1043 View
19 CommentIn discussionComments enabled
In the category:
DEUSG
Hi All,
As a follow up to our DEUSG meeting last week please use this discussion thread for any feedback on additional attributes being added to the MRCM for use in national drug extensions.
Shane Byrnes, if it would be helpful, I’m happy to share the full set of AMT MRCM extensions here. This includes:
Changes to the domain and range for existing international attributes
Additional attributes defined by AMT, along with their intended purpose and usage – for example, Linda discussed 30465011000036106 | Has container type | at the Oslo meeting
It would also be valuable to gather MRCM extensions from other drug extensions wherever they exist.
We could build a shopping list of attributes for other NRCs to consider. If others have similar needs, optional attributes and MRCM extensions could be added to the international edition to make them easy to use - some already exist like "Has supplier".
This would be better than NRCs having to develop their own attributes and MRCM extensions or "side promote" – which can be both time-consuming and complex.
Before I go ahead, could I please get permission to upload files (e.g. images) to this page? I’d like to include visual examples alongside the text, which will make the explanation more concise and accessible.
Here is a summary of the full set of attributes proposed by Canada (including potential opportunities for promotion from the AMT and CMT). NOTE: Our priorities are to agree on the Container type, Subpacks (containered packages), and Multicomponent attributes.
Container and device: Promote the following attributes from the AMT:
30465011000036106 |Has container type| - Grouped
999000081000168101 |Contains device| - Grouped
Either reuse the international attribute 1142142004 |Has pack size (attribute)| or add a new attribute |Has device count| for this - Grouped
999000101000168108 |Count of device type|
Subpacks: Either promote the following AMT attributes
999000011000168107 |Contains packaged clinical drug| – as long as this actually means "Contains containered packaged clinical drug" - Grouped
999000091000168103 |Count of containered package type| - Grouped
Either reuse the international attribute 1142142004 |Has pack size (attribute)| or promote the CMT attribute 62601000087108 |Has subpack count| OR promote the following CMT attributes (if my assumption about the 'contains packaged clinical drug' attribute above isn't correct):
62511000087106 |Contains subpack| - This defines the inner containered packs within a multi-container product - Grouped
62601000087108 |Has subpack count| - The number of this particular type of subpack (ie containered pack) in the outer package. - Grouped
62521000087103 |Count of subpack type| - The number of different types of containered packs
Multi-component MP/MPF/CD: Add the following new attributes (required, but not yet created in the CMT)
|Has medicinal product component| – The component MPs of a multi-component MP - Grouped
|Count of medicinal product component type| – The number of types of MP components (to ensure universal restriction logic)
|Has medicinal product form component| – The component MPFs of a multi-component MPF - Grouped
|Count of medicinal product form component type| – The number of types of MPF components (to ensure universal restriction logic)
|Has clinical drug component| – The component CDs of a multi-component CD - Grouped
|Count of clinical drug component type| – The number of types of CD components (to ensure universal restriction logic)
Additional defining information – Promote the following CMT additional defining attributes (assuming that they don't exist anywhere else)
62591000087100 |Has excipient ingredient| – A non-active (excipient) substance in the drug product
62611000087105 |Is free of excipient ingredient| – A non-active (excipient) substance, which is not contained in the drug product.
62721000087105 |Has dose form characteristic| – Additional product characteristics that relate to the dose form, e.g. “Light” (of powder)
62731000087107 |Has unit of presentation characteristic| – Additional product characteristics that relate to the unit of presentation, e.g. “Single dose”, “Multidose”
92611000087107 |Has dose form after transformation| – Required to distinguish between dose forms that have the same transformation method, but transform into different dose forms, e.g. granules for oral solution, granules for oral syrup.
Non-defining information (additional relationships) – Additional non-defining attributes that others may need as well
62191000087109 |Has registered route of administration|
62701000087104 |Target population| (as a non-defining characteristic)
62181000087107 |Has product registration status|
62151000087104 |Has drug schedule|
62581000087102 |Has drug class|
Non-defining information (annotations) – Additional non-defining concrete values that others may need as well
62711000087102 |Has Global Trade Item Number (GTIN)|
62201000087106 |Has unit of use GTIN|
62161000087101 |Has package GTIN|
62131000087108 |Has WHO ATC code|
62141000087102 |Has last batch on market product expiration date|
After chatting with Dion and Matt (Australia) to clarify the meaning of some of the AMT attributes, I have a few adjustments to make in the proposed list of attributes below:
NOTE: Canada's priorities (in blue) are to agree on the Container type, Subpacks (containered packages), and Multicomponent attributes.
