This wiki has undergone a migration to Confluence found Here

Difference between revisions of "Patient Care January 2016 WGM Orlando, Jan 10 to Jan 15"

From HL7Wiki
Jump to navigation Jump to search
Line 156: Line 156:
Minutes: Michael Tan
Minutes: Michael Tan
*Stan Huff presenting CIMI viewpoints: CIMI has now become a HL7 workgroup.
*Stan Huff presenting [[Media: CIMI_Orlando_HL7_meeting_160110.pptx | CIMI viewpoints]]: CIMI has now become a HL7 workgroup.
*Searching collaboration between CIMI and other WG's that have DCM's.
*Searching collaboration between CIMI and other WG's that have DCM's.
*CIMI produces Clinical Information Models. The thought behind the models is, that systems do not need to know the structure of information from other systems. It uses the information models from webservices to map and share data.
*CIMI produces Clinical Information Models. The thought behind the models is, that systems do not need to know the structure of information from other systems. It uses the information models from webservices to map and share data.

Revision as of 17:31, 27 January 2016

Orlando, WGM - January 2016 Patient Care WG Meeting Approved agenda:

Orlando WGM - January 2016. Patient Care WG Meeting Meeting Minutes

  • Sunday, January 10 - International Council Meeting
- No PCWG meeting

Patient Care WGM, Monday, January 11, 2016

Patient Care Monday Q1

Attendees Jay Lyle, Russ Leftwich, Elaine Ayres, Emma Jones, Michelle Miller, David Pyke, Larry McKnight, Michael Tan

Chair Jay Lyle

Scribe Michelle Miller

Agenda check

Approval of prior minutes (Atlanta)
David/Emma: 0 -2 abstain – 4 approved

Week's agenda Approval for Elaine to be a co-chair designee for the week. Michelle/Emma: 0-0-7

Anesthesia wants to meet.
Stephen suggested to Anesthesia already, but they may only be here Mon/Tues. Suggested times:

  • Monday Q3 (same time as workflow)
  • Tues Q5

SDC wants to discuss Health Concern (and PC wants to discuss goals with ONC) SDC can’t attend existing session. Jay will reach out to SDC with options:

  • Wed Q2 (Emma, Elaine could attend)
  • Wed Q3 (Genomics declined, so FHIR topics such as Questionnaire/QuestionnaireResponse, Clinical Impression)
  • Thurs Q1 check with OO to see if PC needs to attend

Elections Please vote!

Workgroup Health Good job! Keep an eye on

  • Decision making processes
  • SWOT
  • Charter
  • Harmonization - Rob H. was designee
  • 3 year plan – list of projects that are in the register (some are active, proposed, linger); review and see if any can be closed out;

3 yr planning Jay will own cleaning up this week. Send Dave Hamill an email.

  • Detailed Clinical Models Release 1 (320) - not an open project, just needs to be recirculated
  • Care Provision: Assessment Scales Topic (664)– never got published, ask William about that
  • HL7 Care Coordination Service (CCS) (924) – ready for publication; Laura will clean up and submit it for publishing; publish for HL7 members only for 90 days before publish to the public
  • Care Plan Initiative (932) – DAM not published; take all story boards, except one, and move them to the appendix; Laura owns that action item
  • V3 Allergy and Intolerance Clinical Models (1004) – published and available for comments; project is done; close project
  • Care Plan Clinical Models (1085) – check with Laura to see if it can be closed
  • PCWG Develop FHIR resources and profiles for the second DSTU release.

New projects None

FHIR Update from Sunday’s Q4 meeting

  • Maturity Review - what to ballot when (2.1 in May 3 in Sept?) – need work groups to look at resources and find out which resources are projected to have which maturity levels by next release; FHIR health –DSTU comments – check that the changes were made; any outstanding issues?
    • Need clarification in FHIR maturity model to know if there is production use (instead of production systems) and/or what constitutes a connectathon
  • Naming – FHIR could be dropping DSTU label because it isn’t part of ANSI’s process; FHIR is thinking about calling it informally a “working” standard, but Patient Care’s concern is credibility about losing the “draft standard” label. The maturity levels are per resource, not for the entire standard.
    • David will own drafting Patient Care’s statement/response back to FHIR
  • Error list in FHIR build – confirm with Lloyd that he is fixing any errors on Patient Care’s error report

Adjourn 1/11/2016 at 10:33am Eastern

Patient Care Monday Q2

No PCWG meeting

Patient Care Monday Q3

No Patient Care hosted meeting; Patient Care attending FHIR Infrastructure - Workflow meeting

Patient Care Monday Q4

No Patient Care hosted meeting; Patient Care attending FHIR Infrastructure - Workflow meeting

