Pharmacy SIG Meeting Minutes - San Antonio

Pharmacy Working Group Meeting Minutes

Kyoto; Monday 11th May; Q1

Room 673

I. Reviewed and updated WGM Agenda

• Agenda reviewed and amended
• Garry Cruickshank will email updated agenda.

II. Action Items for the week/participants to other meetings

Wednesday

• Q3 - Nicholas Canu re IHE

Thursday

• Q1 - ISO/CEN/HL7 discussion, confirm Tim Buxton
• Q3, 4 - Michael Tan will be a representative at Clinical Statement meeting.

III. Action item list

Short overview of the action list as revised by Gary Meyer. There is more dedicated time for action item ‘work’ later in the meeting.

Kyoto; Monday 11th May; Q2

Room 673

IV. Introduction to the Pharmacy & Medication domain

• Brief walkthrough and discussion related to the structure, meaning and functionality of the models
• Medication Order RMIM
• Medication RMIM
• Overview of the interactions in these domains
• Brief walkthrough and discussion related to the structure, meaning and functionality of the Common Product Model

Kyoto; Monday 11th May; Q3

Room 673

V. Status of the January 2009 ballot reconciliation

• Medication Order Topic

• The Medication Order Topic normative ballot met quorum; met approval with no negatives.
• The Medication Order Topic normative ballot has passed normative ballot.
• Drug Knowledge-Base Query, Release 1

• The Drug Knowledge-Base Query, Release 1 normative ballot met quorum; met approval level but there are two negatives recorded.
• The result of reconciliation of one of the negative comments indicates that there will need to be an addition of one new RMIM to the topic. This would be considered substantive and as a result, there will be the need for an additional Drug Knowledge-Base Query, Release 1 ‘normative level’ ballot.

Note: See Thursday Q2 re: Keith Boone for an updated discussion and resolution.

• Medication Dispense and Supply Event, Release 1

• The Medication Dispense and Supply Event, Release 1 was an informative ballot.
• Negative ballots will be reconciled and changes resulting will be included in the pending normative level ballot.

VI. Approved January 2009 WGM minutes

Motion: Tom de Jong moved to accept; Nicholas Canu seconded that the minutes be approved: vote4 for, 0 against, 0 abstentions

Action: Add action items from Jan 2009 minutes to action item list (Garry)

Action: Add action items from Jan 2009 ballot reconciliation to action item list (Garry)

VII. Medication Order Ballot Reconciliation

• There were no comments to disposition for the Medication Order Topic ballot reconciliation spreadsheet

VIII. Drug Knowledge-Base Query, Release 1

• Disposition of the comments received for the Drug Knowledge-Base Query, Release 1 topic took place. For details, see the ballot reconciliation spreadsheet.
• New requirement from Nicolas (based on ballot comment by Freida):
• Add a ‘flag’ specifying that an allergy test (or some other lab test) has to be performed in order to prescribe the medication (the outcome of the test is not necessarily blocking).

Action: To be discussed in the work group.

Meeting attendees: I believe that the result of further discussion was agreement that the present structure, while perhaps not perfect, would support the use case raised by Nicholas – please confirm or refute.

• Based on ballot comment by Keith Boone, we concluded that it’s not appropriate to limit the medication document response to SPL. Besides that, the SPL reference was no longer valid. We will now create a new R-MIM, with a MedicationDocument class as the query response.

Action: To be discussed in the work group.

Note: See Thursday Q2 for an updated discussion and resolution.

• Based on a ballot comment by Gunther, we propose that we will finish up the current ballot for R1 of Drug Knowledge Base Query, then work on a R2 that is consistent with the Common Product Model. This will be discussed with Gunther over the course of the meeting.

(Add link to the spreadsheet)

Kyoto; Monday 11th May; Q4

Room 673

IX. Medication Dispense and Supply Event, Release 1

• We need to find out why the current Subversion models for the Medication Dispense topic are not the ones in the ballot. E.g., the cardinality of the inFulfillmentOf relationship was changed to 0..* in 2007 (see Subversion, although no note is in the model there either), but this is NOT in the ballot.

Action: Check for correct versions of the models (Subversion vs. ballot) (Hugh).

• We need to review why there is both an inFullfilmentOf relationship with CombinedMedicationRequest and with PriorCombinedMedicationRequest.

Action: Review semantics of inFulfillmentOf relationships during model review. Complete – one removed at time of detailed model review.

• Disposition of the comments received for the Medication Dispense and Supply Event, Release 1 topic took place. Ballot comments (informative) will be taken into account during scheduled dispense model review. For details, see the ballot reconciliation spreadsheet.

