Stream: terminology
Topic: Code system for DEA class code
Mona O (Aug 30 2019 at 13:44):
Looking to provide the DEA class code. Would https://www.deadiversion.usdoj.gov/schedules/ be the appropriate code system to reference? Within MedicationKnowledge does DEA class code align to "regulatory" or "medicineClassification"?
Rob Hausam (Aug 30 2019 at 14:47):
I'm not sure what you're referring to as "DEA class" here. Are you looking at codes for the DEA schedules (1 - 5), or something else?
Mona O (Aug 30 2019 at 14:52):
Yes, DEA schedules (1-5)
Rob Hausam (Aug 30 2019 at 15:42):
I suspect that codes for the DEA schedules likely are already available somewhere, but I'm not certain about that offhand. They're not in the US Edition of SNOMED CT. Pharmacy folks may know. @Melva Peters?
John Moehrke (Aug 30 2019 at 15:48):
wouldn't each schedule be managed as a ValueSet? Managed by whom, I don't know..
Rob Hausam (Aug 30 2019 at 16:05):
I think we're talking here about codes for the schedules themselves (5 codes), not value sets for the lists of drugs that are included in each schedule. Of course, both may be needed. @Mona O can confirm.
Mona O (Aug 30 2019 at 16:25):
@Rob Hausam , correct just looking for the schedules themselves (1-5). The schedule "codes" are classified as:
'1'| 'No accepted medical use'
'2'| 'High abuse, severe dependence liability'
'3'| 'Moderate dependence liability'
'4'| 'Limited abuse potential'
'5'| 'Limited abuse potential, small amounts'
Looking for the appropriate code system to represent. Would this align with FHIR's criteria for code system? https://www.deadiversion.usdoj.gov/schedules/#define
Rob Hausam (Aug 30 2019 at 16:32):
Yes, I think we can use that, but we should do a little more due diligence first to see if we already do have something available for this, and we need to include HTA in the discussion. @Robert McClure @Julie James
Mona O (Aug 30 2019 at 17:25):
@Rob Hausam Thank you! Please let me know if there's already a defined code system for DEA class schedules.
Grahame Grieve (Aug 30 2019 at 22:32):
I recommend against including a # in a canonical URL for a code system
Rob Hausam (Aug 30 2019 at 23:08):
good point
Robert McClure (Aug 31 2019 at 01:28):
I think we need to stop trying to use the url that we happen to find that has the information we need, and use that as the canonical url, when the organization responsible for the content is not working with us. This is a perfect example - the US government is not interested , yet, in supporting this use. I'd rather we make up a canonical url, based on a defined process, and then capture the "source of information" url someplace else.
Grahame Grieve (Aug 31 2019 at 06:13):
that's what we do with sid (system id) when the source doesn't respond. So that wouldn't be a new approach
Julie James (Sep 03 2019 at 11:40):
Gosh, there is a need for some terminology teaching! There is a difference between a "code system" which is a *managed collection of Concept Representations and "a set of things that someone has put in a document" - no matter how official that document. It looks to me as if the DEA list is the latter. Most countries have something similar, and usually an organisation for whom this is important takes responsibility to make the official government "set of things" into a code system that they properly manage; so this might be a pharmacist organisation like AHFS/ASHSP or a terminology organisation like RxNorm/NLM. It's usually not the government organisation that authors the document though, because it's not usually in their mindset. But it would be nice if it was. So, who in the US can or is doing this? What about NCPDP as another possibility?
Julie James (Sep 03 2019 at 21:02):
HI All. We discussed this on the HTA call. And in doing so, I realised that these concepts are already present in a code system - the NCI Thesaurus. So a value set can be drawn from that to meet the requirement as expressed here. No need to invent anything new :) Here is the link for one of the concepts - https://ncit.nci.nih.gov/ncitbrowser/ConceptReport.jsp?dictionary=NCI_Thesaurus&version=19.07e&ns=ncit&code=C48676&key=n2079684485&b=1&n=null
Schedule III Substance (Code C48676) - the other four are also present.
Grahame Grieve (Sep 03 2019 at 21:27):
what are the other 4? it's not clear to me which C48676 actually is?
Rob Hausam (Sep 03 2019 at 22:14):
This is the value set with all 5 codes.
Rob Hausam (Sep 03 2019 at 22:17):
it's described as being for SPL (Structured Product Labeling), but there's no reason it would be limited to that
Rob Hausam (Sep 03 2019 at 22:23):
there's also an NCPDP value set which adds "Unspecified" - and at least a couple of others that mix these with other sets of codes
Grahame Grieve (Sep 03 2019 at 22:25):
so the FHIR equivalent would be
{ "resourceType": "ValueSet", {url, status etc} "compose": { "include": [ { "system": "http://ncimeta.nci.nih.gov", "concept": [ { "code": "C48672", "display": "Schedule I Substance" }, { "code": "C48675", "display": "Schedule II Substance" }, { "code": "C48676", "display": "Schedule III Substance" }, { "code": "C48677", "display": "Schedule IV Substance" }, { "code": "C48679", "display": "Schedule V Substance" } ] } ] } }
Rob Hausam (Sep 03 2019 at 22:28):
yes, that would be it
Mona O (Sep 04 2019 at 13:39):
Thank you for looking into this. For my understanding would this be a valueset that would be owned by FHIR or would each implementer need to create their own valueset for DEA class codes?
Rob Hausam (Sep 04 2019 at 15:33):
I believe it makes the most sense to create a value set for this in FHIR, based on the NCI Metathesaurus codes (essentially as Grahame has above).
Mona O (Sep 04 2019 at 15:57):
Any thoughts to what the URI would be or when it would be defined by?
Rob Hausam (Sep 04 2019 at 16:03):
I presume we would do this in terminology.hl7.org (and ultimately in UTG)? @Ted Klein?
Mona O (Sep 12 2019 at 06:35):
Should I presume value set would reflect the NCI code and the display would link to how they are listed via the below? https://ncit.nci.nih.gov/ncitbrowser/pages/concept_details.jsf?dictionary=NCI_Thesaurus&version=19.08d&code=C48670&ns=ncit&type=relationship&key=null&b=1&n=0&vse=null
C48672 - Schedule I Substance C48675 - Schedule II Substance C48676 - Schedule III Substance C48677 - Schedule IV Substance C48679 - Schedule V Substance
How would you recommend I implement the code, display and reference the value set, since the value set is not currently available on the FHIR?
Rob Hausam (Sep 12 2019 at 12:11):
Are you working on an IG (presumably)? If so, I would define the value set there. Ultimately, I would think this might be defined as part of US Core (I didn't see it there - and it is US realm)? @Eric Haas @Brett Marquard
Brett Marquard (Sep 12 2019 at 14:57):
It would be wise to work this through pharmacy before US Core. i don't remember hearing this request.
Eric Haas (Sep 12 2019 at 18:01):
We limit the codes to the US Core IG to the its use cases and profiles. In the past we have discussed the fact that there is a lot of US Realm specific stuff in the FHIR specification and have decided not to move it to its own realm specific registry or to US Core. I think another good option would be defining in VSAC.
Rob Hausam (Sep 12 2019 at 18:25):
yes, VSAC certainly makes sense, and should be sufficient
Rob Hausam (Sep 12 2019 at 18:27):
we could also reference it directly from NCI - but VSAC would probably be more convenient for most
probably is a good choice
Last updated: Apr 12 2022 at 19:14 UTC
