In order to ensure greater public access to clinical trial data, a system should be in place that provides the desired information while protecting patient privacy and the integrity of the data, according to a report published online Oct. 21 in the New England Journal of Medicine.
Regulatory agencies in the U.S. and Europe have proposed making regulatory data more widely available, sparking a debate over the benefits and potential consequences of sharing participant-level information.
The authors, led by Michelle M. Mello, JD, PhD, of the Harvard School of Public Health in Boston, argued that there are numerous benefits to expanding access to data.
“Although some benefits of data sharing can be achieved to some extent through sharing of summary data, access to participant-level data may enable a much wider range of analyses,” they wrote.
One benefit is that the data would be available to independent researchers, who can perform their own analysis and provide evidence that either backs up or contradicts findings on safety and efficacy. They also can help identify weaknesses in methodology.
Broader access to data may allow for additional analyses beyond the intent of the original study that may lead to significant findings. Additional benefits include the opportunity for participants to learn more about their specific medical conditions and ensuring that the research does not expose participants to unnecessary risks.
Mello and her colleagues, however, also acknowledged there are risks to expanded data sharing. One of the major concerns is the possibility that participant privacy may be breached, making confidential medical information public. Additionally, researchers who are not highly skilled or who are not working with a neutral agenda may use the data.
To maximize benefits and minimize risks, therefore, the authors suggested a system for sharing data from clinical trials of all approved drugs, devices and biologics that meets certain criteria.
They suggested four data expansion models, each focusing on different variables. In a model that is fully open access, data could be downloaded at will. This type of model would allow for transparency and full access, but at the expense of increasing risks.
Another possibility would be to implement a system that requires requesters to submit questions. The data holders or an intermediate body would then analyze the data and provide answers if the questions are scientifically appropriate.
Other possibilities include allowing data holders to decide whether to release data and which data to release, but ensuring there is a way to oversee the process. Alternatively, a model could exist that would utilize a “learned intermediary” to assess requests for data and ensure data are used appropriately.
Despite the risks, “the push for greater transparency in the area of clinical trial data appears inexorable,” they wrote. “The question is not whether, but how, these data should be broadly shared.”
The report precedes the Oct. 23 first-ever meeting of the Institute of Medicine’s (IOM) Committee on Strategies for Responsible Sharing of Clinical Trial Data, according to a press release.