Patient Care Monday Q5: Placeholder for extra meeting if necessary



Patient Care WGM, Tuesday, January 12, 2016

Patient Care Tuesday Q1


Minutes: Michael Tan

  • Stan Huff presenting CIMI viewpoints: CIMI has now become a HL7 workgroup.
  • Searching collaboration between CIMI and other WG's that have DCM's.
  • CIMI produces Clinical Information Models. The thought behind the models is, that systems do not need to know the structure of information from other systems. It uses the information models from webservices to map and share data.
  • The CIMI produces clinical requirements.
  • CIMI's goals are to create a shared repository of detailed clinical information models.
  • The repository is free of license.
  • Models contain formalism:
    • Achtetype definition language (ADL)
    • Archetype modeling language (AML).
    • Models have a binding with a value set.
  • The model can be rich and full, while the implementation only uses a small part of the CIMI model.
  • How should the repository look like?
    • It does not exist yet.
    • It should contain a name, status, version, isosemantic familiy, ownership, approval.
    • What is the relationship between CIMI and the PCWG.
  • Jay draws the landscape of CIMI, DAM's, DCM.
  • What should we do to accomodate the CIMI group? Susan suggests to pick up allergies and intolerances as an example to see how this would work out. Skin intolerances/skin assessment? A project would be started with CIMI as driver and PCWG as collaborating WG to investigate how this would work out. Is the PSS a study or a demo. The study would include a demo. The demo would include an implementer.
  • PCWG would first like to review the PSS. This does create some pressure, because the deadline for PSS is January 22nd.
  • William places a motion to have undertake a pilot to use skin assesment for the collaboration between CIMI and patient care. Steve Hufnagels material is available.
  • Steve Hufnagels PSS will be distributed among PCWG to review this week, before giving approval to co-sponsor the PSS.
  • CIMI will send a invitation for a joint meeting with PCWG for Montreal on Tuesday Q1.
  • action item PC. The PCWG will review the PSS, before the deadline for new PSS and provide feedback to CIMI.

Patient Care Tuesday Q2

Chair: Michael Tan
Scribe: Michelle Miller


Last Name First Name Organization
Cole George Allscripts
Dasgupta Amlan Epic
Dolin Gay Intelligent Medical Objects (IMO)
Fine Steven Cerner Corporation
Goossen William Results 4 Care bv / HL7 Netherlands
Grieve Grahame Health Intersections Pty Ltd
Jones Emma Allscripts
Knee Simon HSCIC
Lyle Jay Ockham Information Services LLC
Majeed Raphael
Mandel Joshua Boston Children's Hospital
Marquard Brett River Rock Associates
McKnight Lawrence Cerner Corporation
Miller Michelle Cerner Corporation
Moore Sean Epic
Nelson Lisa
Pyke David Ready Computing
Shapiro Ron Qvera
Skapik Julia Office of the National Coordinator for Health IT
Torres Andrew Cerner Corporation
Wang Yunwei Intelligent Medical Objects (IMO)
Wright Colin McKesson Corporation
Jorgenson Don Inpriva
Terry May Flatiron


Health Concern vs Problem

Patient Care's view

  • A concern is a higher level abstraction than the problem (ex. concern of obesity) that connect problems into a concern
  • A concern connects events and tracks progress over time (what started with and what it has become) - Example: back pain grows worse into hernia
  • DAM is sufficiently abstract such that it supports knowing who expressed concern; DAM can handle either objective or subjective
  • Patient Care DAM is saying concern is greater than problem; problem can be a subset of concern
  • Concerns are both Problems and Additional Health Concerns - and each is represented in its own CDA section

Structured Doc's view based on ONC's proposal (attachment in Patient Care listserv [1] )

  • Active problems are in the problem concern section
  • The "Additional Health Concerns" section is used for additional concerns, such as concerns from the patient or other care team members
  • Clinician owns the problem list whereas concerns are open to others, such as chaplain
  • Note: Unhealthy family relationships and drug/alcohol abuse could be concerns, but may not be endorsed by patient
  • Clinicians don't need to reconcile additional health concerns, but there is no legal requirement for the clinician to reconcile problems on the problem list - yet they should reconcile a concern of sick parent after parent is deceased
  • Goal is to allow the patient and other care team members to express concerns (e.g. financial instability, homeless) to flow into the longitudinal care plan
  • The MU requirement is to allow health concerns to be entered into the record
  • Intent is that additional health concerns are viewable, clinician considers them, clinician follows them, but clinician doesn't necessary endorse (or agree with) the concern; The clinician should ask patient about concerns, but not the clinician's job to agree or disagree with patient.
  • The Problem concern is more objective whereas the Additional Health Concern section is more subjective
  • The Additional Health Concerns section is not intended to duplicate the concerns in the problem concern section


  • Motion is for Patient Care agrees that the ONC proposal (attachment in Patient Care listserv [2]) is not in conflict with the DAM
    • Jay/Larry: 0-abstain; 0-negative; 22-approval
  • Motion to recommend that SDC modify the ONC proposal: When the document contains both a Problem Concern and a Health Concern section, then the section title will be the "Additional Health Concerns" and keep the same LOINC code.
    • Jay/Brett: 1-abstain; 1-negative; 20-approved