(Add link to the spreadsheet)

Orlando; Tuesday 13th Jan; Q1

(Annex 2)

• Joint with Orders & Observations

Kyoto; Tuesday 12th May; Q2

Annex 2

• Joint with OO; Rx; Pt Safety; RCRIM; PHER

• Common Product Model ballot reconciliation

Kyoto; Tuesday 12th May; Q3

Room 673

X. Agenda change

• Thursday Q2 joint session with RCRIM is cancelled.

XI. Preparing Medication Dispense for ballot

• Session with Gunther Schadow is tentatively scheduled for Wednesday Q4.
• Question about the version of the dispense model in the ballot. Is this the source from subversion? This version is from 17 Sep 2007. There are some cardinalities that have been changed and not found back in the ballot material.
• The original version has a header “CombinedMedicationDispense” that serves no purpose. That was before the choice box was modeled. Tom suggests to discard the header. Canada has implemented the header as a grouper and Garry is therefore reluctant to dump the header.
• Tom recalls that the choice-box was introduced by Wendell for reasons that could not be achieved in the header. The argument was brought forward that the grouper combines the supply information with the administration information. Tom is now convinced that the header should remain.
• Reminder: explain in the next ballot, how to model repetition of dispense. In Canada the dispense are all separate orders.
• We will now review the consistency of the D-MIM and the subversion file.
• Discussion on the cardinality of component2 from substance or supply to header which is now 0 to *. Tom argues if this is ever used. In the worst case it could be different medications under substance or supply.
• Nicolas mentions that an infusion with different compounds is seen as a dispense. Gary says this is actually a magistral compound. (extemporaneous prescription). Nicolas remarks that it could be different medication in one infusion. It is still seen as one component.
• The use case where it is zero is where only an administration is amended while the patient still has an enough supply in house. Nicolas adds that a dispense can be refused.
• Michael also adds the use case, where an initial dispense was issued, but the dosage instructions were not correct. A second dispense note was sent with only altered dosage instructions.
• The CMET participating as consumable has been changed to 220300 Billable Medication.

Kyoto; Tuesday 12th May; Q4

Room 673

XII. Preparing Medication Dispense for ballot

• The comparison with the D-MIM in the ballot shows that the version in the ballot differs very much from the version from subversion. We come to the conclusion that the ballot version is completely screwed up and it is of no use to look at that model.
• We could better reverse the process and start reviewing from the R-MIM and reflect back to the D-MIM.
• Should we have performer next to author in a dispense. The order itself has an author. Currently the dispense has both. Nobody could retrieve the reason why, but the relationship is left intact, because there must have been reasons why that relationship has been added.
• There are two relationships with a subject:
• The recordTarget and
• The subject
• The relation recordTarget is R_PatientPerson (contact). There is an issue with the subject relationship is currently R_Patient (universal ). Action item to take this up with PA.
• Why are there two relationships “inFulfillmentof” necessary? The top one combined Medication Request is to restrictive. The model is simplified to one relationship of “inFulfillmentof” instead of two.
• Technically you could add the complete prescription. This could be a use-case in countries where the prescription is not available in a repository. Decision to add a CMET for prescription.
• The Sourceof recursive relationship hanging of combined medication dispense has been discarded.
• The status code in the header has to be discarded, because the supply and the administration could have different states.
• Questions are raised about the use case of supply events in fulfillment of another supply events. This is simplified to one single relationship (predecessor).

Kyoto; Wednesday 13th May; Q1

Room 673

XIII. Preparing Medication Dispense for ballot

• Reviewed the SubstanceAdministrationRequestRevision portion of the model. Discussed the possibility of replacing this structure with a CMET, simplifying the model. In Jean’s absence, Hugh agreed to spend a few minutes at the beginning of Q2 creating a rough version of Universal CMET that would meet this requirement.

XIV. Feedback from harmonization meeting (Garry)

• Discussed and agreed that the flow for harmonization proposals should be External Proposer e.g. Canada Health Infoway; NHS; NEHTA (this may not be present)  Pharmacy WG (proposal could originate here)  Harmonization Meeting. If changes at harmonization the reverse flow should occur to measure agreement with the changes i.e. Harmonization Meeting  Pharmacy WG  External Proposer.
• All items Pharmacy WG sponsored harmonization proposals that were amended during the Harmonization Meeting were reviewed for acceptability of the changes. There were no issues found.
• One harmonization proposal ‘MedicationGeneralisationRoleTypeCode’ must be resubmitted.

Action: Garry to prepare new harmonization proposal.

