3937 View 10 Comment In discussion Comments enabled In the category: Undefined

This a request for information to support the SNOMED Quality Improvement project.

There are a number classification concepts currently within the terminology which are of the form:

450569000 |Traumatic brain injury with loss of consciousness one hour or more (disorder)|

in addition there seems to be an overlap between the use of the terminology to describe coma v loss of consciousness.

A review of the current practice for management of brain injury (NICE and ACS) suggests that assessment should be based upon the application of the 3 separate elements of the Glasgow Coma Score. and it is these separate scores which should be used as the means of communicating the patients status between clinicians and within the patients notes.

If this is the approved international approach to recording the mental/coma status of head injury patients we would like to inactivate the complex individual statements relating to coma/concussion on the basis that all of the individual elements of the GCS are available within the terminology.

There are 173 concepts in the hierarchy of 127294003 |Traumatic AND/OR non-traumatic brain injury (disorder)| which include the word concussion and the list of these is attached to this discussion thread.

We would be grateful for you views.

Many thanks
Paul Amos

Link to the file https://confluence.ihtsdotools.org/display/shareit/75344051/SOS75eb77401cf0457996f371267eb92189STO/SNOMED%20Loss%20of%20Consciousness%20Concepts.xlsx?version=1

Contributors (5)


  1. Dear Paul,

    I tend to agree that because the conscious state may change over time the sensible method of communicating the snapshot status of a patient with a brain injury is the GCS. 'Concussion' and similar terms are rather imprecise and should therefore be discarded. 450569000 |Traumatic brain injury with loss of consciousness one hour or more (disorder)| can therefore be qualified by both GCS and by time elapsed.

    Kind regards,


  2. Paul

    A couple of things:

    (1) I presume the list posted is more about inclusion of some mention of 'concussion' OR 'consciousness'? Incidentally it has picked up 763310000 | Acute necrotizing encephalopathy of childhood (disorder) which mentions 'consciousness' in its Orphanet-derived definition description but is really a false positive for this exercise.

    (2) >> "there seems to be an overlap between the use of the terminology to describe coma v loss of consciousness" - can you expand on this?

    Do you mean in the terminology or in general use? Is this a tractable problem, and if so is it something that can realistically be 'solved' by the terminology?

    (3) I would be surprised if there is significant support for retaining many of the more extravagant, classification-derived combinations in the set identified, but it would be good to know what inactivation reason would be given, and what historical links back into the active data would be provided.

    (4) James' characterisation of the vocabulary needed for acute management purposes is strong support for a recording approach based on the separate GCS components, however there are other use cases where SNOMED CT might be called upon to provide 'something for the diagnosis box'. Indeed the NHS's Emergency Care Data Set references a number of in-scope concepts in its diagnosis subset:

    127300000 Traumatic brain injury with moderate loss of consciousness (disorder)
    127302008 Traumatic brain injury with no loss of consciousness (disorder)
    127299008 Traumatic brain injury with brief loss of consciousness (disorder)

    As such there may well remain a requirement for a modest set of active combinations that 'summarise' a head trauma/consciousness episode.


  3. Ed,

    Good points which I'll take in turn.

    1. No comment
    2. The term coma should imply a condition of some significant duration cf. e.g. syncope
    3. No comment
    4. 127300000 makes no sense to me until I see that it links to a definition of concussion as being LOC for between 1 - 4 hours and as I've dsaid concussion is a term that I would like to see removed as being of essentially historical interest. However, if TBI terms include no LOC, brief LOC, moderate LOC  and prolonged LOC then these still have utility.



