SNOMED CT Editorial Advisory Group

Page tree


Time:

1330 - 1700 CEST

1230 - 1600 UTC 


Zoom Meeting Details


Join from PC, Mac, Linux, iOS or Android: https://snomed.zoom.us/j/86050046415?pwd=2lizd89CNKv3jiB7rL8LtWXxdpu6cA.1

Password: 429402

 International numbers available: https://snomed.zoom.us/u/kE7xJTxF5

Attendees

Chair:

AG Members

Invitees: Victor Medina 


Observers:

Monica Harry , Stuart Abbott , Karin Drivenes , Yongsheng Gao , Maël Le Gall , Camilla Wiberg Danielsen , Ole Våge , Marte Rime Bo, Erica Culp , Elaine Wooler , Cato Christian Spook , François Macary , Anne Kongsrud , Julie Vindenes Schultz , STEFAN SCHULZ , Helmut Dultinger , Patrick McCormick , Nashar Karim , Julie Boutin , Marjolaine Gagnon , Nindya Widita Ayuningtyas , Jieun Hwang , Marcos Mireles , Hanne Johansen 



Apologies:

Meeting Files:

Meeting minutes:

The call recording is located

here.


Objectives

  • Obtain consensus on agenda items

Discussion items

ItemDescriptionOwner

Notes

Action
1Call to order and role call

Jim Case

This meeting is being recorded to ensure that important discussion points are not missed in the minutes.  The recording will be available to the SNOMED International community.  Joining the meeting by accepting the Zoom prompt declares that you have no objection to your comments being recorded



2

Conflicts of interest and agenda review

Jim Case

None recorded


3Update on "Abnormality of X"

Jim Case 

Discussion page at Re: Proper terming of "Abnormality of X (body structure)".

After reviewing HPO content that generated this issue, SNOMED will be looking to add these concepts as "Structural abnormality of X" for precision for HPO terms that refer to morphology. If the abnormality is also modeled as a congenital disorder, the FSN will include the word "congenital" in the FSN (e.g. "Congenital structural abnormality of bone"). For those HPO concepts that group both functional and structural abnormalities, they will be left in that form, e.g. HPO - Abnormality of the gastrointestinal tract has subtypes "Abnormal gastrointestinal tract morphology" and "Functional abnormality of the gastrointestinal tract".  Early testing of approaches for these grouper concepts has exposed an issue with the Functional disorder hierarchy and will need to be addressed in later tasks.

The focus for this issue is the terming for HPO concepts that represent structural abnormalities.


Discussion:

Seeking consensus for the terming approach. Do we need to add structural to the acquired concepts as well?  There are around 400+ of these already in SNOMED.  There are a large number of inconsistencies in the HPO terming.  Consider an alternative as morphological anomaly?  Would adding these now result in changes in naming down the road.  What do we we do with the concepts that appear as duplicates when added to the International release.  

There are inconsistencies within the terming of these concepts in the International release as well.  Need clearer guidance.  There are also inconsistencies in the use of anomaly, malformation and deformation.  

The meaning of "abnormal" in a clinical sense is a deviation from normal and may not be definable at a level other than abnormal.  The vagueness makes it difficult to resolve.  The downstream impact argues against adding the HPO terms at this time until the issues in the International release are resolved.  

Additional comments from Monique van Berkum :

Decision:

Hold off adding this content until clear editorial guidance is developed that can be applied to both international content and HPO.


3Bypass graft and shunt modeling proposal update

Victor Medina 

The analysis paper: Review of 48537004 |Bypass graft (procedure)| v0.2.0 is attached to this agenda. 

Previous discussion resulted in suggested edits to the following revised definitions:

  • CONSTRUCTION OF BYPASS. Surgical creation of a passage to carry contents
    around a part of a structure in its normal route and back to a structure in its
    normal route.
  • CONSTRUCTION OF BYPASS USING GRAFT. Surgical creation of a passage, using a
    biological or synthetic material, to carry contents around a part of a structure in its
    normal route and back to a structure in its normal route.
  • CONSTRUCTION OF SHUNT. Surgical creation of a passage allowing contents to
    move from one structure to another, not following its normal route.
  • CONSTRUCTION OF SHUNT USING GRAFT (SYN: INTERPOSITION SHUNT). Surgical
    creation of a passage using a biological or synthetic material, allowing contents to
    move from one structure to another, not following its normal route.
  • INTERPOSITION GRAFT. A surgical reconstruction, using a biological or synthetic graft, of a
    structural defect to allow for a tension-free anastomosis.
  • ANASTOMOSIS. Surgical procedure to create a connection between two tubular or hollow anatomic structures.

