Stream: C-CDA
Topic: SNOMED in C-CDA
SueAnn Svaby (Jul 29 2020 at 21:42):
I feel a little silly asking this question but my team wants me to double check. As a general rule, I understand that a binding of SHALL requires that values from the specified value set be used. Is it ever acceptable to use post-coordinated SNOMED codes in C-CDA R2.1? This particular use case if for family history. Thanks!
Lisa Nelson (Jul 30 2020 at 11:42):
@SueAnn Svaby I'm going to reach out to get a couple of other vocabulary people to jump into this discussion. In my experience, I haven't seen any IGs providing guidance on doing this. We do see some examples where a CE data type provides guidance on use of both a code and a translation, but I can't think of any specs that really get into adding qualifiers for post coordinating more complex semantics. @Rob Hausam is very involved in the Term Info project which may offer some additional theory for how and when to do this or not. @Robert McClure may have some helpful view points. @Gay Dolin or @Brett Marquard or @John D'Amore may be able to weigh in on ever seeing this done in the wild. Great questions. Thanks for asking.
Rob Hausam (Jul 30 2020 at 12:03):
This is a little tricky in CDA. The intent of having qualifiers as part of the V3 Data Types R1 CD data type was to allow this, but with that structure the ability to do true SNOMED CT post-coordination is rather limited - you can only add one or more qualifiers directly to a focus concept, and there is no way to use (or replicate) anything more complex that is supported in the SNOMED CT compositional grammar. The V3 Data Types R2 CD data type removed this limitation by making the code a string, which allows a full compositional expression to be used (and that is also how it is done in FHIR), but the R2 data types never made it into CDA. However, adding qualifiers to an existing pre-coordinated concept may be all that you need to do for something like adding laterality or severity to a concept if none of the available pre-coordinated concepts include that (for example). Anything that you do add as a qualifier should be consistent with the SNOMED CT concept model, but there isn't any way that is enforced.
Rob Hausam (Jul 30 2020 at 12:06):
I will add that it certainly may be possible to add the context for family history to an existing concept in some cases. But, again, you would need to make sure it was being done correctly according to the SNOMED CT concept model.
Rob Hausam (Jul 30 2020 at 12:18):
And finally, I guess I'll add that one danger of using post-coordination (in CDA or otherwise) is that no one else may understand what you are doing, as for the most part people have shied away from this (so far) due to the complexity and other issues. So it needs to be part of an implementation guide or otherwise verified that your trading partners will understand and be able to use it correctly.
SueAnn Svaby (Jul 30 2020 at 17:12):
Lisa Nelson said:
SueAnn Svaby I'm going to reach out to get a couple of other vocabulary people to jump into this discussion. In my experience, I haven't seen any IGs providing guidance on doing this. We do see some examples where a CE data type provides guidance on use of both a code and a translation, but I can't think of any specs that really get into adding qualifiers for post coordinating more complex semantics. Rob Hausam is very involved in the Term Info project which may offer some additional theory for how and when to do this or not. Robert McClure may have some helpful view points. Gay Dolin or Brett Marquard or John D'Amore may be able to weigh in on ever seeing this done in the wild. Great questions. Thanks for asking.
Can anyone comment on the correct way to include “No family history of” in C-CDA R2.1 if there is not an appropriate pre-coordinated SNOMED code? Thank you!
Lisa Nelson (Jul 31 2020 at 12:17):
@SueAnn Svaby
This is where you can find the available Family History examples: http://cdasearch.hl7.org/sections/Family%20History Unfortunately we don't have one yet for No family history of XXXX. This would be a very good sample. So, here's what you do. Consider how you might construct this. Look at if/how it gets done in FHIR. Consider how a negated observation is done in C-CDA --we have examples for no known problems that might suggest some options. Then, bring your sample to the Thrusday, noon - 1:00pm Examples Task Force call. It will be reviewed and adjusted/debated, and finally approved. You will have your answer and you will have helped the whole community.
I would be happy to walk you through the process for your first time bringing a sample for review. @Brett Marquard facilitates. It's a super nice group of people from the CDA Community. I highly recommend you give it a try.
The meeting information is on HL7.org in the Conference Calling Center.
SueAnn Svaby (Jul 31 2020 at 16:23):
Lisa Nelson said:
SueAnn Svaby
This is where you can find the available Family History examples: http://cdasearch.hl7.org/sections/Family%20History Unfortunately we don't have one yet for No family history of XXXX. This would be a very good sample. So, here's what you do. Consider how you might construct this. Look at if/how it gets done in FHIR. Consider how a negated observation is done in C-CDA --we have examples for no known problems that might suggest some options. Then, bring your sample to the Thrusday, noon - 1:00pm Examples Task Force call. It will be reviewed and adjusted/debated, and finally approved. You will have your answer and you will have helped the whole community.
I would be happy to walk you through the process for your first time bringing a sample for review. Brett Marquard facilitates. It's a super nice group of people from the CDA Community. I highly recommend you give it a try.
The meeting information is on HL7.org in the Conference Calling Center.
Thank you so much Lisa. I will definitely do that!
SueAnn Svaby (Aug 12 2020 at 17:39):
SueAnn Svaby said:
Lisa Nelson said:
SueAnn Svaby
This is where you can find the available Family History examples: http://cdasearch.hl7.org/sections/Family%20History Unfortunately we don't have one yet for No family history of XXXX. This would be a very good sample. So, here's what you do. Consider how you might construct this. Look at if/how it gets done in FHIR. Consider how a negated observation is done in C-CDA --we have examples for no known problems that might suggest some options. Then, bring your sample to the Thrusday, noon - 1:00pm Examples Task Force call. It will be reviewed and adjusted/debated, and finally approved. You will have your answer and you will have helped the whole community.
I would be happy to walk you through the process for your first time bringing a sample for review. Brett Marquard facilitates. It's a super nice group of people from the CDA Community. I highly recommend you give it a try.
The meeting information is on HL7.org in the Conference Calling Center.
Thank you so much Lisa. I will definitely do that!
@Lisa Nelson How do I get myself on the agenda for tomorrow's Example Task Force call? Thanks!
Lisa Nelson (Aug 18 2020 at 11:16):
@SueAnn Svaby just reach out to @Brett Marquard and ask to be on the agenda.
Last updated: Apr 12 2022 at 19:14 UTC
