Stream: patient empowerment
Topic: Don't use Task its too Generic
Virginia Lorenzi (Jan 30 2021 at 21:23):
I have heard the argument that the problem with using Task for Patient Request for Correction is that Task as a resource is just too generic. Pushing back, you really can't get more generic than Observation. And I see Observation implemented in many of the EHR's APIs but it doesn't support every kind of observation. The EHRs say that. You can say I support these types (which perhaps happen to have terminology codes behind them).
Lloyd McKenzie (Jan 30 2021 at 21:55):
Spinning it a different way, are the EHRs saying they'd prefer to have a specific resource for "patient correction"? If that's too specific, what's the sweet spot?
Dave deBronkart (Jan 31 2021 at 19:47):
Interesting discussion! In reality is the problem that we don't yet consider a patient request for amendment to be "real" EMR work, like any other task would be?
Dave deBronkart (Jan 31 2021 at 19:57):
I'm also wondering - are there other transactions using FHIR that trigger any sort of negotiation?
It happens all the time in finance: customer disputes a credit card transaction. Years ago I crossed paths with the Online Dispute Resolution people, whose ODR.info site has grown into the National Center for Technology and Dispute Resolution. I've been telling them for years that they should watch our space.
Two interesting things happened once laws were passed saying that credit card companies have to eat the damage, to some extent, if they're too lax to prevent fraud. (I may have specifics wrong but that's the gist.) First, the companies got real tight on fraud prevention - all the improvements in card security, PINs, rapid detection of questionable transactions, etc. Second, they got really interested in processing disputes efficiently, and an industry sprung up, specializing in it, which led eventually to ODR.
I won't be surprised if something similar happens in health records over the next 5, 10, 20 years.
Josh Lamb (Jan 31 2021 at 21:12):
It will be a base IG, that does not specify a use case, and the use case specific IGs can be slotted in. We require servers to support the Base IG, and they decide what use cases to adopt.
I am not sure why this in controversial, since it increases patient and provider choice while also meeting the needs of the PE use cases.
I think a call would be beneficial.
Cooper Thompson (Feb 01 2021 at 14:27):
I mean there have been a bunch of discussions about Observation being too generic as well...
Brendan Keeler (Feb 01 2021 at 15:11):
+1 to Cooper. Observation is generic enough that most want to change SMART scopes to work around that and add granularity. So it's not exactly the best example to model after.
Lloyd McKenzie (Feb 01 2021 at 16:09):
Observation being a single resource is driven by the fact that a whole lot of systems don't (and can't) differentiate. The biggest challenge with Observation is that Observation.category got softened to the point where systems can't reliably use it to differentiate internationally.
Cooper Thompson (Feb 01 2021 at 16:45):
I mean if systems can't different between a blood pressure and sexual history or menstruation observation, then they have massive privacy issues they have to deal with when it comes to sharing that data via an API on the consumer side.
Last updated: Apr 12 2022 at 19:14 UTC