It was also proposed that the concepts 360021005 |Bypass - action (qualifier value)| and 424208002 |Shunt - action (qualifier value)| would be inactivated.  However, upon further testing, there were situations where the use of these two concepts were useful.  Alternative solutions are provided.

Discussion:


Decision:

Based on discussion, a review of the proposed model by SI to ensure that all of the proposed RGs are needed to express the meaning of bypass and shunt procedures is needed.  A task that provides examples and the results of the application of the proposed model will be made available to the EAG members to review and provide comments.  A decision to keep or inactivate the unidirectional shunt grouper terms will be made following review of the proposed modeling.


4Endoscopy and endoscopic procedures

Victor Medina 

An analysis document "Endoscopy, endoscopic procedure and related procedures" is attached for discussion.  The consistent distinction between an endoscopy and an endoscopic procedure is inconsistently represented as the initial distinction was made based on whether a procedure other than visualization was performed.  This is a difficult distinction to make as it is unknown when an endoscopy is performed whether an additional procedure will be added (such as a biopsy).  The proposal seeks to eliminate this false distinction as well as propose other modeling changes for this hierarchy.

Discussion:

The history of the distinction between Ednoscopy and Endoscopic procedure was presented.  As the new editorial guidance for this area of the terminology now requires the addition of an inspection RG, as opposed to being implicit for endoscopic procedure may make the need for the current distinction between endoscopy and endoscopic procedure mot, as all endoscopic procedures will contain an inspection RG.  The more constrained use of USING ACCESS DEVICE will be discussed.



Decision:


5 Replacement procedures

Victor Medina 

Jim Case 

An analysis document "Review of the replacement procedure hierarchy" is attached for discussion.  The issues related to the use of a single concept "Replacement - action" to model replacement procedures are discussed an alternative modeling constructs proposed. Consistent use of the Procedure site attributes for Replacement procedures is also discussed. 

Discussion:

Continued to a future call


Decision:


6Conversion procedures

Victor Medina 

Jim Case 

Conversion procedures primarily in the orthopedic space have been historically represented as two separate concepts: i.e. "Conversion from X" and "Conversion to X".  This mimics the approach taken in OPCS4 and was done ostensibly to prevent a large combinatorial explosion of terms that represented both aspects of a conversion procedure.  Recently SNOMED International has received requests for combined conversion procedures (e.g. Conversion from X to Y").  There are currently:

39 "conversion from" orthopedic concepts

44 "conversion to" orthopedic concepts

There are also conversion procedures related to laparoscopy where a procedure is converted from laparoscopic to open, as well as other procedures unrelated to the orthopedic domain.

The question to be answered is whether SNOMED should continue to represent orthopedic conversions using two concepts or whether we would allow combined conversion procedures?

The affected concepts are listed in the attached worksheet.

The number of combined concepts would not be a cross product of conversion from and conversion to concepts (i.e. 39 x 44 = 1716) as many are site specific.  For knee, for example it would be 4 x 6 = 24 combined concepts; for hip it would be 6 x 6 = 36 combined concepts, etc.

Discussion:

Continued to a future call


Decision:



CellulitisJim Case 

Cellulitis is most commonly caused by a bacterial infection; however, there are cases where it can be caused by fungal infections or even non-infectious causes.  The majority of definitions of cellulitis refer to it as a result of a bacterial infection as that is by far the most common clinical presentation.  SNOMED currently has 234 concepts under 128045006 |Cellulitis (disorder)| including a subconcept of 402929007 |Bacterial cellulitis (disorder)|.  Most concepts just state "cellulitis of X" resulting in the lack of subsumption under "Bacterial infectious disease".  It has been noted in the literature that the use of the term "cellulitis" in cases other than bacterial infections is a misnomer.  In many cases the causative agent is unable to be identified, so the disease is diagnosed on clinical signs and response to therapy.  Given the rarity of non-bacterial cellulitis conditions:

  • Should SI infer that the common clinical usage of cellulitis as a bacterial infection is the meaning of "Cellulitis of X" concepts? 
  • If so, would renaming and remodeling be considered a substantive change requiring inactivation and replacement?  
  • If the FSN is changed to "Bacterial cellulitis of X", can the PT of "Cellulitis of X" be retained?