XV. Terminology issues – Administration of Medicinal Substances (AfMS) Harmonization (Garry)

• Discussed that we have not moved AfMS forward through Harmonization as originally planned, though agree that the need is still there.
• Discussed what impact, if any, the Identification of Medicinal Products (IDMP) JIC work might have on any plan to move AfMS to Harmonization
• Agreed that we will:
• Initiate the process to advance AfMS through Harmonization
• Engage in the ISO IDMP CD Balloting process to make our requirements known
• We will discuss the larger engagement in the IDMP JIC process during the presentation by Tim Buxton Thursday Q1

Kyoto; Wednesday 13th May; Q2

Room 673

XVI. Agenda update

• Agreed that we will cancel Q3 meeting. Tom will attend the joint meeting with OO/ ARB and others regarding messages versus documents
• Garry will attend the joint meeting with PHER regarding Clinical Decision Support, with a focus on Immunizations
• It is believed that there is no material for either Medication Identification Services or Continuity of Care Document/Record discussions.
• Open podium for new ideas/proposals will be accommodated in one of the remaining quarters.
• Nicholas will be asked to attend and present his IHE Pharmacy Update in Q4 Wednesday

XVII. CMETs derived from Pharmacy/Medication

• Hugh created a Universal CMET for Substance Administration Request by copying from the Combined Medication Request RMIM
• based on what was currently in the Medication Dispense Event RMIM, an infulfillmentOf relationship was added to refer back to the original substance administration request
• we agreed that the added functionality related to precondition, triggers and goals present in the UV CMET would be reasonable in the dispense context.
• We decided not to copy the recursive precondition relationship on the ObservationEventCriterionClass which was present in the MedicationDispenseEvent RMIM. We were unable to come up with any plausible use case.

Action: Need to get appropriate numbering for the new CMETs (Jean)

XVIII. V2 topics/questions from Australia

• General discussion of V2 pharmacy
• Julia and Chris raised some specific questions
• RXO segment – where should the ‘authority prescription information’ be placed e.g. for methadone. This is not the prescription number.
• There is a prescription number as assigned by the EMR
• There is an authority prescription number, also as assigned by the EMR
• There is an approval number
• The prescriber must get the ‘authority’ as part of creating the prescription.
• Could perhaps use the order number and an order number type

Action: Discuss the requirement to support two additional identifiers with OO (Tom)

• RXO segment – not obvious where the prescription type should be placed e.g. PBS – Pharmaceutical Benefits Script (Public); Veterans; Private

Action: Bring to the attention of OO that Chapter 4 in Version 2.5 refers to a BLG segment that is supposed to be defined in Chapter 6 but it is not. (Tom)

• They will look into FT1 and BLG to see if these segments would support the requirements
• It is not obvious where the ‘Previous Authority’ indicator should be placed (will check the use case)

Action: It is not clear where a ‘valid to date’ should be place. Seek direction on the appropriate placement from OO (Tom)

• Stock supplies – look at OMS
• The specific AU material follows:

Comment form for post publication review of Interim Standards

AS 4700.3-2007 (Project 7356) Implementation of Health Level Seven (HL7) Version 2.5, Part 3: Electronic messages for exchange of information on drug prescription

• It is unclear where data needs to be put. Jared Davison commented that this standard requires different identifiers. More examples and explanations would be helpful.

• Assigning authority is also required. Reference to TR 2966-2007 would be useful.

• RXO Segment -XCN – the standard contains American terms eg.DEA there is a need to define what this means. There must be more interpretation. The usage notes are also unclear – do these come from the American specs?

• Authority indication and real reasons for prescribing the drugs need to be recorded. Where should this information go?

• Jared Davison commented that the Authority Script # and the Authority Approval # need authority codes assigned to them.
• We should make reference to HealthConnect specifications and make this compatible.

• Jared Davison commented that RXO 20 is unexplained we need to refer to HL7 specifications. More explanation needs to be provided here for this to be useful.
• Defined as CE data type in standard.
• This is the indication of why this drug has been given as an authority or a non authority drug. Mark Harrington suggested usage and examples will be very useful here too.
• We need to know Medicare Australia’s requirement/involvement in this area. Mark Harrington commented Ray Fillingham is involved in this area and may be able to Clarify.

• Table 2 in v2.5 needs to be extracted and incorporated into the document. XCN- ordering provider needs to be crossed referenced In TR 2966-2007.ie. <assigning authority> it would be useful to have examples “segment usage examples” perhaps have an EXAMPLES section and reference the table.

• We may need to create a section to discuss authority. There are current messages that Medicare Australia have containing the approval for authority

• Jared Davison commented that this standard is based on theoretical implementation. Most implementations do not follow the standard. It would be useful to map out the fields and segments. Pat Gallagher commented that there would be a large amount of mapping work to be done here. It may be a good idea to consider engaging a consultancy for a quote.
• More detail is required in the usage sections of segment tables. Australianised language/field names are also required. Table 300 in TR2966-2007 needs to be updated.
• This committee needs to determine which name identifiers are relevant.”User defined in table 300 Australian Namespace ID”. Eg. NEHTA AMT codes

Kyoto; Wednesday 13th May; Q3