Researchers are worried that the European Medicines Agency (EMA) is backpedaling on its pledge to open clinical trials data to public scrutiny. Draft documents detailing how the agency plans to make such data available contains small print that may severely limit access and that constitutes a turnaround from earlier promises for openness, academics say.
“This is a stunning and surprising reversal,” says a 16 May statement sent to ScienceInsider by Tom Jefferson, a reviewer at the Cochrane Acute Respiratory Infections Group in Rome, and Peter Doshi, an assistant professor of pharmaceutical health services research at the University of Maryland School of Pharmacy in Baltimore and associate editor at BMJ. (Jefferson and Doshi fought a long battle, supported by BMJ, to obtain trial data for the influenza drug oseltamivir , better known as Tamiflu.)
The European ombudsman, Emily O'Reilly, has echoed the researchers' worries. “I am now concerned about what appears to be a significant change in EMA's policy, which could undermine the fundamental right of public access to documents established by EU law,” O'Reilly said in a statement on 16 May. She urged the agency to justify itself in writing by the end of the month.
EMA did not make the documents public but shared them with representatives of patients, consumers, health care professionals, the pharmaceutical industry, academia, and medical journals a part of “targeted discussions” to finalize the draft policy.
Europe has been leading the clinical trial transparency bandwagon in recent years. EMA vowed to become more transparent in 2010 after the ombudsman sharply criticized the agency for refusing to share data  about two antiobesity drugs with researchers from Denmark. Since then, it has released millions of pages of so-called clinical study reports (CSRs) upon request—including some 20,000 on oseltamivir to Doshi and Jefferson.
In 2012, EMA also announced a system  to “proactively publish clinical-trial data and enable access to full data sets by interested parties” by default, rather than in response to specific queries. The agency consulted extensively with scientists, industry, and patient groups on how best to do this; it missed its January 2014 target for the release of a new draft policy.
Now that the policy is almost ready, it has turned out to be a "U-turn" from EMA's promises, Jefferson and Doshi write. By introducing a number of controls, “EMA is putting a gatekeeping function … that will slow down the process” and hinder access for researchers who want to replicate studies, adds Carl Heneghan, director of the Centre for Evidence-Based Medicine at the University of Oxford in the United Kingdom and a co-founder of AllTrials , a campaign for transparency on trial results.
Scientists also worry that some data in CSRs may be blacked out. Although EMA says that much of the information in the reports is not commercially confidential, the draft policy does list a series of items that may be redacted before the data is shared. This could be necessary in the case of ”novel statistical or other analytical methods,” as well as results about potential off-label uses of medicines, say the draft redaction principles.
“We maintain our stance that it's not commercially confidential information, but it's still information that has commercial value,” says EMA spokesman Martin Harvey-Allchurch. For example, someone could use the information to support a market application outside the European Union's jurisdiction, he says. But Trudo Lemmens, an associate professor of health law and policy at the University of Toronto in Canada, says the changes would put industry in a position to “enter negotiations with EMA about what they want to see hidden and what not.” He calls the plans an "incredible step backward."
On a more fundamental note, researchers who want to access data would have to acknowledge that CSRs are “protected by copyright and proprietary rights” of the company that submitted them to EMA. Ticking that box would create a contractual recognition of rights that are legally “questionable,” Lemmens says: While industry generally considers that it owns the data from the clinical trials data it sponsors, others see the data as a public good.
This debate underlies ongoing litigation between EMA and U.S. company InterMune, which sued the agency for divulging data to competitors . “I would find it highly problematic that while EMA defends its right to make data publicly accessible, they're asking researchers to recognize [industry's] legal claims,” Lemmens says.
EMA's Harvey-Allchurch denies that the agency is reneging on it promises. “We're not going backwards,” he says. “It's about facilitating access.” He stresses that the plans are still open for discussion before EMA presents the revised draft policy to its managing board on 12 June.