The European Medicines Agency has moved to strengthen clinical trial transparency, and the specific measures are worth reading closely, because they reveal more about the problem than about the solution. Responding to a call from 20 health groups, the agency disclosed that it has started sending email notifications to companies and universities when trial results become due, and set up a monitoring system that shows national regulators which of their trials have failed to report results, with a commitment to publicly report on the number of published results by year's end.
Reminder emails and a monitoring dashboard. Those are the interventions, and they are being described by the campaigners who requested them as an unusually fast and positive regulatory response, which is fair. They are also administrative plumbing of the most basic kind, and the fact that they represent real progress against an obligation that became mandatory in 2014 is the actual story here.
A binding law without a mechanism to notice violations
Posting summary results in the EU database became mandatory for sponsors on 21 July 2014, with a deadline of one year after a trial ends, or six months for pediatric trials. This was not a guideline or a best-practice recommendation. It was a legal requirement.
Eight years later, EMA reported that 81% of clinical trials were compliant with that requirement, a figure the agency reached through reminder letters and public naming, including publishing a list of trials whose sponsors did not respond to reminders. Eighty-one percent is a substantial improvement over where it started and it means roughly one trial in five was still not reporting, years after the deadline had passed, under a mandatory regime.
So the constraint was never the absence of a rule. It was that a rule which nobody systematically monitors is closer to a suggestion. What EMA has now built is the capacity to notice, automatically and continuously, when a deadline passes without a filing. That capacity should have been part of the original system, and its arrival twelve years later is a useful reminder that writing an obligation into law and building the machinery to detect breaches are separate projects that often happen decades apart.
The gap between who can see and who can act
There is a structural reason this took so long, and it explains why the monitoring system is the most consequential of the three measures.
Under the EU Clinical Trials Regulation, member states are responsible for authorizing and overseeing clinical trials. EMA maintains the database. That division means the agency with complete visibility into which sponsors have filed has no enforcement authority over them, while the national regulators with enforcement authority historically lacked an easy way to see which of their trials were delinquent.
Visibility sat in one place and power sat in another, which is a reliable formula for a rule that goes unenforced without anyone deciding not to enforce it. The new monitoring system, which shows national regulators specifically which of their trials have failed to report, is a direct attempt to close that gap by routing the information to the people who can act on it.
Whether it works now depends on the national regulators, and that is worth stating plainly rather than assuming. EMA has done what it can do within its authority. If member-state agencies receive the list and take no action, the dashboard becomes a well-maintained record of non-compliance rather than a remedy.
The FDA mirror image
The American experience makes the same point from the opposite direction, and the comparison is instructive because the US has the tool the EU lacks.
FDA can impose civil monetary penalties for failing to report results to ClinicalTrials.gov, in amounts that compound daily and reach into five figures per trial per day. In practice, those penalties have gone almost entirely unused. In March 2026, the agency's action was to send messages to more than 2,200 companies and researchers associated with over 3,000 registered trials that appeared not to have submitted required results, describing it as an opportunity to comply before the agency considered further action.
Put the two regulators side by side and the pattern is clear. Europe had a mandate and no monitoring. America had monitoring and penalties it did not use. Both ended up in the same place, sending reminder letters more than a decade after the underlying obligation took effect, with a substantial share of covered trials still unreported. Enforcement authority without the will to use it and the will to enforce without the visibility to target it produce the same outcome.
What is still missing matters more than what was added
Two items remain outstanding, and both would do more than the emails.
The first is a public dashboard identifying which sponsors are reporting and which are not. EMA's own history is the argument for it, since the rise toward 81% compliance came in significant part from reminder campaigns paired with publishing the names of non-responsive sponsors. Reputational exposure is the enforcement mechanism that has demonstrably worked in this field, particularly for universities and academic sponsors, who are frequently worse at reporting than commercial ones and who are also more sensitive to public listing than to fines.
The second is a quality assurance system to verify that submitted results actually meet the standards the law requires. This is the more subtle problem and it follows a pattern common to disclosure regimes. Once compliance is measured by whether a filing exists, the cheapest way to comply is to file something. A results posting that is incomplete, missing endpoints, or unintelligible satisfies the requirement in form while defeating its purpose entirely. Without a quality check, a compliance rate rising toward 100% could reflect better reporting or simply better box-checking, and the number alone cannot distinguish them.
Why unreported trials are not a paperwork problem
It is worth being explicit about what is at stake, because "trial results reporting" sounds like an administrative matter and is not.
Unreported trials are not a random sample. Studies with disappointing or null results are systematically less likely to be published than those with favorable ones, which means the visible evidence base skews positive relative to the evidence that actually exists. A clinician reading the literature on a treatment sees a filtered subset, and the filter runs in one direction. Meta-analyses built on that literature inherit the bias. Treatment guidelines built on those meta-analyses inherit it again. The distortion compounds at every level, and it is invisible from inside, because the missing studies leave no trace.
There is also an obligation to the participants themselves. People enroll in trials, accept risk and inconvenience, and do so on the understanding that the knowledge generated will be useful to others. A trial whose results are never reported converts that contribution into nothing.
The measure of whether this worked
The number to watch is the one EMA has committed to publish by year's end, and the useful version of it is not the headline compliance rate.
What matters is the reporting rate for trials whose deadlines fall after the notification system went live, compared with those before it, because that isolates whether the reminders changed behavior or simply cleared a backlog. It matters whether the gap between commercial and academic sponsors narrows, since academic non-reporting has been the more stubborn problem and universities respond to different incentives than companies. And it matters whether any national regulator actually uses the monitoring system to take action against a delinquent sponsor, because until one does, the system has demonstrated only that the information can be produced.
EMA has done the part that was within its power, and doing it quickly in response to outside pressure is genuinely to the agency's credit. The remaining question is whether anyone with enforcement authority picks up what the agency has now made visible. A decade of a mandatory rule at 81% compliance suggests that being seen is necessary and has not, on its own, been sufficient.