Health

Half Of European Clinical Trials Haven’t Reported Results

Half of European clinical trials haven’t complied with EU rules on reporting results

  • EU clinical trial rules say results must be reported onto the EU clinical trial register for every trial on the register within 12 months of the end of the trial.
  • 49% of clinical trials on the EU register have not reported results.
  • Companies are complying far more than academics are: 68% of company sponsored trials have reported results v 11% of academic trials.
  • No one has ever been sanctioned for breaking the European rules.
  • The data on the EU register is so riddled with mistakes and inconsistencies that it will be impossible for the European regulator to fully assess compliance for every clinical trial.
  • AllTrials campaign says: there are no excuses for academics’ poor reporting rates, it’s the law to report results in Europe, patients deserve better.

Get The Timeless Reading eBook in PDF

Get the entire 10-part series on Timeless Reading in PDF. Save it to your desktop, read it on your tablet, or email to your colleagues.

Q2 hedge fund letters, conference, scoops etc

european clinical trials
rawpixel / Pixabay

New research published in the BMJ1 today shows that around half of the European clinical trials that should have reported results haven’t. The paper accompanies the launch of a tracking website that shows which clinical trials on the European trials register have reported results and which have not, and displays overall reporting performance for the companies, universities and charities that sponsor trials.

An EU guideline2 says that sponsors of clinical trials must ensure that trials registered on the European Clinical Trials Register since 2004 have posted results onto the register within a year of the trial's end. The online tracker assess which trials are due to have reported results onto the register and flags those which are overdue reporting. Information on the tracker website is updated monthly.

The EUTrialsTracker tool is launched today at eu.trialstracker.net3 The EUTrialsTracker shows today that 48.9% of clinical trials that should have reported results haven’t. Only 51.1% of clinical trials have reported results.

In the BMJ paper researchers at the DataLab, University of Oxford, describe how they built the EUTrialsTracker and share a snapshot of data on trial reporting rates. The analysis in the BMJ shows that:

The majority of major pharmaceutical companies are coming close to complying with reporting rules. The majority of major academic institutes are not.

  • Trials sponsored by companies are significantly more likely to have reported results than trials with academic sponsors – 68% of commercially sponsored trials are reported vs 11% of non-commercially sponsored.
  • This means that 32% of all trials sponsored by pharmaceutical companies haven’t reported results onto the register.
  • And around 90% of trials sponsored by non-commercial sponsors (universities, hospitals, governments and charities) haven’t published results onto the register.
  • Some pharmaceutical companies are reporting all of their trials. The EUTrialsTracker currently shows that eleven major commercial sponsors* are achieving 100% reporting.4
  • The top performing major academic sponsor*, the University of Dundee in the UK, is reporting 82% of its trials but most academic sponsors are reporting less than 50% of due trials.
  • Thirty-two major universities have currently reported results for none of the trials they sponsor.5

Larger sponsors - those with most trials on the register - are significantly more likely to have reported results than smaller sponsors.

European Clinical Trials that completed recently are more likely to have reported results than older trials

The EU Trials Tracker can produce ranked lists of the trial sponsors who report results for the most and least of their trials. The EUTrialsTracker currently shows that:

  • Eleven major sponsors* have reported results for 100% of their due trials. These 11 are all companies.
  • All of the major sponsors who have reported results for 90% or more of their trials are companies. There are 31 of them.
  • Thirty-two major sponsors* have published results for none of their due trials on the register.  All 32 are European hospitals, universities and research institutes.
  • All the major sponsors who have reported results for less than 10% of their trials are non-commercial entities. There are 42 of them.
  • Looking beyond major sponsors to the thousands of organisations who have even one trial on the register, some non-commercial organisations have achieved compliance with the reporting rules. For example the UK medical research charity Cancer Research UK currently has only three due trials on the register but has reported results for all three of them.

The current rules on reporting results are straightforward and clear

In 2012 the European Medicines Agency (EMA) which oversees the European Clinical Trials Register and medicines regulation in the EU announced the new rule stating that results must be reported on the register. Clear deadlines were set for reporting - one year from the end of the trial for most trials, and within six months of completion for trials involving children. After various delays, the final date for sponsor’s compliance was 21st December 2016.

The EU rules are clear that results must be posted directly onto the EU trials register. This is with good reason. Reporting on a register has several advantages over academic journal publication: it is much easier to locate results; it is much easier to track compliance; and registers mandate important information that must be included in the results report. By contrast, there is extensive evidence to show that results reports in academic journals are incomplete6, of poorer quality than registry reports, and less complete on important issues such as adverse events data7.

There are no excuses

The obligation on researchers to report results are ethical, legal, contractual and professional. AllTrials has set out these obligations here8. Nevertheless, researchers continue to make excuses to AllTrials when we ask them why they aren’t reporting. To save us time answering another round of excuses today we here gathered the excuses with some unequivocal answers9.

Why hasn’t anyone been sanctioned for breaking the rules?

Under EU rules2 the EMA is required to publicly flag any trial that is overdue and contact the sponsor. But there has been no flagging of overdue trials on the EU’s register, and to our knowledge no trial sponsor has ever been contacted over non-compliance. We hope the EMA will now explain why they have not done so.

*Major sponsors are those responsible for 50 or more trials on the EU Clinical Trial Register

COMMENTS

Dr Ben Goldacre, Director of the DataLab at the University of Oxford, and lead author on the BMJ paper, said:

“This problem strikes to the heart of evidence based medicine. We cannot make informed choices about which treatments work best, as doctors and patients, unless all results are reported.

We hope that our data will help trial sponsors to move fast and get their houses in order. We have identified the individual non-compliant sponsors, and trials, in order to help them do so.

