FHIR Chat · nl-core-address-official · netherlands

Stream: netherlands

Topic: nl-core-address-official

Martin Grundberg (Apr 05 2019 at 08:51):

Hi
We are at the moment in the process of doing a Swedish profile for patient and also have a need to state that an address is "Official". We see that Nictiz has created the extension nl-core-address-official to state this. We would like to use that extension in our swe core profile instead of creating a new one. We see that it contains information that is specific to a Dutch context, which if this extension should be used internationally, maybe belongs in each countries national profile?

What is your intention with this extension, is it to make it available for others as well, or only to use it in a Dutch context?

Regards
Martin

Alexander Henket (Apr 05 2019 at 09:02):

Good question. We have not yet considered taking this one international, but now that I know that we have a combined interest: I say let's do just that. One of us should create a ticket for FHIR Core on the Address datatype and propose a new core extension. Do you? Do I? Are you at Montreal if possible to discuss if it makes the agenda?

Martijn Harthoorn (Apr 05 2019 at 09:03):

You can always start by publishing the extension in an 'international' context.

Alexander Henket (Apr 05 2019 at 09:04):

What does that mean? Rewrite the contents so it becomes generic enough?

Martijn Harthoorn (Apr 05 2019 at 09:08):

Yes. I mean - you can make the extension generic in a collaborative effort.
Having it adopted as a core extension is a separate concern.

Martin Grundberg (Apr 05 2019 at 11:06):

@Alexander Henket , unfortunately not in Montreal :)
Great that you agree with the approach of having an international extension! Maybe it's better if you create the ticket as it is you that have created the extension? You can at least refer to the same requirement coming from Sweden.

@Martijn Harthoorn , is there a structured way of declaring the intent for an extension? We have also created/are in the process of creating extensions that we really see being international. We also try to keep those generic in terms of metadata.

Martijn Harthoorn (Apr 05 2019 at 11:57):

Extensions do have a "Purpose" field.

Martijn Harthoorn (Apr 05 2019 at 11:58):

But that is a free-text.

Michel Rutten (Apr 05 2019 at 13:58):

'StructureDefinition.jurisduction' should be suitable:
http://hl7.org/fhir/structuredefinition-definitions.html#StructureDefinition.jurisdiction

Alexander Henket (Apr 07 2019 at 17:55):

Hi @Martin Grundberg. I've rewritten our original extension a little to make it more generic. Things that came to mind:

in our context it is completely clear what official registry is meant. Hence a boolean is enough for us. Would that hold in your context too?

the boolean is min=1 and also has a meaningWhenMissing which looks like an error in thinking in the original. The intention of the meaningWhenMissing is to say what it means when you omit the whole extension, but currently it lives on the boolean element inside the extension. I've left that intact for now.

For us it is irrelevant when it became officially registered. Hence there's no timestamping.

Could you check to see if this works for you? address-official.xml

Martin Grundberg (Apr 08 2019 at 07:06):

Thanks @Alexander Henket!

In a Swedish context a person can be officially registered at one (and only one) address. This is the address that e.g. you will receive mail from governmental agencies, police, tax agency, insurance companies etc. It is also the address you will find if you search for a person's address details on various online address search services. It is not the same as a "Home address" as a person can live in two places, and have two "Home addresses", but one of them would be the "official" address.

Regarding your questions:
- Boolean is fine with us as well, it matches our requirements above
- I'm a bit unsure if 0..1 or 1..1 is better. If I interpret it correctly, having 1..1 means you can never be uncertain about whether an address is the official address? That seems to me be a potential use case. Patient gives their address, and later it is found that it is the official address. My feeling is also that restraining 1..1 could be done on a profile level, and leave the extension 0..1 for more flexibility. Thoughts?
-Timestamping is not needed for us either

A question back to you:
- Can you have multiple Dutch official addresses? As mentioned above, Swedes can't, so I suspect we would use a constraint of some sort.

Alexander Henket (Apr 08 2019 at 08:13):

The extension has context Address, so it is available for every element with datatype Address. This would mean that using the extension is sort of relative to the resource. Patient.address and Practitioner.address would have a different connotation.

We assume that if the 'official status' of an Address is unknown, then the extension would just not be used. The 'official status' extension is present when it is a known true|false, hence 1..1 Boolean. Would that work for you too?

We don't a have a 'master file' issue. There's only one official registry for a persons address and you can have only one. There's also only one registry for a business' official address. For businesses the official address maybe have limited use as businesses can have many locations. In practice the official marker is only relevant for private addresses. The person/business registries are not the same registries by the way.

