Table of Contents

1 Glossary............................................................................................................................................... 5

1.1 Domain Terms................................................................................................................................... 5

2 Introduction.......................................................................................................................................... 6

2.1 Purpose............................................................................................................................................ 6

2.2 Audience and stakeholder domain................................................................................................... 6

2.2.1 Input from stakeholders............................................................................................................... 6

2.2.2 Degree of consensus on the statement of problem..................................................................... 6

3 Statement of the problem or need................................................................................................... 7

3.1 Summary of problem or need, as reported....................................................................................... 7

3.2 Summary of requested solution........................................................................................................ 7

3.3 Statement of problem as understood............................................................................................... 7

3.4 Detailed analysis of reported problem, including background.......................................................... 8

3.4.1 Grouper Concepts....................................................................................................................... 8

3.4.2 Morphologic Abnormality........................................................................................................... 9

3.4.3 Modelling Issues....................................................................................................................... 10

3.4.4 Fully Specified Name............................................................................................................... 12

3.4.5 Descriptions Requiring Review................................................................................................. 13

3.5 Subsidiary and interrelated problems............................................................................................. 13

3.5.1 IHTSDO-873 (artf222436) Concept additions to "Cyst of breast"............................................... 13

3.5.2 IHTSDO-33 (artf222501) Add synonym: Breast cyst.................................................................. 14

3.5.3 IHTSDO-89 Congenital Occurrence vs. Congenital Morphology.............................................. 14

3.5.4 IHTSDO-960 Modelling of Disorders of Acquired Body Structures............................................ 15

4 Risks / Benefits................................................................................................................................. 16

4.1.1 Risks of not addressing the problem.......................................................................................... 16

4.1.2 Risks of addressing the problem................................................................................................ 17

5 Requirements: criteria for success and completion................................................................. 18

5.1 Criteria for success/completion...................................................................................................... 18

5.2 Strategic and/or specific operational use cases............................................................................ 18

6 Solution Development..................................................................................................................... 19

6.1 Initial Design.................................................................................................................................... 19

6.1.1 Fully Defined Grouper Concepts............................................................................................... 19

6.1.2 Modelling Issues....................................................................................................................... 20

6.1.3 Descriptions.............................................................................................................................. 20

6.1.4 Morphologic Abnormality......................................................................................................... 20

6.1.5 New Content............................................................................................................................. 21

6.1.6 Significant design or implementation decisions / compromises............................................... 21

6.1.7 Evaluation of Design................................................................................................................ 21

7 Recommendation............................................................................................................................ 22

7.1.1 Detailed design final specification........................................................................................... 22

8 Quality program criteria................................................................................................................... 23

8.1 Quality metrics................................................................................................................................ 23

8.1.1 Quality metric 1........................................................................................................................ 23

8.1.2 Quality metric 2........................................................................................................................ 23

9 Project Resource Estimates.......................................................................................................... 24

9.1 Scope of construction phase......................................................................................................... 24

9.2 Projection of remaining overall project resource requirements...................................................... 24

9.2.1 Expected project resource requirement category..................................................................... 24

9.2.2 Expected project impact and benefit....................................................................................... 24

9.2.3 Indicative resource estimates for construction, transition and maintenance:............................ 24

1 Glossary

1.1 Domain Terms

2 Introduction

2.1 Purpose

The purpose of this content project is to review 441457006|Cyst (disorder) and descendant concepts and evaluate the proposal to incorporate grouper concepts into the hierarchy in order to improve navigation.

2.2 Audience and stakeholder domain

The audience for this document includes all standards terminology leaders, implementers and users of SNOMED CT.

A further significant audience is the community of SNOMED CT terminologists that may be requested to implement the recommended specification this includes the IHTSDO Content Team.

2.2.1 Input from stakeholders

The request to review 441457006|Cyst (disorder) and descendant concepts originated from an internal quality review process commissioned by IHTSDO. There are no direct stakeholders awaiting the outcome of this particular content project therefore appraisal of the proposed solution was sought from within the Consultant Terminologist community.

It is noted that there are approximately 28 requests in the IHTSDO request portal (SIRS) for new concepts that will fall into the scope of this content project.  These new requests are from a variety of National Release Centres and users of SIRS.  However these requests appear to mostly relate to requests for laterality and this issue is being addressed in a separate large-scale content project artf223747 finding/disorder with laterality. 

