Storing data using FHIR - a good idea?
summary
Recommendation is pragmatic: FHIR is designed as an exchange schema not a persistence format, so choose based on requirements. Options include store resources natively (JSON/BSON in Postgres, Mongo/Couch, BigQuery, RDF) and index with document store + Elastic, or map into optimized relational tables for speed and strict integrity. Best practice is a hybrid approach: keep raw FHIR resources, resolve references/codings internally, and extract high-value fields into relational tables for OLAP, constraints, and performance.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 1:24HL7 mission and FHIR design
- 9:31FHIR schema and storage challenges
- 15:51Native storage vs relational databases
- 22:07Hybrid approach for FHIR storage
- 43:00Q&A on hybrid storage
