A wealth of scientific and clinical data is being generated by the HIV research community; the ability to promptly analyze, integrate, and investigate these data sets from different points of view could speed and enhance research and avoid unnecessary duplication of efforts. The Enterprise partners aimed to unite a few major grant-giving organizations to begin harmonization of the way in which their HIV vaccine research is collected, analyzed, and reported. Other stakeholders such as representatives for consortia and individual researchers will be asked to comment on the outcome.
This convening brought together the National Institute of Allergy and Infectious Diseases, the Bill & Melinda Gates Foundation, and the European Commission to discuss and potentially develop policies on data-sharing to be incorporated into their grant agreements. The interim focus was on a few top funders’ with the potential to expand to other funders in the long term.
Format: A virtual convening, facilitated by teleconferences and emails.
Practical challenges for this approach were quickly identified after initial conversations among the organizers. Data-sharing policies are developed based on organization-specific timelines as well as administrative and legal processes, which are difficult (if not impossible) to navigate in a coordinated manner or to synchronize. Moreover, objectives of these policies cannot always be aligned without impacting the long-term scientific strategy of the organization. For example, the Gates Foundation in its CAVD program explicitly aims to build close bridges among the small number of funded laboratories, an objective that calls for extensive data-sharing among a limited number of predefined partners, while NIAID awards a large number of extramural grants to laboratories that cannot be expected to collaborate closely or share primary data. A single policy cannot satisfy both scenarios. It was also pointed out that policies must follow (not forcefully impose) the technological developments in data storage and analysis, and currently only a small subset of data (mostly genomic data) is being productively used on a large scale.
As a result of these discussions, it was decided that proactive harmonization of data-sharing policies is premature at this point and needs to be postponed until future date. Once complete and operational, the Collaborative Data Space project, mentioned in the Background section, should provide some real-world use cases that could greatly inform future discussions in this area.
Currently, funding organizations address the need for data-sharing in their grant-giving policies. However, their requirements on the subject are often vague and are not strictly enforced. Access to data generated from clinical trials is especially problematic due to concerns about the protection of human subjects, while research conducted in collaboration with an industry partner comes with its own negotiated data access rights that may vary from case to case. This state of the field calls for improved harmonization and establishment of shared policies that take into account social, technical, and ethical implications of sharing large data sets.
Facilitation of data sharing can be approached from two different perspectives: agencies who sponsor/fund the research and investigators who generate the research. The Bill & Melinda Gates Foundation, one of the largest funders of HIV vaccine research is currently exploring a “bottom-up” approach to data-sharing by testing a new data-sharing infrastructure accessible to researchers: the Collaborative Data Space (CDS), currently in the development phase. This convening aimed to explore a “top-down” approach where the funding organizations have clear and harmonized upfront requirements for the sharing of data generated by their grantees.