A question arose as to the policy for changing FSNs that are currently plural but have subtypes that are clearly singular. May these FSNs be changed without retirement of the concept or not?
The plurals seem to be used because it "looks better" in navigation, but usually doesn't make sense "ontologically". 276028005|Preoperative procedures | Looks like a reasonable parent when you see it subsuming a bunch of concepts. But Preoperative counselling IS A Preoperative procedure - is more correct. (I've seen a lot of this in substances, depending on what the substance is)
*But... Of course - there's things where specifying things like multiple sites as opposed to single is significant.
Rendering descriptions in singular or plural rendering may influence interpretation of medical record findings. Compare the finding 'Headache' with 'Headaches' on a problem list. Taken at face value these trigger different paths of clinical enquiry. (Acute current versus recurrent/chronic might be my conflation here).
Consider then "metastasis" versus "metastases". 128462008 | Secondary malignant neoplastic disease (disorder) | has active synonyms which include 'Metastases' and 'Metastatic neoplasm' (the latter might cover 'metastasis')
Should one separately and elsewhere qualify these kinds of finding with (for example): 252105006 | Number of metastases (observable entity) | ?
Is there a case for consistency throughout SNOMED? Natalya Noy and Deborah McGuinness write the following
"6.2 Singular or plural A class name represents a collection of objects. For example, a class Wine actually represents all wines. Therefore, it could be more natural for some designers to call the class Wines rather than Wine. No alternative is better or worse than the other (although singular for class names is used more often in practice). However, whatever the choice, it should be consistent throughout the whole ontology. Some systems even require their users to declare in advance whether or not they are going to use singular or plural for concept names and do not allow them to stray from that choice.
Using the same form all the time also prevents a designer from making such modeling mistakes as creating a class Wines and then creating a class Wine as its subclass."
If someone with a stronger background in ontology than myself could comment on the second paragraph with respect to some SNOMED examples I would be very obliged. Thank you.
With regards to the statement "Using the same form all the time also prevents a designer from making such modeling mistakes as creating a class Wines and then creating a class Wine as its subclass.", in fact, the opposite is actually true in a number of cases in SNOMED CT. For example, the concept "Congenital exostosis" has the subtype "Multiple congenital exostoses" and "Congenital cerebral cyst" has the subtype "Multiple congenital cerebral cysts".
This type of hierarchical relationship is also reflected in the Morphologic abnormality subtaxonomy of body structure, so using values such as FINDING SITE = "Multiple embryonic cysts" will cause the classifier to locate the affected concept under fully defined parents using FINDING SITE = Embryonic cyst.
Agree that there is a desire to have FSNs consistent, but there are instances where it is essential to state the plural. While I think we all agree that having singular concepts under a plural parent is incorrect, the related issue is whether having them infer under a singular parent is appropriate.
4 Comments
Jim Case
A question arose as to the policy for changing FSNs that are currently plural but have subtypes that are clearly singular. May these FSNs be changed without retirement of the concept or not?
Matt Cordell
I would think this is reasonable. *
The plurals seem to be used because it "looks better" in navigation, but usually doesn't make sense "ontologically".
276028005| Preoperative procedures | Looks like a reasonable parent when you see it subsuming a bunch of concepts.
But Preoperative counselling IS A Preoperative procedure - is more correct.
(I've seen a lot of this in substances, depending on what the substance is)
*But...
Of course - there's things where specifying things like multiple sites as opposed to single is significant.
Malcolm H Duncan
Rendering descriptions in singular or plural rendering may influence interpretation of medical record findings.
Compare the finding 'Headache' with 'Headaches' on a problem list. Taken at face value these trigger different paths of clinical enquiry. (Acute current versus recurrent/chronic might be my conflation here).
Consider then "metastasis" versus "metastases".
128462008 | Secondary malignant neoplastic disease (disorder) | has active synonyms which include 'Metastases' and 'Metastatic neoplasm' (the latter might cover 'metastasis')
Should one separately and elsewhere qualify these kinds of finding with (for example): 252105006 | Number of metastases (observable entity) | ?
Is there a case for consistency throughout SNOMED? Natalya Noy and Deborah McGuinness write the following
"6.2 Singular or plural
A class name represents a collection of objects. For example, a class Wine actually represents all wines. Therefore, it could be more natural for some designers to call the class Wines rather than Wine. No alternative is better or worse than the other (although singular for class names is used more often in practice). However, whatever the choice, it should be consistent throughout the whole ontology. Some systems even require their users to declare in advance whether or not they are going to use singular or plural for concept names and do not allow them to stray from that choice.
Using the same form all the time also prevents a designer from making such modeling mistakes as creating a class Wines and then creating a class Wine as its subclass."
http://protege.stanford.edu/publications/ontology_development/ontology101-noy-mcguinness.html
If someone with a stronger background in ontology than myself could comment on the second paragraph with respect to some SNOMED examples I would be very obliged.
Thank you.
Jim Case
With regards to the statement "Using the same form all the time also prevents a designer from making such modeling mistakes as creating a class Wines and then creating a class Wine as its subclass.", in fact, the opposite is actually true in a number of cases in SNOMED CT. For example, the concept "Congenital exostosis" has the subtype "Multiple congenital exostoses" and "Congenital cerebral cyst" has the subtype "Multiple congenital cerebral cysts".
This type of hierarchical relationship is also reflected in the Morphologic abnormality subtaxonomy of body structure, so using values such as FINDING SITE = "Multiple embryonic cysts" will cause the classifier to locate the affected concept under fully defined parents using FINDING SITE = Embryonic cyst.
Agree that there is a desire to have FSNs consistent, but there are instances where it is essential to state the plural. While I think we all agree that having singular concepts under a plural parent is incorrect, the related issue is whether having them infer under a singular parent is appropriate.