May 2016 WGM Montreal, Canada: May 8 to May 13

• Back to Patient Care

• Back to Patient Care Work Group meeting Agenda and Draft Minutes

• Back to 2016 PCWG WGM Agenda and Minutes

Montreal (Quebec) Canada, WGM - May 2016 Patient Care WG Meeting Approved agenda:

• Patient Care WGM May 2016 Montreal - version date May 9

Montreal WGM - May 2016. Patient Care WG Meeting Meeting Minutes

• Sunday, May 8 - International Council Meeting

- No PCWG meeting

Patient Care WGM, Monday, May 9, 2016

• Back to Patient Care

• Back to Patient Care Work Group meeting Agenda and Draft Minutes

• Back to 2016 PCWG WGM Agenda and Minutes

Monday Q1

Chair: Jay Lyle
Scribe: Michelle Miller

Attendees

• Ayres,Elaine (NIH/Clinical Center)
• Campbell,Keith (U.S. Department of Veterans Affairs)
• Duteau,Katherine (DDI)
• Jones,Emma (Allscripts)
• Leftwich,Russ (InterSystems)
• Lyle,Jay (Ockham Information Services LLC)
• Miller,Michelle (Cerner Corporation)
• Newman,Shu-Fang (University Of Washington Medical Center)
• Nguyen,Viet (Lockheed Martin, Health and Life Sciences)
• Owens,M'Lynda (Cognosante, LLC)
• Parker,David (Defined IT, Inc.)
• Popat,Amit (Epic)
• Pyke,David (Ready Computing)
• Rehwoldt,Greg (Utah)
• Spielman,Matthew (InterSystems)
• Tan,Michael (NICTIZ Nat.ICT.Inst.Healthc.Netherlands)

Minutes

• Approve last WGM meeting minutes
  • January_2016_WGM_Orlando,_Jan_10_to_Jan_15 -- no comments or concerns

• This week's WGM schedule reviewed and updated
  • Patient Care WGM May 2016 Montreal - version date May 9

• Reviewed FHIR announcements
  • Timeline
    • Thurs, May 12 - Q3
      • What new resources/profiles do you plan to propose?
      • What are your FMM targets for STU3? http://hl7-fhir.github.io/resource.html#maturity
      • Any issues/concerns?
    • Wed, June 1
      • All resource and IG proposals for STU3 have been completed, reviewed by WG and submitted
      • Connectathon tracks for Sept have been proposed
      • Feedback on gForge submitted to FMG
    • Sun, July 17 - Substantive content freeze for ballot -- core resources
    • Sun, July 24 - Total content freeze, start of QA
    • Wed, Aug 10 (midnight) - All QA changes applied
    • Fri, Aug 12 - FHIR ballot opens
    • Fri, Sept 12 - FHIR ballot closes
    • Fri, Sept 16 - FHIR triage complete and ballot content loaded to gForge (or alternate)
    • Sept 17-23 - Baltimore WGM
    • Sun, Dec 11
      • Reconciliation complete/substantive changes applied?
      • Just over 10 weeks
      • Will re-evaluate at Baltimore based on volume of ballot comments
    • Dec 31 - Publish
  • FMM level survey
    • One input, but could have other considerations
    • Take into account volume of responses
    • Prioritized resources -- top 20 -- goal is at least level 3
      • Patient
      • Observation
      • Practitioner
      • Medication
      • Condition -- owned by Patient Care WG
      • Allergy -- owned by Patient Care WG
      • Org
      • Encounter
      • MedOrder
      • DiagReport
      • Immunization
      • MedStatement
      • DiagOrder
      • MedAdmin
      • Bundle
      • MedDispense
      • CarePlan -- owned by Patient Care WG
      • Procedure -- owned by Patient Care WG
      • Conformance
      • Person

• Work group health
  • Mission and Charter updated
    • Motion: Michael/Amit Abstain - 0, Negative - 0, Approve - 10
  • SWOT needs updated
    • Motion: Amit/Katherine Abstain - 0, Negative - 0, Approve - 10

• FMM Level Goals
  • Resource (current FMM level) Goal
  • AllergyIntolerance (1) 3
  • procedure (1) 3
  • Goal (1) 3
  • Careplan (1) 3
  • Condition (2) 3
  • Careteam (0) 2 or 3
  • FamilyMemberHistory (1) 2 or 3
  • QuestionnaireResponse (2) 2
  • Questionnaire (0) 2
  • ClinicalImpression (0) 1 or 2
  • ReferralRequest (1) 1
  • ProcedureRequest (1) 1
  • Linkage (0) 1
  • Flag (1) 1
  • Communication (1) 1
  • CommunicationRequest (1) 1

Monday Q2

• Plenary Session

- No PCWG meeting

Patient Care Monday Q3

Present:

• This session was hosted by O&O. Please refer to their minutes.

