
A lab result means little on its own
A value, a unit and a reference range look like an answer. They are a third of one. What a result means depends on the person it belongs to and on the results around it.
Notes on reading health records, clinical AI that shows its sources, and the standards that let a record travel — from the team building Vitrubo.

A value, a unit and a reference range look like an answer. They are a third of one. What a result means depends on the person it belongs to and on the results around it.

A language model is good at reading and explaining. It is not a calculator. In Vitrubo, every value, unit, range and flag is computed by deterministic code before any model sees it.

An interpretation nobody can check is an opinion. Tying each statement to the clinical guideline it rests on turns it into something a clinician can verify in a minute.

A report that only discusses the results it was given will miss the question it should have asked. Naming the missing tests is part of reading the ones that are there.

A result is a statement about the day the sample was taken. Past a certain age it stops describing the person, and a report that treats it as current will reassure for the wrong reason.

Two acronyms decide whether a health record can travel between systems or stays stuck in the one that made it. A plain-language guide to what they are and why an interpretation layer depends on them.
A walkthrough on your own documents and your own step in the process.