There are currently no restrictions placed on the addition of new content to the hierarchy 441457006|Cyst (disorder) assuming any new requests meet the criteria for inclusion and are not the subject of any other content project (for example pre-coordination patterns).

2.2.2 Degree of consensus on the statement of problem

There is a requirement to update SNOMED CT as part of an ongoing process of content development with the aim of improving the overall quality and useability. This content project is required to improve an area of content identified as a priority for development after quality review commissioned by IHTSDO.

3 Statement of the problem or need

3.1 Summary of problem or need, as reported

In 2011, IHTSDO commissioned the company Computer Science Corporation (CSC) to undertake a quality review of some areas of SNOMED CT.  The resulting report ‘CSC Quality Review Editorial Work Version 1.0’ recommended a ‘rework of the hierarchy 441457006|Cyst (disorder).’

CSC recommended the rework as “the current state of the hierarchy is of concern from a clinical and navigational aspect.”

It is noted that the content of SNOMED CT incorporates contributions from legacy concepts from early terminologies (SNOP, SNOMED International, SNOMED RT, Read 4 Byte Set, Read Version 2 and Clinical Terms Version 3) and also Classifications (ICD9 & ICD10 via Read Version 2) and that this diversity has presented a challenge in organising some of the hierarchies. 

The document also notes that the extent of work to be undertaken may well be influenced by the degree to which SNOMED CT is viewed as an interface terminology, in some areas (e.g. Clinical Data Capture forms) it is possible to constrain pick lists with subsets or a reference terminology whilst in other contexts (e.g. generic problem entry screens) hierarchy navigation has been difficult to avoid completely.

3.2 Summary of requested solution

The report ‘CSC Quality Review Editorial Work Version 1.0’ (page 23) details the following recommendation:

“Under 441457006 Cyst

Concept should be organised with navigational concepts on a system basis e.g Respiratory, Gastrointestinal, Urinary, Genital and also Infective/parasitic

399294002 Cyst of breast should be a child of 89164003 Breast lump not 440299000 Mass of thoracic structure

Children of 399294002 Cyst of breast could be organised on Benign/Malignant

Also could have new child concepts:

Multiple (cf solitary) Complex (cf simple)”

Note: The issue highlighted above with the concept 399294002|Cyst of breast (disorder) has been rectified since the CSC document was produced in 2011.

3.3 Statement of problem as understood

In the international edition of SNOMED CT there is some inconsistency with naming and modelling of concepts that are descendants of 441457006|Cyst (disorder). Also there are a considerable number of concepts that are a direct descendant of 441457006|Cyst (disorder) due to the absence of intervening grouper concepts as seen in other hierarchies such as 123946008|Disorder by body site (disorder) or 362965005|Disorder of body system (disorder).

Navigational grouper concepts are suggested as method of organising this hierarchy into some logical structure that will improve the useability of the hierarchy particularly where it is utilised by the user interface.

3.4 Detailed analysis of reported problem, including background

3.4.1 Grouper Concepts

Using data taken from the January 2016 release, the concept 441457006|Cyst (disorder) currently has 463 descendant concepts. The CSC review proposed the introduction of grouper concepts by body system to improve overall navigation of the hierarchy. 

An example of an area of content where this type of grouper already exists can be seen in the 123946008|Disorder by body site (disorder) hierarchy which has a child of 362965005|Disorder of body system (disorder) and also has site-specific subtypes such as 129565002|Disorder of skeletal AND/OR smooth muscle (disorder).

This approach could be replicated in the hierarchy 441457006|Cyst (disorder) with for example, a new fully defined grouper concept ‘Cyst by body site (disorder)’ with a finding site of 442083009|Anatomical or acquired body structure (body structure). This concept could then have site-specific subtypes such as ‘Cyst of head (disorder). This approach would allow the classifier to assign logical subtypes by finding site and associated morphology using a proximal primitive parent of 64572001|Disease (disorder).

There are obvious areas of content where a fully defined grouper concept might assist with navigation when using a ‘tree walking’ approach to searching. For example 83558008|Cyst of anterior chamber (disorder) and 46210008|Cyst of eyelid (disorder) are direct descendants of 441457006|Cyst (disorder). The addition of a grouper concept ‘Cyst of head (disorder)’ could potentially result in improvement in the structure of the hierarchy by reducing the number of direct descendants from 441457006|Cyst (disorder)

If a further concept ‘Cyst of eye region (disorder)’ was to be added, then the look of the hierarchy might be further improved

However the desire to improve the look of the hierarchy must be balanced again other considerations. For example, is ‘Cyst of eye region (disorder)’ a clinically valid concept or simply a grouper? Or alternatively, by body system is ‘Cyst of visual system (disorder)’ a useful addition? The presence of concept(s) of this nature may be confusing to end users who would need to choose between them.

