Stream: implementers
Topic: SMART EHR Launch and MRN
Brendan Keeler (Dec 15 2020 at 22:52):
When doing a SMART launch from an EHR context, is it common for the MRN to be provided in the access token response?
I was thinking it would always be just the FHIR patient ID and then to get MRN and other patient info by GETting the Patient Resource, but had a discussion where others said they're seeing different things in the wild.
Jenni Syed (Dec 15 2020 at 23:02):
We generally do not recommend that apps use "custom smart context" for things like MRN that will be gathered on a read of patient (or of encounter for FIN)
Jenni Syed (Dec 15 2020 at 23:02):
ie: prefer the standard way and use a custom way for things that are harder
Jenni Syed (Dec 15 2020 at 23:03):
we have apps that don't follow that though :)
Jenni Syed (Dec 15 2020 at 23:03):
if you use the custom way it essentially means you're required to do custom coding for every vendor
Jenni Syed (Dec 15 2020 at 23:04):
Not only that, but for eg, the oid needed to align that MRN to a specific org isn't in our custom context - you can only get that through the FHIR read
Jenni Syed (Dec 15 2020 at 23:04):
so there are tradeoffs...
Josh Mandel (Dec 15 2020 at 23:26):
It is certainly not the intention behind the standard that patient identifiers would be provided in ad hoc ways :-)
Brendan Keeler (Dec 16 2020 at 02:30):
For sure, totally agree.
My guess: for apps living in a hybrid world still doing HL7v2, but hoping to use of SMART for launching given its relative advantages over SAML, it may be they may want a single SSO transaction, avoiding the true FHIR pieces.
From Jenni's answer, it sounds like at least Cerner allows some configuration around custom context extensions. I suppose other EHRs do as well, which might be where the person I was taking to got their info.
Last updated: Apr 12 2022 at 19:14 UTC
