by

As Sharing of Clinical-Trial Data Gains Acceptance, Methods Get Scrutiny

In the quest to improve the scientific transparency of human medical trials, the first hurdle is getting researchers and companies to agree that they really want to do a better job of sharing with one another.

In that regard, after years of pressure from patients and their advocates, there are indications lately that attitudes may be slowly improving.

The next challenge is figuring out how best to share. And that could be just as daunting. Three articles published on Wednesday in The New England Journal of Medicine explore varied approaches to the sharing of patient-level data.

In most of the emerging systems for sharing the results of human medical trials, corporate and university researchers simply post summaries of their findings. For many uses, that kind of broad characterization of a clinical trial is sufficient.

But in other cases, a researcher looking to answer detailed questions or to deeply explore a theory might want to see the patient-by-patient effects of a drug or a medical device that was tested in someone else’s human trials.

That can be more difficult, in large part because of the need to protect the privacy of the patients who participated in a medical trial, or to safeguard the proprietary information of a company.

In one of the three journal articles, authors at GlaxoSmithKline describe their approach. An independent review panel appointed by the pharmaceutical company will evaluate requests from outside researchers seeking access to patient-level data from its medical trials. Any approval of a data release will include terms that protect patients’ identities, and any findings generated by the outside researchers will have to be published.

In another of the articles, a group of researchers in the Immune Tolerance Network, sponsored by the National Institutes of Health, describes treatments for vasculitis, an inflammation of the blood vessels. As part of their report, the researchers have offered open access to their patient-level trial data through a Web site, without the need for a specific research plan or any approval of researcher qualifications. To protect personal identifying information, the researchers have merely left blank some fields in their data sheet.

In the third article in the series, the director of the federal government’s ClinicalTrials.gov Web site, Deborah A. Zarin, offers praise for both approaches, calling them “bold attempts to modify and improve the process by which research is reported.”

Dr. Zarin says the research community as a whole should learn from such attempts to improve the methods of balancing the benefits, risks, and costs of sharing clinical-trial results.

The Glaxo authors draw a similar conclusion, saying they realize that their method, in time, “will clearly need to be modified to accommodate the expectations of other sponsors, academia, and regulators.”

One of the main questions to be answered, the Glaxo authors say, will be whether the sharing of clinical-trial data will be done in a centralized manner or through a more distributed network. The ClinicalTrials.gov site is one option, set up by Congress, but government officials have made clear that it doesn’t have to be the only one.

If industry doesn’t get terms it likes, however, it may be ready to fight back. A report this month in the British newspaper The Guardian describes an internal memo depicting the major industry trade groups as already instigating a behind-the-scenes campaign to encourage patient groups to raise concerns “about the risk to public health by nonscientific reuse of data.”

Return to Top