This might in fact slow down the process of searching for and selecting an appropriate concept to use.  Alternatively end users may choose one concept or the other to record the same disease making disease statistics less reliable. 

Another factor to consider is what types of new concepts might facilitate navigation. Artificial grouping by body regions may seem like a logical solution however it is unlikely that the addition of these concepts would improve the overall useability of the hierarchy. Indeed would this approach be replicated across large areas of the content in the Clinical finding hierarchy or just simply in relation to cysts? Arguably the impact of including navigational groupers would serve only to improve the cosmetic appearance of the hierarchy rather than improve search functionality or the clinical validity of content.

It is noted that the original review by CSC reports that the requirement for navigational concepts was the outcome of clinical review. However ‘tree walking’ is most likely restricted to problem entry screens as noted in the CSC report.  It is questionable how many end users of SNOMED CT are required to use ONLY the taxonomy view of the hierarchy in order to locate their desired concept. 

Use of the search functionality will achieve quicker and more accurate location of a concept that meets the requirement. Indeed it is arguable that navigational concepts lead to extra effort when searching because there is more content in the area and therefore the user is faced with more concepts to choose from.

It is noted that the application of a consistent pattern to 441457006|Cyst (disorder)| and its descendant concepts would assist implementers in the development of intensional value sets. An intensional value set can typically be defined by an algorithm/rule that describes the intension of a concept by stating the superordinate concept and the delimiting characteristics. The primary benefit for implementers is that the value set can be updated frequently with little effort. Value sets defined by intension are characterised by a computable expression that can be refined to an exact list of concepts. For example, an intensional value set definition might be defined as, “All SNOMED CT concepts that are children of the SNOMED CT concept ‘Diabetes Mellitus.’” (http://wiki.hl7.org/index.php?title=Domain_and_Value_Set_Definitions_and_Binding)

The creation of the intensional definition based on the defining attributes of concepts is a reasonable alternative to the creation of groupers, but would also require that the supporting hierarchies that provide the values for the relationship be validated and cleaned up. Application of a consistent pattern to the 441457006|Cyst (disorder)| hierarchy would obviate the need to create the suggested groupers.

3.4.2 Morphologic Abnormality

In the original report by CSC, it is recommended Children of 399294002 Cyst of breast could be organised with grouper concepts of Benign cyst of breast (disorder) Malignant cyst of breast (disorder)

A review of both ICD-O and ICD11 suggests that this proposal is not a viable proposition because the classification of morphologic abnormality does not align with the distinction of ‘benign cyst’ and ‘malignant cyst.’  A cyst is defined as a fluid filled sac and therefore a morphologic structure and not pathology.

449838006|Benign cyst of breast (disorder) has already been added and is now and intermediate primitive parent.  This is suboptimal because the concept does not allow the classifier to logically assign descendants and also cannot be fully defined using an associated morphology.  The proposal to add ’Malignant cyst of breast’ remains the subject of a separate content project IHTSDO-873 mentioned later in this document.

A review of the further content in the hierarchy 367643001|Cyst (morphologic abnormality) identifies that there is some inconsistency that requires improvement. For example, there are instances where the site of the morphology has been included in the morphologic abnormality concept and is used to fully define a disorder concept. As an example 446832006|Benign cyst of ovary (morphologic abnormality) appears to include a finding site which does not align with the editorial guidance https://confluence.ihtsdotools.org/display/public/EditorialGuide/6.6.1+Morphologic+abnormalities+vs.+Findings

Review of both the ICD-O and ICD11 classifications does not support the inclusion of this morphologic abnormality concept into the terminology as a distinct morphology. The concept has been used as the associated morphology to fully define 447249001|Benign cyst of ovary (disorder).  However 449838006|Benign cyst of breast (disorder) has an associated morphology of 367643001|Cyst (morphologic abnormality) and a finding site of 76752008|Breast structure (body structure) and has been marked as primitive. This represents inconsistency in the approach to the modelling of both concepts.

