Research organizations facing a potential new policy on sharing and managing their data want to hear more from the NIH about how it plans to enforce the policy and whether institutes within the agency will impose additional requirements.
The draft data sharing policy, released Nov. 6 by the National Institutes of Health, got generally positive responses from researchers and attorneys who work on scientific regulations and policies.
It would require all scientists to develop a plan for managing and sharing research data as a condition of receiving a piece of the agency’s $32 billion in grant funding. The measure marks the first time the NIH has issued a draft policy for data sharing that hits at every part of the agency rather than specific projects or databases.
One of the questions researchers want answered is whether the 27 various institutes and centers within the NIH will set their own requirements for sharing data, Heather H. Pierce, senior director of science policy for the Association of American Medical Colleges, said.
“The draft policy refrains from being overly prescriptive, which is good for an agencywide policy,” Pierce said. “But I think we need more about precisely what’s going to be left to the” institutes, centers, and offices.
The draft policy sets minimum standards for data sharing, which individual institutes, centers, and offices can build upon. If a research program wanted all scientists to put their data into the same repository, an institute can build that into the terms and conditions of the grant.
That flexibility is the point.
“Given the diversity of our entire research portfolio, it would be very hard to come up with a one-size-fits-all data sharing policy,” Carrie D. Wolinetz, the NIH’s associate director for science policy and acting chief of staff, said in an interview. “We want to create an approach that allows for flexibility.”
However, there’s a danger of “checkbox data sharing” unless the NIH takes steps to evaluate whether scientists took meaningful steps to share their data in ways that other researchers can successfully use, Pierce said.
Does It Have Teeth?
While researchers largely applauded the data sharing push, they questioned how the NIH would enforce the policy given how spotty some grant recipients have been about even reporting their findings.
“What continues to remain unclear is whether there is any teeth to the policy,” Yale University cardiologist and health-care researcher Harlan Krumholz said. Krumholz will provide the researchers’ perspective during an upcoming National Academy of Medicine workshop on clinical trial data sharing.
While failure to meet the data-sharing requirements could jeopardize existing and future grants, the NIH has already had to beef up its clinical trial reporting requirements given low compliance rates from grant recipients required to report back on their findings.
“These results are never known, let alone their data shared—and there has been no consequence,” Krumholz said.
Wolinetz said the agency is “very serious” about enforcement while aiming to be reasonable.
“While we take compliance very seriously, we’re not looking for gotcha moments,” she said. “We’re looking to really move the needle on the culture of data management and sharing.”
‘A Step Backwards’
However, the data sharing requirements come too late in the grant application process after studies have already been greenlit for funding, according to the American Medical Informatics Association. While the NIH pitches that as slashing red tape, Douglas B. Fridsma, the informatics association’s president, said the data-sharing plans should be included earlier in the process and weighed when assessing applications.
“This policy represents a step backwards,” he said. “As proposed, this draft policy imposes a check-the-box requirement that will do little to modernize the 15-year-old NIH data sharing policy, only making the task of managing and leveraging scientific data for supplemental use even more difficult.”
A final data-sharing policy will come out in the next year, and there will be ample time for institutions to comply, Wolinetz said. When the NIH revised its conflict-of-interest requirements in 2011, institutions had a year to comply, and Wolinetz said she expects something similar to happen for data sharing.
“This affects nearly everything we do, and we’re very cognizant of that,” she said. “We want to make sure we are giving the community sufficient time to adapt.”
They suggest that the NIH plans to cover those costs wherever possible, said Valerie H. Bonham, an attorney with Ropes & Gray who was previously a senior attorney in the NIH general counsel’s office.
While the data-sharing policy is part of the NIH’s stewardship of billions of taxpayer dollars, Wolinetz said it’s also important for driving science and making sure the public has access to the research it funds.
“We’re hoping for culture change in which management and sharing of data becomes part of science,” Wolintez said. “We’re hoping for culture change in which management and sharing of data becomes part of science.”