Approved results and observations

Approval is the publication boundaryCopy Anchor Link

The API exposes only results whose BioParse state is ACCEPTED. Drafts, candidates and unapproved results do not appear. Each observation uses the final stored, edited values from the approved result.

Use results:read to list results, read a result and page through its observations. Result metadata includes appointmentId, patientId and appointmentDate so your integration can connect the result to its workflow. v1 has no appointments resource.

Retrieve a resultCopy Anchor Link

1. GET /v1/results lists approved results for the key’s clinic. Use the optional patientId filter to select one patient.
2. GET /v1/results/{id} returns the selected result’s metadata.
3. GET /v1/results/{id}/observations returns a paginated collection of final observations.

Patient, result and observation identifiers are UUIDs. A missing or off-clinic resource returns 404, including a patient supplied as a result filter. An empty observation page uses { "data": [], "nextCursor": null }.

Preserve value precision and meaningCopy Anchor Link

Numeric clinical values are decimal strings. Preserve the original string; converting it to a binary floating-point number can change precision. Values, units, reference ranges and range statuses can be null where the observation type permits it. Do not treat a missing value as zero or an empty unit as a known unit.

referenceRange and rangeStatus describe the persisted observation. They are not a recalculation using the current catalog and do not establish an immutable clinical snapshot. Patient and biomarker names can reflect current records.

Anatomy is structured as bodyRegion, laterality, anatomicalSite and organ, each nullable. Dimensions and scores retain their typed fields. Consult the generated reference for the exact enum and required-nullable contracts before deserializing a response.

Keep the contract boundary intactCopy Anchor Link

The API does not expose source documents, storage URLs, source text, candidate extraction data, doctor notes, prompts or internal report layouts. It has no file downloads or writes. Retrieve the fields you need, apply your own access controls and avoid duplicating clinical values into logs or analytics.

Explore result operations and schemas or learn pagination.