HL7 v3 / CDA
The model-driven HL7 family: the RIM, the domain and message models derived from it, the CDA document architecture, and the vocabulary that binds them together.
Open to browse
The v3 family at a glance
One model, many messages. Unlike v2’s field-by-field segments, everything in v3 derives from a single object model — the RIM. Each clinical domain constrains it into a domain model (D-MIM), refines that into message-specific models (R-MIMs), and serializes the result as XML. The same derivation chain, applied with a document lens, yields CDA.
CDA is the part that shipped everywhere. The Clinical Document Architecture — discharge summaries, care records, the C-CDA templates U.S. regulation leans on — is a v3 R-MIM wearing a document coat. If you work with CDA or C-CDA, the v3 models and data types here are its vocabulary and grammar.
Vocabulary binds by realm. A model names an abstract concept domain (say, a “document confidentiality” slot); a context binding then fixes which code system or value set fills that slot in a given realm — so the same model can carry universal codes in one realm and national ones in another. The three tabs of the vocabulary browser above are exactly those pieces.
How to read a v3 artifact ID
Every v3 artifact — storyboard, model, message type — is named by the same compact grammar, so a single ID tells you where an artifact lives, what kind of thing it is, and which realm it serves:
Artifact type codes
STStoryboardSNStoryboard NarrativeARApplication RoleTETrigger EventINInteractionDMD-MIMdomain modelRMR-MIMrefined modelHDHMDmessage definitionMTMessage Type
The serial is issued per domain for uniqueness — its digits carry no meaning. UV marks the universal realm; national editions substitute a realm code of their own. The final digits version the artifact.