Container and device: Promote the following attributes from the AMT:
30465011000036106 |Has container type| - Grouped
999000081000168101 |Contains device| - Grouped
Add a new attribute |Has device count| for this - Grouped (... or reuse the international 1142142004 |Has pack size (attribute)| as the AMT has done)
We've come to the conclusion that the semantics of this are the same as Canada's 62511000087106 |Contains subpack| attribute, so are happy to use the AMT attribute if promoted. However, we would request a clarification of the FSN to distinguish the term 'packaged clinical drug' from the existing class with the same name in SNOMED Int.'s national drug extension model (which defines the quantity of each clinical drug within, without distinguishing how the inner containers are organised). Possible attributes names include "Contains containered package" or "Contains package in container" or "Has subpack".
CMT: 62601000087108 |Has subpack count| - Grouped
Note: The AMT reuses an attribute like 1142142004 |Has pack size (attribute)| for this, which would be the alternative.
AMT: 999000091000168103 |Count of containered package type|
Multi-component MP/MPF/CD: Add the following 4 new attributes (and reuse the 2 existing international attributes below) with 2 role chains (1) |Contains medicinal product form| o |Has active ingredient| → |Has active ingredient|, (2) |Contains clinical drug| o |Has manufactured dose form| → |Has manufactured dose form|
SCTID |Contains medicinal product| – The component MPs of a multi-component MP - Grouped
SCTID |Contains medicinal product form| – The component MPFs of a multi-component MPF - Grouped
774160008 |Contains clinical drug| - The existing attribute (to be a subtype of |Contains MPF| and |Contains MP|
SCTID |Count of medicinal product type| – The number of types of MP components (to ensure universal restriction logic)
SCTID |Count of medicinal product form type| – The number of types of MPF components (to ensure universal restriction logic)
1142143009 |Count of clinical drug type| – The existing international attribute
Additional defining information – Promote the following CMT additional defining attributes
92611000087107 |Has dose form after transformation| – Required to distinguish between dose forms that have the same transformation method, but transform into different dose forms, e.g. granules for oral solution, granules for oral syrup.
62591000087100 |Has excipient ingredient| – A non-active (excipient) substance in the drug product
62611000087105 |Is free of excipient ingredient| – A non-active (excipient) substance, which is not contained in the drug product.
62721000087105 |Has dose form characteristic| – Additional product characteristics that relate to the dose form, e.g. “Light” (of powder)
62731000087107 |Has unit of presentation characteristic| – Additional product characteristics that relate to the unit of presentation, e.g. “Single dose”, “Multidose”
Non-defining information (concept values)
62191000087109 |Has registered route of administration|
62701000087104 |Target population| (as a non-defining characteristic)
62181000087107 |Has product registration status|
62151000087104 |Has drug schedule|
62581000087102 |Has drug class|
Non-defining information (concrete values)
62711000087102 |Has Global Trade Item Number (GTIN)|
62201000087106 |Has unit of use GTIN|
62161000087101 |Has package GTIN|
62131000087108 |Has WHO ATC code|
62141000087102 |Has last batch on market product expiration date|
Hello everyone, a CRS request has been received from Australia to promote the attribute Has container type (attribute) for Canada.
This request has been discussed internally with the Chief Terminologist, Dr Jim Case, who has commented: Since this attribute is intended for use in defining medicinal products, should the FSN be specific to that use case? There are other use cases for container types that have been proposed (e.g. specimens, non-medicinal products).
My thoughts are that something like something like Hasmedicinecontainer type (attribute) or Has drug container type (attribute) may be required.
Per email from Canada additional information referring to 6 CRS requests we have received from Australia:
To centralize the set of container types to use under 398124009 |Drug container (physical object)|, we have requested Australia to submit these concepts (as well as the AU attribute) for lateral promotion to core:
#970722 -> 999000071000168104 |Package (physical object)| - no longer needed by CA - request withdrawn
#970723 -> 710001000168107 |Pouch (physical object)| - no longer needed by CA - request withdrawn
By "lateral" promotion, I mean going directly from one extension to another extension without going up to the core, which is something we do very rarely. But it sounds like we're getting set up here to move a number of concepts from AU up to the International Edition, which is more usual.