There also appears to be some inconsistency between 123250005|Synovial cyst (morphologic abnormality) and 4125007|Bursal cyst (morphologic abnormality).  These two concepts are siblings in the morphologic abnormality hierarchy with the result that this is perpetuated in the disorder hierarchy.  However 13351007|Structure of bursa (body structure) is a child of 64331004|Synovial membrane structure (body structure) in the body structure hierarchy and therefore 202941002|Cyst of bursa (disorder) could be a subtype of 240205003|Synovial cyst (disorder) were the associated morphology concepts to be correctly aligned.

In the area of ‘congenital cyst’ some of the content is modelled using the associated morphology of 367643001|Cyst (morphologic abnormality) and 246454002|Occurrence (attribute) with 255399007|Congenital (qualifier value) in one role group.  Alternatively it is possible to model the content using 12143007|Embryonic cyst (morphologic abnormality).  This has resulted in inconsistency in the manner in which content is modelled and the classification results.  This is the subject of a separate content project and is discussed further in Section 3.5.3.

3.4.3 Modelling Issues

There are multiple concepts in the hierarchy that are impacted by a stated relationship to fully defined parent(s) for example 61476004|Embryonal cyst of vagina (disorder)

This concept inherits an additional role group due to the stated parent of 52904006|Congenital anomaly of vagina (disorder) and occurrence congenital being missing from the stated role group.  It would be highly beneficial to remove all primitive grouper concepts from the hierarchy in order to allow the classifier to assign logical relationships and prevent the use of stated IS_A relationships to fully defined concepts where possible.

There are also some areas of the content that require improvements to the assigned attribute and value pairs. For example the concept 300923002|Cyst of scalp (disorder) has a fully defined stated parent, incorrect finding site and incorrect associated morphology

This results in suboptimal inferred relationships

Furthermore there are areas of content that are currently marked as primitive and that could be fully defined to improve subsumption of existing content. For example 47639008|Cyst - pilonidal (disorder) is primitive and 85224001|Pilonidal cyst with abscess (disorder) is not currently a descendant concept. This is not what a clinician would expect to see when ‘tree walking’ to search for a concept.

There are specific areas of content that require review as a result of recent changes to the approach to their modelling, for example the concept 236376009|Infected renal cyst (disorder)

There are multiple issues with the modelling of the concept, firstly it is not a descendant of 441457006|Cyst (disorder) as the associated morphology is missing from the modelling.  Secondly the finding site requires refinement.  It is also noted that 236439005|Cystic disease of kidney (disorder) is not a descendant of 441457006|Cyst (disorder) and requires further review.

There are areas where the plural is used in descriptions, for example 204177009|Congenital cysts of the posterior segment (disorder) however the modelling of the associated morphology does not reflect that the disorder is one of multiple cysts despite there being a suitable concept 125291005|Multiple cysts (morphologic abnormality) available.

It is noted that the following concepts are both active

204177009|Congenital cysts of the posterior segment (disorder)

62631009|Congenital cyst of posterior segment of eye (disorder)

However these concepts are siblings and both direct descendants of 441457006|Cyst (disorder).  A stated relationship to a primitive parent along with correctly assigned attribute and value pairs would result in equivalency upon classification and facilitate a ‘clean up’ of the hierarchy where duplication can be quickly identified.

3.4.4 Fully Specified Name

The Editorial Guide states

When possible, the FSN for a Clinical finding should name the morphologic abnormality before naming the site.

Examples:

Inflammation of ampulla of Vater (disorder)

 Here the morphologic abnormality is inflammation and the site is "ampulla of Vater."

| Edema of hand (finding) |   

In this case "edema" names the morphologic abnormality and "hand" names the site.

https://confluence.ihtsdotools.org/display/public/EditorialGuide/7.3.7+FSN+for+clinical+findings

In the hierarchy, there are many examples of conformance to this editorial guidance

399294002|Cyst of breast (disorder)

277333006|Cyst of central nervous system (disorder)

46210008|Cyst of eyelid (disorder)

However it is also noted that many of the descendants of 441457006|Cyst (disorder) do not conform to this pattern

59857007|Branchial cleft cyst (disorder)

195497001|Chylous cyst (disorder)

35893006|Diaphragmatic cyst (disorder)

It is a question of effort versus benefit when deciding whether this is a matter that needs to be addressed.

Recent developments in editorial policy arising from the Event Condition Episode Group have resulted in changes to the naming convention for description for some content https://confluence.ihtsdotools.org/display/public/EditorialGuide/6.1.3+Specific+disorder+types

This impacts on content with a causative agent, for example 233797002|Infected bronchogenic cyst (disorder), would require a change to the FSN to Bronchogenic cyst caused by infection (disorder) although this is currently being addressed as a separate content project.

