By Deborah Borfitz 

November 16, 2022 | There is little doubt about the “data magnetism” of Vivli, an open-access, data-sharing platform integrating clinical trial data contributed by biopharmaceutical sponsors, academic institutions, and nonprofit foundations around the world, according to Rebecca Li, executive director of the nonprofit organization behind its development. Vivli was created as a generalist repository in 2018 and since then has driven 126 new research findings—all publicly disclosed, primarily in the peer-reviewed literature. 

Vivli’s popularity will soon be getting a big boost from a policy change at the National Institutes of Health (NIH), effective Jan. 25, 2023, requiring all research funded by the agency to have a data management and sharing plan. Out of the six supported data-sharing platforms, Vivli is the only one specific to clinical trials that don’t already have a domain repository (e.g., specific to cardiovascular or cancer studies) for consolidating the information, says Li, who is on faculty at the Center for Bioethics at Harvard Medical School. Vivli is cited by NIH as one of the acceptable-use repositories for human research data. 

At present, Vivli houses individual anonymized patient data collected from nearly 7,000 clinical trials, she reports. Data contributors include 27 pharmaceutical companies, together with 17 academic, foundation, and government members. 

“We have one of the richest collections of clinical trial data that exists,” says Li, covering studies from 20 therapeutic areas that include Alzheimer’s disease, cancer, diabetes, and infectious diseases such as COVID-19. The more data that pools together across boundaries, the richer the insights that can potentially be gleaned. 

In an article that recently published in Health Data Science (DOI: 10.34133/2022/9768384), Li and her colleagues briefly describe two case studies of how reuse of data in Vivli could be of value to researchers.  

‘Sea Change’ 

For investigators running trials using NIH funding, “it is going to be a sea change after January 25th,” says Li. In the past, academic researchers tended to share the output of their studies within their own institution—if they shared it at all. “Now, the default will be you need to share it more widely and it should be accessible as soon as possible... [meaning] no later than publication or end of award.”  

The new policy is a “pretty major leap for the NIH... and has some real teeth behind it,” she notes. The trio of expectations is that data sharing “be the default practice, responsibly implemented, and prospectively planned,” which up to now has not been baked into the culture of the academic environment. 

Excepting lab notebooks, preliminary analyses, and physical objects, “anything generating scientific data needs to be shared with external people, not just within your own lab group or institution,” says Li, who previously ran NIH-supported trials at Harvard. 

Vivli has fielded about 621 data requests over the years, 342 of which have been formally approved. Only 62 data requests have been denied, for reasons that include being “out of scope” of the data contributor’s data sharing policy and the data request requiring data that was not collected in the studies. 

A recent analysis has found that secondary analysis of the contributed data in Vivli has been for an assortment of purposes, among them to optimize the design of future clinical trials (14%) and develop new clinical tools and decision aids (70%) or clinical guidelines (4%) and generating artificial intelligence algorithms (8%). In several instances, Vivli data has been used for grants and funding purposes. 

Big-Data Advantage 

Other supported platforms for data-sharing are Dataverse, Dryad, Figshare, Mendeley Data, and Open Science Framework. The NIH “first wants researchers to go to their domain repository,” Li stresses. The agency doesn’t want these general repositories to compete with ones already providing a home for study data in a specific therapeutic area.  

For trials without a home, the big-data advantage of putting information into Vivli is that it can be combined in new and interesting ways with other study data in the same arena, says Li. So much COVID-19 data was being contributed that in 2021 Vivli launched a specialized portal to expedite pandemic-related research. Another portal launches in January specific to HIV/AIDS funded by the NIH. 

Similarly, Vivli introduced a repository specific to antimicrobial resistance (AMR) in June 2022 funded by Wellcome that now includes testing on 69 antimicrobials and 434 organisms, she adds. The AMR Register includes data from several companies with antimicrobial resistance surveillance programs, which are mandatory for approval of new antimicrobial medicines. The platform provides access to surveillance data from companies with major antimicrobial programs: Pfizer (ATLAS), Merck (SMART), GSK (SOAR), Johnson & Johnson (DREAM), Paratek (KEYSTONE), Shionogi (SIDERO-WT), and Venatorx (Global Surveillance program). 

Industry is contributing the raw, participant-level data underpinning their trials for “transparency reasons,” says Li. Contributions by large pharma companies has also been motivated by data-sharing pledges they have made since 2014 through the European Federation of Pharmaceutical Industries and Associations and Pharmaceutical Research and Manufacturers of America. For Johnson & Johnson, this uniquely happens through the YODA Project, a partnership with Yale University that provides an independent process for the release of clinical trial data. 

Under the Generalist Repository Ecosystem Initiative (GREI) of the NIH, Vivli has been hosting monthly webinars to train and educate researchers on FAIR data—meeting the principles of findability, accessibility, interoperability, and reusability—and the importance of data sharing, says Li. “[GREI] is important to the data-sharing ecosystem and we want to make sure everyone is ready for it.”