Dave, your article is nothing if not provocative and insightful. What does it mean for health data to be truly free? At MultiScale, we have been thinking hard about health data: what it means to free it and, most importantly, what is required to liberate it so that those who need it the most can use it to add care value.
To free health data, we first need to free it from where it naturally resides—the electronic health record (EHR). But, at MultiScale we believe health data shouldn’t be removed from the EHR. It serves a purpose there. Instead, health data should be replicated in a secure place with more computing resources and applications.
What is this place? Let’s call it a “Freedom Platform.”
Creating a freedom platform may sound great, but it hasn’t been done until now because it requires thinking about and working with health data in new and inventive ways. Here’s how to build a freedom platform.
Health Data Must Be Transferred
Replicating data has many benefits—not the least of which is relieving the burden from healthcare IT. The MultiScale solution frees data by transferring it from its source production location to the cloud. Data streams to the cloud as it is created in the EHR “ethosphere” e.g. interface engines, data entry, telemetry. The same goes for ERP, Lab IS, and all other relevant data sources.
But transferring is not enough. Some health data enthusiasts think that applying HADOOP processes in the cloud to raw, transferred data is a great way generate new value from the data. This idea has created considerable buzz about data lakes.
But healthcare data is messy. The larger EHR vendors purportedly have upwards of 150,000 data fields, not to mention the countless unstructured data elements in physician freeform note “fields.” This is the “data fog.” HADOOP/map-reduce works for discovery, but it’s not ideal for continuously streaming data for critical clinical and operational decision making.
Health Data Must Be Transformed
Even when transferred, the data is not yet freed if it’s still in raw form. It needs to be unencumbered from its source structure and transformed for meaning and usability.
Free data should not be aged data. Data needed by caregivers now must be transformed now. Those familiar with the “T” in ETL: Extract-Transform-Load realize the traditional batch nature of ETL will not work to free data into the flow of healthcare operations.
The MultiScale Hive platform uses a transformation process that continually processes new data and derives new results and insights from the data—which is itself data. This new data creates critical information.
Health Data Must Be Liberated
The next aspect of freeing data is dislodging it from the transformed databases and event queues where it is stored so that it can be put into the hands of users who need to see it. The most natural way to put data into the hands of users is via apps on their own personal or employer-provided phones.
Once health data is freed and accessible via mobile app, users can subscribe to the apps that provide the data they need to communicate, collaborate and make decisions with their colleagues and peers. When everyone is collaborating and communicating about and with data in real time, the data is truly free—free to meet patient needs promptly.
We call this the TTS Liberation process: Data is Transferred – Transformed – Subscribed to (TTS). TTS Liberation moves data from its transaction source to users who need it, in a curated form. If a machine-learning process determines an event that might cause readmission, that result is pushed to the users who subscribe and act on that event. The TTS Liberation process aligns perfectly with what you describe in your article as the Free, Curate, and Collaborate (FCP) process.
Finding a way to safely liberate health data for access and use in new and meaningful ways is a big step forward. In fact, it creates the foundation for a real-time health system. But it’s only the start for the “Freedom Platform.”
Down the road, MultiScale, like many others, needs to find more ways to address consumer health needs. We need to integrate IoT data sources, consumer-device data, and any data that can be aggregated for better care. For example, in many health systems genomic data is connected as an extension of the EHR, but there is still a long way to go to cleanly integrate it.
The data needed to fulfill the promise of precision medicine is another area of aspiration. Health systems need ways to continually feed scientific discovery-inspired data into their larger data sets. A big step toward realizing those aspiration will be the ability to correlate the latest pharmacogenomic studies tied to genetic markers (genomic, proteomic, etc.) with patient data.
To be fully liberated, all relevant health data needs to flow to the right platforms, be processed for the right uses, at the right time, for the benefit of caregivers, patients and all concerned.
Keep inspiring us Dave.
We have work to do to achieve your data freedom manifesto: Access, Insights, Rights, Flow.
Jim Harding, CEO/Founder, MultiScale Health Networks