Better managing sets of resources
summary
FHIR lacks a clear, reusable way to describe a coherent set of resources created by a clinical session, so current options are ad hoc: linking everything to an encounter, a single Provenance target, a List or a Composition (used mainly as a document header), or stitching things together with GraphQL or graph definition. Proposal: add a new CompositionDefinition resource that declares sections, roles, query/extraction rules (FHIRPath/CQL/URL), confidentiality and workflow expectations; graph definition is Maturity 1 and GraphQL use exists but remains sparse.
AI-written summary · about
chapters (17)
- 0:00Introduction and session setup
- 1:25Background on resource sets challenge
- 3:51Clinical scenario and resource generation
- 5:55Approaches to link resources together
- 10:09Discussion on linking methods
- 13:18Purpose of linking resources
- 17:33Context and human interpretation
- 21:50Composition and clinical workflows
- 24:20Micro encounters and targeted use
- 25:51Describing sets in implementation guides
- 28:31GraphQL and graph definition overview
- 33:30Proposal for composition definition
- 37:21Use cases and feedback
- 39:06Clarifying composition definition
- 41:51Building instructions and extraction
- 45:46Relationship to structure definitions
- 48:24Closing remarks and next steps
