plugin-rosetta¶
plugin-rosetta translates FHIR R4 clinical data into OMOP CDM 5.4 format. It uses LinkML schemas to formally describe both models and to validate output — giving data engineers, clinical informaticians, and data stewards a transparent, auditable FHIR-to-OMOP pipeline.
Mappings are derived directly from the HL7 FHIR-OMOP Implementation Guide FML structure maps, ensuring alignment with the published standard rather than ad-hoc field-by-field decisions.
Supported resources¶
| FHIR R4 resource | OMOP CDM 5.4 table |
|---|---|
| Patient | Person |
| Encounter | VisitOccurrence |
| Condition | ConditionOccurrence |
| Observation (non-measurement) | Observation |
| Procedure | ProcedureOccurrence |
| MedicationStatement | DrugExposure |
| Immunization | DrugExposure |
| AllergyIntolerance | Observation |
Where to go next¶
-
New to plugin-rosetta?
Install the library and translate your first FHIR record in minutes.
-
Auditing the mappings?
See exactly which FHIR fields map to which OMOP columns, with caveats.
-
Integrating the library?
Full API reference for translators, I/O helpers, and validators.
-
Exploring the schemas?
Browse the generated LinkML schema documentation for all OMOP and FHIR classes.
Known limitations¶
Concept IDs are not resolved
All *_concept_id fields in translator output are set to 0. Mapping source
codes (SNOMED, LOINC, RxNorm, etc.) to OMOP standard concepts requires a
running OMOP vocabulary database or API — this is not yet implemented.
The raw source code is always preserved in the corresponding *_source_value
field.
Measurement table not yet supported
FHIR Observations with category = laboratory or vital-signs are
intentionally skipped. They belong in the OMOP Measurement table, which
is planned for a future release.