In that case, do you think we should put a halt to MSSP-4007 and instead expect these additional attributes be listed in CRS requests, promote them to the International Edition and the IE could inherit them from there, rather than having them in their own extension directly? Or (given that CA have withdrawn the request for Package (physical object), are those 3 I listed not under consideration for inclusion in the International Edition and MSSP-4007 should proceed as is?
Therefore as 999000071000168104 |Package (physical object)| has been withdrawn I don't see the 3 you have listed in Blue above. Nor are they in CRS.
For me promoting 30465011000036106 |Has container type (attribute)| through CRS works well as the use case and references can be provided, and Jim was able to document his comment in the CRS ticket so the discussion can be retained in one place on that particular request.
The issue with pausing MSSP-4007 is that Ireland needs these concepts for their work, and delaying international promotion would block their progress.
“Lateral promotion” offers a way forward by allowing Ireland to use the concepts now, without conflict if they’re later promoted internationally. It effectively unblocks them while international promotion of the required content is still pending - or may never happen.
So unless the content Ireland wants to promote is about to be released in the international edition, MSSP-4007 should proceed. This seems like a common scenario where extensions have deadlines and can’t wait for international timelines. Otherwise, their only alternative is to create duplicate concepts in their extension - which we’d all prefer to avoid.
The work here should not affect the lateral promotion of concepts, and including the 3 requested. Therefore this DEUSG discussion board item does not prevent
MSSP-4007
-
Getting issue details...STATUS
from going ahead.
Peter G. Williams - I think "lateral promotion" really is the classic(simple) promotion you describe. A proper "lateral promotion" isn't something I want to think about logistically or technically
As for comment/question "Since this attribute is intended for use in defining medicinal products, should the FSN be specific to that use case? There are other use cases for container types that have been proposed (e.g. specimens, non-medicinal products). My thoughts are that something like something like Has medicine container type (attribute) or Has drug container type (attribute) may be required."
I'm not sure why the attribute couldn't be used for the other use cases (specimens/non-drugs). It means the same thing in all instances "the type of container the thing is in".
"Sudocrem 15.25% cream, 250 g, jar" : has container type = Jar
"Urine specimen in a jar" : has container type = Jar
"250g jar of peanut butter" : has container type = Jar
Now there's a different question as to if these are different types of jars (containers)... But that's about the range (physical object) concepts, not attribute. To be honest I'm not totally sure we need 398124009 |Drug container (physical object)|? Maybe it's necessary to differentiate a regular bottle from a medicine bottle (which is a specific bottle).... But then we (AU) have a bunch of issues like this with the existing container concepts and subheirarchy content... Look at the subtypes of 37284003|Bag, device (physical object)| as an example. Parenteral/enteral solution bag (physical object) is not currently a subtype of Drug container (physical object) ...
This points to a clean up of device/object hierarchy is needed
But I can't see value in multiple different attributes, even if the domains are different.
The reason for my comment was that the definition provided for the attribute was specific to medicinal products. If we would like to generalize the definition to all containers, then my issue goes away.
See Jim Case comment ("Since this attribute is intended for use in defining medicinal products, should the FSN be specific to that use case? There are other use cases for container types that have been proposed (e.g. specimens, non-medicinal products)) and Matt's feedback above.
Note: 398124009 |Drug container (physical object)| exists - with no subtypes.
Re Matt's point: "But I can't see value in multiple different attributes, even if the domains are different. We already have attributes with different domains e.g.ATT 42752001 Due to - HRCM20230531 - SNOMED Confluence"
Does this multi-domain proposal seem acceptable from an ontology and technical aspect?
As I mentioned above, my real issue is that the definition related to this request was specific to medicinal product containers. If everyone is happy with a general HAS CONTAINER TYPE, then the issue goes away.