  4. Thanks James

    regarding (2), the question remains what the 'significant duration' is before 'loss of consciousness' becomes 'coma'. SNOMED CT's current view (based on what content sits beneath 371632003 | Coma (disorder)) is that essentially all 'prolonged loss of consciousness' and about half of the 'more than 24 hours loss of consciousness' content imply the presence of 'coma' at some stage during the situation being described/summarised (not to mention the enigmatic 'prolonged coma'). If 24 hours is agreed clinically as the cut-off then it's possible to reflect that distinction in the data, but perhaps then the slipperiness of language in use might start to cause problems. Would this just be a traumatic boundary, or would any sub-24 hour 'diabetic coma' (which I would hope is most) have to be rebranded a 'diabetic loss of consciousness'.

    regarding (4), the standard answer on how terminology content is precisely defined stems from section 1.3 of Alan Rector's paper here (nearly 20 years ago!), where the distinction between 'terminology knowledge' and 'inference beyond the scope of terminology' is made. The reality is messier - a growing number of concepts are accompanied by text definitions that may richly describe the essential and optional features of a syndrome, but in general relative stratifications (mild, moderate, severe, high, low), 'grades' and 'levels' named but require some external authority or reference to 'anchor' their meaning.


  5. James and Ed, many thanks for your contributions so far:

    With regard to Ed's comments:

    1. I simply printed out all of the inferred subtypes of 127294003 |Traumatic AND/OR non-traumatic brain injury (disorder)| which include the word concussion as examples and therefore I expect that some will have been missed due to incomplete/incorrect modelling and may be found elsewhere and there are conditions other than head injury in which the state of consciousness is referred to e.g. diabetes.
    2. This is a real issue which needs to be addressed consistently and will require some international agreement on the definitions of 'loss of consciousness', 'coma' and 'concussion' - I would be grateful if the Anaesthetic CRG and others could take this on and work towards formulating or adopting a consensus view.
    3. We will share these with you for comment before implementing
    4. This relates to my comments in 2 above but in addition, as a terminologist I would like to avoid the use of relative stratifications, grades and levels which do not have a clearly defined and internationally agreed meaning

  6. Sorry I've been a bit slow in contributing to this discussion, but just returned from holiday.

    From my Intensive Care point of view, little practical use is made of the terms coma or loss of consciousness - everything gets described in terms of a patients Glasgow Coma Score and its components. However I note Ed's comments about the diagnosis subset terms for recording a more general statement of a patient's neurological state following injury.

    The literature definitions of coma generally refer to a GCS of 8 or less - some also note a time component, most commonly GCS or 8 or less for greater than 6 hours. Definitions of loss of consciousness and concussion are far less precise than this.

    As Paul notes in his most recent post, there would be value in getting to consensus definitions of these terms (but I agree with James that concussion is a relatively obsolete concept).  However we would need to engage clinicians in quite a number of other specialities to get any progress - neurology, neurosurgery, emergency medicine and acute medicine are just some of the ones that immediately come to mind.

  7. Hi Andrew,

    It would be really helpful if you could take forward the discussion regarding consensus on the definitions of coma and loss of consciousness as I think we need these to adequately provide concepts to cover all aspects of the management of patients who have intracranial insults which result in altered consciousness.

    We have recently had a further query regarding concepts in this domain so it would be timely for us to address the remaining issues.

    Would this be possible?

    Many thanks


  8. Paul,

    Thanks for the request. We can certainly discuss this when the Anesthesia CRG meets in April, but as I had mentioned in my previous comment, the problem will be engaging enough of the other specialities who would have a clinical interest in this area but don't have an active presence in the SNOMED community of practice. Guess it also be useful to try and enlist some help from Jane Millar and Charles Gutteridge in trying to engage relevant stakeholders.

    Best wishes


  9. Andrew,

    That would be very helpful. I will discuss with Jim whether we need to make any of the changes suggested above ahead of a conclusion on the coma/concussion issue.



  10. Ed Cheetham , with regards to  your item "3." above, I have had discussions with Jeremy Rogers on the value of the classification based terms with ranges, as well as the "without X" terms.  He has indicated that they have been negligible in use within the UK and would only be valuable as part of a discharge summary, if at all, rather than clinical recording.  Unless there is a hue and cry from this group (or others that will be informed), these classification terms are scheduled for inactivation in the July 2020 release as part of the quality improvement initiative project.