Question on vaccinations in SAM2 not seemingly supported by SNOMEDCT VaccineCodes

[hansdekeersmaecker]

After checking the list of SNOMEDCT values in https://ehealth.fgov.be/standards/fhir/vaccination/ValueSet-be-vs-vaccine-code.html I had a question on some vaccinations present in SAM2 and the codes to use for them.

Vaccinations under 'Diverse antibacteriële vaccins' in the tree structure of BCFI: https://www.bcfi.be/nl/chapters/13?frag=11833 Example: https://samviewer.digile.be/nl/sam/ampps/137322-02

Vaccinations under 'Vaccin tegen RSV' in the tree structure of BCFI: https://www.bcfi.be/nl/chapters/13?frag=mps-8904013 Example: https://samviewer.digile.be/nl/sam/ampps/661565-01

These don't seem to be supported by a SNOMEDCT code, but can be administered currently in Belgium. I suppose these are considered as 'other' currently, but was expecting a SNOMEDCT code for all current vaccinations in SAM2. Will there be an addition for them in the future?

[HanneVuegen]

Kind reminder for this question: is it currently on the roadmap of SAMv2 to add a mapping to snomedct codes? Or what is the proposed solution?

Is it an acceptable risk if we leave this to the enduser to register the correct Snomed code for a vaccine (VaccineCode) to a CTi extended/CNK code for a product (AdministredProduct)?

[bdc-ehealth]

This might be a question for Jeroen De Wilde (RIZIV/INAMI)

[KarlienHL7Belgium]

@KarlienHL7Belgium will put this on the agenda of the meeting with SAMv2, NRC might have to be included

[bdc-ehealth]

@mlambot

can you provide 2 SNOMED CT codes for the vaccines mentioned by @hansdekeersmaecker

[mlambot]

The RSV vaccine concept is 51311000087100 |Vaccine product containing only Human orthopneumovirus antigen (medicinal product)|. This concept was created in October 2023, the vaccine having been approved only in summer 2023 in US. This was after the Vaccine care set value set was created and the latest BE extension 20231115 depends from the July 2023 edition of the international SNOMED CT Core so this concept doesn't exist yet for the current BE extension. It will be present in the 20240515 BE release though so it can be added to the BE vaccination refset.

I see that in the BCFI there are two vaccines, one against subtypes A and B of the virus and one against only subtype A so maybe it's worth asking for that distinction to SNOMED international. Is this distinction important/wanted in clinic? I suppose yes like the distinction for the HPV vaccines or meningococcal serotypes. If you confirm the need of the distinction, I can ask the BE NRC to send a request to SNOMED international for two more refined vaccin concepts, one for vaccine against RSV subgroup A and B, and one for vaccin against RSV subgroup A only.

[mlambot]

Now bronchovaxum isn't a vaccine. It's a bacterial lysate acting as an immunostimulant. So it's a drug or a complement, that has its place in the medication line but having it in the vaccination card is debatable since it's not really a vaccine. It's a decision for the vaccination user group to take, to include things that are not strictly vaccines in the vaccination care set. It could open the door to many things ending up in the vaccination care set.

I don't find readily a concept to represent that bacterial lysate in the International edition. Maybe I don't find the right keywords, maybe it doesn't exist. I've sent a question to SNOMED international. If the concept doesn't exist then the BE NRC will need to make a content request submission and/or to create it in the BE extension.

[mlambot]

Answer from SNOMED drug support group UK member

From: Stuart Abbott stuartabbott@nhs.net Sent: 02 May 2024 15:23 To: Drug Extension User Support Group deusg@ihtsdo.org Cc: LAMBOT, Marie-alexandra (CHU Saint-Pierre) marie-alexandra.lambot@stpierre-bru.be Subject: Re: Question about immunostimulant bacterial lysate

Hello Dr Lambot, I don't think these are well represented in SNOMED CT if at all. I suspect this is due to bacterial lysates being mainly a European product, but not a UK product. Until recently there was therefore no need to develop content to include them. The closest I can come up with is that taking the view that a 'bacterial lysate' is a type of 'bacterial antigen'.

There's an EU document: https://www.ema.europa.eu/en/news/bacterial-lysate-medicines-respiratory-conditions-be-used-only-prevention-recurrent-infections which states: "Bacterial lysate medicines are made from bacterial cells that are broken down and are intended to stimulate the immune system to recognise and fight infections." With an 'antigen' defined as "Any substance that causes the body to make an immune response against that substance. "

It could therefore be inferred that any bacterial lysate product is a type of 836368004 |Vaccine product containing bacteria antigen (medicinal product)| A vaccine meaning 'a substance used to stimulate immunity to a particular infectious disease or pathogen, typically prepared from an inactivated or weakened form of the causative agent or from its constituents or products.'.

I can't find anything in SNOMED CT that does cover that, and to be honest calling them a vaccine product feels somehow wrong.

The only other place, though still with no detailed content might be: 81759008 |Immune enhancement agent (product)|

Stuart Abbott UK NRC

[mlambot]

Dear Bruce, you who knows so much about immunologic pathways, would you have an opinion about how to model those extracts like the broncho-vaxum products? Would you accept them in the vaccine hierachy and as consequence in the vaccination card of a patient or would you be more inclined as I am to consider this as aspecific immunomodulation, to be recorded in the general medication list of the patient but not in the liste of received vaccine products (immunization ressource in FHIR)?

