Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

Scribe: Riki Merrick / Lorraine Constable

  • Joint with II, CG, BRR
  • Agenda Review
  • Same session for Syndey for CG, II and BRR - YES!
  • Healthcare Product Recap
  • Cancer specifications for Genetics (M-codes, v2 IG, and FHIR)
  • Clinical Genomics
    • close to publishing Genomics reporting Implementation guide
    •  will Changes to observation in R5:
      • will replace current extensions
      in 
      • in R4 that will result in R5 deprecations 
    • technical correction on examples
    • questions about searching and computational observations
  • II
    • Specifics around DocumentReference and Media for tomorrow
    • co-sponsor updates
  • BR&R
    • no updates


Notes:

  • Healthcare Product recap
    • Slide from Hans
    • Resources owned by OO need clarity of scope boundaries to other resources
    • Need to up FMM levels up and improve statements in introduction
      • Focused on darker green
      • Less on lighter green
      • Orange and mauve don’t have resources yet
        • NutritionIntake
        • NutritionProductIngerdientLable
        • BiologicalylDerivedProductRequest – or does MedicationRequest cover this space
      • There have been several gForge trackers to reference BiologicallyDerivedProduct
      • There is a word document summarizing the introductions to all the covered elements – expect distribution for review in the next few weeks
      • MedicationStatement – transitioning to MedicationUseStatement
        • Based on
        • Becky will meet with Rx WG to make MedicationUseStatement more generic, if possible
      • How do resources around
        • when attached to patient =
          • MedicationRequest
          • DeviceReuqest
        • Move medication without a patient
          • suppyRequest
        • Difference between Substance and SubstanceDefinition
          • substanceDefinition is about the characteristics of the substance
          • Substance is instance in a specific "barrel" before used in instance of medication
          • MUST have series of use cases to clearly describe the boundaries between these use cases
            • Ideally like in ClinFHIR define the use cases and list all the
          • FHIR is not being clear about the information models we are using – we are using the resources
          • In OO related resources we will have a lot of definitional resources, while – BRR owns the substanceDefinition
            • We are trying to define patterns for resources that are similar
            • What is the difference between definition and knowledge in the resource names?
              • Knowledge would be interpretation / comment on something
              • definitional is defining attributes that don’t change
              • the definition patterns in the CDS area are actually more knowledge may need to discuss this on larger organizational level
            • What all is comprised by biologicallyDerviedProduct – why 2 boxes?
              • Use Supply request for tissue (not likely, because the supply request will be for a specific patient when it is identified) vs blood product (can deliver to blood bank on ward, before it is assigned to patient)
              • May need definitional resource here as well, when you want to describe what kind rather than a specific instance
              • Is breast milk a biologicallyDerivedProduct?
                • Had discussion around that, only used as expressed breast milk – so treat similar to food
                • Is there need to test compatibility, then consider biologicallyDerivedProduct
              • Calls are Mondays 3 – 4 PM – ADD LINK to confluence
            • CG update:
              • Publishing Clinical Genomics Reporting FHIR IG soon after 2 cycles
                • The will replace all extensions in observations along with valueset, so need to deprecate these – as well as the examples
                  • One of those is a very bad example – could we do that as technical correction in R4? = example 3 = ADD LINK:
                    • This has only a performer, just text for a note, but no observation.value
                    • Deadline for technical corrections is next week Friday = 9/20
                  • A lot more examples will be in the Clinical Genomics reporting guide – could we point to those from here?
                  • Is there a way to label the examples and extensions to indicate that they will be deprecated in R5 – just like it was done for the profile with
                  • Lorraine will follow up with Grahame
                  • CG will need to get the list of items these corrections apply to
                  • And then send the publication request for technical correction to Wayne
                • In the old deprecated profiles extenstion.geneticGene there is valueset binding to HGNC
                  • This is owned by OO – can we do that for R5? Yes – enter a gForge items – ensure that this one MUST make it into
                • Have observation profile that groups observation should we reference just the grouper resource, or do we need to have the observations that are part of the groups
                  • DiagnosticReport vs observations for these kind of groups
                  • DiagnosticReport can include nested panels
                  • Make examples for this and send to zulip chat and send that link to the OO co-chairs
                • Tumor mutational burden – is a computational result an observation
                  • In observation method is codeable concept – but since this is often a longer text that is described in a protocol but may include alterations for this instance – create a tracker for this
                  • What is the reason for method cardinality 0..1 – should be 0..* maybe
                    • Use case is listing the analyzer, software used there – in the V2 world there is OBX-18 for equipment
                      • So device, but that is also 0..1 – should be 0..*
                      • Bring this up with FHIR-I to get feedback on what we can do normative resources
                    • Could we use a reference to SOP / protocol = planDefinition?
                      • Make method a choice:
                        • Codeable concept (existing)
                        • Embedded text
                        • Reference to planDefinition (or something more detailed)
                      • OR create an extension for NEW element methodProtocol to cover these situations
                    • For publication – how to make sure the editor is listed – Jamie has worked hard on that
                      • IG Publisher should have a section for editors/ authors etc – Riki will make that gFOrge
                    • II
                      • II Co-Sponsors updates:
                        • 1453:
                          • ADD the link here
                          • Set up
                        • 1010:
                          • Catalog project – still ongoing and getting resource proposal approved – (catalogEntry)
                          • Focus so far has been lab and pharmacy
                          • Calls are Fridays 12 -1 PM EDT
                          • II would like to have connectathon track in May 2020
                          • LINK TO WIKI that has the use cases
                          • II to add their use cases here
                          • Schedule a call specifically for II topics – Francois to do
                        • ImagingStudy is not good at referencing a specific image of the study
                          • Would you need to get down to the region – may be use DICOM instead of FIR resources
                          • Also linking to specific observations in
                        • IHE has used documentReference to replace XDS
                          • Why not use DiagnosticReport for this with the actual observations
                          • It is used as an indexing for FHIR artifacts
                          • Next cycle will need to work on radiology reports
                            • Once we clean up HL7 base resources ow to provide feedback to IHE
                            • Discussion in OO tomorrow Q2
                          • category and diagnosticReport.category
                            • In observation the list of values is very broad
                            • In diagnosticReport has more granular list than observation, but is missing quite a few terms
                            • It is exemplar binding – could we make it preferred?
                            • Regenstrief is encouraging to use LOINC parts form the class part for category
                              • Will publish this hierarchy and include a category for labtests as a class part in next LOINC
                            • Category will be used for different purposes, so category can be repeated, but it will be used very differently by different organizations for the same purpose – in practice, so is it worth the work
                            • If we can provide codeable representations for that
                            • One reason for difference was that diagnosticReport was used for surgical report, though that code is not in that value set

Adjourned at 12:20 PM EDT

Q3

Chair:  ???

Scribe:  Riki Merrick

...