirements. This leaves 88,406 loosely regulated or unregulated medical studies that were registered worldwide. Assuming that these registered studies make up only 69.3% of all such studies actualy conducted (as 30.7% of all such studies remain unregistered), over 39,000 studies were not registered in 2018. While the baseline figure (95,494 studies) of this ballpark estimate double counts studies registered on more than one registry, the estimate of total unregistered studies (>39,000) is nonetheless likely to be conservative as it is extrapolated from UK clinical trial (non-)registration ratios. Trial registration became mandatory in the UK over five years ago, and awareness of the importance of trial registration tends to be far higher among researchers, institutions, funders and public bodies in Britain than among their peers in other countries.
Note: TranspariMED would like to thank FDAAA Trials Tracker co-designer Nick DeVito of the EBM Data Lab for providing the number of trials subject to US legal registration requirements in 2018. Registration data for 4,409 trials subject to FDAAA was first submitted during 2018.