Goal: Experimental proof of concept, using FHIR to extract large quantities of data from OpenMRS.
Sub goal: Best practices about (1) how to send this patient level data to OpenHIM, and (2) improvements to make the FHIR module more efficient, and (3) a test validation plan (verifying the patient observations got correctly into the FHIR store).
4 Chunks of Work to be Done:
- OpenMRS datasource (OMRS instance)
- Extraction engine (Intellisoft - atomfeed; OMRS - take from EIP team’s debezium approach)
- Sending data through OpenHIM mediator
- HAPI FHIR server
- Validation process to check accuracy
Mentors: Kaweesi, Moses
Fellows: Ayesh, Cliff
A Note about Overlap with the FHIR Squad
FHIR Squad now needs to know what their next squad priorities are (since FHIR2 release is done). This squad has demands from (1) the ITECH implementation & SHR in Haiti, (2) the Micro Frontend architecture, (3) the Notice D work.
Have to (1) get a set of patients out of OMRS in FHIR format and (2) somehow calculate TX_PVLS. (1) is mostly handled by FHIR2, concern will be the size of the data request and needing to speed it up. Taking a stream of FHIR data out of OMRS.
How the PLIR team can help: work on the FHIR module to make the extraction more efficient at handling large datasets (current: handles 10-50 things/time; want 100,000’s of obs handled)