I created a ticket for MRCM to summarize the request for the attribute 30465011000036106 |Has container type (attribute)| at
ATF-1703
-
Getting issue details...STATUS
The domain is restricted to << 781405001 |Medicinal product package (product)|. This attribute is only applicable to packaged products. 763158003 |Medicinal product (product)| should not include information about packages. The MRCM would need to be expanded in the Australia extension to support their use cases and model. Matt CordellDion McMurtrie
The range is restricted to < 49062001 |Device (physical object)|. There are more specific concepts such as 706437002 |Container (physical object)|, 398124009 |Drug container (physical object)|, 434711009 |Specimen container (physical object)|, and 732990009 |Container (unit of presentation)|. If the range were restricted to 398124009 |Drug container (physical object)|, we would need to make this attribute name more specific, as mentioned by Jim Case to support different use cases that might require different ranges. The current implementation of MRCM limits us from assigning different ranges to ‘has container type’ for different domains. Therefore, the high-level range 49062001 |Device (physical object)| should be used.
We may argue that concepts 706437002 |Container (physical object)|, 398124009 |Drug container (physical object)|, 434711009 |Specimen container (physical object)| are not needed for this model. The attribute has represented the intended use or function of devices as a container. The value of this attribute can be any device that can be used as a container. These concepts are devices grouped by their function as containers and things could be contained within them, e.g. medicinal products, specimens, food, etc. A device can have different functions or intended uses, e.g. measurement. Groupers can be helpful in certain contexts, but they can also be contradictory in others. We need to review whether we should retain these container concepts and their usage as suggested by Matt Cordell
The attribute is not grouped and the attribute cardinality is 0..1. This means that we can only have one container type for a medicinal product package (MPP). A package has different containers that need to be modeled by 'MPP' contains multiple 'MPPs' that each contained MPP has a container type. Because the container types have been covered by the contained MPP, do we still need 'Has container type' be modeled in a multi-container product? What would be the value, 49062001 |Device (physical object)| or 706437002 |Container (physical object)|? We need confirmation about the agreement on the modeling of the subpacks. Linda BirdDion McMurtrie
Yongsheng Gao It seems like you're encountering the issues we've been trying to highlight for a number of years.
The Domain isn't just "Medicinal product package" (Though that might seem logical) - because drug extensions want to talk about container types elsewhere in the model. "Amoxicillin 500 mg injection, vial" "Paracetamol 1 g/100 mL injection, bottle" Similar concepts exist in a number of extensions and they use the "Has unit of presentation" attribute for this (in compliance with the INT model)If we consider a concept like ""Paracetamol 48 mg/mL oral liquid, 100 mL, bottle" Is this a "Medicinal Product" or "Medicinal Product Pack"?In AMT we say this is a Medicinal Product Pack. Other extensions say it's a Medicinal Product. We just use "Has pack size" "Has Container type" on MP and MPP concepts instead of "Has presentation size" and "Has unit of presentation". Essentially means the same thing. Extensions are already implementing the model in different ways because of the ambiguity/confusion.
The range is restricted to < 49062001 |Device (physical object)| - because of there are targets outside this range. e.g. 469844003|Ampoule|, 68276009|Bottle|, 705586008|Jar| etc. We (Extensions) can't do much about the quality of the Physical Object heiarchy, and can't wait for these to be addressed through CRS submissions. We can submit the ones we know about. But until we're confident 706437002 |Container (physical object)| subsumes everything required the range is set a bit wider.
I don't think we want to rush into getting rid of groupers. Whilst "any container" could hold a specimen. "Specimen container" are designed to - e.g. may be sterile. Similarly for some drug containers.
Container types might still be useful for multipacks. You can have 6 bottles in a box. Maybe you don't care if it's a box or something else, and can just say "Pack" or nothing at all. As above (and through the years of discussions) the requirements for drugs extensions are varied - and there's no value in making the model too restrictive. There are lots of changes we would like to make with AMT, but we have requirements based on our regulator and reimbursement legislation etc that determine much of what is needed.
Regarding the “groupers”, I still find them useful - though I don’t really think of them as groupers, but rather as more abstract classes, depending on the concepts in question.
To me, true “groupers” are things like "Finding of sensation by site" - concepts that don’t carry much clinical meaning and mainly exist for hierarchical navigation or organisational purposes. They’re not really intended for use beyond that.
But something like Container is genuinely useful when I need to refer to a broad class. It’s exactly the kind of abstraction that lets SNOMED CT play to what I see as one of its greatest strengths - supporting the right level of precision for the task at hand, without forcing unnecessary specificity.
