Mapping to SNOMED when mapping to ATC creates ambiguity?

[annabel-uzl]

Hi, following the discussion in https://github.com/hl7-be/allergy/issues/101 and the business rules, we're trying to follow the rules for the code element:

Si on parle de substances 🡺 Référence SAM V2 (ATC)
Si on parle de produits 🡺 Référence SAM V2 (CTI-extended 🡺 CNK 🡺 ATC)
Tout ce qui n’est pas médicament (liste des aliments et substances non médicamenteuses) : Value Set CausativeAgent 🡺 SNOMED CT qui sera dans un premier temps un Value Set limité fourni par le CSCT qui sera validé par le NRC avant publication.

If there is an allergy to medication ==> map to ATC

One of the internal codes we have in UZL is "Antibiotica, cefalosporines" (ENG: Antibiotics, cephalosporins), however, there is no ATC code for all cephalosporins. There exists only a code per generation (like a class?).

• If we take ATC code J01D we also have carbapenems;
• We could put multiple ATC codes in de code-element to take all classes (from 1st to nth generation) (overkill?);
• Or, we could use SNOMED (764147003 |Cephalosporin (substance)| or 764299009 |Medicinal product containing cephalosporin and acting as antibacterial agent (product)|)

Is it an option to map to SNOMED when mapping to ATC creates ambiguity?

[mlambot]

And indeed this is why we had proposed to use SNOMED CT sometimes for drug (classes) and that we have asked for an official SNOMED-SAMv2 mapping. Which is supposed to partly exists, made by Karel de Loof (ex-AFMPS and now working at Digile), however I'm not sure how we can get access to it. I thought we could get the SNOMED codes using his SAM browser here https://samviewer.digile.be/ but I see can't seem to get one for cephalosporines when I try it now. I'll mail Karel and let you know if he has something for us. Else I'm as curious as you as to how the authorities will handle that problem. We also use very broad codes in our EHR for drug allergies sometimes and it is the SNOMED CT concepts from the substance hierarchy. We have not to this day found it possible to use the SAM in our EHR allergy module in Xcare since the clinicians don't want to record a higher degree of detail most of the time, can't actually record something better that a class when encoding past data out of the vague memory of a patient. They find the SAM too granular for most use-case. The just want to record allerty to "penicillins", "cephalosporins", things like that since once you're allergic to one you're allergic to the whole family.

[annabel-uzl]

Okay thanks! I don't know if there is still a business WG about allergy @annenerenhausen ? Otherwise, this might be something to be discussed at the next (technical) WG (this thursday)? (Although I cannot attend this week but if this can be discussed - great - otherwise I'll come back to it next week)

[bdc-ehealth]

@karel@digile.be

From: Karel De Loof ***@***.be Sent: 24 April 2024 19:46 To: LAMBOT, Marie-alexandra (CHU Saint-Pierre) ***@***.be Cc: Bart Decuypere (eHealth) ***@***.be; 'An Crepel' ***@***.be Subject: Re: Fw: [hl7-be/allergy] Mapping to SNOMED when mapping to ATC creates ambiguity? (Issue #119)

Hi Marie-Alexandra,

I hope you're doing fine. Sorry for the delay in my answering, but I already want to give you a quick answer:

The SAM viewer should indeed show all content that is in SAM. The BCFI/CBIP was/is indeed including SNOMED CT codes for the substances it uses as ingredients of its VMPs (Virtual Medicinal Products). For example, if you follow in SAM a specific package (AMPP or Actual Medicinal Product Package) to its AMP, to its VMP, to its VMPC(s) and their Virtual Ingredients, you'll arrive at a substance.

As an example for the package Cefepim Fresenius Kabi inj./inf. oploss. (pdr.) i.v. [flac.] 10 x 2 g: https://samviewer.digile.be/nl/sam/ampps/376275-02 It is linked to the VMP cefepim 2 g poeder voor oplossing voor injectie/infusie (i.v.): https://samviewer.digile.be/nl/sam/vmps/6148#medical It contains a Virtual Ingredient cefepim: https://samviewer.digile.be/nl/sam/substances/10043 And at this level you can see in SAM the SNOMED CT code 96048006 for Cefepime

However, I'm not sure if the SNOMED CT link for the newest substances was maintained by the BCFI/CBIP after I left. I took some samples and I'm afraid it doesn't look like it...

There is, a project proposal by the UCL and the Terminology center of FOD volksgezondheid to include (part of) SAM in the Belgian SCT extension, and, vice versa, to let SCT concept codes flow back to SAM (and to keep it up-to-date). We (Digile) have in association with Open Terminology Server (https://www.openterminologyserver.com/) just put ourselves forward as a candidate for the technical developments. Mainly because I really believe in the importance of keeping these things moving forward.

If it could be arranged, I'd be happy to provide consultancy. I'm a little short on time right now, but I'd be happy to discuss this further.

Have a nice evening, Karel

[KarlienHL7Belgium]

this has to be taken up with FAGG/SAMv2, will be done by @KarlienHL7Belgium together with issue 101