3.4.5 Descriptions Requiring Review

There are a number of concepts in the hierarchy 441457006|Cyst (disorder) that may be considered ambiguous and require review.

For example 83558008|Cyst of anterior chamber (disorder) does not explicitly state anterior chamber of eye although arguably this is the only conclusion by default.  A change to Cyst of anterior chamber of eye (disorder) would disambiguate the FSN but is not a minor change and therefore is not allowable.  Although it would be preferable to avoid inactivation of concepts where possible, in some case this is likley to be required in order to disambiguate descriptions and ensure that the content of the hierarchy is clear and unambiguous.

Similarly examples of content that use ‘OR’ but not ‘and/or’ require review and inactivation to align with editorial policy.  As an example 195835003|Pharynx or nasopharynx cyst (disorder) does not conform to naming guidance, is a primitive concept and is probably derived from classifications and is unlikely to be of clinical utility. 

3.5 Subsidiary and interrelated problems

There are related artifacts noted in the content tracker that could be resolved alongside this content project.  This content project also overlaps with other ongoing content projects.

3.5.1 IHTSDO-873 (artf222436) Concept additions to "Cyst of breast"

The following information is taken from the content tracker https://jira.ihtsdotools.org/browse/IHTSDO-873

These concepts have been added:

449838006|Benign cyst of breast (disorder)

449837001|Complex cyst of breast (disorder)

449836005|Multiple cysts of breast (disorder)

“With outstanding question for submitter: For Malignant cyst of breast - We are unable to find the term "Malignant cyst of breast" in the literature or in the major classification systems. Neoplasms of the breast with a "cystic" component are usually referred to by the name of the neoplasm, e.g., encapsulated (intracystic) papillary carcinoma, or by the name of the neoplasm with a notation that "cystic" change/degeneration is present, e.g., medullary carcinoma with cystic degeneration. Neither is, however, a "malignant cyst," i.e., a cavity lined by malignant epithelium. Do you still want to add "Malignant cyst of breast (disorder)"?

Add as requested - Multiple cysts of breast (disorder), Complex cyst of breast (disorder), Benign cyst of breast (disorder) - Will be added as children of "Cyst of breast (disorder)". Note: Benign cyst of breast (disorder) cannot be fully defined, as the morphologic abnormality "benign cyst" does not exist in SNOMED CT. The children that already exist (e.g. Simple cyst of breast) will be made stated children.”

‘Malignant cyst of breast’ cannot be fully defined due to the issues noted in the tracker as documented above. 449838006|Benign cyst of breast (disorder) is a released concept that currently acts as a primitive grouper concept.  It is recommended that the request for Malignant cyst of breast be declined and consideration given to the inactivation of the released concept 449838006|Benign cyst of breast (disorder) as this is an intermediate primitive concept that cannot be fully defined with an associated morphology and therefore impedes the action of the classifier where a structured approach to relationship grouping is implemented.

3.5.2 IHTSDO-33 (artf222501) Add synonym: Breast cyst

Add synonym breast cyst to concept 399294002|Cyst of breast (disorder). Not changed - Word order variants are considered outside the scope of the SNOMED CT International Release (see IHTSDO document SNOMED CT Style Guide: Term Conventions Spelling, Style, and Allowable Changes).  It is recommended that this tracker item be closed.

3.5.3 IHTSDO-89 Congenital Occurrence vs. Congenital Morphology

The area of ‘congenital cyst’ could be improved by the addition of a grouper concept ‘Congenital cyst (disorder)’ to improve both navigation and the look of the user interface. However the modelling of content in this area is not consistent and a grouper concept would not achieve the goal of grouping all existing content without further modifications. For example 31339007|Congenital cyst of vulva (disorder) uses occurrence ‘congenital’ and associated morphology ‘cyst’

However 87047005|Congenital cyst of esophagus (disorder) uses the associated morphology 12143007|Embryonic cyst (morphologic abnormality) without role grouping of occurrence.  There is also a stated fully defined parent of 69771008|Congenital anomaly of esophagus (disorder) which results in the second inherited role group with the associated morphology of 21390004|Developmental anomaly (morphologic abnormality).

The area of congenital disease is the subject of another large-scale content project and the clean up of this area should remain part of the congenital disease revision project.

