Data logistics system enabling real-time pathogen surveillance. Built for the Seattle Flu Study.

FHIR

Adopt FHIR R4 as a standard for enrollment data.

Motivation

  1. Simpler to focus on a single standard where specs are clear.

  2. Decouples ID3C from the peculiarities of the Seattle Flu Study (towards the vision of a more general data system).

  3. Promise of greater interoperability in the future from other data providers (HealthKit/ResearchKit/ResearchStack, EPIC and other EMRs, …).

Challenges

  1. FHIR is complex.
  2. It may be hard to avoid special-casing our needs.

Work

  1. Writing a conversion from Audere's documents to FHIR (or get Audere to do this?)
  2. Writing a conversion from REDCap surveys to FHIR (or using existing tools?)
  3. Writing a conversion from clinical records to FHIR (or pulling these directly from the EMR?)
  4. Write a single ETL routine to handle all these sources of FHIR documents

FHIR resources we'd primarily use would be:

  • Patient (sex)
  • Encounter (age, site, place)
  • Questionnaire

Prior art

  • FHIRbase
  • REDCap on FHIR — export FHIR documents from REDCap, paper and slides available, not sure about code.
  • Dynamic data pulls (DDP) on FHIR — a way to have EMR records flow into REDCap (and then onto ID3C)
  • FHIRCap?