Patient Care Tuesday Q3

Present: Laura Heermann Langford - Chair, Elaine Ayres - Scribe Attendees: Patty Craig, Ewout Kramer, Eivind Kristiansen, Oyvind Aassve, Donna Quirk, Amlan Dasgyli, Daivd Dunering-Stuekl, Justin Schiirle, Casten Quinlan, Senlthil Nacimuthen, Joe Kunisch, Bruce Bray, Larry McKnight, Martin Rosner, Peter Hendler, March Hadley, Mark Kramer, Bryn Rhodes, Julia Skapik, M'Lynda Owens, Muhammad Asim, Andrew McIntyre, Kensaku Kawamoto, Yan Heras, Tony Schweth, Latasha Archer, Daniel Lanphelt, Ann Phillips, Juliet Rubini, Charis Markle, Non Hall, Nancy Orvis, Graham Grieve, David Sattara, Matt Burton, Michael Tan, Lenel James, Richard Esmond, Claude Nanjo. (41)



  • Care Plan and Order Set Resource (CDS)
  • Updates to C-CDA and updates to CQI products - allergy content
  • QRDA and exchange of provider information necessary for quality reporting

Care Plan and Order Set Resource

The Care Plan takes a multidisciplinary point of view. An order set would contribute to the overall plan for the patient. In the care plan model - the care plan is in the middles with goals, risks, problems and the health care team are associated with the care plan. With an activity will see orders. an order can be a single order or an order set. So, from a care plan view, not repetitive to have a resource for order set. An order set is a natural component of a care plan. Orders may need to be elevated within the model to be more visible.

Are there descriptions or definitional items that assist with building order sets? Can discuss the various levels of a care plan. Can have goals related to all issues or one issue. Can have an entire care plan around one condition, or topic. Plans can be nested. The FHIR Care Plan was created without necessarily aligning within the Care Plan DAM. Now harmonizing the two. Will the FHIR Care Plan go through changes in the next release? This is unknown, but could be likely. For the proposed order set resource -- from a care plan perspective would point to the order set resource.

Care plan vs. care process model? Could use the care plan model as a care process model. However, current model addresses data only, not authority or process. Is this a point in time or over time plan? It is designed to be both a static plan snap shot as well as a longitudinal plan. Designed to be fluid between care team members. Is it definitional -- e.g. for a patient with Diabetes do this specific plan of care -- can use it in this way. It can also be customized to each individual patient. For a protocol (form) plan of care, may not have preferences or barriers. The model is designed to be flexible and adaptable to various care models. For exchanging or sharing plans, how might this work. The standard does not mandate consistency but encourages it.

Looking at the order set build in FHIR - the meta data is harmonized with HQMF reference. The library data elements applies logic based on condition (not the FHIR resource). The orderable items are references to other resources like a med order with any number of items. Are these prospective or retropective -- probably neither. This is a point in time. In the RIM have mood codes, in FHIR have deinitional resources as well as prospective vs. retrospective states. CDS is the "owner" of the order set resource.

A care plan may have a dependency on the order set -- to keep up the lifecycle of the care plan. Can you use order sets without a care plan? Can an order set be a proxy for a care plan? Actual care benefits from the context of the care plan, rather than a list of orders. This enhances the coordination of care. The order set is actually similar to a care plan template. Future -- include CDS/CQI for Thursday FHIR calls. Same schedule for PC/CDS/CQI/FHIR for Tuesday Q3.

CQI and Allergy Topics The issue is the consonance between C-CDA version and derivative QRDA products. Note that the development cycles are not the same. CQI asks that if PC/SDWG makes a change that will impact the QRDA, changes should be submitted to CQI as well.

CQI White Paper to look at use cases for QRDA Data from provider to provider related to quality measures. Are there any standards for continuity of care data outside of QRDA? Orders a mammogram but returned data may not be sufficient for quality reporting measures. The Care Coordination Service may address this issue (Patient Care and SOA).

Patient Care Tuesday Q4

Chair: Elaine Ayres
Scribe: Michelle Miller

Last Name
First Name Organization
Aassve Oyvind HL7 Norway
Ayres Elaine NIH/Clinical Center
Brown Jeff American Society of Clinical Oncology
Burgess David Laboratory Corporation of America
Campbell Keith U.S. Department of Veterans Affairs
Couderc Carmela Intelligent Medical Objects (IMO)
Doerr Lisa McKesson Corporation
Gonzaga Zabrina Lantana Consulting Group
Grain Heather eHealth Education
Hausam Robert Hausam Consulting LLC
Hoggle Lindsey Academy of Nutrition & Dietetics
Jones Mark Orchard Software
Kreisler Austin Leidos, Inc.
Kristiansen Eivind National ICT
Lyle Jay Ockham Information Services LLC
McClure Robert MD Partners, Inc.
McKenzie Lloyd Gevity (HL7 Canada)
McKnight Lawrence Cerner Corporation
Merrick Ulrike Vernetzt, LLC
Miller Michelle Cerner Corporation
Nanjo Claude Cognitive Medical Systems
Nelson Lisa
Owens M'Lynda Cognosante, LLC
Parker Craig Intermountain Healthcare
Patterson Dennis Cerner Corporation
Quirk Donna Lexington Medical Center
Rhodes Bryn Database Consulting Group
Schueth Tony Point-of-Care Partners, LLC
Shabo Amnon Philips Healthcare
Solbrig Harold Mayo Clinic
Tan Michael NICTIZ Nat.ICT.Inst.Healthc.Netherlands
Taylor Sheryl National Institute of Standards and Technology
Wang Yunwei Intelligent Medical Objects (IMO)
Warner Jeremy Vanderbilt University Medical Center
Nguyen Viet Systems Made Simple
Haug Peter Intermountain Healthcare
Boumann Steven McKesson Corporation
Gerlach Chuck McKesson Corporation
Richesson Rachel Duke