It is proposed that the current top level concept 128045006 |Cellulitis (disorder)| be retained in order to group the non-bacterial types under a common header.

References:

https://jamanetwork.com/journals/jama/article-abstract/2533510 "Cellulitis is a bacterial infection of the skin, presenting with poorly demarcated erythema, edema, warmth, and tenderness."

https://www.njmonline.nl/cntpdf.php?t=i&id=210#page=6 "Cellulitis is a bacterial skin and soft tissue infection which occurs when the physical skin barrier, the immune system and/or the circulatory system are impaired."

https://www.medical.theclinics.com/article/S0025-7125(21)00047-X/abstract "Cellulitis is a skin infection typically precipitated by entry of bacteria through a breach in the skin barrier"

https://www.clinicalkey.com/#!/content/clinical_overview/67-s2.0-7dabbf3c-ea31-4472-8aea-72dc2b003cab"Cellulitis is a spreading, acute bacterial skin infection of the dermal and subcutaneous tissue that presents clinically with skin erythema, calor, induration, and tenderness"

https://pmc.ncbi.nlm.nih.gov/articles/PMC5466351/ Fungal orbital cellulitis

https://pmc.ncbi.nlm.nih.gov/articles/PMC10058517/ Aseptic orbital cellulitis


Discussion:

Continued to a future call. Comments from EAG members below indicate a preference for maintaining the distinction between cellulitis of unspecified cause (whether bacterial or not) as there are many instances in which the offending organism is not identified.  This would be consistent with the open world view that the absence of a statement in an ontology does not imply that the statement is false in the real world (i.e. not bacterial).  This would leave the subhierarchy of 402929007 |Bacterial cellulitis (disorder)| as only subsuming concepts where the specific bacteria or bacterial class is specified in the FSN .

The problematic concept 130021000119100 |Cellulitis of nasal mucous membrane (disorder)|, which is modeled with a finding site that is not consistent with the text definition on 128045006 |Cellulitis (disorder)|, is being investigated as a potential erroneous concept as no authoritative references related to this concept have as yet been located.



10

AOBEAG



11Next meeting

Jim Case 
















15 Comments

  1. John Snyder

    Item 5: Replacement Procedures

    • Agree with proposal to inactivate 282089006 |Replacement - action (qualifier value)| and use two RG (i.e. removal and introduction) as the preferred modeling.
    • Prefer to avoid the use of GCI as it can have unintended consequences at the extension level and in some cases may facilitate the use of incorrect modeling in extension concepts to facilitate proper classification
    • Under Item 2: Proposed model for the replacement of a body structure - "second RG" section.
      • Can procedure site - indirect be consistently modeled using <<2969000 |Anatomical space structure (body structure)| instead of the "the structure where the replacement would be introduced to"? I interpreted this quoted statement to mean (body structure), but the document points out that using the body structure concept implies the introduction is occurring in the (body structure) and not in the space occupied by the body structure, unless I am misunderstanding the text and/or intent.
    • Any gaps identified in the <<2969000 |Anatomical space structure (body structure)| subhierarchy identified during remodeling should require the creation of new concepts to fill those gaps as part of the scope of work. This will help to minimize the requests for new concepts as national extensions work to update their modeling for content in this arena.

    This response may change upon hearing discussion on this topic during the meeting and will be edited appropriately.

  2. John Snyder

    Item 4: Endoscopy and endoscopic procedures

    Agree with General proposal for the << 363687006 |Endoscopic procedure (procedure)| hierarchy review.

    Request clarification regarding the consistent use or removal of RG containing "Insertion - action" when modeling Intraluminal endoscopy concepts using either a natural orifice or pre-existing surgically constructed opening (e.g. gastrostomy). I could be misunderstanding or have missed it, but I don't see an explicit statement to address this under either proposal item 5a (page 14) or Proposal item 9b (page 15).

