How Harlan Krumholz got big pharma to open its data

Harlan Krumholz, M.D., has street cred. In scientific circles, the Harold H. Hines Jr. Professor of Medicine (Cardiology) and professor of investigative medicine and of public health (health policy), is known as a gadfly who calls out pharmaceutical giants for—among other things—sitting on clinical trial data that might hurt their bottom line. So when Krumholz came up with a plan to collaborate with big pharma to release such data to other scientists, his colleagues told him it would never happen. “People said that industry would never share data. If they did share data, they would never share them with me because they’d think I’d be too tough on them,” Krumholz recalled.

But it did happen. Twice.

First Medtronic allowed Krumholz to farm out their clinical data to two independent research groups. They came back with slightly different analyses, confirming Krumholz’s belief that transparency can only be good for science, physicians, and patients. And early this year the Yale University Open Data Access (YODA) Project, which Krumholz directs, entered into an agreement with Johnson & Johnson. Under the agreement the company relinquishes control of its clinical trial data to YODA, which sends the data to two independent research groups for analysis. The company will initially release pharmaceutical trial data, followed by data from clinical trials of medical devices and consumer products.

“I have not done this alone,” Krumholz said. His collaborators include Joseph Ross, M.D., M.H.S. ’06, assistant professor of medicine, Cary Gross, M.D., professor of medicine, the British physician and academic Richard Lehman, M.D., and YODA staff Beth Hodshon, Jessica Ritchie, and Kat Blount. Bioethicist Ezekiel Emanuel, M.D., Ph.D., leads YODA’s steering committee. “The progress we have made is a tribute to the companies that have been willing to lead a new approach. We forged a vision of what could be accomplished together.”

Krumholz talked with Yale Medicine in February about the agreement with Johnson & Johnson and his career in outcomes research.

How did you get involved in health policy and outcomes research?
When I was a Yale undergrad, I had the opportunity to do an internship at the office of rural health service in North Carolina. They had set up a series of clinics run by nurse practitioners and connected to general practitioners at a distance. My job was to learn about those community clinics and do some surveys about what the challenges and obstacles were. Before that I had worked in basic science labs and thought that that might be a direction that I would take. This [internship] turned me more toward thinking about the public and society and patients, and how you could generate knowledge that would have a practical impact on people.

What is outcomes research?
It’s saying we want to know at the end of the day whether we’ve helped people. Have patients experienced a good outcome? Are they in a better place than they would otherwise be?

What was the problem you wanted to solve with YODA?
It turned out in our research that half of clinical trials are not published within three years—and many of them are never published. If science is to be progressive and self-correcting, it is critical that the data be available for multiple groups to look at—to draw their own conclusions, to put the results in public view, and to allow dialogue to take place.

Why are studies not being published?
It is not entirely clear. We saw that studies that are sponsored and owned by industry are now published at a slightly higher rate than previously. We did an audit of NIH and saw again that half the studies at three years were not being published. This is part of our culture—people run to finish studies; then they move on to the next thing. They may lose interest. They may get results that didn’t excite them. Maybe there is a profit motive and the results aren’t what they were looking for.

How did YODA begin?
We were trying to figure out how we could partner with industry and show a way forward. Prior to that, litigate and regulate were the two primary ways to get data to be more widely shared. There was a third path, which was to partner. It turned out Medtronic was having some problems with a product. I approached the chief science officer and started a huddle with their leadership. I said you’d probably hire a consultant, hire someone to look at the data, do an internal review, circle the wagons, and try to defend yourselves. How about a different path? What we’re going to do is share all the data assets related to these products; we’re going to partner with an independent academic organization; and we’re going to allow them to enlist two expert groups to do reviews.

Why did Johnson & Johnson agree to participate in YODA?
At Johnson & Johnson there seemed to be sincere interest in this third path. They have a credo that is built on trust and serving the public. And they had had some issues with some of their products. They wanted to be leaders in setting the standard for the future.

How will YODA benefit patients?
There should be a more complete picture of the balance of risks and benefits of the drugs and clinical strategies that lie in front of them. There should be more confidence in the scientific system—that it’s not just one group that’s maybe getting paid by another group that has a vested interest. It’s a more trustworthy, more credible system that is producing information better, faster, and cheaper.

Watch an interview with Harlan Krumholz here.