Transforming Estonian health data to the Observational Medical Outcomes Partnership (OMOP) Common Data Model: lessons...

Area of Science:

• Health Informatics
• Data Standardization
• Real-World Evidence Generation

Background:

• Electronic health records (EHR), claims, and prescription data are vital for research but exist in disparate formats.
• Standardizing these diverse data sources is crucial for large-scale health research and generating reliable real-world evidence.
• The Observational Medical Outcome Partnership (OMOP) Common Data Model (CDM) provides a standardized framework for observational health data.

Purpose of the Study:

• To describe a reusable transformation process for converting EHR, claims, and prescription data into the OMOP CDM.
• To document challenges encountered during data transformation and present implemented solutions.
• To facilitate efficient data harmonization for future research.

Main Methods:

• Utilized Estonian national health databases containing comprehensive resident data.
• Developed and demonstrated a transformation process using a 10% random sample (150,824 patients) from 2012-2019.
• Converted data from three databases into OMOP CDM version 5.3 and validated using open-source tools.

Main Results:

• Transformed over 100 million entries into standard OMOP concepts with an average mapping rate of 95%.
• Achieved mapping rates exceeding 90% for conditions, observations, drugs, and measurements.
• Primarily used SNOMED Clinical Terms as the target vocabulary for standardization.

Conclusions:

• Successfully created a reusable transformation process for converting national health data into OMOP CDM.
• The developed process enables efficient data harmonization from multiple sources.
• This work supports researchers in generating higher quality real-world evidence through standardized data.