  3. John Snyder

    Item 6: Conversion procedures

    What is the use case and/or technical limitation driving the requests for combined procedure concepts? 

  4. Victor Medina

    John Snyder Thanks for your comments. 

    As stated in the review: 8. Lack of consistency in use of |Method|: Surgical insertion versus |Method|: Insertion. Currently, the intraluminal endoscopic procedures are considered as non-surgical, in consequence, Insertion procedures through Intraluminal endoscopic procedures should be modeled with 257867005 |Insertion - action (qualifier value)|.

    Intracavitary endoscopic procedures are considered surgical procedures, so in modeling Insertion procedures related to these types of endoscopic procedures, the 425362007 |Surgical insertion - action (qualifier value)| should be used.

    1. John Snyder

      Victor Medina Thanks for the response:

      Does the modeling for intracavitary endoscopic procedures need to change if for example the gastrostomy creation takes place during a different temporal episode than the endoscopic procedure that uses the gastrostomy for access purposes? 

      1. Victor Medina

        John Snyder I'm not sure I fully understand your question.

        From what I understand: Those are two different procedures with different purposes. For an Endoscopic Percutaneous Gastrostomy, the Endoscopic procedures should be considered and modeled as an Intraluminal Endoscopic Procedure with additional model for the creation of the Gastrostomy. In the current model for 235159007 |Percutaneous endoscopic insertion of gastrostomy tube (procedure)|, the Inspection using an Endoscope is not modeled. However, it should be considered an Intraluminal procedure from the Endoscopic perspective.

        The Endoscopic inspection through the Gastrostomy should be considered an Intraluminal endoscopic procedure through a non-natural orifice and the proposal is to add an additional Role Group with the Procedure approach attribute stating the introduction of the endoscopy through the Gastrostomy.

        Hope I've answered your question.

  5. John Snyder

    Cellulitis:

    May the following be included in the scope of work:

    • Evaluate the continued need for 717857009 |Acute cellulitis (morphologic abnormality)| or should this be inactivated and the modeling updated to include the clinical course attribute at the concept level. Stating both, as we see on 1197179009 |Acute cellulitis (disorder)|, seems to be unnessarily redundant unless it serves a purpose during classification?
    • If these changes are made, will concepts modeled with 190280006 |Phlegmon (morphologic abnormality)| 1332475002 |Necrotizing cellulitis (morphologic abnormality)| continue to classify properly?

    I will defer to the SME's on the EAG, however, this change seems too substantive to allow existing concepts to be renamed and allow for the synonym of of "Cellultitis of <x>" to be added to "Bacterial cellulitis of <x>" concepts by default.

    I agree with the purposal to retain 128045006 |Cellulitis (disorder)| as Concepts like 240780005 |Mucormycotic gangrenous cellulitis (disorder)|, which is a fungus, will be displaced from its current position without it.

    1. Jim Case

      The vast majority of cellulitis cases are acute in terms of sudden onset; however, there are instances of recurrent cellulitis (https://link.springer.com/article/10.1007/s11908-014-0422-0).  This would indicate that it would be preferable to represent the acute and/or chronic nature of the disorder using the CLINICAL COURSE attribute as opposed to precoordination into the morphology.  

      1. John Snyder

        This is purely a technical exercise, as I do not want to go down the semantic meaning of "acute" rabbit hole.

        Would like confirmation that the classifier is interpreting the following three potential stated logical definitions as equivalent so that classification is completed properly.

        Concept A:

        • Clinical course: Sudden onset AND/OR short duration
        • Finding site: structure of <x>
        • Associated Morphology: Acute Cellulitis


        Concept B: 

        • Clinical course: Sudden onset AND/OR short duration
        • Finding site: structure of <x>
        • Associated Morphology: Cellulitis


        Concept C: 

        • Finding site: structure of <x>
        • Associated Morphology: Acute Cellulitis


        As long as these three logical definitions are all treated as equivalent by the classifier, then it is likely that the preferred modeling will be that of Concept C because it requires fewer stated attributes. If the classifier does not see these three logical definitions as equivalent, then we need to standardize to one logical definition. 


