Transforming German Health Data for Research
Author Information
Author(s): Melissa Finster, Maxim Moinat, Elham Taghizadeh
Primary Institution: Fraunhofer Institute for Digital Medicine MEVIS
Hypothesis
How can the German Health Data Lab's claims data be standardized into a Common Data Model for better research access?
Conclusion
The ETL process successfully standardizes health data, improving usability for research and facilitating cross-border studies in Europe.
Supporting Evidence
- Field coverage of 92.7% was achieved for Format 1.
- Data Quality Dashboard showed 100.0% conformance for Format 1.
- Mapping coverage for the Condition domain was low at 18.3% due to invalid codes.
- Format 3 achieved a field coverage of 86.2%.
Takeaway
This study shows how to change health data into a common format so researchers can use it easily, helping them work together better.
Methodology
An Extract, Transform, and Load (ETL) pipeline was developed to convert health data from two formats into the OMOP Common Data Model.
Potential Biases
The use of fictional data may introduce biases that do not exist in real datasets.
Limitations
The study used mock data, which may not accurately reflect real-world scenarios, and some information was lost during the transformation process.
Participant Demographics
The study involved health data from approximately 3.4 million insured individuals in Germany.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website