We would hope that public pressure and the ethical need for trial reporting should be enough. But the European Clinical Trials Regulation is coming into force next year. It will bring substantial financial penalties for sponsors in breach of reporting requirements. All trial sponsors - especially universities - must get their house in order now.”

Síle Lane, head of international campaigns and policy, Sense about Science:

“The new tracker shows us that some companies and universities are taking this issue seriously. But there’s a vast difference between the best and the worst. There’s no excuse for failing to report results.

So, why should Helsinki University be allowed to run one more trial, while they have not posted results for any of their 12 currently overdue trials? Or the University of Nottingham, which has reported results for just 1 out of their 17 overdue trials on the register?

Why should these institutions be allowed run any more trials on patients? Why are ethics committees giving them permission to run more trials? Why are funders paying for them?”


Notes:

AllTrials is the global campaign for all clinical trials to be registered, with their results reported, joined by 750 organisations and 90,000 people worldwide. Working together we have: put clinical trial transparency on the agenda at the highest levels including the UN and WHO; audited the transparency policies of the world’s largest pharmaceutical companies and research funders; persuaded them to put stronger measures in place; persuaded the EMA and the FDA to bring in tough new laws and rules on reporting. The AllTrials campaign is run day-to-day by Sense about Science the UK charity that campaigns around open discussion of research and evidence. www.AllTrials.net

The EU TrialsTracker was built by the DataLab at the University of Oxford, which is headed by Dr Ben Goldacre, a co-founder of the AllTrials campaign. The DataLab is a multidisciplinary team of clinicians, academics and software engineers: they build interactive data-driven tools for the NHS and academia, as well as pure academic papers.

  1. Compliance with requirement to report results on the European Clinical Trials Register: a cohort study and web resource. Ben Goldacre, Nicholas J DeVito, Carl Heneghan, Francis Irving, Seb Bacon, Jessica Fleminger, Helen Curtis, Open Knowledge International.
  2. European Commission guideline (2012/C 302/03) Guidance on posting and publication of results-related information on clinical trials in relation to implementation of Article 57(2) of Regulation (EC) No 726/2004 and Article 41(2) of Regulation (EC) No 1901/2006.

Section 4.7 specifies EMA’s role in publicly flagging overdue trials.

  1. The eleven major commercial sponsors achieving 100% compliance include Boehringer Ingelheim, Almirall SA, Gilead Sciences Inc, Genentech and LEO Pharma.
  2. The 32 major academic sponsors with 0% compliance include the Karolinska Institute, University of Copenhagen, Manchester University NHS Foundation Trust, University of Amsterdam, Hospitals of Paris, Erasamus University and the European Institute of Oncology and the National Cancer Institute Italy.
  3. https://systematicreviewsjournal.biomedcentral.com/articles/10.1186/2046-4053-1-60
  4. https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1001566 (For adverse events claim) https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4650202/ (Outcomes discrepant b/w registry & journal outcomes)
  5. http://www.alltrials.net/find-out-more/why-this-matters/obligations-to-report/
  6. To save us losing time to another round of excuses, here they all are, with some unequivocal answers:

A LIST OF REBUTTALS TO BAD EXCUSES FROM TRIALISTS

‘Asking for perfect compliance is unreasonable.’

Lots of companies and some small non-company sponsors have 100% compliance.

‘Some of the trials you flag as unreported have been reported in academic journal papers.’

The EU rules are clear that results must be posted directly onto the EU trials register, in a standard format. This is with good reason. Reporting on a register has several advantages over academic journal publication: it is much easier to locate results; it is much easier to track compliance; and registers mandate important information that must be included in the results report. By contrast, there is extensive evidence to show that results reports in academic journals are incomplete[1], of poorer quality than registry reports, and less complete on important issues such as adverse events data[2].

In any case, the researchers checked a small sample of non-compliant trials to see if they really are reported in academic journals, even though this is inadequate: half were not2.

‘The researcher who ran the trial doesn’t work at this institution anymore.’

This is a tougher one. The institute has failed to ensure adequate reporting by its staff and now can’t remedy that easily. If an institute is listed on the register, it remains responsible for reporting results, just as it would with, for example, a grant report. Before giving this excuse, we think sponsors should show that they have attempted to track down the results and the researcher to remedy the situation. And universities should urgently put policies in place, to make sure this situation doesn’t arise again in the future!

‘These trials are from 10 years ago/an abandoned trial/not interesting... so the results wouldn’t be useful for anything now.’

The law in Europe is very clear: every trial on the European clinical trials register must report results directly onto the register. This is for good reason. Trials with an “uninteresting” or “negative” result provide information about what doesn’t work.  Terminated trials that have already collected some data on some patients can contain important insights. Research on abandoned treatments can shed light on other drugs in the same class, on adverse event risks, or on new uses of older treatments.

‘There’s no money left in the research grant for anyone to spend time reporting results.’

This is just flat out against the law, and usually against the terms of the funding. It’s also unethical. Trials are conducted to inform patient care. If you don’t report the results of your trial, you might as well have never done it. We question whether trialists or institutions should be conducting further research, if they are willing to make this excuse.

‘There’s no career incentive to report negative results.’

This is unethical, and increasingly untrue. Major funders and ethics approval bodies are starting to scrutinise the reporting history of researchers applying for new approvals. Some professional bodies now view non-reporting as misconduct.

‘But this is my research. I own the results.’

This is true. For research conducted on yourself and yourself alone. If you have conducted a trial on patients, then you have obligations to those patients, to your funders, to the law, and to all patients.


[1] https://systematicreviewsjournal.biomedcentral.com/articles/10.1186/2046-4053-1-60

[2] https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1001566

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4650202/