Minutes:

• An order does not represent a fulfillment request. To fulfill you need to post the resource with a tag.
• Alternative is to use an operation or message.
  • Or use a Task resource.
• The name order will be changed to request, such as MedicationRequest.
• The term request would also refer to referrals, procedure and appointment.
• Lloyd has displayed a request pattern slide.
• The attribute category represents the different states of a request.
• The request can be nested.
• Also the request have response resource as answer.
• The task resource has been developed.
• Besides that there is also and OrderSet and Protocol resource.
• How to use transaction reversal in combination with task is not yet done.
• Test in Connectathon where a patient with a Iphone was requested to schedule a lab-test and after confirming the plan set out to make a ask for the lab. In the end the physician was instructed by a task to look at the results.
• Discussion on the word or versus request. Some want to word order back again, but this is discussion that has already been dealt with.
• Need for examples of use cases. These seem to have been worked out for certain use cases.
• Discussion whether dosage instructions should be a datatype.

Patient Care Monday Q4

Present:

• This session was hosted by O&O. Please refer to their minutes.

Minutes:

• Continuation of the FHIR workflow discussion.
• Using the requisition-id as the grouper.
• Tasks itself has different states. What is going on with the task? There are separate business logic internal flows ( for example checking the authority of the order requester), the state of the tasks could be on hold.
• The relationship between sub-tasks and supertasks is not been worked out yet.
• Has the sequencing been sorted out. No.
• Can it be integrated with BPMN? Graham doesn’t think so. When you have a high volume process where the steps are well defined BPMN will work, but it does not work with ad-hoc flows.
• There are recommendations on the FHIR wiki when to use each method of workflow:
  • A tag is kind of throwing it over the wall.
• There is some resistance from some IT vendors who do not feel comfortable with the idea that orders are not actionable.
• HL7 has the ambition to write implementation guides to give guidelines for implementations.
• The attribute category represents something like “mood” or “stage”.

File:FHIRworkflow.pptx

Patient Care Monday Q5: Placeholder for extra meeting if necessary

Present:

Minutes:

Patient Care WGM, Tuesday, May 10, 2016

• Back to Patient Care

• Back to Patient Care Work Group meeting Agenda and Draft Minutes

• Back to 2016 PCWG WGM Agenda and Minutes

Patient Care Tuesday Q1

Chair: Jay Lyle
Scribe: Emma Jones

Attendees

• Heerman-Langford,Laura (Intermountain)
• Matney, Susan (Intermountain)
• Maclean, Andrea (Infoway)
• Huff, Stan(Intermountain)
• Lyle, Lyle (Contractor)
• Linda Bird (Intermountain)
• Demarmels, Marco (HL7 Switzerland)
• Ryan, Sarah (Eckhorn Int Services)
• Zaidi, Shirin (Gevity Consulting)
• Popit, Amit (Epic)
• Nanjo, Claude (Cognitive)
• Rehwoldt, Greg (Deloitte)
• Allen, Kurt, W (Penrod)
• Esmond, Richard (Pen Rad)
• Newman, Shu-Fang (Univ of Washington)

On phone

• Patrick Langford
• Ken Kawamoto

Minutes: Agenda:

• PSS - investigative action - looking at a project to take existing patient care concepts and represent it in CIMI
  • PSS thru DESD (response) -
    • need to review dates,
    • choice of data, artifacts (need a list),
    • draft output patterns - some are in CIMI, others may be needed,
    • draft output of the composition - how large will CIMI get? medicalPsychiatric condition will be left out for now. CIMI will need to look at conditions/findings. Aim is to have the composition ready for review Sept 2016
• Review of scope (the below are specific to ulcers)
  • Skin assessment model
  • Braden - note made that Braden Q is proprietary and request compensation to be used (ONC is aware)
  • review example ADL
• Need to add FHIR as next phase