Martin Grundberg (Apr 08 2019 at 11:51):

There is no issues in Sweden either regarding master files, this is a nationally held official address. And all healthcare organisations will be integrated with an external MPI that contains information about the (nationally) official address. So setup seems very similar to yours.

We were thinking of including this extension as part of the national core profile, so hence it would always be used when communicating Patient data. But I guess we could just put 0..1 on the extension itself in the patient profile and the boolean could be 1..1?

When it comes to organizational addresses I dont think we will use this extension there, but need to look into that.

Grahame Grieve (Apr 08 2019 at 21:41):

This same notion appears in multiple other countries as well. Enough to justify a 'registered' code in the standard address use value set, I think

Alexander Henket (Apr 09 2019 at 17:53):

Alright, I've made it official: GF#20743 -- Added your name in the ticket Martin.

Alexander Henket (Apr 09 2019 at 18:00):

@Grahame Grieve In V3 we mark names as official using name/@use = 'OR'. This is not supported by the FHIR datatype, but perhaps that would skip the need for the extension.

Thomas Tveit Rosenlund (Apr 10 2019 at 10:37):

We have the same extension defined in the Norwegian base profile for address:
https://simplifier.net/HL7Norwayno-basis/no-basis-address-official

However, I think @Grahame Grieve is on to something with the suggestion of registering a code in the address use CodeSystem. I think that would be a more elegant solution.

Alexander Henket (Apr 10 2019 at 16:16):

I've added Grahame's suggestion to the ticket GF#20743 and the notion that I would be sympathetic to that solution. The extension is not the goal; fulfilling the requirement is

Jens Villadsen (Apr 10 2019 at 23:10):

+1 for denmark

Martin Grundberg (Apr 12 2019 at 08:25):

The solution suggested by @Grahame Grieve works very well for us in Sweden as well (better than original one)!

Alexander Henket (May 11 2019 at 14:16):

I see the issue was reassigned from FHIR-I to MnM and MnM has a documented concern in their minutes, not yet in the issue https://confluence.hl7.org/display/MnM/MnM+Minutes+WGM+201905+Montreal.

concern is what if it is orthogonal to 'home' and 'work' (ie. is it the 'official' home address, 'official' work address?)

That sounds like a valid concern in a sense, but never stopped us before in V3. In V3 we only marked that the official person registry has this address connected to this particular person. Our registry only contains primary home addresses (one and only one, even though you might have multiple properties to your name). Not sure how that plays out in other countries.

The same address may or may not be a 'work-from-home' type of work place, but that goes into a different registry (Chamber of Commerce). You could, but we never do, mark Organization.address, Location.address or Practitioner.address etc with the proposed code/extension.

So maybe to distinguish home-official/work-official/billing-official we may need multiple codes instead of just one to exactly that end?

Martin Grundberg (May 11 2019 at 15:27):

The concepts of “home official”, “work official” etc does not exist in Sweden, or is at least not in practical use.

The address is official, whether you live there or not. Of course, you could argue that if you live at the official address it should be home+official, but I think that will add unnecessary confusion. You may not even know if the patient lives there, but it does not matter . Of course normally it would be assumed that you live there, and if you live at some other place that would more be considered some type of temporary address, or you have moved and are waiting for that to become your new official address. If you live at several places one of them would still be your official one.

So, in a Swedish context at least, information about Home, Work etc is not applicable if the address is Official.

Alexander Henket (May 13 2019 at 06:51):

Somewhat similar for us. The address is considered to be your primary address, suitable for official snail mail for example. If you own property in other places that's fine, but the primary address is where your tax mail will go for example. Considering that it is as civilian registry the thing it most most resembles is a home address, but the home/work status is really irrelevant.

Jens Villadsen (May 13 2019 at 12:57):

same for DK - there's one primary address for your snailmail. We do however have a national project that leaves room for the actual use of temporary addresses if you eg. live somewhere else for a period (week or month I would guess)

Grahame Grieve (Sep 03 2019 at 21:30):

we just made further comment on GF#20743. Further comments sought

Alexander Henket (Sep 11 2019 at 09:19):

A holiday and work-swamped days prevented me from seeing this earlier. I've added a couple of cents that are hopefully useful. Is this to be discussed in Atlanta? I'll be in Atlanta next week.

