2016-10-24--26 - OBSERVABLE Meeting

• Remember recording!

• Check GotoMeeting number before next session

1. None declared

• If needed

• See attached slide deck
• Review of the document IHTSDO-979 Vital Signs Observables and the ontology available here.
• Should there be a vital signs concept or a refset? See discussion here.

Recommendations

1. The Observables group recommended that due to ambiguity in what is classed as a Vital Sign, that the Vital Sign grouper be retired. Users who want to group vital signs can use the Ref Set Mechanism.

2. That all non-aggregate Vital Signs should be time aspect "Single point in time." The decision changed later in the meeting to "Do not add the time aspect unless it is clear that the observable is Point in Time."
   Noted: Jim Campbell and Scott Campbell opposed this and Mary and Raj abstained form the vote.

3. Retire “peripheral body temperature” to navigational concepts.

4. Heart Rate and Respiratory rate processes (Check with Daniel K regarding this recommendation). The names put forward were recommended by the group, unopposed.

5. Blood pressure model. There where two possible valid models proposed.
   
   1. Inhere in the cardiovascular structure
   2. Inhere in the fluid, with inherent location the vascular system.
      The group recommended that model should be a) however it was noted that should multiple sufficient sets be implemented, then both versions of blood pressure could be used. 
6. Heart rate:

   The group resolved to use "arteries" for the direct site of the pulse

7. The version of the model for Peripheral Oxygen Saturation that was proposed by Daniel Karlsson was recommended (unopposed.)

8. The version of the model for Ankle brachial pressure index was recommended (unopposed)

• Resolution of LOINC-mapping issues

1. Query from RII on modelling of Reagin in serum from Donor. Should it be donor?
    Resolution: We need to send this one back to RII. We should use the supertype for inherent location. 
2. Red Blood Cell Morphology. Should property be modelled as a "Morphology"?
   Morphology not consistently used in LOINC.
   Resolution: these should be modelled as property. = Morphology
3. Erythrocyte sedimentation rate...should this be a process?
   Resolution: Model like LOINC as a velocity with sedimentation moving to technique.

• Next cand. baseline late feb, early march 2017
• All LOINC parts from lab. LOINC → outsourced, RFP
• Current discussion about physiological measurements and vital signs

Cancer checklist observables

• Review of issues related to cancer checklist observables, see attached document.

1. Structure of Lymph node with micrometastases
   Resolution: This question will be discussed offline with Daniel Karlsson. 

2. Size of largest metastatic deposit in lymph nodes 
   Resolution: This is correctly modelled.

3. Status of vascular invasion by tumor (observable entity).
   Resolution: Remove Towards attribute and property as ordinal.

4. Status of tumor treatment effects
   Resolution:Change the FSN to make it about the tumor and change the inherent location to lymph node.

5. MLH1 Promoter hypermethylation
   Resolution: Model as shown

6. Protein as a proxy for gene impression.
   Resolution: Towards will be replaced by direct substance and technique = immunohistochemistry

Glycosylated haemoglobin

Glycosyalated Haemoglobin is used incorrectly as a synonym for Glycated Haemoglobin in the model of this Evaluation Procedure.

Resolution: Change FSN to Haemoglobin A1C and model component = glycated Haemoglobin. Add Glycosylated and Glycated Haemoglobin as synonyms,