1 Comments
The NIH Data Science Strategic plan drives us towards having accessible, well-organized, secure, and efficiently operated data resources to maximize the value of data generated from NIH funding. To meet the mark as we move forward, data need to be interoperable, interconnected, harmonized, standardized, and shared where and when appropriate. One way we hit the mark is through encouraging researchers to adopt Common Data Elements (CDEs).
CDEs foster rigor, facilitate data sharing, and allow multiple datasets to be integrated. They also help make data more FAIR (Findable, Accessible, Interoperable, and Reusable). Many different CDEs are currently in use and can vary across research disciplines, so we would encourage researchers check out databases like the NIH CDE Repository for examples, tools, and other related resources.
Through a recently released Request for Information (NOT-LM-21-005), we seek your thoughts on how you use CDEs, potential challenges to their adoption, and how NIH might facilitate and incentivize their use to help us plan future CDE-related efforts.
Do you use CDEs? How have they benefited your work? Did you face any barriers, and how were they overcome? What resources or tools would make it easier for you to use CDEs? Can the NIH CDE Repository be enhanced? Please tell us.
We seek general feedback on CDEs regardless of the research topic or disease area. That said, we are especially interested in their use in COVID-19 research. Systematic and consistent data on study participants, for instance, could be collected across multiple COVID-19 sites with CDEs. And since CDEs allow data to be pooled, strengthen their statistical power, and facilitate reuse, we might learn more about coronavirus disease as a result.
We look forward to hearing your thoughts. Comments are being accepted electronically here until May 10, 2021.
CDEs are hugely important and I welcome the NIH focus on their use, particularly related to COVID.
From an informatics and computational perspective the more ontologically/clinically annotated and machine readable CDEs are, the closer we get to automated curation of harmonized research assets. The shift towards automating harmonization pipelines is a critical transition in making biomedical data more reusable.