SQL on FHIR: Workshop on Transforming FHIR Data into Queryable Tables

Details

The workshop focused on methods for handling complex FHIR data structures for analytical purposes. Nikolai Ryzhikov (CTO of Health Samurai) explained the process of using the SQL on FHIR specification to flatten nested FHIR resources and transform them into queryable tables. The session covered building a complete analytics environment using tools like Docker, PostgreSQL, Grafana, and Jupyter notebooks. Specifically, participants learned how to create view definitions—flattening FHIR data into tables—by utilizing key functions such as `column`, `where`, `forEach`, `select`, and `union`. This capability allows for the construction of interactive dashboards in Grafana, displaying patient demographics, condition patterns by age group, and medication patterns. SQL on FHIR is an open initiative welcoming contributions from implementers and users. Although Health Samurai's engine is currently in preview, this technology significantly contributes to standardizing healthcare data analysis by offering integration into data science workflows and portability across various database environments.