Long Term Observations
We solved for a scenario where a clinician is looking a multiple conflicting blood pressure observation results and needs to determine which is the latest. Think of a patient that presents into an ER with a broken arm and then follows up with their primary care physician. Both encounters will generate blood pressure observations.
In this demo, we are the case nurse who is looking at a blood pressure centials app. Unfortunately, the observations are conflicting. Which is the lastest correct observation? It looks like someone fat fingered one of the data values. It also looks like our latest data is in conflict - we can mark the correct entry.
We will mark the correct entry into the blockchain. This represents a single source of truth across all data sources.
Our team leveraged 4 main components, the UI, the FHIR Servers for the data, the blockchain API, and the infrastructure to run these components. We leveraged a SMART on FHIR application to present the user interface, HL7 FHIR's API and data model to access off chain data, and Monax's aka Eris tools to interface into the block chain. We put a smart contract on the Observation FHIR data model resource.
FHIR is HL7's latest data and api standard for interoperability.
Some of the discussion items that we talked about is how much data to store in the chain, how to handle historical data to handle analytic functions.
While the example is a bit stretched, the combination of these technologies presents an opportunity to ensure data consistency, provenance, and immutability.