Data Access Committees (DACs) must answer important questions about access to data, but often have to interpret complex and ambiguous inputs to make those decisions. Learn how the Data Use Oversight System (DUOS) makes it easier for DACs to make these decisions.
How DUOS streamlines Data Access Requests
Currently, when DACs receive data access requests, they must decide if the proposed research use is within the bounds of the data’s use limitations dictated by its consent forms.
Unfortunately, data use limitations are often described with unique language across the various consent forms in which they appear (diagram left). Thus a DAC is left to either attempt to interpret the consent form or receive the original IRB’s interpretation (ex. NIH Institutional Certification, Broad Data Use Letter) to determine the official use limitations.
On the other hand (diagram right), researchers' data access requests are often narrative scientific proposals of varying levels of depth and specificity.
The inconsistency and lack of clarity in the terms used to describe data use limitations and research proposals make it difficult for DACs to answer the question of “Is the proposed research within the bounds of the data use limitations?”
To resolve this issue, the Global Alliance for Genomics and Health created a common vocabulary for data use limitations and proposed research, called the Data Use Ontology (DUO; Lawson et al., 2021). The ontology is not only a standardized series of terms and definitions describing data use but is also computer-readable.
DUOS leverages the Data Use Ontology by enabling Data Submitters to describe their data use limitations with DUO terms, and Researchers to describe their research purposes with DUO terms. The result is that DACs using DUOS can compare data use limitations and research purposes using the same vocabulary of terms.
Additionally, with the use limitations and proposed research in DUO terms, DUOS can enable an algorithm to compute the comparison of the data use limitations and proposed research in an attempt to replicate the decision the DAC would make. Through testing, the DUOS algorithm has seen >90% agreement with DACs (Cabili et al., 2021). Currently, DACs are able to leverage the algorithm as a decision-support tool, reviewing the DUOS algorithm’s suggested decision prior to logging their own decision (Rahimzadeh et al., 2022). If the DUOS algorithm proves to consistently decide as the DAC would, DACs may choose to use the DUOS algorithm to automatically respond to data access requests.
Next steps
To learn more about how DUOS supports Data Access Committees, read Frequently Asked Questions for DACs.