RegulatedAuthorization: subject can be PackagedProductDefinition

[rlindstrm]

In several countries (e.g Portugal) marketing authorisations are issued per package, not per medicinal product. The core IDMP model also allows marketing authorisation to be linked to a medicinal product or a packaged product.

ePI currently only allows MedicinalProductDefinition for RegulatedAuthorization.subject. I would recommend adding reference to PackagedMedicinalProduct in addition.

[cander2]

I would advise against that. PackagedProductDefinition is meant to describe the packaging itself and should not be used in place of a product.

For those countries, it should have one medicinal product, one authorization, and one packaged product definition per authorized presentation.

[joofio]

How would you distinguish the medicinal products then? They would be the same

Craig Anderson @.***> escreveu em seg., 27/11/2023 às 23:24 :

I would advise against that. PackagedProductDefinition is meant to describe the packaging itself and should not be used in place of a product.

For those countries, it should have one medicinal product, one authorization, and one packaged product definition per authorized presentation.

— Reply to this email directly, view it on GitHub https://github.com/HL7/emedicinal-product-info/issues/101#issuecomment-1828723966, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADMJVUDWVUD36LFHWMOXI4TYGUHKVAVCNFSM6AAAAAA736PW72VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMRYG4ZDGOJWGY . You are receiving this because you are subscribed to this thread.Message ID: @.***>

[cander2]

Each medicinal product would have a unique number.

Like here you could have the MA number https://www.ema.europa.eu/en/documents/all-authorised-presentations/suliqua-epar-all-authorised-presentations_en.pdf

The PackagedProductDefinition only represents the packaging by itself which is the last three columns.

[rlindstrm]

I would advise thinking twice before knowingly going against ISO IDMP core specification and EMA IG.

It's very unlikely that marketing authorisation procedures would be changed because of ePI specification.

[cander2]

Important to remember that this is a global IG (i.e., not just capturing an EU/EMA centric viewpoint) and also important to note that this is not just meant for regulators only. It is also for industry, healthcare and patients.

[rlindstrm]

I understand your point, but you're not just going against EMA, but also against IDMP data model.

The consequences may arrive much later, because the technical workarounds are likely to cause some confusion. For example, let's take MedicinalProductA, and 3 packages A10, A20, A30, all of which have been assigned a separate Marketing Authorisation.

The structure of this data in FHIR would be as follows:

What ePI proposes now seems to be this (and it would make it impossible to understand which authorization goes for which package):

And I'm not sure if I got it right what you were proposing, but it sounded like creating MedicinalProductDefinitions with fake identifiers to keep one MP per Package? So, something like this maybe?

You cannot do the third option without creating fake IDs, because ePI allows multiple MedicinalProductsDefinitions in one Bundle, so you need to forcefully alter the data to keep links clear between a package, authorisation and product. But when you create fake IDs for MPDs, it would become difficult to manage the rest of the data.

[joofio]

I wont even go through the idmp compliance. If there are use cases of products being authorized through the package, we should allow it. Not only this example if from europe, but if we want to make this ig global, i believe we have to be flexible.

[cander2]

Thanks for the images. This helps. Do you have time to chat this week?

I'm free (EST): Tuesday 0900-1000 Wednesday 0700-0800 Thursday 0600-0900 Friday 0700-0800 (or after 0900)

Maybe working through a real-life example will help me. Think we are already on the same page with cardinality and I suspect we're headed in a similar direction for structure but there are a few things I am not following. Need to understand this better so I can write a description in the IG that ensures ePI can work across markets.

Regarding the images, definitely not 2. And definitely wasn't thinking of fake identifiers. More like a variation of image 1 where MA is above MPD plus the MID and APD in there.

[joofio]

@cander2 i believe we should change this. do you agree?

[rlindstrm]

Hi @cander2 I'm mostly available on Friday. Contact me on Zulip to organise the call. :) I think I can find some examples from UNICOM (will ask @joofio if I don't).

[cander2]

@joofio , @rlindstrm are the images above aligned with this: http://hl7.org/fhir/medication-definition-module.html?

Similar to Rutt's diagram, this link also has PPD linking to MPD. But the images don't have the MID. Is MID left out just because it was a quick image or do you mean that we would have a pack pointing to the MPD with no MID at all?

[cander2]

That would mean these diagrams need to be updated as well right? http://build.fhir.org/ig/HL7/emedicinal-product-info/steps-to-create-epi3.html#steps-to-create-type-3-epi-dcoument

[joofio]

The diagrams are not aligned, but I see them as examples. Nevertheless, the base specification is. It enables several resources to be subject of the RegulatedAuthorization.

As for the second diagram, it needs to be updated yes.

[cander2]

What would this change look like? RegulatedAuthorization.subject would point to MPD or PPD?

Do we have an example of an ePI that has an authorization for just a PPD alone without any MPD?

[joofio]

to one or the other.

Karvea is like that I guess, https://www.ema.europa.eu/en/documents/all-authorised-presentations/karvea-epar-all-authorised-presentations_en.pdf

[cander2]

@cander2 i believe we should change this. do you agree?

ok, let's make the change then.

[joofio]

https://github.com/HL7/emedicinal-product-info/pull/102

[rlindstrm]

What would this change look like? RegulatedAuthorization.subject would point to MPD or PPD?

Do we have an example of an ePI that has an authorization for just a PPD alone without any MPD?

Even if the authorization is for PPD, the MPD still always exists. So one MPD typically has several PPD-s, but the authorization can be for MPD or for each PPD separately. On your diagram, just one arrow should be added to show that RegulatedAuthorization.subject can be PPD.

MID is referenced from the PPD. The same MID could be referenced from different PPD-s. So for that part, everything remains the same.

[cander2]

Ah ok, thanks Rutt. This part "...the MPD still always exists." is key for me. I was thinking it would lead to PPDs by themselves.

I think I follow now.

[cander2]

@joofio, @rlindstrm, we ok to close this one since Joao implemented the change?

[adamzkover]

I would suggest to update the diagrams before closing this issue. Some readers, like me, might begin there to see what links where, before (or instead of) checking the resource profiles...

[cander2]

ok, let's leave this open until we update the IG