Problem statement
The ways current interoperability standards handle negation or propose to do so,

  • are inconsistent & may not support transformations
  • may not support representation requirements (e.g., patient does not have ebola)
  • may not support computation requirements (e.g., description logics constraints)

Motion to approve problem statement passes: Viet/Jay: 3-abstain; 0-negative; 36-positive

Goal statement
Draft: Provide concrete requirements from clinical experts to interoperability solution designers to enable these designers to ensure that their solutions meet requirements, including.

  • information requirements & examples ("assert absence of symptom"; "no rash")
  • impl/communication requirements (how do I represent "no rash")
  • transform requirements (?)
  • [other, e.g. DL negation concern]

Do we agree that we need a consistent approach to negation except where requirements prevent it?
FHIR design is driven by implementation requirements.
Assume consistency until told about exceptions, but is consistency is prioritized higher than other things that are good to have, such as quality
Default position is that you can assert semantics

Example Use Cases
Clinicians on FHIR - via the process of testing FHIR resources, Clinicians on FHIR will identify issues that warrant additional analysis; create use cases to test negation

Atlanta (Oct 2015) Clinicians on FHIR discussed the inability to communicate negated family member history

Note: Negation can be handled via terminology or the information model.

Clinical certainty (no known vs not known) -- categories of negation

Consider regulatory requirements (no known problems)

There are some older project scope statements (PSS), but need to see if they are still appropriate to determine if a new PSS is needed or not. Suggested that we create a new PSS.

Motion: Create PSS to define negation requirements Keith proposed amendment to include several approaches and evaluation of approaches to solve the issue Note: CQI should be involved since they recently discussed and may have some resolutions Vote: Jay/Rob 0-abstain; 0-negative; 37 positives

Contributors going forward will include SD, OO, FHIR, CQI, CIMI, Vocab
Keep the same WGM session in May 2016, with the addition of CQI

Patient Care WGM, Wednesday, January 13, 2016

Patient Care Wednesday Q1

Present: Chair: Laura Heeermann Langford, Scribe: Laura Heermann Langford, Presentor: Mike Padula, Attendees: Russ Leftwich, Gaye Dolin, Larry McKnight, David Pyke, Masoud Hosseini, Sylvia Thun, Dominik Bremman


  • Review of Child Health's work re: Essential information for Child with Special Needs.
  • Discussion of how this document is useful beyond pediatric patients. But could be used with others with special needs, chronically ill, repetitive visits etc.
  • It is becoming apparent how important the style sheet is to the overall use of the document. Style sheet is not currently part of the scope. Explored options of participating in the style sheet competition - or perhaps hold a separate competition just for this project.
  • Comment re: the use of coded data and textual data. There may be a need to have more textual data.
  • Discussion on how much and methods to use to provide more guidance on the style sheet. If we could get more information out it is believed providers could and would encourage their vendors to implement it.
  • Method of delivery (elelctronic/paper) is also an important part of the overall workflow and design. It is reality that the use of electronic tools is targeted, but for some instances (camp) a paper document may also be needed.
  • The relationship between this document and the Care Plan is tight. This document may be the shortened "tweeted" version of the care plan. Would like to explore how the content for each is entered and reused to accommodate the needs of both.
  • Next steps
    • clean up the storyboards to include a representation of the output (the data elements used and what would be included in those data elements) include things like "the parents provide access to xxxx".
    • make one storyboard about the paper document - (reality of the need)
    • explore companion guides.... is this a companion guide?

Patient Care Wednesday Q2

Present: Elaine Ayres, Susan Matney, Martin Hurrell, John Walsh, Lori Singer, Larry McKnight

Minutes: This session was a brief update from the Anesthesia WG. The Anesthesia WG is continuing to work on the Anesthesia DAM and hope to ballot within the next six months. The DAM includes use cases, UML models and SNOMED CT coding. At this time, the co-chairs of the Anesthesia WG have encountered funding difficulties for co-chairs. The group will continue to work on their projects but the co-chairs have spoken with HL7 HQ about alternative administrative options for the group. Medical Devices has expressed an interest in supporting ongoing Anesthesia work. A final resolution is still pending, but Anesthesia will keep PC updated on status.

