Feedback about MedicationDispense

[qligier]

About the MedicationDispense structure definition:

• The ch-emed-ext-prescription extension has the same use as authorizingPrescription, we should use the latter (and fix its type).
• What's the use of the ch-emed-ext-pharmaceuticaladvice here? Is it for the reference to PADV concerns? If so, we should probably use supportingInformation or detectedIssue.
• substitution.wasSubstituted and substitution.type are mandatory and contain the same information, which can lead to contradictions like true + none or false + equivalent. Is there something we can do here? Can we go against wasSubstituted's 1..1 cardinality from the FHIR resource? Otherwise, we could remove none from the value set, make type optional and constraint its use on true only.

[ziegm]

1. In the currently used extension (ch-emed-ext-prescription) the identifier of the document and the identifier of the entry are mapped. The reference authorizingPrescription is used to reference the MedicationRequest from the PRE document, the info about the document would be missing.

2. Yes, if there is a PADV document in an earlier use case step. Usage of the extension (ch-emed-pharmaceuticaladvice) due to the same reason as point 1.

3. 1..1 for wasSubsituted is given through the base spec. However, with the last proposal, it would no longer be possible to map when a drug may not be substituted (none = No substitution occurred or is permitted). Logical content validation cannot be done via the profile, in my opinion.

[qligier]

1. Of course, I missed that detail.
2. Indeed.
3. You're removing 'None' from the binding of substitution.type? That's not a bad idea; it also requires changing the cardinality back to 0..1.

[qligier]

Also, we don't have a profile equivalent to the "Pharmaceutical Advice Concern Entry Content Module" (not the "Pharmaceutical Advice Item Entry Content Module"). The pharmaceuticalAdvice extension can't point to anything, it should be dropped.

Edit: I misremembered the hl7:Supply specifications.

[ziegm]

17.11.2022 Telco PJ/OE/MZ:

1. Geklärt, keine Anpassung notwendig.
2. @ziegm letzten Kommentar nochmals mit Quentin klären, was genau gemeint ist.
   -> 18.11.2022: ebenfalls geklärt, keine Anpassung notwendig
3. Need to clarify VS binding for a future version

[ziegm]

Open point (discussion Nov 22):

3. Need to clarify VS binding for a future version

Earlier discussion to point 3

QL: substitution.wasSubstituted and substitution.type are mandatory and contain the same information, which can lead to contradictions like true + none or false + equivalent. Is there something we can do here? Can we go against wasSubstituted's 1..1 cardinality from the FHIR resource? Otherwise, we could remove none from the value set, make type optional and constraint its use on true only.

MZ: 1..1 for wasSubsituted is given through the base spec. However, with the last proposal, it would no longer be possible to map when a drug may not be substituted (none = No substitution occurred or is permitted). Logical content validation cannot be done via the profile, in my opinion.

QL: You're removing 'None' from the binding of substitution.type? That's not a bad idea; it also requires changing the cardinality back to 0..1.

Overview

MedicationDispense.subsitution 0..1

element	FHIR card.	CH EMED card.	description	value
wasSubstituted	1..1	1..1	Whether a substitution was or was not performed on the dispense	true/false
type	0..1	1..1	Code signifying whether a different drug was dispensed from what was prescribed	ActSubstanceAdminSubstitutionCode (preferred)

• E equivalent (and subcodes) - Substitution occurred or is permitted with another bioequivalent and therapeutically equivalent product.
• F formulary - This substitution was performed or is permitted based on formulary guidelines.
• N none - No substitution occurred or is permitted.

Proposal

Should we add a tab 'Substitution' under the section 'Guidance' and describe (incl. examples) the different use cases?

• DIS: MedicationDispense.substitution.wasSubstituted/type
  Whether a substitution was performed on the dispense (this issue)
  • Substitution was performed -> true & equivalent
  • Substitution was not performed -> false & none
• PRE: MedicationRequest.substitution.allowedCodeableConcept
  Whether substitution is allowed or not
  • Substitution is allowed -> equivalent
  • Substitution is not allowed -> none
• CARD: MedicationStatement.extenstion:substitution
  Whether substitution is allowed or not (issue from STU3 ballot)
  • Substitution is allowed -> equivalent
  • Substitution is not allowed -> none
• MTP: MedicationStatement.extenstion:substitution
  Whether substitution is allowed or not
  • Substitution is allowed -> equivalent
  • Substitution is not allowed -> none

[ziegm]

12.01. PJ, QL, OE, MZ

• remove "none" from the VS and change the binding to required
• type 0..1
• wasSubstituted = false => no type
• no guidance -> short/description

[ziegm]

@pjolo @qligier
Please review the changes:

• Profile MedicationDispense
  • ValueSet
  • Constraint
• Examples: was substituted, was not substituted
• Change log

[qligier]

The current constraint ((wasSubstituted = true and type.exists()) or (wasSubstituted = false and type.exists().not())) makes the type mandatory if it was substituted. Do we have said it was optional in that case?
I would have used the simpler one wasSubstituted = true or (wasSubstituted = false and type.exists().not()).

[ziegm]

The current constraint ((wasSubstituted = true and type.exists()) or (wasSubstituted = false and type.exists().not())) makes the type mandatory if it was substituted. Do we have said it was optional in that case? I would have used the simpler one wasSubstituted = true or (wasSubstituted = false and type.exists().not()).

I have adjusted this accordingly.

[qligier]

Looks good to me, thanks!

[pjolo]

For me, the changes are ok.