3.5.4 IHTSDO-960 Modelling of Disorders of Acquired Body Structures

This tracker issue has been identified through the SNOMED CT and ICD-11 work. Investigation is required to determine the modelling approach that should be used for the disorders of acquired body structures. For example an ulcer located at the point of an anastomosis. The finding site should identify the anastomotic site not just the anatomical site. Other disorders of other acquired body structures should also be considered.

4 Risks / Benefits

Achieving a workable solution to this content project is desirable because an identified area for quality improvement will be addressed.

End users will benefit from resolution of this content project with a solution that ensures that current content aligns with editorial policy and future content will do the same. Application of a consistent approach to the modelling of both descriptions and stated relationships will also ensure that duplicate concepts are identified and inactivated.  

For those involved in translation, the implementation of a consistent approach to descriptions is likely to reduce anomalies in translated content and may assist with an algorithmic approach to translation. 

By addressing this content and the existing new requests, overall quality and consistency of content will be improved. 

Achieving a workable solution for this content project will also be of benefit to the IHTSDO content team, as the disposition of new requests will be consistent and more efficient.

4.1.1 Risks of not addressing the problem

The project risk is calculated by utilizing the Project Risk Profile assessment tool as detailed on Page 11 of “A Guide to Stakeholder Engagement” https://confluence.ihtsdotools.org/pages/viewpage.action?spaceKey=CONSULTANTTERMINOLOGISTGROUP&title=Stakeholder++Engagement++Resources

The risk level is assessed as a score of 0, which is low risk.

If this problem is not addressed the current situation of inconsistent inclusion of content, descriptions and concept modelling will be perpetuated. User queries for these concepts will consequently be more difficult to formulate, as there will be continue to be inconsistency in the way new content is modelled. Clear guidance will assist the content team to add and model new content in a consistent manner. It is highly desirable to fully define new content in order to harness the benefits of using a classifier, optimise subsumption and eliminate equivalency.

4.1.2 Risks of addressing the problem

Whilst improvement to the quality and consistency of the content is of benefit to all users, the implementation of proposed changes will result in amendments to released descriptions, modifications to an existing hierarchy and inactivation of some released content. This may impact implementations of SNOMED CT.  However the potential negative impact of these factors may be mitigated by

1. Provision of release notes when the changes are implemented
2. Ensuring that all inactivated concepts maintain a historical relationship to an active released concept that the end user can implement in place of the existing concept.
3. Updating editorial guidance to detail all changes that are made and how content of this nature will be added to SNOMED CT in the future.

5 Requirements: criteria for success and completion

5.1 Criteria for success/completion

All concepts that are in scope for this project

1. Will be clearly and consistently represented in SNOMED CT.
2. Will have stated IS_A relationships, attributes and values assigned consistently to improve classification results.
3. Duplication of active concepts will be rectified.
4. Primitive concepts will be fully defined where possible.

5.2 Strategic and/or specific operational use cases

This content development project aligns with one of the core principles as outlined in the IHTSDO Strategic Direction 2015-2020 - Quality focused content development with adherence to strict editorial rules.

6 Solution Development

6.1 Initial Design

The purpose and scope of this content project is to:

1. Review the hierarchy 441457006|Cyst (disorder) and descendant concepts.
2. Identify existing primitive concepts that can be fully defined to improve classification of content.
3. Review the content in 367643001|Cyst (morphologic abnormality) for duplication or areas for improvement.
4. Review descriptions of concepts in hierarchy 441457006|Cyst (disorder) for conformity to editorial guidance (for example plurals in descriptions), duplication of descriptions, errors and ambiguity.

It is recommended that the following steps be taken in order to rectify the issues identified as in scope for this content project.

6.1.1 Fully Defined Grouper Concepts

Sub-concepts of 441457006|Cyst (disorder) could be logically ordered using fully defined grouper concepts to organise the content.  This could be achieved using a new concept ‘Cyst by body site (disorder)’ with a pattern similar to that seen in 123946008|Disorder by body site (disorder) and subtypes. 

However it is recommended that the proposal to add logical grouper concepts to subtypes of the hierarchy 441457006|Cyst (disorder) be rejected. The content in this area should be allowed to grow by request. When a request for a new concept is accepted it should be on the basis that there is a justification for inclusion and a clear definition of the requirement.

The additional work required to incorporate artificial groupers into an existing hierarchy without clear value is not justifiable and does not appear to be a priority for users that are not part of the CSC review.

Changes to the hierarchy are likely to be disruptive to end users and it is questionable whether they would achieve the original intent of making it easier to find a concept in SNOMED CT.  For example if a new ‘grouper’ concept is added; Cyst of hand (disorder), does this mean cyst of the entire hand, some part of the hand, the finger, the skin of hand, the skin of finger? The end user would need to decide the answer to these questions before making a selection. 

It is a more sensible approach to add content on the basis of that which is clinically valid and requested for inclusion rather than trying to pre-empt what may or may not be useful.  It is acknowledged that 441457006|Cyst (disorder) has a large number of direct descendants however this is not widely reported by users to be a problem in itself and grouper concepts to fix this may improve the overall appearance of the hierarchy but are unlikely to improve usability or be used in the clinical record. 

Where a request is received that meets the requirements for inclusion (understandable, reproducible, useful) this concept can be added to the hierarchy regardless of whether it is groups other concepts or not.

6.1.2 Modelling Issues

There are multiple concepts in the hierarchy 441457006|Cyst (disorder) impacted by having a stated relationship to fully defined parent(s).  For example 61476004|Embryonal cyst of vagina (disorder) inherits a second role group as a result of its stated parent.  This factor impedes the classifier from assigning descendant concepts unless the modelling of descendants also includes incorrect modelling of stated relationships, attribute and value pairs.

It is recommended that all descendants of 441457006|Cyst (disorder) be reviewed and their stated relationship changed to a proximal primitive parent and a consistent relationship group structure, to ensure that inferences are correct based on the grouping of fully defined parents. This is likely to require remedial work to fix improper inferred relationships after classification.  One area where this is likely to require much review is the area of congenital cyst and as noted this is being addressed in a separate content project.

Where content is currently marked as primitive this can be reviewed with the intention of making the concept fully defined.  This aim may achievable using existing anatomy or content from the morphologic abnormality hierarchy, however it may be necessary to request changes to the anatomy hierarchy where new content is identified to be required.

There are areas of the hierarchy where content appears outdated or veterinary specific such as 47932002|Cystic mucinous hypertrophy of gallbladder (disorder). An overall clean up of the hierarchy will result in the inactivation of inappropriate concepts and improve searching capability by removing superfluous content.

6.1.3 Descriptions

Much of the content in the hierarchy 441457006|Cyst (disorder) does not conform to the editorial policy on using the pattern ‘Cyst of X (disorder).’  With the exception of concepts where the existing description is ambiguous or erroneous, it is NOT recommended that fully specified names be changed to align with this pattern. 

The benefit of changing to a prescribed pattern of FSN versus the inactivation of released FSN with the associated implementation burden is not justifiable. This excludes those concepts that will require a change to FSN as a result of the recent changes to editorial policy arising from the Event Condition Episode Group for specific disorder types. https://confluence.ihtsdotools.org/display/public/EditorialGuide/6.1.3+Specific+disorder+types

6.1.4 Morphologic Abnormality

In the original report by CSC, it is recommended concepts be organised by Benign cyst and Malignant cyst. A cyst is defined as a ‘fluid filled sac’ and is therefore a morphologic structure and not pathology. It would not be possible to add the concepts Benign cyst and Malignant cyst to the morphologic abnormality hierarchy in any meaningful way and their inclusion would not align with other classifications such as ICD-O.

It is recommended that the concept 449838006|Benign cyst of breast (disorder) be inactivated as it is primitive, an intermediate grouper and cannot be fully defined.

The proposal to add ’Malignant cyst of breast’ remains the subject of a separate content project IHTSDO-873 however there is no evidence found to support the inclusion of this concept and this again would represent an intermediate primitive concept were it to be added.

The CSC recommendation to add new concepts ‘Multiple cyst of breast’ and ‘Complex cyst of breast’ has already been implemented. 449837001|Complex cyst of breast (disorder) is a primitive concept that cannot currently be fully defined.  449836005|Multiple cysts of breast (disorder) has also been added as a primitive concept however this concept can be fully defined using the associated morphology 125291005|Multiple cysts (morphologic abnormality) and it is recommended that this change be implemented.

6.1.5 New Content

It is recommended that the hierarchy 441457006|Cyst (disorder) be updated to include any content that is included in ICD-11 and ICD-O but not currently represented in SNOMED CT.  Work to align SNOMED CT with ICD-11 and ICD-O is already ongoing as part of the process of content development.

6.1.6 Significant design or implementation decisions / compromises

Addressing any inconsistency in released concepts may require changes to way the concept is modelled or changes to descriptions. This is likely to result in changes to inferred relationships after classification. As noted above, in order to minimise disruption to users, it is recommended that changes to descriptions only be made where the content is outdated or erroneous, or falls into the scope of a separate content project. 

The process to improve the overall modelling and classification results of the content in the hierarchy with result in improvement of the quality of the content of the hierarchy and therefore will most likely benefit users of SNOMED CT. It is probable that the proposed changes will uncover equivalency between released concepts and this will need to be managed using inactivation and historical relationships. The impact of the proposed changes on end users is difficult to assess however the benefits of improvement in the quality of existing content without large scale inactivations is intended to mitigate any associated implementation burden. 

6.1.7 Evaluation of Design

6.1.7.1 Exceptions and Problems

Generally where a concept has a mixture of descriptions and is also inconsistent in the way in which its attribute and value pairs have been assigned, it would be considered to be ambiguous.  This would lead to concept inactivation.  The recommendations of this content project do not include large-scale inactivation of existing inconsistent content.  However to mitigate for this, implementation of changes will require careful review and analysis by the content team during implementation.

7  Recommendation

7.1.1 Detailed design final specification

1. Review the hierarchy 441457006|Cyst (disorder) and descendant concepts.

a)    Reject proposal to add navigational grouper concepts to the hierarchy.

b)    Continue to add content to this hierarchy upon request subject to justification, definition and the proposed new concept meeting the criteria to be understandable, reproducible and useful.

1. Identify existing primitive concepts that can be fully defined to improve classification of content.

a)    Retire stated IS_A relationship to fully defined parent and implement the proximal primitive parent approach along with a consistent relationship group structure.

b)    New anatomy content may be requested to support the approach of fully defined content in this area where necessary.

c)    Overall clean up to retire existing content that is outdated, ambiguous, and erroneous or duplicate.

1. Review the content in 367643001|Cyst (morphologic abnormality) for duplication or areas for improvement.

a)    Align the content with ICD-O.

b)    Retire content that is outdated, ambiguous and erroneous or duplicate.

1. Associated tracker items:

a)    IHTSDO-33 (artf222501) Add synonym: Breast cyst

b)    IHTSDO-873 449838006|Benign cyst of breast (disorder) has already a released concept and is now an intermediate primitive parent.  This is suboptimal because the concept interferes with the actions of the classifier in logically assigning descendants and also cannot be fully defined using an associated morphology. 449838006|Benign cyst of breast (disorder) should be inactivated and the request to add ’Malignant cyst of breast’ should be declined. 

8 Quality program criteria

8.1 Quality metrics

8.1.1 Quality metric 1

8.1.2 Quality metric 2

9 Project Resource Estimates

Estimated project size is small and can be implemented over one release.  

9.1 Scope of construction phase

The proposed solution is detailed in Section 6 of this document with recommendations detailed in Section 7.  The scope of the initial phase of this content project will be to review all existing content in the hierarchies 367643001|Cyst (morphologic abnormality) and then 441457006|Cyst (disorder) in order to align the relationship groups and descriptions as set out in Section 7 where possible.  This would represent the first phase of implementation to be undertaken in one release cycle.

The second phase will be to review all existing requests for new content currently in SIRS and would be undertaken in the subsequent release after completion of phase one to ensure that the released changes are implemented and allow time for stakeholder feedback.  

9.2 Projection of remaining overall project resource requirements

9.2.1 Expected project resource requirement category

This content project in isolation does not require project management and can be implemented by the content team. However, implementation of this content project would be resource intensive and requires support for batch editing changes in the tooling. Quality review of the changes will be required by the content team and this will be a labour intensive activity.

9.2.2 Expected project impact and benefit

The implementation of this project will benefit end users by improving the quality of existing content in more than one hierarchy.

9.2.3 Indicative resource estimates for construction, transition and maintenance:

367643001|Cyst (morphologic abnormality) has 78 descendants and 441457006|Cyst (disorder) has 463 descendants that will require review and changes to be made where necessary.  Likely resource requirements for this are as follows:

541 existing concepts to be reviewed and changes made to align with new policy at a rate of 20 concepts per day would mean that one terminologist could implement the required changes in approximately 5 working weeks whilst incorporating time for peer review.  This estimate could be substantially reduced if the data changes were to be supported by batch editing processes in the existing tooling.

Existing SIRS requests (approximately 28 requests) are to be reviewed and added to SNOMED CT. 

Maintenance is not required assuming that going forward new content follows the recommended approach for term construction and modelling.