Last year, the U.S. National Institutes of Health (NIH) delivered a stern warning to two in-house clinical researchers who had broken an important rule. They had failed to submit the results of two clinical trials they had overseen to ClinicalTrials.gov, a database meant to inform the public about human studies and their results. The reporting requirement has often been ignored, but this time the agency took an unprecedented step: It told the scientists it wouldn’t approve any more of their research until they fell in line.

After that warning and other agency actions, the pair complied, well after the one-year deadline.

The episode, described in a Government Accountability Office (GAO) report published in April, adds to other, systematic changes NIH has recently undertaken to ensure that the more than $6 billion in clinical trials it funds annually, along with their results, are visible to scientists, physicians, patients, and ultimately taxpayers. Transparency advocates say the tougher stance is beginning to pay off. For example, GAO also found that between July and November 2022, the agency brought 235 extramural researchers into compliance with registration and reporting requirements.

“We really do like some of the changes that the NIH has made. We think that’s a really great start,” says Navya Dasari, a lawyer who until recently headed efforts by the nonprofit lobbying group Universities Allied for Essential Medicines to increase transparency of clinical trial results.

Candice Wright, lead author of the GAO report, says NIH “should be ensuring compliance [with the policy]. It exists for a reason.”

“NIH should be ensuring compliance [with the policy]. It exists for a reason.”

Under a 2007 law, sponsors running many clinical trials of drugs and devices—including those funded by NIH—are required to register them on ClinicalTrials.gov within 21 days of enrolling the first volunteer. The results generally must be submitted to ClinicalTrials.gov within one year of when key data are collected on the last participant. The law directs NIH to shut down funding to any institution whose researchers are not up to date.

But NIH has done little to enforce the requirements, even after it put in place a new policy in 2017 that expanded them to cover all NIH-funded trials and media reports began to throw a spotlight on problems.

As recently as August 2022, the U.S. Department of Health and Human Services’s Office of Inspector General found that just 35 of 72 NIH-funded clinical trials due to report their results in 2019 and 2020 had done so in a timely manner—and that 25 had not submitted them at all.

NIH has recently taken steps to bring those numbers up. They include having both the funding institute and the Office of Extramural Research contact tardy investigators to bring them into compliance. GAO, too, noted that extramural investigators are now required to show NIH proof of trial registration and results reporting before filing the annual progress reports necessary to receive their grant’s next year of funding.

Michael Lauer, NIH’s extramural research chief, credited the agency’s changes when he gave updated numbers for 530 extramural trials required to report results in 2020, 2021, and 2022. In a March blog post, he reported that fully 96% of these trials had reported results to ClinicalTrials.gov. Only 37% had met the one-year deadline, however, and in 2022 the median for tardiness was 400 days.

“What if I told you that 18% of trials had not obtained informed consent? You’d probably be appalled.”

“Clearly, we still need to improve, and we are committed to taking this challenge head on,” Lauer wrote on the blog. “Moving forward, you will see increased communication from us and, if needed, enforcement actions to get us to where we need to be.”

NIH’s critics say the agency still needs to do more. The GAO report also found that 16% to 18% of trials are registered late—a number that did not budge from 2019 through 2022. (The numbers are worse for pediatric trials, a recent study reported.) The tardy performances included NIH’s own institutes, led by the National Cancer Institute, where 81 trials were registered late in that period.

Deborah Zarin, who directed ClinicalTrials.gov from 2005 to 2018, argues that trial registration and results reporting is as important as getting a research volunteer’s informed consent to participate in a study. “What if I told you that 18% of trials had not obtained informed consent? You’d probably be appalled,” says Zarin, who is now at Harvard University and Brigham and Women’s Hospital. She and others note that the information is needed for many reasons, from making sure two research groups don’t repeat the same trial to revealing failed trials that often aren’t published so others can steer away from those approaches.

Till Bruckner, a policy analyst who founded TranspariMED, a campaign aimed at ending evidence distortion in medicine, calls NIH’s recent actions “an improvement.”

But Bruckner thinks NIH should pull funding from entire institutions that have a track record of poor compliance with the requirements. “If NIH would just once crack down properly on institutions, not only on individuals, that would send such a strong signal that going forward, 95% of the problem would be solved.”

Originally published in Science.