Like you, I had also initially thought that access restrictions had to do with misalignments between the majority of patients' desires to empower accelerated discovery and a culture of paternalism that seeks to protect vulnerable patient populations. Over the past year, we've piloted accessing and depositing pediatric data from various sources and under various process in Cavatica including published datasets stored in dbGAP, the EGA, unpublished data provided by consortia such as the CBTTC and PNOC, data provided by investigators directly, and even data provided by patients themselves! We have not found recurrent support for the presumed paradigm of "locked behind Data Access Committees (DAC) for reasons of privacy and consent." Indeed, while many pediatric partners were extremely supportive of data-sharing, in the few instances where sharing failed, the limitations of access or use we documented were of the proprietorial type--i.e. perceived ownership. Thus, even when data are actually deposited for secondary use under the management of a DAC, access and use can remain informed by unaligned forces of data ownership that can restrict the discovery process.

For clarity, upon publication, many journals require the deposition of data for reproducibility of results and secondary use. Two main data download repositories for published genomic data include dbGAP and the EGA. The US government has tried to mitigate data ownership as a determinant of access for secondary use of data in dbGAP and has recently provided updated guidance on genomic data sharing policies ( https://gds.nih.gov/ ). Furthermore, Data Access Committees within the NIH that determine access to deposited data attempt to provide uniformity in decisions (https://gds.nih.gov/04po2_1dac.html ) that by and large do not support practices of data ownership (though some have suggested embargo periods permitted by the NIH require review). However, there exists a lack of uniformity in access between dbGAP and the EGA. In contrast to dbGAP, access to data deposited in the EGA (https://www.ebi.ac.uk/ega/about/access)--the major European repository where both US-based and international researchers also deposit data as part of the academic publication process--is controlled by DACs largely within the depositor's own institution. As such, some have identified such platforms as largely unregulated in terms of the mandate for access, with little recourse for a requester whose request is ignored or rejected.

In recognition of such failures of empowered FAIR access (Findable, Accessible, Interoperable, Reusable), the issue in both the pediatric and adult realms is indeed becoming more visible. Currently, patients largely lose the capacity to direct the use/access of their data in the context of research (as opposed to clinical data—see previous question’s answer). However, they do not lose their capacity to advocate and mandate for such FAIR practices of research data. Indeed, patients, as part of the consenting process, have every right to inquire and inform the mandate of data sharing. However, in our experience, in contrast to the indifference that may be encountered by researchers seeking access to data from the less-than-willing data generating researcher or institution, few researchers or institutions can long sustain a practice of no, limited, or delayed access to research data when confronted by an informed patient population asking the right questions in a visible and transparent manner.

However, as will be addressed in a subsequent question--even when DAC approval is provided, to date, file-system-like repositories including dbGAP/EGA supporting mere data download as a use-engagement strategy fail to meet modern definitions of data-empowerment for collaborative use and accelerated discovery. The biomedical data enterprise is still in its infancy when it comes to the routine use of cloud-based architectures like precisionFDA that co-locate compute and storage while supporting transparent and auditable collaborative use of data.