A Briefing Note will be discussed at the next Editorial Advisory Group: Changes to 706437002 |Container (physical object)| and the physical object MRCM
Thanks for the summary Yongsheng Gao! With respect to the |Has container type| attribute:
The CMT is using a domain of << 781405001 |Medicinal product package| for this ... so as long as these are included in the international domain, we're covered
The CMT is using a range of << 398124009 |Drug container (physical object)| ... so as long as these are included in the international range, we're good
The CMT does not group this attribute, so we're okay with this approach.
The CMT uses a cardinality of [0..1] on MPPs in general - but we have exactly one ([1..1]) container on each of our "Containered clinical drug" concepts (which sits below "Packaged clinical drug" in the model). Note: Allowing a container to be added on a multi-container product enables the ability to define both the outer container type (on the multi-container MPP) and the inner container types (on the contained MPPs), if required.
19 Comments
Dion McMurtrie
Shane Byrnes, if it would be helpful, I’m happy to share the full set of AMT MRCM extensions here. This includes:
Changes to the domain and range for existing international attributes
Additional attributes defined by AMT, along with their intended purpose and usage – for example, Linda discussed 30465011000036106 | Has container type | at the Oslo meeting
It would also be valuable to gather MRCM extensions from other drug extensions wherever they exist.
We could build a shopping list of attributes for other NRCs to consider. If others have similar needs, optional attributes and MRCM extensions could be added to the international edition to make them easy to use - some already exist like "Has supplier".
This would be better than NRCs having to develop their own attributes and MRCM extensions or "side promote" – which can be both time-consuming and complex.
Before I go ahead, could I please get permission to upload files (e.g. images) to this page? I’d like to include visual examples alongside the text, which will make the explanation more concise and accessible.
Linda Bird
Here is a summary of the full set of attributes proposed by Canada (including potential opportunities for promotion from the AMT and CMT).
NOTE: Our priorities are to agree on the Container type, Subpacks (containered packages), and Multicomponent attributes.
OR promote the following CMT attributes (if my assumption about the 'contains packaged clinical drug' attribute above isn't correct):
Linda Bird
After chatting with Dion and Matt (Australia) to clarify the meaning of some of the AMT attributes, I have a few adjustments to make in the proposed list of attributes below:
NOTE: Canada's priorities (in blue) are to agree on the Container type, Subpacks (containered packages), and Multicomponent attributes.
Nicola Ingram
Hello everyone, a CRS request has been received from Australia to promote the attribute Has container type (attribute) for Canada.
This request has been discussed internally with the Chief Terminologist, Dr Jim Case, who has commented: Since this attribute is intended for use in defining medicinal products, should the FSN be specific to that use case? There are other use cases for container types that have been proposed (e.g. specimens, non-medicinal products).
My thoughts are that something like something like Has medicine container type (attribute) or Has drug container type (attribute) may be required.
Yongsheng Gao re MRCM.
Peter G. Williams
Just to join some dots here, we've also got MSSP-4007 on the board, which is a request for a lateral promotion from AU to IE for the following:
CC Elena Ilyukhina
Nicola Ingram
Hello Peter G. Williams
Per email from Canada additional information referring to 6 CRS requests we have received from Australia:
To centralize the set of container types to use under 398124009 |Drug container (physical object)|, we have requested Australia to submit these concepts (as well as the AU attribute) for lateral promotion to core:
#970722 -> 999000071000168104 |Package (physical object)| - no longer needed by CA - request withdrawn
#970723 -> 710001000168107 |Pouch (physical object)| - no longer needed by CA - request withdrawn
#970724 -> 287011000036106 |Blister pack (physical object)|
#970727 -> 277011000036103 |Tube (physical object)|
#971421 -> 117941000036106 |Tub (physical object)|
#971420 -> 30465011000036106 |Has container type (attribute)|
Peter G. Williams
By "lateral" promotion, I mean going directly from one extension to another extension without going up to the core, which is something we do very rarely. But it sounds like we're getting set up here to move a number of concepts from AU up to the International Edition, which is more usual.
In that case, do you think we should put a halt to MSSP-4007 and instead expect these additional attributes be listed in CRS requests, promote them to the International Edition and the IE could inherit them from there, rather than having them in their own extension directly? Or (given that CA have withdrawn the request for Package (physical object), are those 3 I listed not under consideration for inclusion in the International Edition and MSSP-4007 should proceed as is?
