Skip to main content
Most EHR-provided “sandbox” patients are little more than UI demos — they rarely resemble the volume, structure, or messiness of production data. Use the resources below to validate your ingestion, normalization, and downstream analytics pipelines before you onboard live connections through Fasten.

​
How to use synthetic patient data with Fasten

  1. Choose a dataset that matches the workflow you want to mimic (FHIR Bulk exports, claims-heavy data, pediatrics, etc.).
  2. Convert it to NDJSON if necessary - Fasten’s /ehi_export endpoint provides FHIR resources in newline-delimited JSON (ndjson) format, with one resource per line. The datasets shared below are in a variety of formats, and may require some manipulation.
  3. Load it into your pipeline tooling - e.g., drop ndjson files into the same object storage bucket that you store Fasten exports in, or POST bundles against your local test endpoint.
  4. Run the exact conversions/validations you expect to run for real patients: schema validation, de-identification, transformations, downstream notifications.
  5. Swap the input source with Fasten’s /ehi_export payloads when you are ready for end-to-end testing; no code changes should be required if your fixtures match the contract.
Keep the original archive of each dataset in version control or object storage so test runs are reproducible and easy to diff across releases.

​
CCDA

​
IPS