IHTSDO Content development – fast track (simple/single changes)

Word version documents located at the bottom of this page. 

Review of Classification of Some Military Service Findings

(artf235830 – 'Move History relating to military service from findings to situations')

Version Information

Document review

Table of Contents

Statement of problem as requested or initially identified

IHTSDO Content Tracker Description:
"The concept 365549006 – 'Finding of history relating to military service (finding)' is misplaced and some of the children are also history concepts. These need to be moved to the situation hierarchy."
Tracker details at date 20160613:

Relevance to International edition

Consideration of the issue falls within scope of the international edition. The content describing military service findings uses a nomenclature in common use world-wide and is consistently understood. There are no obvious country-specific references.

Related changes impacted by this content development request

No related changes are identified as the proposal and solution are tightly constrained within a limited sub-hierarchy.

Agreed scope statement

This is limited to the small subhierarchy '365549006 | Finding of history relating to military service (finding)>>'. This has 16 subtypes:
365549006 | Finding of history relating to military service (finding)
266963001 | Exempt from military service (finding)
276088008 | Conscientious objector (finding)
276090009 | Medically unfit for service (finding)
276089000 | Non-combatant (finding)
10987541000119104 | History of military deployment (finding)
236305007 | Joined military forces (finding)
266964007 | Left military service (finding)
276085006 | Bought out of service (finding)
276087003 | Deserted from military service (finding)
276086007 | Dishonourable discharge (finding)
276083004 | Failed service training (finding)
276082009 | Medical discharge from forces (finding)
276084005 | Resigned commission (finding)
224355006 | Served in armed forces (finding)
419482003 | Served in armed forces in combat (finding)
224357003 | Served in peace time (finding)
The issue is specifically those concepts describing patient 'history' but can be taken to be seeking a review of all findings here that describe a past event. Speculatively, users may have problems where they cannot locate all 'History of clinical finding' concepts grouped in the situations hierarchy. The issues can be split out:

• an issue with the grouper

• issues with the subtypes

  • an issue with the concept 'History of military deployment (finding)' because it specifies 'history of'

  • possible issues with the rest of the hierarchy because they describe past events.

Problem: "The concept 365549006 – 'Finding of history relating to military service (finding)' is misplaced."

The top grouper code has been identified as needing to be moved to the Situation hierarchy. It is unclear why this should be moved as:

1. the concept is clearly termed to operate as a grouper and could not, without some determination, be used in a clinical record to state anything meaningful

2. it is not of a format allowing for classification in situations without changing the term and hence it would need to be retired and replaced; but unless grouped content was being moved to situations for which the replacement was a grouper then providing a grouper in Situations doesn't have a separate use case and so a move depends rather tenuously on significant subtype content being moved. However, groupers while often helpful are not obligatory.

Problem: "Some of the children are also history concepts."

The tracker item does not identify specific subtypes that are candidates for the context/Situations hierarchy. Reviewers of this paper suggested that anything with a verb in the past tense in the FSN is also a candidate for the situation hierarchy as it implies "in the past". There is an IHTSDO backlog intention to review findings that should have context definitions and move them accordingly but it is hard to justify treating concepts here in isolation of the clarification of default findings context and a more systematic programme of moves. On review of the initial work proposal it was argued that heuristics for Clinical Findings FSNs might include the avoidance of phrases with context implications such as 'history of' but this was countered by this already being editorial policy as all new 'History of…' concepts are automatically placed in the situation hierarchy. However, this reliance on policy seems to miss the step where the concept addition is reviewed for the appropriateness of using 'history of' in the term in the first place.
'History of military deployment (finding)' is the prime example here because, at first inspection, it seems to have an obvious context-dependent format. Other concepts here also have an apparent past tense ('done' or 'past') format leading to an initial appearance of past context.
Speculatively, provider organizations or systems suppliers/vendors may wish to align all 'history of…' content together in the Situations hierarchy for ease of subsetting etc. This is a reasonable expectation but it ignores some of the distinctions to be made both grammatically and clinically and might arguably be a task left to implementers in menu/refset/subset work.

Considerations of classification semantics.

