ahdis / ch-elm Goto Github PK

Change status of IG from Draft to Release for version 1.2.0

ch.fhir.ig.ch-elm#1.0.0 /StructureDefinition-ch-elm-document.html

Change minimal cardinality for Practitioner from 1..1 to 0..1

Franz Auf der Maur, BAG

will be managed by FOPH to check speciment code if it sent but not requried

• Name for HIV/AIDS: https://fhir.ch/ig/ch-elm/guidance.html#personal-data-patient-name

Invalid GLN or social security should not give an error and should pass validation.

This error comes from CH Core, will be fixed with in hl7ch/ch-core#257

Bundle.entry[3].resource: Constraint failed: ch-elm-patient-name-representation-initial-snomedct: 'If Observation.code is a mapped to initials in http://fhir.ch/ig/ch-elm/ConceptMap/ch-elm-results-to-foph-patient-name-representation then patient.name.first and given can must have one character' (defined in http://fhir.ch/ig/ch-elm/StructureDefinition/ch-elm-observation-results-laboratory-strict)

example has a full name and should not fail

ch.fhir.ig.ch-elm#1.0.0 /

Some general feedback from Epic:

-The specification is clearly written. The definition of MustSupport is excellent - refreshingly simple to understand compared to many other implementation guides. We're also interested to see the use of imposeProfile, as this is a relatively new concept and few other implementation guides use it yet.
-The main concern we have is with complexity, and therefore difficulty of implementation.
For example:
-Varying name format by organism type.
-This is rare in a FHIR specification, and will require work from vendors and/or
laboratory organizations to develop.
-In some cases in other countries, we've seen the use of a centralized middleware
layer to do some translation instead of having every implementer maintain this logic,
allowing implementers to use standard FHIR profiles.
-Varying codeset by diagnosis.
-This is complex and potentially requires workflow updates to ensure laboratory
users document the required information.
-Could this information be sent as a standard DiagnosticReport (like the example of a
microbiology report at https://hl7.org/fhir/STU3/diagnosticreport-micro1.json.html,
or using the microbiology format with methods outlined in Epic's documentation at
https://fhir.epic.com/Specifications?api=998?)
-FHIR is seeing global success due to its implementation of REST with standard and reuseable healthcare resources and OAuth. This spec is trialing FHIR documents with bespoke FHIR resources using mTLS instead of OAuth. What's the reason for doing that, rather than using what has been successful elsewhere?

Katie Reynolds, Epic

there are two different validation approaches for ch-elm.

the regular validation checks that the message is correct but it allows that in case of a new infectious diseases a new message could be added and still pass through validation. in addition some business rules are not enforced.

the strict validation checks that all the message conforms to the published implementation guide, e.g a new infectious diseases code will throw an error in validation.

recommendation: check during implementation that your implementation is conforming to the strict version, test in production regular validation.

canonical url's for validation:

ch.fhir.ig.ch-elm#1.0.0-trialuse /profiles.html

Weekly vom 2.10: It would be nice to have the cardinalities (0..1), (0..*) etc defined in the overview for the link Fig. 6: Resource Overview, ahdis will add this to next version of the ig

Oliver Egger, ahdis

ch.fhir.ig.ch-elm#1.0.0-trialuse /StructureDefinition-ch-elm-patient.html

FHIR element patient.gender should be 1..1 (required)

Franz Auf der Maur, BAG

ch.chmed.emediplan#3.0.0-ci-build /

test

Michaela Ziegler, ahdis ag

ch.fhir.ig.ch-elm#1.0.0-trialuse /api.html

Aber ich fände es auch schön, im Sequenzdiagramm klar zu zeigen, dass die Einlieferung und das Pollen zwei unterschiedliche Requests sind. Das sieht man im Ersten nicht bildlich heraus, sondern muss den Text lesen.

Feedback BAG

Oliver Egger, ahdis ag

FullUrl wird in Bundles gesetzt, um die einzelnen Ressourcen miteinander zu verlinken, es gibt zwei Arten wie fullUrl's angebeben werden können:

• Via absolute URLs wie Du in obigem Beispiel herausgefunden hast.
• Via uuids or oids (siehe Beispiel hier: https://fhir.ch/ig/ch-emed/Bundle-2-7-MedicationCard-UUIDfullUrl.json.html), denke wir sollten auch im CH ELM noch so ein Beispiel hinzufügen .