Patient Care Wednesday Q3

Present:Jose Costa Teixeira, Emma Jones (Scribe), Michael Tan, David Pyke, David Parker, Vassil Peytchen, Massoud Hosseiric, Andrea Pitkus, Maribeth Gagnon, Mark Jones, Carolyn Knapik, Elisabeth Berger, David Burgess, Lloyd McKenzie, Ilcon Kim, Sungchud Bae, Joon Hyun Song, M'Lynda Owens, Larry McNight, Rob Hausam, Lorraine Constable, Swapna Bhatia, Jennifer Brush, Greg Gustafson, Elaine Ayres, Chris Melo, Dennis Patterson, Andrew Tores, Michell Miller, Jenni Syed, Tim McNeil, Laura Heermann (Chair), Hans B, Sheryl Taylor, Riki Merricke

Minutes: Goal: discussion of resources where PC and OO have touching boundaries

  • Update on ordersets and careplan (out of CQI)*
  • Observation and clinical impression for capturing clinical notes*
  • Update on the negation topic*

Observation and documents - Michelle presented the background to this issue

  • How to represent notes using observation resources?
    • structure notes
    • Annotations
    • Pure commentary notes
  • Can we use document reference?

The following discussion occurred

  • Clinical impression was intended to document the thought process that leads to a diagnosis - not meant to be used for notes
  • What should be used? -
    • composition resource is like CDA header. It gives the metadata about the document. Used when the system hasn't bother to encode the data
    • Document is always a bundle resource and starts with the composition
    • Doc ref resource is an indexing resource - points to fhir or other structures. Provides the metadata but would not contain the document content
    • A FHIR document sections are other resources. Can have observations and bundle them up with other things.
  • Observations may not be a great fit for narrative data
  • Observation is a point in time statement made about a general statement
  • Diagnostic report points to observation. Having a big blob of text does not make a document
  • Blob of text can be used in the right context - in what context do we deal with documentation?
    • documentation does not mean document. Need to use the appropriate resource for the best fit
  • Need a way to define what these things are in a glossary
  • Examples in patient care minutes Notes -
  • Does observation implied the subject is a patient? If the patient is the subject, should annotation be used because it may not need to be
  • The intent of observation is to be made on anything {patient, group, defice, location} - subject
  • Annotation is appropriate if everyone can modify the resource
  • Where ever we end up putting it, need to be clearly named
  • Physician will be likely to give the note some sort of structure. Attending may need to annotate the resident note - using an observation doesn't feel like the right fit
  • May need to understand how use cases fit into this - e.g. Pathologist uses templates that fits in a bigger report
  • Narrative goes in the narrative of the resource and apply the code - list of allergy will fo in the narrative of a list resource. If there is a narrative and you want it fully encoded FHIR will support this.
  • This is not intuitive for implementers - even for ones participating in HL7
  • From Lloyd - according to Graham, annotation is a separate resource than observation. It already exists. Clinical impression should be used. It's the summary of everything.
  • Clinicians thinks an observation is a synthesis of the provider thoughts.
  • When annotations are associated with other resources, when do we split it and when to we keep it together? RIM made this mistake.
  • Clinical impression is not the only resource intended to contain freeText - every resource is intended to contain freeText.
  • If clinical impression is intended as a summary of everything - if that's the case, it can be used for every note type - H&P and progress notes etc.
  • Freetext yes or no does not do it. Clinical impression has other properties in it that isn't appropriate for use with everything
  • Point made that clinical impression was not meant to be used in this context
  • Clinical impression cam eabout as a collector - samilar to orderset. Don't need to build an artificial place to put these notes.
  • Use cases are on the patient care wiki - these are the places we need to capture data and can't use FHIR to do this
  • Somehow we have to figure out how the improve the problem -

Next Steps

  • Small group to work on it
  • Patient care can continue to work on this - we will take a stab at it to come up with a way to document a note


  • Update on Negation
  • Resource owner: DS
  • Resource: orderset
  • There is a potential overlap between orderset and care plan
  • OO has strong interest in orderSet
  • FMG have not made a ruling on where it sits. Would like to see the draft proposol
  • Related to a broader topic for protocol. It's also a topic for the FHIR workflow project. Will end up with a resource or set of resource. DS has not come to FMG to rule on this yet.

OO, PC, RCRIM will care about it.

  • Timeline - workgroups are asked to let FMG know how comfortable with going forward with a May ballot - see Lloyd's slides for time line.
  • Per Lloyd, This is not a candidate for May ballot.
  • Would it make sense to play with these resource to test out the real use of these resources
  • They've already been to draft for comment ballot. Created profile of the basic resource that went for a draft for comment ballot.

Negation - PC report out

  • Voted that this is a problem
  • Ended up talking about an outline for capturing the use cases. Need to come up with a full list of requirements that we need to solve
  • Desire a need for consistency - not completely clear on the scope of the consistentcy
  • Progress was made
  • Want to ground the discussion and solution on real world use.
  • Will address as part of the Clin-on-FHIR testing

Will keep this spot at next wgm - need to add CDS and RCRIM for the orderset discussion.

