Canadian FHIR Baseline Profiles - Draft Vision, Scope, Principles - for Review and Feedback

Hi Thomas,

Here is the link as originally provided by Michael:

docs.google.com/presentation/d/1w98EOZh1ZDfyJfF6V6Jav7xyIXQ25NgyJQNZjTGTsuQ/edit?usp=sharing

Andrea

Hi Andrea,

Could you please post the presentation with all comments that was presented in today's HL7 council meeting on infocentral?


Best Regards,
Thomas Zhou (AH)

I think we can make a lot of decisions about terminology bindings in a way that's independent of use-case. We can say that nationally, we'll identify drugs by CCDD code, LOINC+PLCOCD for Observations and conditions and allergies by SNOMED CT without limiting ourselves to particular use-cases. It may be that specific use-cases will need to refine those value sets (for example a diabetes registry might want to mandate support for specific codes or constrain exactly how SNOMED, LOINC + PCLOCD is used to represent diabetes-specific information, but that shouldn't stop us from being able to make a base blanket use-case independent declaration about which codes will be used where. To a large extent, we've already done that in the v3 space.

The only exception might be the 'billing' use-case which is so tied to jurisdictional requirements, but even there, billing codes can easily co-exist with CCDD, LOINC/PCLOCD and SNOMED codes given that the FHIR structures allow transmission of more than one coding.

I think Andrea's comment is coming from an assumption that, while the FHIR specifications are intended to serve a number of interactions by a variety of applications and care setting contexts, the "value" of interoperability is better understood where specific use-cases provide the necessary context to understand the terminology bindings needed supported the use case.

I strongly endorse this approach. We actually have "infostructures" in Canada that need to be interacting with a spectrum of applications at the Point-of-Service but aren't. It seems to me that, while there has been integration of lab and drug content to certain GP applications and in viewers, we are still not seeing the sharing of encounter information and observations that are occurring outside the acute care domain. The whole idea of the Shared Health Record (I am presuming this is still a desirable thing?) was that applications would be putting encounters and encounter-based observations in that shared record.

I must admit -- I think there is a core interoperability use case that would have a huge and positive impact. I'd advocate for supporting the "conversation" between a point of service (POS) application and a shared digital health infrastructure that allows the POS to:

1. Resolve the identity of a subject of care ("which Derek are you?")
2. Retrieve the health summary for a subject of care ("what is Derek's current state of health?")
3. Record observations, actions and orders related to a care encounter ("what happened during this encounter with Derek?")

There are a lot of things we could focus on... but I think we should focus on them after we are able to do this. This very non-fancy, non-sexy information exchange has huge value. Let's go to scale with this very-doable thing first... then let's get fancy and sexy.