ehealthsuisse / ch-atc

The profile CH:ATC defines the audit trail consumption requirements for the EPR in Switzerland which a community has to provide for a patients audit trail.
MIT License
2 stars 1 forks source link

Some comments #7

Closed qligier closed 9 months ago

qligier commented 1 year ago

For all AuditEvent profiles: Can we get better description of entities? Patient, Healthcare professional, Group, Document are not really descriptive :) E.g. The patient who is being notified, The HCP that has been added to a group, The group in which a HCP has been added.

In http://fhir.ch/ig/ch-atc/ValueSet-PolicyAuditEventType.html: ATC_POL_INCL_BLACKLIST has the description "Assign a Healthcare Professional to Blacklist", I would use "Include" for consistency with the chosen code and ATC_POL_EXL_BLACKLIST.

In atc-doc-create-rep-pat:

<!--  Document  -->
<what>
  <identifier>
    <type>
      <coding>
        <system value="urn:uuid:2e82c1f6-a085-4c72-9da3-8640a32e42ab"/>
        <code value="IHE XDS Metadata"/>
        <display value="XDSDocumentEntry.uniqueId"/>
      </coding>
    </type>
    <system value="urn:ietf:rfc:3986"/>
    <value value="urn:oid:1.2.3.4.5"/>
  </identifier>
</what>

urn:uuid:2e82c1f6-a085-4c72-9da3-8640a32e42ab is not a system, but could be a code. IHE XDS Metadata is not a code. urn:uuid:7edca82f-054d-47f2-a032-9b2a5b5186c1 as a system maybe? https://gazelle.ihe.net/XDStarClient/rimihe-templates/rimihe-XDSDocumentEntry.html

source.observer.identfiier: Why not use source.site, as mapped by FHIR/IHE? Annex 5, 2.2 §4.2.1 has errors in this example: the OID is displayed in source.identifier, which does not exist.

In atc-doc-read-ass-hpc:

<!--  Document  -->
<role>
  <system value="http://terminology.hl7.org/CodeSystem/object-role"/>
  <code value="3"/>
  <display value="Report"/>
</role>

Report is badly suited here, because the document is not any kind of persistent document created as a result of the event, it is only being read. Domain Resource?

In atc-pol-create-acc-right: HCPs are System Objects, not Persons any more? :) Same for atc-pol-create-rep

In HpdAuditEvent: I would argue that the patient here has the role of Subscriber, not Patient, because they have no direct relation to the event, but that would be nitpicking.

oliveregger commented 12 months ago
  1. For all AuditEvent profiles Can we get better description of entities? tbd: can we describe general for the group more specific?

  2. concerning the textual changes (ATC_POL_INCL_BLACKLIST) in the ValueSet they were provided by the FOPH (including translations), they should provide feedback if we can do chagnes

  3. XDSDocumentEntry.uniqueId

<ns4:ExternalIdentifier id="urn:uuid:df9df3fc-4da7-4603-a8fe-c361676f579d" identificationScheme="urn:uuid:2e82c1f6-a085-4c72-9da3-8640a32e42ab" objectType="urn:oasis:names:tc:ebxml-regrep:ObjectType:RegistryObject:ExternalIdentifier" registryObject="urn:uuid:a3767774-f91b-4d2c-9a17-ce8a2c96e6e2" value="1.3.6.1.4.1.12559.11.1.2.2.1.1.3.479">
                <ns4:Name>
                    <ns4:LocalizedString value="XDSDocumentEntry.uniqueId"/>
                </ns4:Name>
            </ns4:ExternalIdentifier>

maybe we could just add?

      <coding>
        <system value="urn:oasis:names:tc:ebxml-regrep:ObjectType:RegistryObject:ExternalIdentifier"/>
        <code value="XDSDocumentEntry.uniqueId"/>
      </coding>
  1. the OID is displayed in source.identifier is incorrect -> fix with a proper test id

other structural changes proposition: the profiled AuditEvents are in production, I don't think they should be changed, even if there would be structural problems within them. tbd.

oliveregger commented 11 months ago

Telco DK/MS: will change the ExternalIdentifier but not make changes to structural message as it is already used in that way.