Patient Care Wednesday Q4

Present: George Cole Greg Gustafson Nancy Orvis Margaret Ditloff Russell Leftwich Rob McClure Russel Ott Brett Marquard David Parker Sheryl Taylor Rob Hausam Elaine Ayres Bruce Bray Michelle Miller Serafina Versaggi Emma Jones Michael Tan Ted Klein Gay Dolin Jim Case Tom de Jong Susan Matney Jay Lyle Sylvia Tun Susan Barber Marla Albitz Melva Peters Heather Grain Sandy Stuart Carmella Condere M'Lynda Owens Donna Quirk Dan Rutz Lindsey Hoggle Larry McKnight


David Parker presented VA/DoD JET project to identify substances commonly used in allergy records by frequency using 18 million VA & DoD records.

Slides and results will be posted

There are ~200,000 terms in the set of allowable terms (i.e., as defined in C-CDA)

There is very heavy skewing of terms: most of the data is supported by a small number of terms.

RxNorm mapping supports ~70% correlation VA DoD; 85 if IN MIN PIN only

Classes: sulfa 70%, opioids 10.

Group proposes that RxNorm Active Ingredient (IN, MIN) is the functionally relevant level to report

Non-drug substances represented by SCT & UNII. SCT has a better level of granularity but lacks some terms; UNII has more terms but granularity does not support allergen reporting.

The group recommends

  • sharing medications as RxNorm IN/MIN
  • using a very short Class list
  • submitting gap concepts to SCT & remove UNII from the value set


International difficulties. We could use SCT with national drug extensions.

IDMP defines similar levels to RxNorm. 11238 how to identify a substance.

Larry McKnight presented a similar analysis of data from Cerner, using a frequency count from a 45MM patient database

See File:Allergy-counts.xlsx

Even in “standardized” data we see some duplication of terms

Deduplicated first, then mapped to SCT & RxNorm

SCT seems to works better than NDF-RT for classes; NDF-RT also changes.

Counted 40 ways to say “no known drug allergies”

~3% were text entries

Procedure flags are another class: these are relevant at different points of the care process

They are often recorded in the negative as they are used to clear a procedure

How information is captured matters, as patient reports are likely to be less precise.

Foods & environmentals can be much more vague; we need to understand what is the right level of granularity

Next steps: looking for an owner.

Russ Leftwich provided the clinical moment. Recent reports on overrides: 91%. 90% of Penicillin records don’t react when challenged. Drug rash is not immunological. Reactant is often metabolite; which can’t be tested for. Most cross reactions are not immunologic: true immunology is too specific.

What should Patient Care do next?

Rob McClure encouraged the group to think more broadly about the process and how to address problems like false positives and the litigious system that encourages them.

It was suggested that clinical practice does not line up with data standards: domain analysis should assert what is needed to the standards process.

Brett Marqhard moved "To form a minimum universal realm list of substances for allergy and intolerance lists, including negation.

Rob second"

Vote: 8 abstentions, 0 against, 24 for; motion carries.

Patient Care WGM, Thursday, January 14, 2016

Patient Care Thursday Q1

Present: Michael Tan; Susan Matney; Steve Hufnagel, Laura Heerman, David Pyke, M'Lynda Owens; Rick Geimer; Peter Jordan; Lloyd McKenzie; Sunchul Bae; Andrew Torres, Michelle Miller; Dennis Paterson; Larry McKnight; Elizabeth Berger; Viet Nguyen; Jay Lyle; Elaine Ayres; Russell Ott; Emma Jones (Scribe); Donna Quirk; Margaret Dittloff; Mark Shafarman; Galen Mulrooney; Jackie Mulrooney


Topic: CIMI work

Susan Matney presented Model for the skin and wound assessment

Discussion Topics:

  • Skin and wound assessment was put out as a challenge grant by ONC
  • Followed a model development process
  • Assessment and skin risk assessment DAM were thoroughly vetted
  • LOINC codes can be cross walked to observations
  • LOINC used for the questions and SMOMED used for the answers.
  • See LOINC for the panel - See [3]

Skin Inspection Class

  • Jay - most DAM does not bind semantics. This was done here. Proposal is for patient care and CIMI to work together to do this kind of work. What is the vehicle to use?

Is this the right thing to do? If not, what should be done?

  • Susan - what should the scope be? If skin assessment is used for the process then FHIR can used it.
  • For the effort, is this the scope or is the scope the bigger project
  • Suggestion to take an agile approach. Start with the skin inspection and add more pieces.
  • Larry - everything is encoded which does not reflect reality. Suggest using something else that have variation in the data types.
  • Susan - may need to have a high level user story that inspects the skin and include wound along with measurement.
  • Jay - suggestion is to make sure we cover all the data types.
  • Larry - also suggest that calculations are included - like braden scales
  • Jay - include negation
  • Steven- what would be the end product? Normative standard?
  • Laura - this does sound like an experiment. Patient care is looking at what to do as a workgroup.
  • Steven- if an experiment will fit in his PSS.

