Get Oura data as FHIR R4

The Oura API returns Oura's own JSON. A clinical system, an EHR integration or a research pipeline usually wants FHIR instead. Open Health Sync pulls a person's Oura data and serves it as FHIR R4 Observations: heart rate, SpO2, HRV and the other measurements with UCUM units and LOINC codes where one applies, each night as a sleep session, and each day's readiness, sleep and activity scores as a daily summary. The output is checked with the official HL7 validator in CI.

You can try the FHIR output in five minutes with synthetic Oura data, then connect a real ring.

1. Start an instanceCopy Anchor Link

Follow the quickstart to Start the stack: clone the repository, create .env with a fresh encryption key, and start it with Docker Compose.

2. Try it with synthetic Oura dataCopy Anchor Link

ohs seed --demo writes 90 days of invented data from six sources, Oura among them (sleep with stages, night heart rate, HRV, breathing rate, and a daily summary with readiness, sleep and activity scores), and prints a demo API key and user id:

docker compose --env-file .env -f deploy/docker-compose.yml exec ohs /ohs seed --demo
export OHS_URL=http://localhost:8080 OHS_KEY=ohs_sk_... OHS_USER=...   # from its output

Read the sleep sessions as FHIR (LOINC 93832-4):

curl -s "$OHS_URL/v1/users/$OHS_USER/fhir/Observation?code=http://loinc.org%7C93832-4&view=all&limit=3" \
  -H "Authorization: Bearer $OHS_KEY" | jq '.entry[].resource | {id, effectivePeriod, source: .meta.source}'

Each entry is an Observation whose meta.source ends in /fhir/provider/oura for the ring's nights, with the stages and the sleep score as components. view=all keeps every source's record; without it the server returns the merged view, where a better ranked source of the same night replaces the ring's (why).

The daily summaries, with the readiness score:

curl -s "$OHS_URL/v1/users/$OHS_USER/fhir/Observation?code=https://github.com/soma-intelligence/open-health-sync/fhir/CodeSystem/record-type%7Cdaily_summary&view=all&limit=3" \
  -H "Authorization: Bearer $OHS_KEY"

And the person as a Patient, with every Observation, Device and Provenance in pages:

curl -s "$OHS_URL/v1/users/$OHS_USER/fhir/Patient/$OHS_USER/\$everything?limit=50" \
  -H "Authorization: Bearer $OHS_KEY"

Follow the next link of each Bundle until there is none. The full mapping, the search parameters and example resources are in FHIR R4.

3. Connect a real Oura ringCopy Anchor Link

Oura data reaches the API through an Oura API application that you register yourself, in a few minutes:

1. In the Oura developer portal, create an application. Set its redirect URI to <OHS_PUBLIC_URL>/oauth/oura/callback and copy the client ID and secret.

2. Add them to .env and restart:

   Copy Text

   OHS_OURA_CLIENT_ID=...
   OHS_OURA_CLIENT_SECRET=...

3. Create your own user, then open Oura's consent page:

   Copy Text

   docker compose --env-file .env -f deploy/docker-compose.yml exec -T ohs \
     /ohs admin personal setup > ~/.ohs.env
   chmod 600 ~/.ohs.env
   python3 recipes/connect.py oura

The first data arrives a few seconds after you allow access, and the history follows in the background. Rings from Gen3 on need an active Oura Membership for the API to serve their data, and a new Oura application works for up to 10 people until Oura approves it. The Oura connector guide covers the scopes, webhooks and errors.

Then read your own data with the same calls, using OHS_URL, OHS_KEY and OHS_USER from ~/.ohs.env (set -a; . ~/.ohs.env; set +a). With only the ring connected, every Observation comes from Oura.

What maps to whatCopy Anchor Link

Records also come as Open mHealth data points (Open mHealth) and in the REST JSON of the API, from the same stored rows.

NextCopy Anchor Link

• Add other sources: the same FHIR routes serve every connector.
• Get your data as CSV or Parquet for analysis.
• Hand a person all their data in one zip, FHIR Bundle included.