Data model harmonization for the All Of Us Research Program: Transforming i2b2 data into the OMOP common data model.

Area of Science:

• Biomedical Informatics
• Health Data Standards
• Clinical Data Warehousing

Background:

• The All of Us Research Program (AOU) requires robust data interoperability for its nationwide cohort of one million patients' EHR and genomic data.
• Standardizing EHR data to the Observational Medical Outcomes Partnership (OMOP) model is crucial for AOU, but challenges arise from differing data models.
• The i2b2 platform, utilized globally, offers an ontology-driven approach adaptable for data reconfiguration, previously demonstrated for transforming data into the PCORnet model.

Purpose of the Study:

• To leverage existing i2b2 infrastructure and ontology investments to support the AOU OMOP data pipeline.
• To develop and validate a high-performance i2b2-to-OMOP data transformation method for the AOU program.
• To provide an open-source solution for research projects needing to transform i2b2 data into the OMOP common data model.

Main Methods:

• Utilized the ARCH-OMOP ontology and OMOP concept mapping dictionary to drive the i2b2-to-OMOP transformation.
• Demonstrated and validated the transformation approach within the AOU New England HPO (NEHPO).
• Employed AOU tools for structural compliance verification and the OMOP Achilles tool for data self-integrity checks.

Main Results:

• Successfully transformed a synthetic i2b2 patient dataset into OMOP, confirming structural compliance with AOU standards.
• Transformed a subset of data from the Partners Healthcare data warehouse into OMOP format.
• Validated the transformed data's self-integrity using a checklist and OMOP's Achilles tool.
• The developed i2b2-to-OMOP transformation is actively used for sending NEHPO production data to AOU.

Conclusions:

• The developed i2b2-to-OMOP transformation method effectively supports the All of Us Research Program's data pipeline.
• This open-source tool, driven by a standardized ontology, facilitates data interoperability and can be adopted by other research initiatives.
• Leveraging existing i2b2 investments and ontologies provides an efficient pathway for data standardization to OMOP.