Online lab testing that comes back read, not as a PDF.

Start with the flags: a result outside its reference range is marked, but a single flagged value rarely means much on its own. What matters is the group it belongs to, how it compares with an earlier draw, and which medication or condition might explain it. The live reports on this page show what that reading looks like; if your provider uses Vitrubo, your report is built the same way. A report is context to bring to a doctor, not medical advice.

A plain-language report: flagged results grouped into a few connected stories in four zones, needs attention, worth watching, back to normal and always normal, each with what it means and what to discuss with a doctor. Trends for every marker across previous draws, and the sources behind each statement one tap away. Delivered as a branded PDF, in the portal or inside the provider's app.

The same analysis in the clinician view: full reasoning, connected conditions and medications, tests to consider, and a drafted comment beside every flagged result with the guidance it rests on. The clinician reviews, edits and signs. Vitrubo drafts; it does not decide.

Yes. Every result is mapped to LOINC and its units made comparable, so a lab test online from one partner laboratory and an older hospital panel read on the same line. That is what makes online lab testing useful over time: each later draw is read against the ones before it.

Whatever they already return. PDF, HL7, FHIR R4 or JSON through the API are all read into one FHIR R4 record; a photo of a paper test result is read as well. Blood, urine and allergy panels arrive as values; hospital records, prescriptions and notes are read as context.

A HIPAA business associate agreement for US deployments, a GDPR Article 28 processing agreement and Standard Contractual Clauses. Encryption in transit and at rest, and identity never enters the prompt: the models see values and an internal ID, never a name.

Every statement links to named medical guidance, so the basis for a claim can be checked rather than trusted. Numbers, units, ranges and flags are computed in code; several AI models cross-check the same case before a draft is shown. In a clinical program roughly 100,000 anonymised hospital records were read on premises, and on Stanford MedAgentBench Vitrubo completed 600 of 600 tasks across five runs. The clinician still signs.

No. It does not diagnose, select or recommend treatment, or replace a clinician or laboratory scientist. It drafts a reading, organises the history and names the tests to consider. The decision stays with the professional; for the customer, the report is educational context to bring to an appointment.

Yes. White-label PDFs and a role-based portal carry your logo, headers and disclaimers; or take JSON through the API and render the report in your own product. Comment style, reference ranges and answer style per audience are configured for each deployment.

An AI chat grounded only in the person's own record. A customer who has just received a lab test online can ask what a flagged value means, whether it has moved since the last test and what to ask their doctor; the answers come from their results and the guidance behind them, not from the open web. It is switched on per deployment.

With a call, then sandbox and API access. Your team runs its own anonymised lab tests through the pipeline, your clinician validates the drafts, and nothing is committed at that stage. Production follows when you are ready, inside your perimeter or a private cloud.