Grahame Grieve (Sep 11 2019 at 11:25):

no sure when it will next be discussed

Grahame Grieve (Sep 11 2019 at 11:38):

anyway thanks

Alexander Henket (Feb 03 2020 at 00:20):

@Giorgio Cangioli : https://jira.hl7.org/browse/FHIR-20743

Giorgio Cangioli (Feb 03 2020 at 00:43):

Thanks @Alexander Henket , unfortunately too late to contribute to the discussion :-).

Giorgio Cangioli (Feb 03 2020 at 00:43):

do you know when the extension will be released ?

Alexander Henket (Feb 03 2020 at 13:01):

Ah I totally missed that it was already on and agreement was reached. @Brian Postlethwaite could probably tell when this on the todo list: https://jira.hl7.org/browse/FHIR-20743. The answer could be "we'd welcome contributions" :-)

Giorgio Cangioli (Feb 03 2020 at 22:37):

I confess it is not fully clear to me why it is a choice boolean country code..how do you say that address this is not the official Italian address (= residenza) ?

Giorgio Cangioli (Feb 03 2020 at 23:58):

official-Address.xml

Just a first cut... this how I've interpreted the text..

Giorgio Cangioli (Feb 04 2020 at 00:00):

but I beleive that having boolean required and jurisdiction optional (allowing both) might be more useful

Alexander Henket (Feb 04 2020 at 22:09):

I don't think we discussed the use case of saying "not the official address in jurisdiction X". Indeed you would need a complex extension with 2 parts something like:

<extension url="http://hl7.org/fhir/StructureDefinition/official-Address">
    <extension url="boolean">
        <valueBoolean value="false"/>
    </extension>
    <extension url="country-or-region">
        <valueCodeableConcept>
            <coding>
                <system value="urn:iso:std:iso:3166"/>
                <code value="IT"/>
                <display value="Italy"/>
            </coding>
        </valueCodeableConcept>
    </extension>
</extension>

I think the boolean part would be required (1..1) and the country-or-region part optional or even repeating (0..*)?

Giorgio Cangioli (Feb 04 2020 at 22:33):

Yes this is what I think is useful, but unfortunately it is not what the resoultion says ... but maybe we can ask to revise it...:-)

Alexander Henket (Feb 04 2020 at 23:04):

@Lloyd McKenzie https://jira.hl7.org/browse/FHIR-20743 discusses how to mark Address as 'official'. As far as I can see the resolution has not led to an extension just yet. @Giorgio Cangioli, added a use case to the decision reached in the ticket that could be considered:

The ticket proposes:

<extension url="http://hl7.org/fhir/StructureDefinition/official-Address">
    <valueBoolean value="true"/>
or
    <valueCodeableConcept>
        <!-- CodeableConcept e.g. iso code -->
    </valueCodeableConcept>
</extension>

This would not allow to be specific about non-official addresses. You could solve that by making the extension one level deeper:

<extension url="http://hl7.org/fhir/StructureDefinition/official-Address">
    <extension url="boolean">
        <valueBoolean value="false"/>
    </extension>
    <extension url="country-or-region">
        <!-- CodeableConcept e.g. iso code -->
    </extension>
</extension>

Would that be something that could be considered somewhere?

Lloyd McKenzie (Feb 05 2020 at 08:45):

Sure. Submit a new tracker item and reference the original. That said, sending a flag with false seems odd to me. Wouldn't it make more sense to just flag the ones that are true?

Alexander Henket (Feb 05 2020 at 22:42):

@Giorgio Cangioli since you raised this option: could you look into Lloyds question/suggestion?

Giorgio Cangioli (Feb 05 2020 at 22:56):

I'd expect that from the absence of the extension you cannot infer that the address is not an official address.

Alexander Henket (Feb 05 2020 at 22:58):

Yes that is correct, but is there a use case to explicitly say "not an official address for Italy". If there isn't then the limitation does not need solving.

Giorgio Cangioli (Feb 05 2020 at 23:18):

In italy we have two types of home addresses (that in general are coincident) ("domicilio" where you actually live and "residenza" were you officialy live). Residenza is important for knowing which region will pay for your care...

Giorgio Cangioli (Feb 05 2020 at 23:19):

If I want too say this is the "domicilio" my idea is to use official=false....

Giorgio Cangioli (Feb 05 2020 at 23:21):

so if no one care about the fact that this is (or is not) an Italian official address we can stay with the current proposal and I can ignore the jurisdictional part of the extension..

Giorgio Cangioli (Feb 05 2020 at 23:21):

otherwise I'll create a new tracker item

Giorgio Cangioli (Feb 07 2020 at 03:24):

added a jira tracker item J#25938

Last updated: Apr 12 2022 at 19:14 UTC