Skip to content

Bulk FHIR Overview

Overview

Bulk FHIR ingests happen asynchronously, where we trigger an export request and then poll a status URL until the job completes. Once its completed, FHIR resources are fetched and uploaded to GCS, to then be processed by FHIR Store.

Some key aspects we need to be aware of within the realm of bulk FHIR exports are:

  • Authentication
    • Our ingests largely rely on custom classes based off of dlt
      • most common are client id, client secret and client_id private_key ones
    • we can customize new ones if needed
  • Status fetching
    • Some API's do not retrieve completion marks (X% Processed), so we just wait and give some retries.
    • Future improvement could be to kill retrieve task instantly if isn't processed.
  • Checkpointing and multithreading

  • parameters > IMPORTANT: CHECK WHICH PARAMETERS ARE VALID FOR EACH EHR PROVIDER BEFORE IMPLEMENTING

    • _since - usually using the format YYYY-MM-DDTHH:mm:ss Note : Not all EHR providers support it.
EHR PROVIDER Support email(s) DOC PAGE COMMENTS
Athena Athena Doc Supports appointments and _since, robust API. Ingests take really fast in comparison to other EHR providers
Eclinicalworks fhirfighters@eclinicalworks.com Eclinicalworks Doc Exports run overnight. Our experience was that it takes extremely slow to process backfills, and cancelling exports is unreliable on their end, so the job keeps running.
ModMed support@modmed.com ModMed Doc No special remarks on this one, backfills take long to process.
PracticeFusion help@practicefusion.com PracticeFusion Doc Backfill takes a while, supports _since even though documentation doesn't display it
Relimed chopper@relimedsolutions.com, rkasula@relimedsolutions.com Relimed Doc Custom endpoints, may need extra config
EyeMD Jayneen.Lemus@eyemdemr.com EyeMD Doc Eye care specific resources
Veradigm VeradigmConnect@veradigm.com Veradigm Doc No active integration yet on this one, their API seems to be very problematic
Epic Epic Doc No active integration yet on this one.