• Inheritance Map review
  • Discuss the use of evaluation Vs observation
    • observation is direct observation. Evaluation involves a thought process - per OpenAir
    • the distinction is that the thing is done on a patient vs lab things. With things done on the patient can have measured high and low values rather than normal/abnormal
    • Need to consider other types of assessment/examination that might fall under this category - e.g. blood pressure and heart rate. May need to compare that modeling. Modeling was done by CIMI.
    • SNOMED bindings came from the nursing group working on this at the time it was CAP
    • Need to take into consideration the status
    • Stan: Root level bindings and code level bindings - in the end this is what's more useful. the root level binding should be to something that captures all the semantics. The code says what the value means - name value pairs. The semantic of the whole model is defined by those things. To be consistent, the binding for the SBP will be a compositon of these things. Intermountain has made a unique ID of the model and uses it as a concept. Therefore will have the whole thing together with all it parts defined and have it's own semantic meaning.
    • Linda: agree with Stan. Will understand the meaning of the model itself and the structure which has its own hiercharchy. Separate to that is the clinical concept that is the focus o fteh information recorded. Stan agrees.
    • there are different approaches to terminology bindings.
    • Skin assessment panel with all these observations that have the various observations
    • Code binding will use LOINC for that is used now.
    • complete is static -may need to change if needed
    • More discussion as to how to model. the example observation came from the lab
  • Model binding for the value for this instance. Note made that the logical models have attributes that are fixed that will not be included in the implementation.

Patient Care Tuesday Q2

Present:

Minutes:

Patient Care Tuesday Q3

Present:

Minutes:

Topics:

• Adverse reactions
• Negation
• Condition status
• Care Plan
• Consolidation of clinical workgroups.

Adverse event resource

• RCRIM: Lise Stevens and Mead Walker, have done preliminary work, based on ICSR v3.
• Issue is the scope of the resource to be usable for other settings.
• This session is a restart of the initiative. Goal for today is get this project restarted.
• Questions:
  • ICSR was built for adverse event reporting. What do we want now? Do we want to consider to use it also beyond regulatory reporting.
  • After starting with Allergy/Intolerance they quit this approach and started with ICSR reporting that is sent to FDA.
  • Nobody opposes to limit the adverse event to substance events
• Question to Lloyd about the FHIR planning and deadlines.
  • Need content to be frozen by 15 July.
  • PC should give the statement of maturity of the product.
  • It will be published by the end of the year.
• Team is now set.
• Explanation of Mead’s concept model.
  • Identifier
  • Type
  • Subject
  • Date
  • Status
  • Outcome
  • Description
• Field “ severity” is not available.
• According to Julia, FHIR already has a value set for outcome.
• How should you do the profiling? Excluding attributes and remove them from the profile?
• This model will be discussed in the RCRIM group.

Negation

• What are the thoughts of CQI about negation? The PSS for the negation project is to gather the requirements on the negation. Solutions is not the focus.
• CQI will definitely be involved. What is the direction of PCWG for looking for the solutions.
• In radiology: They have a matrix for each modality to decide what is normal. ( for example the size of the hart). The question is, how to decide what normal is in other context.
• Rob Hausam has a slide presentation to explain.
  • Proposal 1
    • Align renaming of resources
    • AllergyIntolerance substance into AllergyIntolerance code.
    • The binding to the value set remains the same.
    • Create a new allergy-intolerance-negated-code value set.
  • Proposal 2
    • Follow the OpenEHR solution. Split it out in 2 separate resources that have only positive statements.
    • Discussion whether solution 1 would not lead to 2 profiles.
  • This discussion will be continued in Q4.

Consolidate workgroups.

• Discussion to group workgroups together because of overlapping topics. This is a topic CVS, CQI are working on, but PC could also consider to be involved.
• CIC might also consider participating to gather some clout.
• Laura remarks that this is a recurring topic that returns every 8 years. In the end the burden of the WG becomes so large, that it would split into smaller groups again.
• External parties have difficulty finding the right WG or the need to address different WG.
• Lloyd suggests that the need to be in touch of the progress is to have a conference call once a month to report what each WG has done.

