SNOMED International is seeking input on how navigational concepts e.g. 4730003 |Mosquito-borne infectious disease (navigational concept)| are being used.
|Date||Requested action||Requester(s)||Response required by:||Comments|
|4 December 2019||Provision of input on the use of navigational concepts||Please post your final responses in the Country response table below. Discussion comments can be made as comments.|
- Type your task here, using "@" to assign to a user and "//" to select a due date
|It is to early to say how vendor might use these concepts. It is not unthinkable that they could be used sensibly, but my guess is that it will vary very much from implementation to implementation.|
I've always struggled with the navigational concepts, and how they should be used...
There's a lot of probably"useful" concepts in there, but implementing them in a meaningful way seems to be to complex and difficult such that it becomes difficult. I'm not aware of anyone using them in AU.
I get the impression they are clinically relevant concepts, but too difficult to model/outside model, so they've been relegated to the navigational hierarchy. Very similar concepts in the proper hierarchy now...
For example, Arbovirus infection (disorder). This could be a supertype of "Mosquito-borne infectious disease". There's not much difference between these two concepts.
CSIRO's shrimp browser creates it's own 'navigational type' things, to just group stuff (example) - I think they autogenerate when a concept has too many subtypes.
Our internal tool (Lingo), also uses a few arbitrary "concepts" so that tree view/taxonomy doesn't crash. But these are build into the tool. (0-1, A-E, F-J etc...)
Maybe, they could "be refsets"? But I don't like creating more implementation hurdles.
Use of navigational concepts in value sets seems extremely marginal (and possibly erroneous). I only found 2 value sets in which navigational concepts are used:
|Sweden||2020-01-07||As far as we know no navigation concepts are used and (so far) we recommend our users against using them. No system in use that we know of can make use of navigation hierarchies.|
I agree with Australia about the complexity of using them. They are outside the model in a separate hierarchy. If we need a concept that is a navigational concept than we make a new concept in the model and don't use the navigational concept.
In our guidelines we recommend not to use them.
|Canada||2020-01-21||We did not receive feedback from our SNOMED CT Community. We are not aware if the navigational concepts are being used in systems. Like AU, we have always struggled with the navigational concepts, and how they should be used... we also think there's a lot of probably"useful" concepts in there, but implementing them in a meaningful way seems to be THE difficulty. From the end user, if SI would provide more useful guidance, this would probably be very helpful for them to "organise" their drop down lists.|
|Member countries without a CMAG rep|
|Date||CMAG Response||Next steps|
Due date has been extended to Monday 13th Jan.
Thanks to those who have already responded
I'm unable to add comments above, but we have usage of approximately a quarter of the concepts in this hierarchy. Much of it is very low level, with the exception of e.g. 394617004 | Result (approx. 5 million a year), 160237006 | History/symptoms (3/4 million a year), which are presumably being used as a template header for the relevant information section. Much of the remaining content with usage (which is all pretty low) looks to be either related to classifications-type coding, or laboratory finding groupers.