Brett Norman, Politico Pro: September 16, 2016
Drug companies and research institutions will have to publicly report more clinical trial data, including results that show their products or experiments failed, under new policies rolled out Friday by HHS.
The new rules address the fact that researchers have routinely flouted requirements to report data to ClinicalTrials.gov. A final rule from FDA and a complementary policy from NIH clarify which trials must be reported and project a new federal commitment to crack down on violators – threatening fines of $10,000 per day or disqualification from NIH funding.
The widespread failure by institutions to report the data is “simply unacceptable,” NIH Director Francis Collins said on a call with reporters.
“This is fundamentally an ethical issue,” FDA Commissioner Robert Califf said. “This is about maintaining the trust that we have with [clinical trial] participants … that if it doesn’t benefit them, it will benefit others.”
A 2014 review of the results of 400 clinical studies found that 30 percent had not been published or shared with ClinicalTrials.gov four years after they had been completed. A study earlier this year found that 43 percent of the trials conducted at 51 academic medical centers went unpublished two years after the trial was completed.
The new requirements, which will take effect in January, clarify that all NIH-funded studies must be published on ClinicalTrials.gov. The FDA will also require publication of studies being done on drugs and devices even if they have not yet been approved. Most studies must be reported within a year of completion, but those involving unapproved products could wait up to three years.
FDA and NIH are also expanding the range of information that must be published, including the demographic data on study participants, any adverse events and the original plan for statistical analysis. The latter is meant to combat so-called p-hacking – when the original hypothesis of an experiment fails and researchers massage the data to come up with some other positive finding that is statistically significant.
“We’ve had a problem where people do the trial and don’t like the result, so they do another analysis,” Califf said.
FDA and NIH decided not to require a lay summary of trial results that would enable the public – not trained in statistics – to more easily digest the information that is posted, angering some consumer advocates.
“That is an outrageous decision,” said Diana Zuckerman, president of the National Center for Health Research. “Without it, clinicaltrials.gov will not be providing useful information to patients.” The data won’t be useful to physicians either unless it is summarized, she said.
Collins and Califf said they had been uncertain how to keep “bias” from creeping into the summaries, adding that advocacy groups should be able to harness the relevant data and make it available in a way that would be more useful to patients. […]
Read the original article here.