Patient Care Tuesday Q4

Present: Jay Lyle - Chair, Elaine Ayres - Scribe (total of 50 attendees) Andrea Pikus - IMO Lloyd McKenzie - Gevity Michael Tan - NICTIZ Lindsey Hoggle – Academy of Nutrition and Dietetics Margaret Dittloff – Academy of Nutrition and Dietetics George Cole - Allscripts Joey Cole - Intermountain Matthew Spielman – Intersystems Michael Lawley – CSIRO Keith Campbell – VHA Brian Wrighte – Mayo Clinic Esteban Aliverti – Cognitive Ron Shapiro – Quera Chris Chute – Hopkins Tom De Jong – HL7 NL Keith Allen Martin Rosner – Philips Senthil Nachimuthu – 3M Marla Albitz Rita Menial Sean Moore – Epic Michael Donnelly – Epic Daniel Rutz – Epic Greg Gustafson – Penrad Emma Jones – Allscripts Scott Robertson – Kaiser Permanente Sue Thompson – NCPDP Shelly Spiro – Pharmacy HIT David Burgess – Labcorp Craig Parker – Intermountain Michelle Miller – Cerner Dennis Patterson – Cerner Oyrind Aassave Sherry Taylor – NIST Claude Nanjo – Cogmedsys Susan Barber – State of TN M’Lynda Owens – Cognosante Ted Klein – KCI LLC Heather Grain – Ehe Australia Rob McClure – MDPartners Bruce Bray – HSC Utah Rob Hausam

Agenda:

1. OO: multiple answers to a question in a questionnaire
2. Negation
   1. PSS: asking MnM for participation; CQI & PA had expressed interest. Also need to articulate how to engage with existing design & guidance -- design constraints based on legacy environment are requirements.
   2. Plan
      1. Collection of use cases
      2. Classification of use cases
         1. Including cases out of scope to clarify boundaries
      3. Identification of principles
      4. Identification of what can be provided to design groups / how feedback can be addressed
   3. Review classification current state; solicit gaps

Minutes:

1. OO question re multiple answers to a question in a questionnaire – based on current coding in observation and observation component.

The group tried to recall why the invariant was included. The argument against – will the observation code will be misused?? Need to understand the boundaries around the component observation scenario.

Currently in Observation you have a code and components also can have a code, but the code at component cannot be the same as the observation.Code (invariant), so when a question has more than one answer - Example Alcohol type? = parent code as well as component code with 2 answers (beer and wine) currently not allowed.

QuestionaireResponse treating each answer as its own observation, so we want to remove the invariant.

Discussion: Looking for the argument to have the invariant – will this open up to use component for other things that we don’t want (need to clearly define the boundaries between code / component / related observation) – component should be used for items that cannot stand on their own.

Will take back to OO for Q1 Wednesday – PC feedback: yes we need more than one answer to the same question

2. Negation Project Scope Statement – went to DESD. Need to add MnM as per DESD feedback. This is a design pattern project. Issue is dealing with current design patterns in current standards vs. new recommendations. Key issue is not to use the concept of negation. The word negation is creating ambiguity. Need to first develop use cases of how the absence of findings are used, and then classify these use cases.

The spreadsheet on the wiki is located under the Patient Care site under the Negation project.

URL: http://wiki.hl7.org/index.php?title=Negation_Requirements

Spreadsheet: http://wiki.hl7.org/images/1/18/NegationUseCases.xlsx

Spreadsheet has a category column: examples are:

a) Finding absent (exam) as opposed to not known

b) Finding absent (instrument)

c) Anatomical deficit

d) Anatomical deficit, congenital

e) Anatomical deficit, surgical

f) Condition absent – common in pathology. A previous history of MRSA, would note – absence of MRSA. Part of medico-legal response.

1. 1. 1. Part of the condition resource

1. 1. 2. In family history resource – my father had X but my mother did not

g) Condition refuted

h) Procedure not done

i) Encounter not held

Nesting of questions and answers make negation even more complex. What rules apply to nested observation structures? Need to look at use cases for example cancer reporting forms.

A finding/assertion may be absent, but multiple findings may be absent. A batch may have a mixed assertion pattern as well. Context conduction (a RIM artifact) may represent this issue, but perhaps not.

Noted that classification categories in spreadsheet are quite detailed. What about more general patient observations such as “no next of kin” or other types of observations in the social history?

Conditional prohibition – e.g. no food until pain subsides. Functional deficit – no vision in right eye Patient preference to abstain – assertion of patient not to breastfeed.