From: Bruce Goldberg bjgoldbergmd@outlook.com Sent: 02 May 2024 17:02 To: LAMBOT, Marie-alexandra (CHU Saint-Pierre) marie-alexandra.lambot@stpierre-bru.be; Stuart Abbott stuartabbott@nhs.net Subject: Re: Question about immunostimulant bacterial lysate

All, these agents represent mixtures of lysates from several bacteria. I would not call these antigens which usually refer to substances that result in specific adaptive immune responses. The substances in bacterial lysates are PAMPS (pathogen-associated molecular patterns) which are conserved molecular structures found on pathogens that non-specifically activate innate immune responses by binding to pattern recognition receptors such as mannose binding lectins, C-reactive protein, toll-like receptors, etc. Bacterial lysates are usually called immunostimulants or immunomodulators. I think they should be placed under 372558009 |Substance with immunomodulator mechanism of action (substance)|.

Bruce.J.Goldberg, MD, PhD Physician director, SCPMG Regional Allergy-Immunology Laboratory Los Angeles, California

[mlambot]

BE-NRC request entered under https://jira.ihtsdotools.org/browse/BERMP-794

[HanneVuegen]

Could it be an option to use ATC as a fall back codesystem for vaccineCode?

As soon as a vaccine code in SAMv2 is made available (CNK, ATC,...) the SNOMED code should also be available (preferably even mapped). Is that realistic? Because if that isn't the case it won't be possible to register vaccinations for that specific vaccine... unless we allow ATC codes as well.

FYI: the European specifications for vaccinations also allow both snomed and ATC.

[mlambot]

Making a SNOMED CT concept for a vaccine takes I'd say maximum a half day so this isn't the limiting step. What can be is that you need to publish it in your extension/reference national server for people to be able to use it. Again that can be worked around since the NRC is setting up more frequent releases and there is always the possibility to have emergency releases like happend for Covid vaccines. So the true need is for the NRC to be aware that a new code is going to be needed. A new market authorization doesn't pop up out of thin air. All it would take is actually for the federal medicament agency to send notice to the BE NRC of the drugs that are about to be approved. The NRC could then have the concepts representing them ready for release once they get to the market. They could even release the new drug concepts before, when drugs are on trial use in the country under temporary or test authorizations. SNOMED international has a rule about not creating drugs that are only in trial phase in the Core but each country is free to have its own rules in its own (drug) extension. So having the SNOMED CT codes at the same time as the ACT ones for new vaccines isn't a terminology issue but a talk-together-between-federal-instances issue. A process needs to be set up.

I would put forward my old notion again. All this would be nicely settled if one would make a law that in Belgium the pharmaceutical companies are obliged to give the stated SNOMED representation of their active ingredient and medicinal product to the Belgian NRC as part of their authorization for market demand file. They are the best placed to know what their drug is as substance, where it and their product fits on the general drug ecosystem of parents and children. It would relieve the NRC of a lot of work they don't have currently the means to tackle. It would put the responsibility of the meaning of the concept on the people who best know it and have already the legal responsibility of the substance. And I would go fruther. I would have it made mandatory that the drug companies make available the side effect notice of their drugs in SNOMED CT concepts. That they give a SNOMED equivalent of their users notice so that could be used in decision support without third parties having to take time and legal risks doing those mappings.

Companies have money. And a limited number of products. This is a drop in the sea for them. They know best their products. And if we make it mandatory to give the SNOMED representation of drug/product/side effects as part of the notice, then any update of the notice will see them update this also. All drug data would then always be up to date in SNOMED CT. It is the only way to make the process sustainable for a country, and for Europe too. Not one NRC who'd have to update thousands of drugs definition, not software vendors having to manage thousands of side effects updates, each drug company fully responsible to deliver the digital version of their drug/product/notice.

I've been proposing it for three years everywhere. It makes utter sense. If we did a POC of this with one big corp, it would give them an advantage on the market, in safety, in making customers happy and then all the others would follow, so as not to be left behind, so competitive is that market. We actually had one company that was interested to try back before covid but back then we didn't have the authoring tools at the CSCT and you can't create medication products without a true SNOMED authoring tool so it stayed in the dream box. But now with medication line getting ready, vaccination going live, it'd say it's the time to go dig that out of the idea box. That's just my 2 cents.

[mlambot]

Setting aside the philosophical discussion about dialogue between industry, AFMPS and NRC, practically, there is a field/element in the data model for the actual product that is administered in the care set, called "Product", where you can put the ATC/CNK/CTI-extended code.

The SNOMED CT code used in the VaccineCode element is for the disorder one vaccinates against. If you don't have a SCT code in the list that correctly represents the organisms/diseases the vaccin is against, you can use the main vaccine code, or the closed vaccine codes (we could add to the value set groupe codes of vaccine against bacteral, vaccine against virus) and add the missing detail in the notes, like the rule for the allergy substances if your exact substances isn't in the refset. I believe there is a code for "other" in that value set, that may be used too in the VaccineCode field if it wasn't suppressed in the process somewhere. If it was then we fall back on the business rule of using the closest incomplete concepts and adding the missing flavour in notes, regularly looking at those notes as input for missing content to be created and crreating what is missing as soon sa a gap is discovered. So not having the exact SNOMED CT code for vaccination against a specific agent in VaccineCode doesn't block vaccination registration not the use of the exact CTI/CNK/ATC product code in Product.

[KarlienHL7Belgium]

1. short term solution: have a Snomed code for vaccin --> ongoing: NRC and Marie-Alexandra will set these up (through Snomed int'l) - @mlambot : do you have an idea by when this will be done ? can you please specify which ones will be added in a short term
2. mid term solution - having vaccine in SamV2: need for process how to do this in a structured way - WG thinks that RIZIV should be the owner (together with SAMv2 board)- contactperson: Jeroen De Wilde

[bdc-ehealth]

see #155 for a partial solution of this issue.