Observable and Investigation Model Project

Page tree

Sarah Harry

Owner
1760 View 4 Comment In discussion Comments enabled In the category: Undefined

Options for modelling cycle threshold (Ct) results that are reported alongside the DNA/RNA presence results for microorganisms, reviewing existing SNOMED options and examining LOINC and NPU/IUPAC representations.

Contributors (3)

4 Comments

  1. Sarah Harry

    Dear Observables Group,

    Looking at microbiology reports I see a good few 'Ct' (cycle threshold) values for various organism RNA/DNA detections but I'm not seeing these described in SNOMED and I might assume a reason for this.  LOINC has these represented and uses LP188772-0 ThreshNum (threshold number) 

     

    e.g. SARS-CoV-2 (COVID-19) RNA [Cycle Threshold #] in Respiratory specimen by NAA with probe detection

     

    I would have just used 'number' as a property but oddly there isn't one!   In SNOMED we only have:

    118589004 | Threshold (property) (qualifier value)

    Which seems to be one half of 'threshold presence' property.

    I cannot see a scalar number property, only vector ones.  260299005 | Number (qualifier value) seems directed at other uses and is thus not available as a property.  (I regret the use of a coherent quality/property as a mere header for numbers but that's an aside.)

    To me this concept of Ct value for NAAT result is metrologically a number, 'threshold' just being added onto 'number' as extra qualification but maybe 118589004 | Threshold (property) (qualifier value) is OK? I cannot find a pure 'number' property except for 'entity' types.  This is almost a parameter setting perhaps?  

    The LOINC concept seems to 'expect' a pure ordinal number but the notional 'amplification cycles' UoM seems to be included in the property? I think I'd expect not a 'pure' number but to see a UoM of '(amplification) cycles' if there is such a thing in IUPAC?  The only 'cycles' concept in SNOMED seems to be a pharmacy one.

    Just for completeness, I did consider whether the Ct could be a flavour of presence, an adjunct to the pos/neg presence observable: e.g. Presence of x by NAA at Ct 25 cycles (obs entity).  It is worth considering but it would be highly combinatorial. I find in lists, the test Ct is a separate line so the information model could and should connect them IMHO. This seems more scalable anyways.

    It would be useful to have a sense of the SNOMED editorial line on this topical issue (or a ruling on its in/out of scope-ness) if people think it's in order because, even in SNOMED expression terms, it is not obvious to me how this would be represented/modelled.  I'd be grateful for a view and would be happy to do a brief presentation in a group meeting if appropriate.

     Thanks.

  2. Jim Case

    Ct represents the number of cycles required to reach an assay minimum threshold level of detection, so it is a number.  I do not think that including the number of cycles in the observable is useful as each amplification method or device has its own determined level based on the optimization of reagents.  The number of cycles certainly indicate the "relative" amount of nucleic acid in the sample (but not necessarily the relative amount of nucleic acid in the patient due to sampling error).  

    As to whether we really need a new property, if all we are reporting is a number (count?) and we have the proper method, do we really need a new property?  Also, many/most PCR reports are simply reporting out Pos/Neg or Detected/Not detected.

  3. Sarah Harry

    Thanks Jim.  Whether reported or just recorded internally is a little moot here.  I've read the arguments for and against but what you write is fine.  I would still expect there to be a property that supports a simple number value as there must be other instances where this is needed (in machine parameters for example).  The only one I can see is 410681005 | Count of entities (property) (qualifier value).  Would this be the one to use, taking 'entity' in its broadest possible meaning? Daniel Karlsson?

  4. Daniel Karlsson

    By the definition of entity from M-W below (my higlight), I believe that entity is broad enough. Anything that isn't an entity would not be countable. So, yes, this would be the one to use. One could argue that plain "Count" should be a synonym.

    Definition of entity

    1a: BEING, EXISTENCE especially : independent, separate, or self-contained existence

    b: the existence of a thing as contrasted with its attributes

    2: something that has separate and distinct existence and objective or conceptual reality

    3: an organization (such as a business or governmental unit) that has an identity separate from those of its members