FHIR Chat · Spec notes · CARIN IG for Digital Insurance Card

Stream: CARIN IG for Digital Insurance Card

Topic: Spec notes


view this post on Zulip Josh Mandel (Oct 09 2021 at 04:42):

Just reviewing http://build.fhir.org/ig/HL7/carin-digital-insurance-card and will share a few notes here.

  • http://build.fhir.org/ig/HL7/carin-digital-insurance-card/branches/master/StructureDefinition-C4DIC-PlanBeneficiaries-extension.html defines a complex extension with 1 subscriber and 1+ beneficiaries, but there's no need for these to be grouped in a complex extension, since you can't have more than one subscriber in a Coverage resource. Furthermore there is already a Coverage.subscriber (and subscriberId) as core data elements, so no need to repeat them in a complex extension (or any extension). Suggestion for a cleaner structure:

    * Define one extension called "additionalBeneficiaries", and allow it to repeat; it should be a valueReference(Patient), and if you want to use contained resources for the patient, that's fine. Or for this use case just populate Reference.display (since you only allow a single HumanName anyway) and Reference.identifier (for your "memberId" type identifer).

    • Use top-level Coverage properties for Subscriber details.

view this post on Zulip Josh Mandel (Oct 09 2021 at 04:43):

None of the extensions have a "Context of use" defined (so they're... valid on any element of any FHIR resource, which surely isn't intended).

view this post on Zulip Josh Mandel (Oct 09 2021 at 04:46):

Definitions on http://build.fhir.org/ig/HL7/carin-digital-insurance-card/branches/master/StructureDefinition-C4DIC-CardImages-extension.html seem to be copy/pasted. I don't fully understand the use case. Is this complex extensions intended to repeat to describe multiple images each of which has its own highlights, colors, etc? If so, it shouldn't define terms like "Insurance card background color", which would appear to be a whole-card concept. And if not, why is it a complex extension rather than a flat collection of simple ones?

view this post on Zulip Josh Mandel (Oct 09 2021 at 04:47):

I can't find any language about what this IG intends to accomplish, or how. Use cases at http://build.fhir.org/ig/HL7/carin-digital-insurance-card/branches/master/Use_Case.html#use-cases and security information at http://build.fhir.org/ig/HL7/carin-digital-insurance-card/branches/master/Authorization_Authentication_and_Registration.html#smart-on-fhir-application-launch seems to be copy/pasted from a general-purpose "consumer access" guide (maybe CARIN BB).

view this post on Zulip Josh Mandel (Oct 09 2021 at 04:48):

http://build.fhir.org/ig/HL7/carin-digital-insurance-card/branches/master/StructureDefinition-C4DIC-Coverage.html should not require "profile" to be populated on resource instances.

view this post on Zulip Josh Mandel (Oct 09 2021 at 04:50):

http://hl7.org/fhir/R4/datatypes-definitions.html#CodeableConcept.coding states there can be multiple codings but all need to be identical? Should use slicing, or if it's intended to allow just one (which doesn't seem right to me) impose an upper bound of 1.

view this post on Zulip Josh Mandel (Oct 09 2021 at 04:52):

There needs to be some explanation about why the profile seems to anticipate multiple beneficiaries (see comments on extension above), where the core data structure is designed around just one (1..1). I'm concerned that the resource is being used in a way that's not intended, and that the approach will break down when fields like "costToBeneficiary" have no way to state which (of several) beneficiary they refer to, since the core model expects exactly one.

view this post on Zulip Josh Mandel (Oct 09 2021 at 04:53):

http://build.fhir.org/ig/HL7/carin-digital-insurance-card/branches/master/StructureDefinition-C4DIC-Coverage-definitions.html#Coverage.class.type should not have a required binding; additional classes should be able to use diverse vocabularies.

view this post on Zulip Ryan Howells (Oct 18 2021 at 18:25):

Thanks @Josh Mandel + other members of the team to chime in here @Mark Roberts @Cille Kissel Watkins @Pamela Maklari @Gail Kocher

view this post on Zulip Pamela Maklari (Oct 19 2021 at 04:30):

For the concern related to multiple beneficiaries and the approach for "costToBeneficiary" fields and core model expecting exactly one - I think that the amounts reflected (deductible, copay, etc.) will always be the same for the subscriber and all members. @Josh Mandel @Mark Roberts @Cille Kissel Watkins @Gail Kocher have you seen any instances where this would vary among the subscriber/dependents covered under the same policy?

view this post on Zulip Josh Mandel (Oct 19 2021 at 04:43):

I'm not familiar with this domain; just trying to provide feedback on the coherence and clarity of the proposed models

view this post on Zulip Josh Mandel (Oct 19 2021 at 04:44):

If there are multiple beneficiaries required, is this being fixed in r5 and in the fix, how are costs described?

view this post on Zulip Gail Kocher (Oct 19 2021 at 14:32):

@Pamela Maklari I've not seen where family plans have different deductible/co-pay/coinsurance for different members of the family. My experience is the amounts are for the family not specific to individuals within the family on the coverage

view this post on Zulip Cille Kissel Watkins (Oct 19 2021 at 14:34):

Thanks @Josh Mandel . Please note, most of the updates we've been working through are available on a branch here: https://build.fhir.org/ig/HL7/carin-digital-insurance-card/branches/20211011/index.html. You'll see this has a lot of more of the context as well as general cleanup/clarifications that were missing from the first build.

I agree with Pam's comments. CostToBeneficiary information is for the "plan" at large.

So you are aware, we have been discussing how to simplify the extensions for beneficiaries in our public meetings. The overwhelming majority of payers involved so far see benefit in representing all the beneficiaries like they are shown on the physical insurance card in one section on the Coverage resource, but there might be a way to avoid the "complex" extension.

No one from our team has been involved in R5 development as far as I'm aware.

view this post on Zulip Josh Mandel (Oct 19 2021 at 14:41):

Thansk @Cille Kissel Watkins. It'll be important to log issues for any aspects of the R4 models that you find you have to "work around" (and not just "augment") to meet your needs. Max cardinality on the beneficiary certainly sounds like such a place.

view this post on Zulip Mark Roberts (Nov 02 2021 at 14:29):

@Josh Mandel Hello Josh. Our next CARIN IG for Digital Insurance Card Meeting is Thursday November 4th at 1 pm ET if you'd like to join to discuss further. Zoom: https://leavittpartners.zoom.us/j/93218109773?pwd=VWRFMHYyUU1JeFErUWZyWGpzRDVWZz09
Password: 052128

view this post on Zulip Josh Mandel (Nov 02 2021 at 20:53):

Alas, I can't make this slot


Last updated: Apr 12 2022 at 19:14 UTC