The Centers for Medicare and Medicaid (CMS) released all of its Medicare billing data for individual physicians last week. The week before that—March 31st—pharmaceutical and device companies submitted their first round of data to CMS to comply with the Physician Payments Sunshine Act. Indeed, the drive for transparency has touched almost all aspects of healthcare in recent years, including clinical trial data.
Pharmaceutical companies have become more open about releasing drug data in recent years, pushed by large U.S. legal settlements and growing public pressure in the U.S. and in Europe.
A recent article on Nasdaq.com reported that "[a]round half of all trial results go unpublished, according to current estimates, including a 2009 paper that looked at 677 studies conducted in different countries. A 2012 paper found just 45% of 635 U.S. National Institutes of Health- sponsored trials were published within 30 months of completion."
Times have changed since 2009 and even 2012. In December, we reported that GlaxoSmithKline (GSK) promised to make detailed data from its clinical trials available to individual researchers. Roche has also announced that it would provide researchers with access to analyzable data from Roche clinical trials through GSK's portal for data sharing, which is being used by GSK, Boehringer Ingelhiem, and ViiV Healthcare (a joint venture between Pfizer and GSK). Starting January 2014, Sanofi also began using the portal.
The GSK website is supposed to make available data from clinical trials to independent researchers. Scientists would have an opportunity to request access to anonymized patient-level data, develop their own findings about safety and effectiveness, and conduct further research.
Earlier this year, Johnson and Johnson announced a clinical sharing agreement with Yale School of Medicine.
There is general agreement that clinical data sharing can improve the efficiency of clinical trials and increase public confidence in clinical research. But some stakeholders are understandably concerned about who controls access to data, the purposes of disclosure, and safeguards to protect all parties. The Association of the British Pharmaceutical Company (ABPI) agreed that the information should be publicly available, but stated that there needs to "be a process involving the company in identifying what elements should not be disclosed due to data privacy concerns or to protect commercially confidential information."
The Wall Street Journal (WSJ) recently conducted an interview with Peter Doshi, assistant professor of pharmaceutical health services research at the University of Maryland School of Pharmacy and an associate editor at the British Medical Journal. WSJ noted that Doshi "has been at the forefront of a campaign to prod drug makers to move faster."
Doshi stated that there has been much more discussion about releasing data in the past few years than ever before. Furthermore, companies have taken concrete steps towards accomplishing it, though he is not ready to celebrate just yet: "What we have now are a set of policies and procedures and software platforms that need to be tested and see if they deliver what they need to deliver, not just what they promise."
WSJ noted that GSK has gotten "a lot of praise" for being the first company to make strides in providing clinical trial data. Doshi agreed that "GSK does deserve credit for being the leader – for doing it first and doing it earlier. The company was the first to come out with this corpus of policies and a framework for providing a scheme for access to data, although it's in a controlled environment. They're not the only ones now, but when you read the media coverage these days they're still singled out as doing something dramatically different than other companies. I don't think that's the case anymore." As we noted above, many companies are becoming more transparent, and several have used GSK's own platform.
In terms of off-label uses, Doshi stated that (emphasis added):
"We know off-label use is widespread, but these trials are not on the menu of trials that can be requested [the companies are offering only trials that were submitted to the FDA]. It's tricky. You can request whatever you want, but off-label trials are being treated differently. When they will become unavailable is unclear. In terms of the way their policies are written, those are not trial that they are intent on providing access. It's a major, identifiable gap, because off-label use is so common. But if one wants data from randomized, controlled clinical trials that addresses the actual ways in which drugs are being used off label, well, they are out of luck.
"Another issue is that there are trials for on-label use that are not being offered. Sanofi is only making data for certain trials for products that were approved since January of this year. But of course, there are Sanofi products being used today for which there were trials conducted in past years. Those approved products are being prescribed and used today were based on clinical trials from the past, but the relevant data is not accessible via the Sanofi policy."
Doshi speculated that the reason for Sanofi and Novartis' cut-off date could be "cost and their interpretation of patient consent. They may feel the patient consent prohibits them from making detailed data available. It could also have to do with the fact that the older clinical study reports were written under the assumption that they would be treated as secret and they are not comfortable revisiting that assumption."
GSK's cut-off goes back to 2000. Doshi noted that GSK's "Paxil antidepressant is an interesting example. The data actually originates with SmithKlineBeecham [which merged with Glaxo in 2000]. It's really SmithKline's data, but to GSK's credit, they're making it available beyond the written policy, which only covers access for data from the merger. Technically, the Paxil data should not be available, but they're doing so and that's a good thing."
However, full transparency can lead to issues of mistakes in data. Doshi notes that with clinical trial data, "[i]f this raw data wasn't recorded properly, there may be inaccuracies. So a key question becomes whether [corporate] policies need to also address [access to] the case report forms, which is the raw data recorded by the trial investigators." Doshi contended that policies do not address access: "Policies should address this because if they're really going to be supportive of external double checking, then the people who will be doing the checking need to be able to access all of the data. But right now, a deeper analysis of the accuracy of the trials and certain events pertaining to participants are largely out of view, according to most of the policies."
Doshi stated that GSK's sharing of raw Paxil data is "the only instance" he knew of where researchers asked for case report forms and the company was willing to provide it. Wall Street Journal asked if GSK had set a precedent with this decision. "I hope so," Doshi responded. "I don't want to see it treated as a one-off exception. But there is always the risk that high-profile cases become exceptional events and so the average researcher cannot access the same level of detailed data or there are sufficient hurdles in terms of time and energy to access. If we're trying to see a change, we need to look at what happens with run-of-the-mill requests."
Doshi also answered questions about "gatekeepers" of the clinical trial data.
What do you think of Johnson and Johnson's decision to rely on the Yale Open Data Access project to vet researcher requests for trial data?
All of the companies are using a gatekeeper. J&J chose an academic group. The others have agreed on a panel that is considered to be external. They use words like 'independent' to describe the panels, but there are instances of financial ties.
Are you suggesting the gatekeeping concept may prove unworkable?
Look, I think these are all experiments. It's fantastic they are happening now and we have to let them happen to see how they work out over time. We need to see how much time and effort is involved in gaining access to data. Every gatekeeper request and every requirement [made by a gatekeeper] may be a hurdle. We have to see if they're working as efficiently as we want them to do so over time. So I don't have an answer right now and I don't think anyone does.
The interview ended by Doshi speaking about the Food and Drug Administration's (FDA) role. "It's vital to also keep regulators in our sights. The fact is that the FDA is in possession of more clinical trial data than any single company, and policies allowing easy third-party access to the data in FDA's holdings could single-handedly create sweeping changes. That is what is happening in Europe, with the European Medicines Agency's new transparency policies, and we need to see the FDA, which holds far more data than EMA, take a similarly active interest in data transparency."
Read the entire Wall Street Journal Interview here.