IOM report on sharing clinical trial data met with skepticism

Oncology researcher Dr. Vinay Prasad says he often feels frustrated conducting studies comparing the optimal use of marginally beneficial but expensive cancer drugs. Prasad, based in Washington, D.C., often runs into roadblocks when he makes information requests to drug manufacturers about studies they have conducted or funded.

“Medical science has been hijacked by commercial interests and proponents of new things,” said Prasad, who was so passionate during an interview about the issue that he was nearly yelling into the telephone. “You consistently get either no answer or an excuse.”

Prasad is not alone in his frustration. Data from many clinical trials remain unpublished and unreported long after the investigations are completed. Drug companies often cite the need to protect confidential business information.

Of 182,330 clinical trials and observational studies listed on the website maintained by the National Library of Medicine and the National Institutes of Health, ClinicalTrials.gov, only 15,899 have posted results.

When researchers from the Cochrane Collaboration, which provides systematic research reviews, looked for data on the safety and effectiveness of various treatments for flu symptoms, they said it took 4½ years to obtain the complete list of trials conducted by manufacturers.

Last week, the Institute of Medicine urged major changes in data-sharing practices to make it easier for independent researchers to compare therapies.

Some experts have hailed the report—Sharing Clinical Trial Data: Maximizing Benefits, Minimizing Risk—as a big step toward greater research transparency.

In presenting the report, IOM leaders recognized stakeholders' hesitancy to share research, but urged the adoption of policies to guide the registration and release of study results, particularly for NIH-funded trials.

“Responsible sharing of clinical trial data will advance the science that underlines the clinical foundation of good care,” said Dr. Bernard Lo, chair of the IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data.

Progress has been made, with several drug and device manufacturers voluntarily entering into collaborations to share research over the past few years. The IOM report moved the bar by suggesting specific timetables for releasing adverse-event summaries and analyzable data sets.

But ethicists, patient advocates and researchers remain skeptical. They cite the lack of an enforcement body to audit and ensure compliance, language that still leaves certain chunks of data inaccessible and unclear guidance on who will be responsible for sifting through all of the data.

“The problem is larger than simple access,” said Peter Doshi, an assistant professor of pharmaceutical health services research at the University of Maryland.

He's also associate editor at The BMJ, which last year published Cochrane's review of the two major studies on Tamiflu, and posted more than 150,000 pages of influenza antiviral data on the Web.

“There is still nothing that requires people do anything with” the data, he said.

The Cochrane studies ultimately concluded that government stockpiling of Tamiflu drugs wastes money and that the antivirals' efficacy is vastly overstated.

The IOM recommended that data should be shared as a normal operating procedure and that a full analyzable set of data from a study be made available no later than 18 months after a study's completion, with some exceptions.

Complete data on newly approved products or drugs should be made available within 30 days of regulatory approval, the IOM stated.

A third IOM recommendation calls for implementing strategies to encourage data-sharing, including creating technological platforms and boosting financial support. The report acknowledged the lack of an enforcement body to ensure compliance.

Some drug and device manufacturers already have sought out collaborative data-sharing opportunities. GlaxoSmithKline has a Web portal called ClinicalStudyDataRequest.com in which a number of major drugmakers participate.

Johnson & Johnson and Medtronic share data through the Yale Open Data Access (YODA) Project. Last week, that Yale School of Medicine project announced Johnson & Johnson had agreed to share clinical trial data for its devices and diagnostics, making it the first devicemaker to broadly share such data with a medical school.

A major challenge for projects such as YODA has been accessing data for products already on the market.

Data from the 1980s and 1990s may not even be in a digital format, and the language in informed consent forms could pose legal challenges, said Dr. Joseph Ross, an associate professor of medicine and member of the YODA Project.

The IOM has done little to address this concern.

For now, Prasad said he has not noticed a “real-world difference” in accessing the data he requests.

Manufacturers “don't tell us the truth about products. They tell us the best version, in the best light,” he said. “This should really be about what's best for” patients.

Follow Sabriya Rice on Twitter: @sabriyarice