Je nachdem wie man die fullUrl setzt müssen dann die Referenzen entsprechend angegeben werden, siehe exakte Beschreibung hier:
https://build.fhir.org/bundle.html#references

in Kurzform: "subject" : { "reference" : "Patient/1Pat-DM" } in der Composition wird dann aufgelöst im Bundle mit fullUrl http://test.fhir.ch/r4/Patient/1Pat-DM
bei einer uuid wäre es dann "subject" : { "reference" : "urn:uuid:9616e879-c4d0-411e-88c3-e001cb7a8fcd"} in der Composition wird dann aufgelöst im Bundle mit fullUrl "urn:uuid:9616e879-c4d0-411e-88c3-e001cb7a8fcd"

Für Specimen.type wird das BAG ein ValueSet zu Verfügung stellen

Modulo 11 verfahren

consider removing the version for the google search:

• Test site:fhir.ch/ig/ch-elm/1.1.0

does not return any result:

see ahdis/matchbox#197

If we test the ConceptMap CH ELM Expecting Specimen Specification To Results Completion Vs and intentionally send a specimen code that is not allowed per valueset ("geni_spec"), then the document is correctly rejected, but the error message is misleading ('Must have a resolvable URL conforming to the CH-ELM Bundle.') and says nothing about the actual problem why it is not conform.

the invariant itself is not sufficient to get the failing validation information and we should make it more explicit with introducing slice constraint:

1. https://fhir.ch/ig/ch-elm/StructureDefinition-PublishDocumentReference.html
   add slice for min 1 contained resource Bundle with profile http://fhir.ch/ig/ch-elm/StructureDefinition/ch-elm-document

2. https://fhir.ch/ig/ch-elm/StructureDefinition-PublishDocumentReferenceStrict.html
   constrain slice above for min 1 contained resource Bundle with profile http://fhir.ch/ig/ch-elm/StructureDefinition/ch-elm-document-strict

we could simplify then also the constraint that is only checks that url resolves and remove the conformsTo

as soon as the igs where we are depending on are published, change the version to a fix one (not current):

• ch.fhir.ig.ch-lab-report: current
  • hl7.fhir.eu.laboratory: current
• ch.fhir.ig.ch-core: current

Meeting 01.11.2023:
Maximumkardinalitäten einschränken bei definierten, verwendeten Elementen; bspw. AHV-Nummer Pat. (so dass nicht mehrere Angaben gemacht werden können und nicht differenziert werden kann, welcher Eintrag genommen werden soll)

ch.fhir.ig.ch-elm#1.0.0-trialuse /

Turtle als Austauschformat für Beispiele / Profile entfernen

Diskussion Telco 3.10.2013

Oliver Egger, ahdis ag

ch.fhir.ig.ch-elm#1.0.1 /

integration of a new concept map defining a valueset of valid interpretation-codes for each observation.code (leading code)

Franz Auf der Maur BAG

attachment.url points to the containted bundle, however it is not checked if the contained bundle is conforming to ch-elm-document requirements

ch.fhir.ig.ch-elm#0.1.0 /

Test2

oliver

update description to be conform to those of the FOPH:

CodeSystem-ch-elm-foph-patient-name-representation.json
This CH ELM resource defines the codes which identify possible patient name schemas to be used.

ValueSet-ch-elm-foph-patient-name-representation.json
This CH ELM resource defines the codes which identify possible patient name schemas to be used.

Meeting 14.11.2023:

Wir sehen 4 Datentypen vor:
String-100
String-50
String-25
String-10

Bei String-Elementen, wo ein Coding/System dahintersteht wie z.B. bei GLN, AHV-Nummer oder ISO-Country Code sehe ich eine String-Begrenzung als nicht erforderlich, weil die korrekte Angabe durch das dahinterliegende System/Regel geprüft wird.

ch.fhir.ig.ch-elm#1.0.0 /api.html

Release 4 of the HL7® FHIR® standard.

Link the correct Release: https://hl7.org/fhir/r4/index.html

Michaela Ziegler, ahdis ag

We do not except (and actually discourage) that source systems send the meta.profile in the instances.
However the example bundles show for each resource according to which profile they are valid.

We should remove them, it should. be confiugrable in sushi, see https://chat.fhir.org/#narrow/stream/179252-IG-creation/topic/Examples.20shouldn't.20have.20meta.2Eprofile.2C.20text

Chlamydia trachomatis: https://fhir.ch/ig/ch-elm/guidance.html#collection-material

Eine Änderung am bestehenden Leitfaden hat sich im Zusammenhang mit der Patient-Identifikator für Fürstentum Liechtenstein ergeben (patient.identifier:oasi-li). Wir hatten bis anhing die FL-Version der AHV Nummer als System hinterlegt. Dieses fällt weg und wird ersetzt durch die ID-N Nummer des Fürstentums Liechtenstein. Die OID der ID-N Nummer ist 2.16.438.10.20.10.10.1.2.1.

Die ID-N Nummer (ID für natürliche Personen) ist eine 13-stellige Zahl OHNE Prüfziffer. Sie wird im Gesundheitsbereich des FL breit eingesetzt.

Separates Tab mit CS/VS (error code and description as text)

Meeting 01.11.2023:

Address.line explizit spezifizieren (Profil):

• Strasse und Nummer
• Postfach

Beispiele entsprechend anpassen.