Topic: Steven Hufnagel presentation of his investigative study

  • Common Health Interoperability Model and practitioner's Guide for HIE Interoperability
  • How to take CIMI detailed clinical model and come up with a common logical model.
  • This is a top down approach. CIMI is a bottom up approach.
  • How to work with a low hanging fruit such as skin assessment.
  • Engineering perspective is to build a tooling environment using all these modeling so an implementer can build an interface that is consistent across implementers.
  • Schedule for this - Jan - Sept 2016 - investigative study, may-aug 2016 demonstrative
  • Will have a practitioner guide that will be draft during the project. Once project completed will become a text book.
  • Will have 2 cycles of DSTU and plan to go to normative ballot in 2019.
  • Request for vote for Patient Care as co-sponsor
  • CIMI has approved an investigative study to do an experiment.
  • Jay - this is ambitious. CIMI wants to demonstrate that their modelling can represent patient care requirements. Don't want to tie simple demonstrations with things that may not be understood. We may need a simple PSS for the CIMI modeling work.
  • Steven- Has worked with EHRwork group doing. Only request is to co-sponsor.
  • Laura - what does co-sponsoring mean?
  • Steven- Can work directly with CIMI
  • Laura - as co-sponsoring means commitment on PC side. PC could take our clinical content, provide to CIMI to model and use that for FHIR
  • Jay - we're the clinical group that is responsible for the content that goes in the clinical resource but we are not responsible for the resourc - it's FHIR resource.
  • Galen Mulrooney - how much of the work is the open group doing and how much is HL7 doing?
  • Steven - they provide the architectural information and HL7 provide the clinical content
  • Laura - CIMI has signed off on the PSS for this so the part with open group has to be worked out with CIMI. The question for PC is if they are willing to take on this project and how to work with the Skin assessment project. What we have is a Question for patient care to decide if we can make the decision to work on this as co-sponsor now.
  • Steven - Motion made for patient care to sign up as co-sponsor. No second. Motion died.
  • Elaine - comment that patient care is stretch very thin
  • Laura - motion for a call with CIMI to discuss the CIMI topic and need to think about how we can move forward. Viet second. no further discussion. 7 abstained. O opposed. 11 For

Topic: Healthconcern and linkage using FHIR

  • Michael Tan - discussion started in Altlanta. Lloyd volunteered to come up with a way to link conditions. Question to Lloyd if further thoughts given about this.
  • Lloyd - some small ideas. Not a lot to share right now. Has been thinking to have an interim target before next working group meeting to get something done.
  • Michael - this can be used by PA for episode of care. This linkage functionality can also be used to link episodes of care.
  • Laura - we need this documented somehow. This can be used to help Lloyd out with what we're asking for.
  • Jay - we need to document all the things that are represented in the DAM that is missing from FHIR
  • Laura - the complications maybe dealing with multiple servers. Might be useful or interesting to explore how the existence of this meets the use cases.
  • Michael - we need our list of what we think is missing. PA need to do the same
  • Lloyd - are there examples of where the gaps are?
  • Jay - can do this from the DAM. Showed sample use cases (concern as attribute of diagnosis, concerns related to each other, temporal changes in disorder, concern as expressor - i.e. family member asserts concern)
  • Larry - discussion on the structuring of the condition in the DAM
  • Laura - this topic will continue in Q3.
  • Lloyd will join PC in Q3

David Tao MU Goal Topic

  • Group: MU 3 is suspended at the end of 2016.
  • Susan read notice from CMS that MU will be replaced with something else.
  • Laura - in this case this is a non-issue. no need to discuss
  • Emma - as implementers some folks will continue to do development so PC need to at least review David's proposal
  • Laura - Motion to discuss David's proposal later on future conference calls. Michael second. No further discussion. 5 abstain. 0 opposed. 14 Approve
  • Keep this quarter next WGM?

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

Present: Elaine Ayres Kai Heitmann Mathias Aschloff Russel Ott M’Lynda Owens Larry McKnight Steve Fine Donna Quirk Vinayak Kulkarni Jeremy Warner Ammon Shabo Austin Kreisler Zabrina Gonzaga Gay Dolin Rick Geimer Andrew Lin Lisa Nelson Brian Scheller William Goosen Anoop Sheth Michael Tan David Pyke Jay Lyle Russ Leftwich (briefly)


Elaine shared a ballot comment on severity

C-CDA 2.1 has six values for severity. PC proposes reducing it to mild, moderate, & severe.

We highlight this one because it may be a patient safety issue.

SD proposed deprecation; PC is amenable to it.

Austin suggests VSAC be in the loop.

SD will move forward and let PC know when so that PC can participate.

Subsequently, SD co-chairs agree to ask VSAC to change it.

Jay: Motion to create a new version of the severity value set that uses the three identified values (mild moderate severe) and does not include the other values.

William amendment: absent? Not a severity observation.

Larry seconds motion

Austin proposes amendment to dispose explicitly stating that the value set is not a patient safety issue.

Larry agrees.