These concepts are in an employment finding hierarchy and not in a clinical finding hierarchy. The findings 'failed', 'deserted' 'left' inform the subject of the record's current status and are verbs describing what the record subject did or had done to them outside the terminology role/actor domain rather than indicating past clinical states or acts on which the terminology rests. Examples of this format abound;
250171008 | Clinical history and observation findings (finding)
<<365854008 | History finding (finding)
<<309018005 | Referred to service (finding)
413170007 | Removed from depression register (finding)
445060000 | Left against medical advice (finding).
162660004 | Asthma resolved (finding)
1755008 | Old myocardial infarction (disorder)
<<309039003 | Discharge status (finding) e.g. 183665006 | Discharged from hospital (finding)
183535009 | Referred to endocrinologist (finding)
278838006 | Released from prison (finding)
224342005 | Acquitted from criminal charge (finding)
428922000 | Found in automobile driver seat (finding)
224327007 | Passed exams (finding)
The argument for removal to Situations hierarchy opens up possibilities for both procedure and finding modelling and these are both considered. One might argue that 'referred to endocrinologist' can be converted into a situation with a neat associated procedure (not finding note) but that would change an instantiated meaning from 'this is a referred patient, referrer not stated' to 'I have made a referral for this patient' and the author considers that the clinician might not want or might not be able to say that. Grammatically, there is a distinction to be made in the examples 'failed', 'deserted' 'left' etc. between the past participle of a (usually transitive) verb referring to a procedure and a verbal adjective referring either to a noun or clausally by extension to a person; even though the word is the same in both cases. For example, 'Failed test' might be a type of test or a type of failure. The distinction is not always worth making or is often moot but purifying apparent past participle verb forms from the findings hierarchy would need a much more high level use case and is probably unrealistic.
A possible litmus test of whether concepts are appropriate to 'situation' modeling is to parse the term for it's context-neutral 'associated procedure' or 'associated finding'; i.e. the concept to which the temporal and other context can be applied (and that should in principle require the addition of such a defining concept where it does not exist). If the concept fails this test it is perhaps more likely to be appropriate as a straight finding and requires further thought. Such a test undermines the case for context modelling (but note this test may not align with editorial policy once this has been elucidated in this area; it is merely a working test from the author of the paper). 'History of military deployment (finding)' fails this sense test as a theoretical 'Deployment of serving member of armed forces (procedure)' seems an unlikely procedure to find in SNOMED CT and 'On military deployment (finding)' a perhaps better but still unlikely finding to add to SNOMED CT.
This further applies to most content in the hierarchy. For example in the case of 'Failed service training' the theoretical procedure 'Failing service training (procedure)' is not a valid or useful clinical utterance even subject to various context definitions. A situation concept would have to be framed along the lines of 'History of having failed service training' which is semantically the same thing as the existing finding but the derived finding would be 'state of having failed service training' or somesuch. This seems over-worked. Contexualizing this in the past does not add anything and indeed adds an unwanted possibility that this happened in the past but is not necessarily now true which is again a nonsense. These concepts represent current and permanent perspectives of the subject informed by past/historical events and are not designed as a pre- or post-coordinated record of a past event, act or finding in the context of a clinical history such that it can be interrogated for references to operations or diseases. This content cannot easily be modeled with an associated finding that isn't nonsense or semantically duplicating the situation term it defines and this further supports deferring any decisions on such content until a more general review is undertaken.
A further consideration is that one risks filling the situations hierarchy with semantically complex but unmodelable findings that are then deprived of the richer classification found in the Findings model and the move therefore degrades the content. The question would need to be asked as to whether migrating content from Findings to Situations is enhancing the terminology in use.

Clinical considerations.

