Achieving Data Liquidity: Lessons Learned from Analysis of 38 Clinical Registries (The Duke-Pew Data Interoperability Project.

BACKGROUND: To assess the current state of clinical data interoperability, we evaluated the use of data standards across 38 large professional society registries. METHODS: The analysis included 4 primary components: 1) environmental scan, 2) abstraction and cross-tabulation of clinical concepts and corresponding data elements from registry case report forms, dictionaries, and / or data models, 3) cross-tabulation of same across national common data models, and 4) specifying data element metadata to achieve native data interoperability. RESULTS: The registry analysis identified approximately 50 core clinical concepts. None were captured using the same data representation across all registries, and there was little implementation of data standards. To improve technical implementation, we specified 13 key metadata for each concept to be used to achieve data consistency. CONCLUSION: The registry community has not benefitted from and does not contribute to interoperability efforts. A common, authoritative process to specify and implement common data elements is greatly needed.

Duke Scholars

Author James Enlou Tcheng Medicine, Cardiology