        1. Jim Case

          Thanks John,

          We will need to look at the various ways Cellulitis concepts are currently modeled.  I agree that conflating clinical course with morphology in this case is not the optimal approach.  A more subtle issue is the correct assignment of the clinical course value (i.e. 385315009 |Sudden onset (qualifier value)| vs. 424124008 |Sudden onset AND/OR short duration (qualifier value)|)

  6. Victor Medina

    John Snyder 

    About your first comment for Item 5: Replacement Procedures - Under item 2. You understood the statement correctly. This is a difficult to tackle issue.

    While many devices are introduced into cavities, other are introduced into non-cavitary structures (e.g. 105111000220109 |Replacement of dental implant prosthetic screw (procedure)| where the Procedure site - indirect is 28035005 |Tooth, gum, and/or supporting structure (body structure)| or in the case of the replacement of heart valves where the Indirect site is the Fibrous valvular annulus).

    If this model is agreed, a one-by-one review would be needed, and as you commented, we may need to create new concepts for modeling these procedures. Even though, modeling of such spaces/structures may be complicated for some "introduction" procedures (e.g. arthroplasty)


  7. James R. Campbell

    In reponse to: "Should SI infer that the common clinical usage of cellulitis as a bacterial infection is the meaning of "Cellulitis of X" concepts?", speaking as a clinician I would respond No since we are taught that the condition is basically an inflammatory disorder and can have many etiologies.  I believe the preponderance of bacterial infection in the literature reflects the case load brought to the physician for attention and that should not cloud SNOMED science. Following John's comment on "acute cellulitis", I would further wonder out loud whether "Cellulitis" is necessary as a morphology since cellulitis is reserved clinically for the subset of inflammatory conditions of soft tissue structures?

    1. Feikje Hielkema-Raadsveld

      https://pmc.ncbi.nlm.nih.gov/articles/PMC9074255/ Cellulitis is an infection most commonly caused by bacteria and successfully treated with antibiotics. However, certain patient populations, especially the immunocompromised, are at risk for fungal cellulitis, which can be misidentified as bacterial cellulitis and contribute to significant morbidity and mortality.

      If SNOMED equates cellulitis with bacterial cellulitis, it will contribute to such misidentification. Obviously antibiotics will not cure a fungal infection so this misidentification threatens patient health. 

    2. Jim Case

      Cellulitis describes a particular morphology, so I think it is a worthwhile concept as it is well understood clinically. 

  8. Victor Medina

    After our discussion in Oslo. Several changes were made to the proposal for modeling of the Bypass and Shunt procedures (Item 3):

    • In "3: Model proposal by type of procedures, a. Model for construction of Bypass, iii. Add a grafting RG when a << 260667007|Graft (substance)| is used." :
      • After review, the use of this additional Grafting RG would be eliminated from the proposal.
    • In "3. Model proposal by type of procedures, b. Model for construction of Shunt, v. ..For these procedures, use two additional RGs with Method=<<129325002|Introduction-action(qualifiervalue)| for each involved structure...":
      • After review, the use of these additional Introduction RGs would be eliminated from the proposal.


    For the Endoscopy procedures (Item 4), the Procedure site - direct (attribute) within the Incision RG for Intracavitary endoscopic procedures:

    • Aligned to the Body structure used in the model for Laparoscopy (procedure): 59380008 |Anterior abdominal wall structure (body structure)|. The Procedure site - direct (attribute) for the Incision RG in 14671001 |Thoracoscopy (procedure)| , will be corrected to 78904004 |Chest wall structure (body structure)|. The current Procedure site - direct (attribute) is 51185008 |Thoracic structure (body structure)|, which is incorrect.
    • In the case of 13714004 |Arthroscopy (procedure)|, modeling will be reviewed based in the anatomy of each involved joint and the existence of a 32551008 |Structure of articular capsule (body structure)|. When the joint has a capsule, it will be the target value for the Incision RG. When the joint does not have a capsule, the target value for the Incision RG would be the specific subtype of 39352004 |Joint structure (body structure)|.

    This changes and will be available for review and comment in a task which will be announced soon.