Despite progress in data transparency, the FDA still keeps its data secret
History shows that hiding clinical trial data can be deadly.
Vioxx is a well-known example of how the US drug regulator withheld important information about the harms of the drug for over three years, before it was withdrawn from the market and tens of thousands of people died as a consequence.
Numerous initiatives have been launched over the past two decades to improve access to trial data after it became evident that what was reported in peer-reviewed journals was often cherry-picked and misleading.
Eminent scientists have succeeded in gaining access to trial data from the European and Canadian drug regulators, but a recent analysis published in the Journal of Law, Medicine, & Ethics, found that the US FDA still lags behind others when it comes to data transparency.
Europe ahead of the pack
Drug regulators have traditionally been the guardians of a treasure trove of trial data which they kept hidden from the public. But, over a decade ago, the efforts of Danish professor Peter Gøtzsche turned that on its head.
Gøtzsche and his PhD student were studying the effects of an anti-obesity drug and requested the trial data held by the European Medicines Agency (EMA).
“We already had good evidence that the efficacy and harms of drugs were incompletely reported in the medical journals, so by asking for the regulatory data for the anti-obesity pills, we were convinced it would get us closer to the truth”, said Gøtzsche.
At first, EMA denied their request, saying that it needed to protect commercially confidential information, but Gøtzsche was undeterred. He made a formal complaint to the European Ombudsman.
After an arduous 3-year process, the Ombudsman accused EMA of “maladministration” for refusing to share its data - it was a serious and embarrassing charge, so EMA had no choice but to capitulate.
In 2013, EMA announced that it would provide public access to regulatory data – which included study reports, protocols and the raw anonymised patient data in statistical programmes enabling anyone to independently scrutinise the data for all new drugs that it approved.
It was a bitter-sweet moment for Gøtzsche.
“I was satisfied with the outcome, but I also felt a bit betrayed. When EMA praised itself for being transparent, it conveniently omitted telling the public that it was basically forced to make the decision because of my efforts and that of the Ombudsman,” said Gøtzsche.
“I’ve been around a while to know that this is exactly how the drug industry operates. They cover up their failures while praising themselves for what others force them to do,” he added.
Millions of pages containing trial data have since been released. Interestingly though, this remarkable feat has gone largely unrecognised and the response from the research community has been rather tepid.
Gøtzsche suspects it’s because analysing regulatory documents is complex and requires experience to decipher regulatory data – skills that few researchers have.
“It is a huge job to do systematic reviews of clinical study reports held by drug regulators, but it is the difference between producing reliable reviews or merely “garbage-in, garbage-out” reviews,” said Gøtzsche.
Since then, Gøtzsche’s group showed this was the case for reviews of antidepressant drug trials.
When they compared data from medical journals to that from regulatory documents, they found major discrepancies such as underreporting of harms, including suicide attempts and aggressive behaviour.
Canadian regulator in the cross hairs
Following the landmark policy change in Europe, researchers believed it would help unlock regulatory documents elsewhere that were historically kept hidden from the public.
In 2016, Peter Doshi, professor at the University of Maryland and senior editor of The BMJ requested the release of unpublished clinical trial data relating to antivirals for the treatment of influenza (Tamiflu, Relenza) and three human papillomavirus vaccines from the Canadian drug regulator, Health Canada.
After some resistance, Health Canada agreed to allow Doshi access to the documents but imposed a confidentiality agreement that would prevent him from making his findings public.
When Doshi refused to sign the confidentially agreement, his request for access to the trial data was denied, so he filed a lawsuit in a federal court seeking a judicial review of the regulator’s decision.
Remarkably, in 2018, in the case of Peter Doshi v. Attorney General of Canada, a federal court judge ruled in favour of Doshi and in the public’s interest, ordering Health Canada to hand over the trial data for independent scrutiny.
It was hailed a “major victory” for transparency and after the win, Doshi told The BMJ, “For me this case has always been about something larger than my specific request. It is about the principle of transparency. If my case sets a precedent and Health Canada begins making clinical trial data available to others—promptly, and without imposing confidentiality agreements—that will be the real victory.”
Notably, the Canadian drug regulator has gone one step further than EMA by proactively releasing data for not only approved drug submissions, but also “unapproved, and withdrawn drug and biologic submissions…Class III and IV medical device applications.”
What about the US FDA?
The US FDA houses the largest known repository of clinical trial data in the world, but it doesn’t proactively share it.
In 2018, the FDA launched a new pilot program to proactively publish clinical study reports from the pivotal studies of nine recently approved drugs – but the agency put an end to that program in March 2020.
“It is just so typical of the FDA, which is very beholden to industry, and which some have dubbed the Foot Dragging Agency when it comes to the public interest,” said Gøtzsche.
Now, the only mechanism to ascertain regulatory data for FDA-approved drugs is to submit Freedom of Information Act (FOIA) requests, a lengthy process which often results in heavily redacted documents of limited value.
A study by US researchers, analysed the FDA’s willingness to release data, compared to other regulators, EMA and Health Canada.
They found that between 2016 and April 2021, EMA released data for 123 unique medical products, while Health Canada released data for 73 unique medical products between 2019 and April 2021.
In stark contrast, the FDA only proactively disclosed data supporting one single drug that was approved in 2018, clearly demonstrating that the agency has failed to keep pace with the European and Canadian regulatory bodies.
The problem of data secrecy within the FDA has been especially evident during the pandemic. Recently, I reported in The BMJ that the agency had failed to disclose covid-19 vaccine ‘safety signals’ derived from post-marketing data.
Also, the non-profit group, Public Health and Medical Professionals for Transparency had to sue the FDA for access to trial documents used as the basis for licensing Pfizer’s covid-19 mRNA vaccine. Initially, the agency wanted 75 years to release all the data but a Federal Court Judge rejected its request, ordering the release of the documents at a rate of 55,000 pages per month, taking approximately 8 months.
Given the widespread use of this important public health intervention, and the billions of dollars in public funds used to conduct vaccine research and development, these data should have been made publicly available immediately.
Data secrecy has undermined the health care system by subverting the allocation of scare resources and eroding public trust. The damage done to people’s confidence in vaccines, and medicines more broadly, will be felt for generations and likely to harm public health.
Maryanne Demasi, reports is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.