In spreadsheets have a mapping for similar statements – and looking at representation in V3 and FHIR resources to create harmonization across standards. (not in project scope but will help design teams).

On wiki there is a page that looks at negation principles that have been collected during the project discussions. For example – no double negatives. http://wiki.hl7.org/index.php?title=Negation_Principles

Should negation be done in the model or in the terminology (pre-coordinated or post coordinated expression).

Depending on the code system the answer may vary. In SNOMED CT negation does not work in the description logic. Need to put negation into the information model. The pre-coordination does not work in SNOMED CT.

Negation in the code field as a post-coordinated expression will violate logic. In the information model it will work.

SQL supports negation – put where to you put it? If in that query it works. If you allow it in other settings, it does not work.

Head injury without loss of consciousness, and head injury with loss of consciousness. This is an observation result with a presence or absence stated in a separate place based on query systems (e.g. ICD-9 or ICD-10).

A query for loss of consciousness should return correct information.

Key is to represent negation in the information model, not the terminology. (IHTSDO)

For free text entry – the user interface becomes critical. Mom thinks there is an allergy but Dad disagrees. Should codified statements be separated in the model or for all use cases? Even with NLP parsing text, will need a consistent model for data representation analysis.

Not just negation itself, but what are you negating? There is some context.

Asplenia – (absence of a spleen) can be congential or surgical. These require different representations.

Plea is to keep negation out of the terminology, and put in the model. This should become a design principle.

MOTION: Keith Campbell/Michael Lawley

Motion to write a principle (Keith Campbell) Negation will not be allowed in a coded expression in the code field of a resource but will be allowed in a specified field of the resource. This field can be populated with a code from the terminology system.

Rewritten after the meeting as

Concept codes (or expressions) used in coded data element fields should not include semantic negation (including, e.g., finding absent, procedure not done, etc.). Information models should provide other means for asserting these modifiers, whether as a separate field in the information model, which may itself be populated with a code or expression, or in some other way.

Need a canonical representation for interoperability. The user interface may need to be pre-coordinated. In the reasoning, data entry needs to be separated (transformed) for data exchange. Should not be reported clinical statements in a classification.

How would this play out in the allergy and intolerance resource? In the condition resource – can add pre-coordinated context codes – would need to represent via the status.

MOTION to table: Keith Campbell/Kurt Allen Motion is tabled by acclimation

Summary: Bring use cases, review spreadsheet. Call is Wednesdays at 11 AM ET. Topic will be continued on Wednesday, May 11 during Patient Care Q4.

Patient Care WGM, Wednesday, May 11, 2016

• Back to Patient Care

• Back to Patient Care Work Group meeting Agenda and Draft Minutes

• Back to 2016 PCWG WGM Agenda and Minutes

Patient Care Wednesday Q1

Present:

Minutes:

Patient Care Wednesday Q2

Present:

Minutes:

Patient Care Wednesday Q3

Present:

Minutes:

Patient Care Wednesday Q4

Present:

Minutes:

Patient Care WGM, Thursday, May 12, 2016

• Back to Patient Care

• Back to Patient Care Work Group meeting Agenda and Draft Minutes

• Back to 2016 PCWG WGM Agenda and Minutes

Patient Care Thursday Q1

Present:

Minutes:

Patient Care Thursday Q2 (PCWG, SDWG, Template Joint meeting)

Present:

Minutes:

Patient Care Thursday Q3

Present:

Minutes:

Patient Care Thursday Q4

Present:

Minutes:

Patient Care WGM, Friday, May 13, 2016

• Back to Patient Care

• Back to Patient Care Work Group meeting Agenda and Draft Minutes

• Back to 2016 PCWG WGM Agenda and Minutes

Patient Care Friday Q0: PCWG Co-Chairs meeting

NOTE: co-chairs to determine whether this will be moved to Thursday Q-Lunch

Present:

Minutes:

Patient Care Friday Q1

(1) Template WG meeting

- PCWG representative(s) to attend

(2) Agenda: open

- Possible: Clinical Connectathon

Present:

Minutes:

Patient Care Friday Q2

Agenda: open

Possible: Clinical Connectathon

Present:

Minutes:

Patient Care Friday Q3

Agenda: open

Possible: Clinical Connectathon

Present:

Minutes:

Patient Care Friday Q4

Agenda: open

Possible: Clinical Connectathon

Present:

Minutes: