W3C

– DRAFT –
FHIR RDF

29 January 2026

Attendees

Present
David Booth, Erich Bremer, EricP, Gaurav Vaidya, Jim Balhoff, Ken Lord, Tim Prudhome
Regrets
-
Chair
David Booth
Scribe
dbooth

Meeting minutes

Security alert: Does it affect our code?

advisories/GHSA-2xpw-w6gg-jr37

tim: Here is the Python used in that repo: https://github.com/search?q=repo%3Aw3c%2Fhcls-fhir-rdf++language%3APython&type=code

tim: It only affects python

tim: I have a suggested related to @Gaurav Vaidya (RENCI) ’s work w3c/hcls-fhir-rdf#123 (comment)

Adding prefixes for the CodeSystems that are part of the “core” spec

ericp: It is only for a DOS attack that would not affect us. Easiest to do nothing.

dbooth: oky

IRI stems

tim: I have a suggested related to @Gaurav Vaidya (RENCI) ’s work w3c/hcls-fhir-rdf#123 (comment)

Adding prefixes for the CodeSystems that are part of the “core” spec

tim: Example usage of a code from a “core” CodeSystem:

The fhir:status of this example: https://build.fhir.org/diagnosticreport-example.ttl.html

I think it should at least have a fhir:l value, even if not using a prefix there

tim: Some CodeSystems are part of core FHIR, with a couple of different bases
… Many are used in required bindings in the core spec.
… Should we do prefixes for these core CodeSystems?
… One reason we're doing it is for type assertions, and we're not currently doing that for examples that use the core CodeSystems

gaurav: We can put IRI stems on any CodeSystem we want, if we respect balloting rules.
… Makes sense to do that even if we're only formalizing it. Grahame likes to just use a hash separator. We could propose to formalize that, but not sure if it will resolve to an RDF document.
… FHIR core systems are maintained by HL7. I think Grahame would be ok with the hash approach.
… For terminology.hl7.org we're recommending that people dong' tpu IRI stems on CodeSystems at all, because it only has NamingSystems now.
… But it has CodeSystems for HL7 CodeSystems.
… We now have the mechanism to do this. Just need to identify which ones we want IRI stems for, then talk w HL7 to come with IRI stems that we want.

tim: At least w SNOMED and LOINC, each concept has an obvious IRI for each concept/term, but it's not clearly the case for these internal core systems.

tim: Each code within the CodeSystem would not have its own URL.

gaurav: Interestingly https://hl7.org/fhir/codesystem-diagnostic-report-status.html#diagnostic-report-status-registered does resolve. Does resolve to the individual value >.>

tim: One use of this is to add a type assertion. We're only doing it for LOINC, SNOMED and MeSH. We should do it for any codes.

Official docs here https://build.fhir.org/rdf.html#representing-concepts

gaurav: I think the internal ones all start with hl7.org and end with code-system.

gaurav: Be warned that there are a lot of internal code systems: https://hl7.org/fhir/terminologies-systems.html#

tim: Diagnostic report status points to http://hl7.org/fhir/ValueSet/diagnostic-report-status

gaurav: We should not manually add IRI stems to individual CodeSystems, because there are 400+ of them. So we should use a rule.

gaurav: E.g. IRI stem should be “http://hl7.org/fhir/version-algorithm#”

ken: The largest hl7 valueset is about 150 terms, other than examples.

gaurav: CodeSystem pages for hl7.org/fhir have anchors for the codes.
… But the anchors include the CodeSystem names.

dbooth: redundant.

tim: https://hl7.org/fhir/codesystem-diagnostic-report-status#code this isn’t so bad

gaurav: We should propose to put this into the rdf.html page

ACTION: Tim to post on zulip asking Grahame about formalizing adding a hash to hl7 internal CodeSystems to make IRI stems, and putting this rule on the rdf.html page

gaurav: http://hl7.org/fhir/reference-handling-policy

gaurav: https://jira.hl7.org/browse/UP-755
… Mark changed it to a pro forma change. It still cannot be moved forward. Jessica Bota is looking into what's wrong.

gaurav: https://github.com/gaurav/hcls-fhir-rdf/blob/8e9385145b04ad68734105442ce2353371cdf6d2/data/utg/iri-stems-in-hl7.md
… Also working on a markdown to magically making the IRI stem table of IRI stems

R6 testing

tim: The R5+6 files are ready for testing now: w3c/hcls-fhir-rdf#213 (comment)

ericp: No progress yet.

FHIR ontology

ken: Eric Jahn and I are working on the use case, regarding housing insecurity
… Hope to share something in a couple weeks.

ADJOURNED

Summary of action items

  1. Tim to post on zulip asking Grahame about formalizing adding a hash to hl7 internal CodeSystems to make IRI stems, and putting this rule on the rdf.html page
Minutes manually created (not a transcript), formatted by scribe.perl version 248 (Mon Oct 27 20:04:16 2025 UTC).

Diagnostics

Succeeded: s/ACTION: Dbooth to move those python file to an archive.//

No scribenick or scribe found. Guessed: dbooth

Maybe present: dbooth, gaurav, ken, tim

All speakers: dbooth, ericp, gaurav, ken, tim

Active on IRC: dbooth