CC Elena Ilyukhina Shane Byrnes
Nicola Ingram
Peter G. Williams Elena Ilyukhina My understanding is for this DEUSG project we are prioritising the ones in Blue above.
Therefore as 999000071000168104 |Package (physical object)| has been withdrawn I don't see the 3 you have listed in Blue above. Nor are they in CRS.
For me promoting 30465011000036106 |Has container type (attribute)| through CRS works well as the use case and references can be provided, and Jim was able to document his comment in the CRS ticket so the discussion can be retained in one place on that particular request.
Dion McMurtrie
The issue with pausing MSSP-4007 is that Ireland needs these concepts for their work, and delaying international promotion would block their progress.
“Lateral promotion” offers a way forward by allowing Ireland to use the concepts now, without conflict if they’re later promoted internationally. It effectively unblocks them while international promotion of the required content is still pending - or may never happen.
So unless the content Ireland wants to promote is about to be released in the international edition, MSSP-4007 should proceed. This seems like a common scenario where extensions have deadlines and can’t wait for international timelines. Otherwise, their only alternative is to create duplicate concepts in their extension - which we’d all prefer to avoid.
Nicola Ingram
Peter G. Williams Dion McMurtrie Elena Ilyukhina the Content team discussed this issue internally earlier.
The work here should not affect the lateral promotion of concepts, and including the 3 requested. Therefore this DEUSG discussion board item does not prevent MSSP-4007 - Getting issue details... STATUS from going ahead.
Matt Cordell
Sorry I missed the meeting last night.
Peter G. Williams - I think "lateral promotion" really is the classic(simple) promotion you describe. A proper "lateral promotion" isn't something I want to think about logistically or technically
As for comment/question
"Since this attribute is intended for use in defining medicinal products, should the FSN be specific to that use case? There are other use cases for container types that have been proposed (e.g. specimens, non-medicinal products). My thoughts are that something like something like Has medicine container type (attribute) or Has drug container type (attribute) may be required."
I'm not sure why the attribute couldn't be used for the other use cases (specimens/non-drugs). It means the same thing in all instances "the type of container the thing is in".
Now there's a different question as to if these are different types of jars (containers)... But that's about the range (physical object) concepts, not attribute.
To be honest I'm not totally sure we need 398124009 |Drug container (physical object)|?
Maybe it's necessary to differentiate a regular bottle from a medicine bottle (which is a specific bottle)....
But then we (AU) have a bunch of issues like this with the existing container concepts and subheirarchy content... Look at the subtypes of 37284003|Bag, device (physical object)| as an example.
Parenteral/enteral solution bag (physical object) is not currently a subtype of Drug container (physical object) ...
This points to a clean up of device/object hierarchy is needed
But I can't see value in multiple different attributes, even if the domains are different.
We already have attributes with different domains e.g. ATT 42752001 Due to - HRCM20230531 - SNOMED Confluence
Jim Case
Matt Cordell
The reason for my comment was that the definition provided for the attribute was specific to medicinal products. If we would like to generalize the definition to all containers, then my issue goes away.
Nicola Ingram
Hello Yongsheng Gao 30465011000036106 |Has container type| is now a request in CRS:
https://request.ihtsdotools.org/#/requests/preview/971420?fromList=true
See Jim Case comment ("Since this attribute is intended for use in defining medicinal products, should the FSN be specific to that use case? There are other use cases for container types that have been proposed (e.g. specimens, non-medicinal products)) and Matt's feedback above.
Note: 398124009 |Drug container (physical object)| exists - with no subtypes.
Re Matt's point: "But I can't see value in multiple different attributes, even if the domains are different. We already have attributes with different domains e.g. ATT 42752001 Due to - HRCM20230531 - SNOMED Confluence"
Does this multi-domain proposal seem acceptable from an ontology and technical aspect?