The difficulty in crafting neutral procedures or findings to model Situation content is not merely one of sense but of utility and validity. The Situations hierarchy is designed to provide context for clinically useful statements about largely clinically relevant findings and diagnostic and interventional procedures. That is not to be puritanical and exclusive of non-clinical utterances from decision trees but merely to set some admittedly rough limits to the extent such content should be usefully modeled or defined. 'Deserted from military service' for example is potentially useful where psychological evaluation takes place and where welfare and health decisions might need to be made but it is not a clinical finding of itself in the sense that it does not describe the nature of the problem engendered by the history which is otherwise merely inference). It does not obviously drive or inform clinical decision making directly but is just part of an informative clinical picture. This seems an arbitrary line to some degree but many 'non-clinical (if there is such a thing) findings could be subject to moving to situations if they describe states based on past events and simple life event findings from history-taking and as previously stated, this analysis of content for purification has yet to be undertaken.
Ostensibly 'History of military deployment' is more clearly a statement affecting differential decision making or diagnosis than other concepts here. However, further analysis finds it inadequate to support an obvious decision point just as something like for example 'History of moving house' would not merit a context definition. 'Deployment' of itself cannot be taken as standing as a risk indicator which is what seems to be driving the change request unless it is for neat classification reasons; the risk is undefined – does this trigger something related to post-traumatic stress disorder or other mental health issues, infectious disease investigation, welfare, pastoral or financial issues follow-up? Deployments may be internal (as an aside, drone operators now act remotely from the battlefield and may not leave the home country at all so physical deployment is not the entire picture anymore) or to countries without significant disease or other risks. The deployee may not see/go through combat.
There are more obvious ways of describing risks and issues. 365457007 | Foreign travel history finding (finding)>> would be more useful in evaluating a patient risk than deployment per se. The UK extension for example has a concept in this hierarchy 506931000000109 | Recent travel to disease affected area (finding). If the issue is PTSD then there are for example:
419482003 | Served in armed forces in combat (finding)
224356007 | Exposure to combat (event)
10995991000119109 | History of combat and operational stress reaction (situation) which is modelled with Associated finding: Combat fatigue (61157009)
…all of these seem to serve more directly than 'History of military deployment' in history taking to inform decisions. Indeed, it is hard to see what 10987541000119104 | History of military deployment (finding) adds to the terminology at all that isn't already perfectly well covered by 419482003 | Served in armed forces in combat (finding) or its 'in peacetime' relative or merely 224355006 | Served in armed forces (finding) etc since the geographical location or combat/non-combat role isn't specified and no obvious clinical inference can be drawn as it stands. In fact a current 'On military deployment' would be conceivably useful while its past form seems irrelevant. It is informative but not directly definitive of actions to be undertaken in the presence of a health or welfare problem. There are no clear justifications for moving the content and clear editorial support for leaving this ex-extension concept as a finding, if only because it is ambiguous and of questionable value. It is worth noting that in the case of 'History of combat and operational stress reaction (situation)' above there is a clear neutral clinical associated finding; i.e. a stress reaction - a sound modelable clinical concept. Whether that is the same as combat fatigue is another matter.

Identification of additional changes

A very small and incidental change was identified in review:
224357003 | Served in peace time (finding), if it is in this hierarchy at all (and it should be as the context is clear), should be a subtype of 224355006 | Served in armed forces (finding) alongside 'Served in armed forces in combat'. The absent text '…in armed forces…' is no impediment to this. As an incidental change this has not been elaborated in this Fast Track document, the thrust of which is to reject the proposal. It is an incidental review finding.

Solutions proposed

For the reasons expounded above, the header concept, 'Finding of history relating to military service (finding), should not be altered, moved or retired.
Although a case may be made for moving some content to the Situations hierarchy e.g. 'History of military deployment', the existing editorial guidance for past context requires further elaboration before this work can be undertaken on a consistent and reliable basis.

Stakeholder input

The proposed solution is yet to be shared with any initiators of the request but as the solution is to reject rather than add or move content then this is not required. It is not known if the originators of the content reviewed have expressed concern over the current classification.

Impact assessment

No changes are proposed.

Risk assessment

The content meets current editorial guidelines as a finding concept and is usable as it is. There are therefore minimal risks involved. The concept is seemingly being viewed as an outlier to a 'History of clinical finding catalogue/subset but isolating the concept from other military service findings may also cause difficulties.

Approval process

Priority

☐Very high
☐High
☐Medium
☒Low

Specify the basis for the above priority assignment

Not applicable. The solution is to reject. Therefore no priority is assigned.

Content editing

Not applicable.

Details of content changes

 
Not applicable.

Manual quality check

 
Not applicable.

Automated quality check

 
Not applicable.

Publish to release branch

 
Not applicable.

Word copy version 0.1

Word copy version 0.2