Add simplified graphics to guidance to support the description about the ConceptMap mechanism

Meeting 21.03. mit BAG

ConceptMap die Zusatzmaterial verlangt -> BAG-seitige Änderung wird noch erfolgen

TODO: Alle Terminologie-Instanzen updaten bevor der Publikation

Unification of spelling for FOPH/Foph

We always need a postcode (address.postalCode) and the city (address.city) from the client, regardless of whether the postcode and city come from the practitioner or the organisation. Can a rule be defined that ensures that we always receive a postcode and a city, regardless of whether it comes from the practitioner or the organisation?

Important: If city and postalCode are included in one of the two resources, they may be missing in the other resource.
If only the Practitioner is listed, address.city and address.postalCode must be filled in
If only the organisation is listed, address.city and address.postalCode must be filled in
If both resources are listed, address.city and address.postalCode must be completed in one of them (regardless of which one)
If city and postalCode are filled in both, this is ok.

ch.fhir.ig.ch-elm#1.0.1 /Bundle-1bDoc-NeisseriaGonorrhoeae.json.html

Example has reference to Specimen/1Spec-Specimen which is not included in the bundle. Should those point to Specimen/1bSpec-Specimen?

Oliver Egger, ahdis ag

ch.fhir.ig.ch-elm#1.0.0-trialuse /ValueSet-ch-elm-lab-specimen-types-chlamydia-trachomatis.html

Der Name/URL und allenfalls der Inhalt (?) muss angepasst werden, da das VS Werte von der Gruppe "geni-spec" enthält und nicht Erregerspezifisch ist.

Michaela Ziegler, ahdis ag

Meeting 21.03. mit BAG:

ch.fhir.ig.ch-elm#1.0.0 /StructureDefinition-ch-elm-organization-orderer.html

die auftraggebende Organisation muss künftig via BUR und/oder UID identifiziert werden. BUR und UID sollen Kardinalität 0..1 erhalten. GLN auf 0..1 belassen. Die ZSR brauchen wir nicht.

Franz BAG

ch.fhir.ig.ch-elm#1.0.1 /

test

test

ch.fhir.ig.ch-elm#1.0.0-ci-build /StructureDefinition-ch-elm-patient.html

country MS 0..1
country.extension:countrycode 1..1

Michaela Ziegler, ahdis ag

As decided in the EU LAB meeting from 13.10.23 the DiagnosticReport.identifier will be required if the Composition.identifier is required, and it should be the same. Because we are in the document context for CH ELM the Composition.identifier is required, so we have to require the DiagnosticReport.identifier aswell.

See also hl7ch/ch-lab-report#52

16.02.2024: checked with the current builds of EU LAB and CH LAB-Report (and CH Core) -> no conflicts with EU LAB

Meeting 21.03. mit BAG

Neue Regel für gewisse LOINC codes bezüglich Patientenname -> Infos folgen BAG-seitig

https://build.fhir.org/ig/ahdis/ch-elm/guidance.html#personal-data-patient-name

ch.fhir.ig.ch-elm#1.0.0 /Patient-Pat-AntoineDupont.json.html

The examples should have the adress.line also specified (in this example: Rue de la république 18"), even the IG requires in addtiiton the extension streetLine and streetNumber.

Oliver Egger, ahdis ag

https://fhir.ch/ig/ch-elm/guidance.html#collection-material
In all other cases, the Specimen.type has to be defined as fixed text: “Material declared by LOINC system axis”.

• Specimen.type.text remove fixed text as it is checked with the new added rules (ConceptMaps)
• Specimen.type.coding remove binded VS CH ELM Lab Specimen Types because through the VS in the ConceptMap more specific ones are given (see also #26)

ch.fhir.ig.ch-elm#1.0.0-ci-build /ValueSet-ch-elm-lab-specimen-types.html

Je nach ValueSet Mechanismus braucht es dieses VS nicht mehr.

Michaela Ziegler, ahdis ag

Wenn der Antrag auf die Reduktion der Minimumkardinalität von Specimen.type im Ballot von EU LAB angenommen wird, könnte das in CH ELM analog angepasst werden und der "fixedValue" für Specimen.type.text wäre nicht mehr notwendig.

https://jira.hl7.org/browse/FHIR-43027

Meeting 01.11.2023:

Allfällige Möglichkeit zur Dokumentation der Meldezeiten im IG:

• ConceptMap: führende Codes -> canonical URL der ActivityDefinition
  • ActivityDefinition mit Meldezeit und evtl. weiteren Angaben

we would like to search the ig by site.

this is not supported directly with the IG Publisher. we could adapt however the Swiss IG Template that a search box can be added to an external search engine with the url "fhir.ch" and searching that the ig name is within the site parameter.