Elaine moves to table until Russ can join.

HL7 does not currently have a patient safety officer; it would go to the chair.

There is discussion about how and whether there is a patient safety issue.

Russ came to the meeting: Yes, the patient safety issue is the fatality not the value set.

Abst: 6; neg: 1; for: 17

ASCO presented a comment on why a lesion organizer may be important to cancer treatment plans

A patient may have many lesions with different locations, sizes, symptoms, responses to treatment. Need to trace individual lesions over time.

There is a comment on cancer document IG

SD: this is new material.

PC: this is a modeling issue, not just a document design issue. It may belong in the concern model.

SD: IHE may have some analogous sections already defined.

SD needs to conduct the ballot comment vote, so we did not vote on the comment.

Goals David Tao sent a Goals position statement; PC & SD to set up a discussion.

Clinical Status

This has been deprecated, and PC believes it is needed.

Problem status removed from problem. Prior decision: a group needs to analyze the interplay the relationships among these statuses.

FHIR addresses this in the Condition resource.

There are three distinct status properties: observation (has the observation been completed?), concern (is it of concern now), clinical (active, resolved, remission, relapse). First two might follow RIM semantics.

Resolution: PC to create a 2.1 DSTU comment for values we want in problem status. SD can determine how to manage it.

Patient Care Thursday Q3

Present: Michael Tan; Jay Lyle; Harold Solbrig; Emma jones; Susan Matney; David Pyke; Larry McKnight; Steve Fine; Andrew Torres; Lloyd McKenzie


What open items remain for DAM completion?

  • Episode of care ?
  • Structure of subconcerns
    • Lesions - possibly
  • Kind - yup
  • Public health - nope
  • Reconciliation - yup
  • Status – not here, not ‘condition’
  • Instance diagrams – yup
  • Linkage: does condition resource support concern domain

We’ll start with reconciliation as it may affect the linkage decision.

Reconciliation of concerns means accepting new, rejecting new, or “merging.”

A merge is not a rollup (used internally to compare similar collections).

A merge results in a new resource containing data from both sources, but it must allow the merger to make decisions, e.g., the older one may have the correct name.

Larry points out that merging results in subconcerns; the merged concern remains available to the viewer as a subconcern.

Jay will detail out the merge cases as instance diagrams.

Emma: lists that are sent lists contain health concern id so folks don’t double-reconcile.

Condition label is on FHIR condition, not list item.

Lloyd: linkage relates conditions only; event containment will remain by reference.

“Condition has both concern and event data” (actually, only “evidence”).

Concern relationship: proposed as linking resource. Not version specific.

  • Linkage is a resource with a time stamp; add timestamp to concernRelationship.
  • If concern semantics are in list, then you have multiple instances of it if you have multiple lists.
  • Not to mention concerns that are not conditions.

Patient Care Thursday Q4



Patient Care WGM, Friday, January 15, 2016

Patient Care Friday Q0: PCWG Co-Chairs meeting

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



Patient Care Friday Q1

(1) Template WG meeting

- PCWG representative(s) to attend
  • Jay Lyle
  • Michael Tan (scribe)

  • Vadim Peretokin adding MDHT templates to HL7 Templates DSTU. Nehta is starting with this tool. This tool does not have import, export functionality. Vadim has to discuss with NEHTA first.
  • Kai wants to organize a tutorial about templates. The issue is, that it is not allowed to promote ART-DECOR. You could give a tutorial about templates using ART-DECOR or Trifolia as a platform. Rick Geimer will send some screens shots from Trifolia to Kai to make it more neutral. The education board is rather hesistant to organize tutorials on tooling.
  • Best practice documentation. Adding templates DSTU to the best practices use cases.
  • Lisa has a problem sorting out the projects in the wiki. Should cleanup the wiki site. Motion to redesign the wiki page to reflect current usage better. Action Kai.
  • PSS set up for a project "templates hardening" to be able to test the schematrons. A test suite would contain sample files to test against the Schematrons. Could SD contribute with sample files? Templates WG and other groups are invited to create sample files and test alternative schematron files. for CCDA 2.1.
  • Structured definition represent profiles in FHIR. Prototyping underway to express CDA template definitions also in Structure Definition . Grahame Grieve demonstrated some work he has done for NEHTA, based on datatypes R1. This would give an opening to put a CDA in a FHIR resource. Previously CDA could only be done with extensions.
  • There is a formal Project CCDA-1 on FHIR to experiment with CDA in FHIR. In the current demo from Rick the transformation requires a manual mapping. Spring 2016 ballot of the CDA in FHIR. Fall 2016 ballot.
  • Requirement to have tooling for FHIR templates.

(2) Agenda: open

- Possible: Clinical Connectathon



Patient Care Friday Q2

Agenda: open

Possible: Clinical Connectathon



Patient Care Friday Q3

Agenda: open

Possible: Clinical Connectathon



Patient Care Friday Q4


No Clinician-on-FHIR session (to be concluded Friday Q3)

PCWG will meet if there are topics of interest