Page History

•  Approve minutes from 06/27/2016
  
  

Discussion on the Use of “+” and “&” in LOINC Components and their representation in SNOMED CT as part of the LOINC/IHTSDO alignment

Disjunctive components

Daniel presented an update on the number of disjunctive components (~814). The plus sign has a different meaning in susceptibiltiy tests vs. other observables and were not included.

Presented 4 alternatives for representing these disjunctive concepts. Review of the comments on the discussion page. It was suggested by GRE that these might be handled through the use of an extension to handle this specific case without putting the content in the core.

The default has been to add these concepts using combination concepts and these are used in other projects (e.g. allergies).

PAM: What is the use case and how will this be used in clinical practice and do these add value? These may be useful in the laboratory environment, but not necessarily in reporting to clinicians/patients. Will this set a precedent for adding additional disjunctive concepts?

BGO favors keeping them in the core as primitive and asserting the children directly.

Current options:

1. Model these as primitives in the core.
2. Create an extension that can be used in conjunction with the core.
3. Make the 700-800 observable primitive, no component.

Suggested that option 3 might be used in the immediate short term to address the need to get the preview out.

Are numeric ranges separated by dash, em-dash, or "to", or "from"?

e.g. |128131000119102|History of low birth weight status, 2-2.5kg (situation).
Is there supposed to be a space between the number and the unit (e.g. 2mg or 2 mg)?

Existing content does not use a space, but use of a space has been standard editorial policy.

Should we use a reference such as AMA or Chicago Manual of Style (and there are probably others) to provide guidance?

Discussion: Use of dash and em-dash has caused problems in the past with release files. It is better to actually represent the words, e.g. "2 to 2.5". To improve searching, it would also be better to separate them by a space. This would require revision of a number of existing concepts.

Question: Should each number in the range have a following unit?

Question: Should the same number of significant digits be represented in the FSN?

Final representation: History of low birth weight status, 2 to 2.5 kg (situation).

Additional representative examples will be posted on a new discussion page for testing.

Adding diagnostic imaging of multiple body sites and procedures

Examples:

Ultrasonography of abdomen and ultrasonography of pelvis with transrectal ultrasonography (procedure)

Ultrasonography of pelvis and obstetric ultrasonography with transvaginal ultrasonography (procedure)

Ultrasonography of knee and doppler ultrasonography of vein of lower limb (procedure)

Discussion: Are these convenience order sets or have a particular importance to be kept as a unit?

Is this similar to a laboratory panel?

CMT has been submitting same modality with different body parts?

This opens the door to a potentially very large set of combined procedures. What is the line that can be drawn to include concepts like this, but does not open the flood gates.

Consider adding them where the procedure types are related in some way, or the sites are related.

Summary of previous discussion: (see 2016-01-22 Editorial AG Conference Call)

• Current FSN naming conventions result in "non-user-friendly" descriptions. Creation of another description without the semantic tag adds no value
• Many existing FSNs would be ambiguous without the addition of the semantic tag.
• Users often do not request "preferred terms" so rely on editors to provide appropriate descriptions.
• Historically, many FSNs were glorified preferred terms
• This is an issue that only affects English versions as translation provide terms of use in their languages
• Challenge to editors is when to create a different preferred term if the FSN term is not user-friendly.
• This has an impact on acceptance of word order variants as preferred terms.

Assumptions:

• Semantic tags provide value in disambiguating concepts in different hierarchies with the same description (e.g Swab (physical object) vs. Swab (specimen))
• Current FSN naming conventions do not always provide user friendly descriptions, so duplication of these descriptions adds no value.

Discussion: Bruce provided a brief discussion on the recommendations for naming of the ECE patterns for the combined disorders.

GRE: Prior to the WB the matching descriptions for these terms were automatically created. There is a history of some terms not having matching descriptions for the FSN.

BGO supports the continued use of the matching description and the creation of a "user-friendly" description. This may or may not be the preferred term. Consensus was reached on this.

Naming pattern for FSN:

• X co-occurrent with Y --→ X and Y
  • Asthma co-occurrent with allergic rhinitis --→ Asthma AND allergic rhinitis
  • Asthma co-occurrent with chronic obstructive lung disease
    • Asthma AND chronic obstructive lung disease
    • Asthma-chronic obstructive lung disease overlap syndrome
• X co-occurrent and due to Y --→ X and Y due to Y
  • Intestinal obstruction co-occurrent and due to hernia --→ Hernia AND intestinal obstruction due to hernia

•  Consensus on options: 1) require matching description (SCA give error) or 2) matching description optional (SCA gives warning)
•  Bruce Goldberg To take this naming issue back to the ECE for discussion and recommendation.

During the congenital disease revision project, it was determined that the differentiating feature of Developmental anomalies vs. acquired anomalies was the process/origin of the anomaly. Historicvally this resulted in the creation of the Developmental anomaly (morphologic abnormality) subhierarchy which essentially duplicates content within the Morphologically abnormal structure hierarchy with only the addition of "Congenital". This has caused substantial issues with the Congenital disease revision project. The following is proposed:

1. In order to inactivate the “Congenital X (morphologic abnormality)” subhierarchy, must have a way to identify the non-congenital morphologies as a result of a developmental process.  The new proposed model is that the range for PATHOLOGICAL PROCESS be expanded to include “Pathological development process”(existing concept not in approved range).  This relationship would only be used where the abnormality is the direct result of abnormal development.  The primary use would be for children of “Congenital malformation”.
2. Congenital deformities, which may not be the result of a developmental process (mechanical or traumatic), would not have the PATHOLOGICAL PROCESS relationship added.
3. Concepts that currently have the ASSOCIATED MORPHOLOGY = Developmental anomaly would be replaced with a relationship group of ASSOCIATED MORPHOLOGY =  “Morphologically abnormal structure” and PATHOLOGICAL PROCESS = “Pathological developmental process”.

This has been preliminarily tested and no unexpected, untoward or erroneous inferences were identified.

Idiopathic disease is currently a primitive concept that is inconsistently used. As a working definition, Idiopathic = "relating to or denoting any disease or condition that arises spontaneously or for which the cause is unknown." While it is arguable that Idiopathic is a "process" per se, one could equate it to "unidentified process" that has clinical value. There are currently around 200 concepts with Idiopathic in one or more descriptions (IHTSDO-799).

Propose to test the extension of the range of PATHOLOGICAL PROCESS to include

Space Actions

We are meeting for a full day in New Zealand. Need substantive topics to move forward.

AG Member travel FAQ: https://ihtsdo.freshdesk.com/support/solutions/folders/4000008052

•  Solicit additional agenda topics for Face-to-face meeting
•  Verify travel requests have been submitted

JCA is traveling Sept 10-23. Meeting scheduled for Sept 26