Researchers conducting human studies of new treatments have significantly increased their public reporting of results in the last two months, the National Institutes of Health disclosed Wednesday.
The improvement suggests researchers in academia, industry, and the government may be responding to public pressure, after years of routinely failing to follow a federal law requiring they report results to an NIH database called ClinicalTrials.gov.
Deborah Zarin, director of ClinicalTrials.gov, attributed part of the improvement to a STAT investigation published in December, which for the first time named the academic medical centers and drug companies with the worst records of compliance with the reporting law. She also credited the agency’s own outreach to researchers and training efforts.
“Anecdotally, we have heard from certain institutions that they have been energized by the STAT article,” Zarin said. “Naming names and calling attention to it has an effect.”
STAT found that many prestigious medical research centers — including some top recipients of NIH research funding — flagrantly violated reporting requirements that became mandatory in 2008. Many reported results to ClinicalTrials.gov after the legal deadline — or failed to report at all — in 95 percent to 100 percent of relevant trials. Overall, drug companies also performed poorly, but academic institutions did worse — findings that NIH Director Dr. Francis Collins called “very troubling.”
The data released by NIH show that results for 290 more studies were posted to ClinicalTrials.gov in December 2015 and January 2016 than in the same period a year earlier, a 25 percent rise in new submissions, and a 6 percent increase in reporting of corrected results for trial findings that had previously been submitted but not approved by NIH.
Zarin would not say which organizations mentioned posting results in response to the STAT article, but in interviews last fall, officials from Stanford University, Memorial Sloan Kettering Cancer Center, the University of Cincinnati, and other research centers said they were redoubling their efforts to report results after being informed of STAT’s findings. Those institutions were among the worst offenders of the reporting law.
A study published Wednesday in the medical journal BMJ found similarly lax reporting by leading academic institutions. The BMJ report showed that 51 top academic and nonprofit institutions routinely flouted the requirement to post study results on ClinicalTrials.gov. They reported required results within two years of a study’s completion — a year beyond the legal deadline — only 13 percent of the time.
The study also found that researchers published findings in medical journals within two years just 29 percent of the time.
“The lack of timely reporting and publication fundamentally impairs the research enterprise, violates the commitment made by investigators to patients and funders, squanders precious time and resources, and threatens to compromise evidence-based clinical decision making,” the authors wrote.
Millions of patients and medical professionals use ClinicalTrials.gov to compare the effectiveness and side effects of drugs and devices used to treat deadly ailments. The failure to report results can mislead practitioners and skew research directions.
Dr. Harlan Krumholz, coauthor of the BMJ study and professor of medicine at Yale University, said in an interview that “somehow we’ve developed a culture where the final step in the research process, reporting the results, became discretionary” — a personal choice with little oversight and no consequences for violators. That failure of the system “corrupts the medical literature,” he said.
The BMJ study relied on data collected in September 2013, while STAT used more recent information, downloaded from ClinicalTrials.gov last fall.
Some researchers have argued that publishing results in peer-reviewed journals is a superior form of reporting to posting findings in the government’s public database — and should be sufficient.
Krumholz rejected that argument. Not only did most researchers fail to publish in journals in a timely fashion, he said, but ClinicalTrials.gov — unlike journal articles — is easily accessible to the public at no charge and designed to be a centralized repository of clinical data.
By law, NIH may withhold grant funding after disclosure lapses, but it has not done so in a single case. That might change later this year, after the agency issues new rules meant to clarify reporting requirements.