Mapping FHIR to OMOP Using Open Source Tools
summary
Demonstrates an open source pipeline that takes FHIR STU3 transaction bundles from a HAPI test server, runs a configurable FHIR-to-OMOP converter (deployed locally or on Google Cloud), and writes OMOP CDM v6 tables into BigQuery using OHDSI vocabularies and the concept_map feedback loop. Shows actual pain points: medication administration vs dispense/request, observation vs measurement domain choice, missing person demographics. Converter produces usable CDM output but requires iterative reconfiguration of mappings and vocabulary lookups.
AI-written summary · about
chapters (7)
- 0:00Session introduction
- 1:09Project overview
- 6:33History and goals of OMOP
- 12:05OMOP data model structure
- 25:37Incremental ETL pipeline
- 31:13FHIR and OMOP mapping challenges
- 37:01FHIR conversion and Q&A
related talks
21mOMOP to FHIR Bulk Data API used for MENDS
23mThe CDISC lab semantics in FHIR Implementation Guide in action
36mTransforming HL7 version 2 to FHIR
42mSQL on FHIR
35mMix, Match, Merge
19mTurning data into better healthcare: data for hemophilia
41mBuilding a FHIR app with Java
16mThe story of FHIR data exchange and storage combined
