As I read through the report from the Health IT Standards Committee Architecture, Services and API Workgroup today I was impressed. Dave McCallie and Arien Malec have done an excellent job of putting some real meat on the bones of the national interoperability roadmap.
What I really like about the SMART/FHIR emphasis is that passing patient context is core to the vision. When you’re dealing with pluggable apps in the EHR this is essential.
This got me thinking about native apps and how they should share context. All kinds of context.
Let me explain with an example. Pretend I’m a physician who is using my iPhone app as a tricorder. In today’s world I’d be using multiple apps and hardware attachments to the phone to do this. How should these apps share context such as measurements and derived data?
Yes, we would treat them as babies suckling on the EMR, but that isn’t quite what I have in mind. I’d prefer to see cooperating entities exchanging data as peers rather than subservient to large iron.
iOS doesn’t really have any context like this today. Sure, there are extension points and the new API for that, but nothing medical. Likewise Android is bereft of this capability.
As apps become more complex and need to pass data between themselves medical standards must emerge to allow for data exchange within the phone. Each app must be a first class citizen sharing data with it’s peers. In many cases, there won’t be an EMR to mediate the exchange, just a group of apps that participate to compute a result and then possibly elect to pass the data to the EMR or some other repository.
Here’s the basic idea:
- apps pull data and patient context from the EMR (or not, if no EMR is present)
- they gather data of their own
- they exchange data and compute new data
- final data is sent to a repository for storage.