Jim Case
Nicola Ingram
As I mentioned above, my real issue is that the definition related to this request was specific to medicinal product containers. If everyone is happy with a general HAS CONTAINER TYPE, then the issue goes away.
cc: Yongsheng Gao
Yongsheng Gao
I created a ticket for MRCM to summarize the request for the attribute 30465011000036106 |Has container type (attribute)| at ATF-1703 - Getting issue details... STATUS
The domain is restricted to << 781405001 |Medicinal product package (product)|. This attribute is only applicable to packaged products. 763158003 |Medicinal product (product)| should not include information about packages. The MRCM would need to be expanded in the Australia extension to support their use cases and model. Matt Cordell Dion McMurtrie
The range is restricted to < 49062001 |Device (physical object)|. There are more specific concepts such as 706437002 |Container (physical object)|, 398124009 |Drug container (physical object)|, 434711009 |Specimen container (physical object)|, and 732990009 |Container (unit of presentation)|. If the range were restricted to 398124009 |Drug container (physical object)|, we would need to make this attribute name more specific, as mentioned by Jim Case to support different use cases that might require different ranges. The current implementation of MRCM limits us from assigning different ranges to ‘has container type’ for different domains. Therefore, the high-level range 49062001 |Device (physical object)| should be used.
We may argue that concepts 706437002 |Container (physical object)|, 398124009 |Drug container (physical object)|, 434711009 |Specimen container (physical object)| are not needed for this model. The attribute has represented the intended use or function of devices as a container. The value of this attribute can be any device that can be used as a container. These concepts are devices grouped by their function as containers and things could be contained within them, e.g. medicinal products, specimens, food, etc. A device can have different functions or intended uses, e.g. measurement. Groupers can be helpful in certain contexts, but they can also be contradictory in others. We need to review whether we should retain these container concepts and their usage as suggested by Matt Cordell
The attribute is not grouped and the attribute cardinality is 0..1. This means that we can only have one container type for a medicinal product package (MPP). A package has different containers that need to be modeled by 'MPP' contains multiple 'MPPs' that each contained MPP has a container type. Because the container types have been covered by the contained MPP, do we still need 'Has container type' be modeled in a multi-container product? What would be the value, 49062001 |Device (physical object)| or 706437002 |Container (physical object)|? We need confirmation about the agreement on the modeling of the subpacks. Linda Bird Dion McMurtrie
Matt Cordell
Yongsheng Gao It seems like you're encountering the issues we've been trying to highlight for a number of years.
"Amoxicillin 500 mg injection, vial"
"Paracetamol 1 g/100 mL injection, bottle"
Similar concepts exist in a number of extensions and they use the "Has unit of presentation" attribute for this (in compliance with the INT model)If we consider a concept like ""Paracetamol 48 mg/mL oral liquid, 100 mL, bottle"
Is this a "Medicinal Product" or "Medicinal Product Pack"?In AMT we say this is a Medicinal Product Pack. Other extensions say it's a Medicinal Product.
We just use "Has pack size" "Has Container type" on MP and MPP concepts instead of "Has presentation size" and "Has unit of presentation".
Essentially means the same thing. Extensions are already implementing the model in different ways because of the ambiguity/confusion.
e.g. 469844003|Ampoule|, 68276009|Bottle|, 705586008|Jar| etc.
We (Extensions) can't do much about the quality of the Physical Object heiarchy, and can't wait for these to be addressed through CRS submissions.
We can submit the ones we know about. But until we're confident 706437002 |Container (physical object)| subsumes everything required the range is set a bit wider.
Similarly for some drug containers.
Maybe you don't care if it's a box or something else, and can just say "Pack" or nothing at all.
As above (and through the years of discussions) the requirements for drugs extensions are varied - and there's no value in making the model too restrictive.
There are lots of changes we would like to make with AMT, but we have requirements based on our regulator and reimbursement legislation etc that determine much of what is needed.
Dion McMurtrie
I think Matt Cordell has pretty much covered it.
Regarding the “groupers”, I still find them useful - though I don’t really think of them as groupers, but rather as more abstract classes, depending on the concepts in question.
To me, true “groupers” are things like "Finding of sensation by site" - concepts that don’t carry much clinical meaning and mainly exist for hierarchical navigation or organisational purposes. They’re not really intended for use beyond that.
But something like Container is genuinely useful when I need to refer to a broad class. It’s exactly the kind of abstraction that lets SNOMED CT play to what I see as one of its greatest strengths - supporting the right level of precision for the task at hand, without forcing unnecessary specificity.
Nicola Ingram
A Briefing Note will be discussed at the next Editorial Advisory Group:
Changes to 706437002 |Container (physical object)| and the physical object MRCM
See attached 2025-07-29 SNOMED Editorial Advisory Group Conference Call
Linda Bird
Thanks for the summary Yongsheng Gao! With respect to the |Has container type| attribute: