From the Director: Clinical Trial Data Sharing, the Law, and Reality

Elisa Hurley

Though not a new issue, clinical trial data sharing has over the past several months taken center stage within the medical research community. In December 2015, a STAT article showed there was substantial under-reporting to ClinicalTrials.gov. One month later, a French clinical trial left one man dead and European scientists calling for more data transparency. Five days after the deadly incident, the International Committee of Medical Journal Editors (ICMJE) issued new proposed clinical trial data sharing requirements. While many agree that increased data sharing is an ethical imperative given the risks that clinical trial participants are asked to undertake, the question many institutions are now struggling with is how they can meet that imperative with their current resources and while protecting the privacy of study subjects.

In January 2016, one man died and five other subjects were sent to the hospital after taking a drug during the phase I portion of a French clinical trial sponsored by Bial. In response, the British Pharmacological Society called for “improve[d] early access to data from catastrophic clinical trials.” The Society pointed out that with respect to the drug used in the French clinical trial, “A better understanding amongst the scientific community of the nature of the drug, and the way in which it was dosed, would be helpful, even at this early stage, in preventing further harm to volunteers in other clinical trials.”

One month earlier, in the United States, STAT released their findings that academic institutions and companies frequently violate a federal law that requires trial results be reported to ClinicalTrials.gov. ClinicalTrials.gov, which was created in 1997, is run by the National Institutes of Health (NIH). The FDA Amendments Act of 2007 (FDAAA) expanded the registration requirements and led to the creation of the ClinicalTrials.gov results database for the public. The database includes study outcomes and adverse events that were reported during the course of a trial.

It is important to note that while FDAAA covers all clinical trials that meet its requirements, irrespective of funding source, FDAAA does not apply to phase 1 trials. FDAAA also does not apply to trials overseas if they were not conducted under an Investigational New Drug (IND) Application or Investigational Device Exemption (IDE) or if the product was not manufactured domestically and exported for the trial.

In 2014, the Department of Health and Human Services (DHHS) issued a Notice of Proposed Rulemaking (NPRM) to expand “the scope of clinical trials required to submit summary results (under FDAAA) to include trials of unapproved, unlicensed, and uncleared products.” DHHS noted that expanding the submission requirements would provide additional information to future research subjects and reduce the number of clinical trials being inadvertently duplicated.

In 2014, NIH also proposed a policy on ClinicalTrials.gov registration and results submissions that would apply to any NIH-funded clinical trial, regardless of study phase, type of intervention, or whether they are subject to the applicable FDAAA registration and results submission requirements. The proposed policy states “there is an important ethical dimension to dissemination of clinical trial results because individuals who volunteer to participate in such studies, and who may assume risks, trust that what we learn will contribute to generalizable knowledge about human health.”

FDAAA has monetary penalties for non-compliant entities and the law directs the NIH to withhold funding if federal grantees are not complying with the law. However, the responsible agencies (FDA and NIH) have never penalized researchers or institutions for not posting data. Four years ago, Congressional representatives wrote to FDA and NIH citing their concern with a study that found substantial under-reporting to ClinicalTrials.gov. The lawmakers also asked if NIH had adequate resources and authority to enforce the reporting requirements. In April 2012, NIH responded that it did.

In their analysis, STAT found that industry had a better reporting rate than nonprofits and universities: “Results from academic institutions arrived late or not at all 90 percent of the time, compared with 74 percent for industry.” These findings were confirmed in a more recent BMJ study as well.

Several university officials told STAT that they lacked the time and administrative funds to fully comply with the law. In a March 2015 NEJM article that also reported low rates of results reporting to ClinicalTrials.gov, the authors noted one estimate that preparing results summaries for a single trial for ClinicalTrials.gov took between four and sixty hours. The authors also note that “it is possible that the NIH and other funders have been unable or unwilling to allocate adequate resources to ensure timely reporting.”

Despite under-reporting and enforcement problems, ClinicalTrials.Gov seems to be popular with the public. As of October 2015, the website received approximately 65,000 visitors each day and 207 million page views each month.

In another sign of growing consensus around mandatory data sharing, the ICMJE, which counts many of the top-tier medical journals as members, released its proposed data sharing requirements in January 2016. ICMJE “proposes to require authors to share with others the de-identified individual-patient data (IPD) underlying the results presented in the article…no later than 6 months after publication” and “proposes to require that authors include a plan for data sharing as a component of clinical trial registration.” ICMJE is accepting feedback on its proposals until April 18, 2016.

The ICMJE’s requirement to share data at the individual patient level has implications for IRBs and human research protections. For one thing, IRBs need to ensure that the informed consent process includes appropriate and understandable information about the data sharing plan and the conditions under which information will be shared. For another, the sharing of individual patient data, even de-identified, also entails that IRBs need to ensure there are adequate measures in place to protect individuals’ privacy and the confidentiality of their data.

Clinical trial data sharing serves to maximize research subjects’ contributions to science and enhances the chances that the risks they undertake are offset by future benefits. It also increases transparency around biomedical research and may, as a result, protect future potential subjects and serve to foster much-needed public trust in the research enterprise. There is no question that mandatory clinical trial data sharing poses significant operational issues for institutions, investigators, and IRBs. It may also raise additional ethical questions around adequate protection of subject privacy. The work it will take to address these issues should not be diminished. But there is also no question that data sharing is an idea whose time has come.

I invite you to share your perspective on clinical trial data sharing. Has the news of the French clinical trial and under-reporting to ClinicalTrials.gov changed your views on data sharing? How?

Should there be federal funding to help institutions report to ClinicalTrials.gov?

What other issues does mandatory data sharing raise?