First Collaborative Datathon Between HL7 and AMIA: Informatics in the Age of FHIR
On March 26th, the first HL7-AMIA Datathon: Informatics in the Age of FHIR was held in conjunction with AMIA's 2017 Joint Summits on Translational Medicine in San Francisco, CA.
While this datathon was similar to an HL7 FHIR Connectathon, it was geared towards clinical informaticists and designed to provide them with the technical skills to evaluate and use HL7's Fast Healthcare Interoperability Resources (FHIR®) standard to support research and discovery.
It was an intense day of hands-on, heads down development, working directly with fellow participants as well as HL7 FHIR experts.
An Introduction to HL7 FHIR
While the majority of the attendees had heard of HL7 FHIR, the Datathon was their first hands-on experience using the standard.
Attendees got a crash course introduction to FHIR that included several tools and approaches to explore the standard further. As one of the event organizers, we selected many of the tools and approaches specifically to help make FHIR even more accessible to beginners.
The tools and approaches we used to introduce FHIR to attendees included:
SyntheticMass – A Realistic Synthetic Patient Clinical Data Set
To address the challenge of creating sample data for use in testing solutions and in proof of concepts, the MITRE Corporation has been working on a synthetic patient population simulator.
SyntheticMass is an open-source simulated Health Information Exchange (HIE) populated with one million realistic “synthetic residents” of Massachusetts. The data is free from cost, privacy, and security restrictions. It can be used without restriction for a variety of secondary uses in academia, research, industry and government.
SyntheticMass also provides a FHIR-based API layer that can be used to access the data stored within it. We used this FHIR server extensively during the Datathon. SyntheticMass proved to be a great data source for use during the Datathon as there was no concern about personal health information (PHI) and the database contains a lot information to explore.
Watch this video of Jason Walonoski's presentation about SyntheticMass from HL7's FHIR Applications Roundtable in March.
clinFHIR - A Tool to Play with FHIR
While HL7 FHIR is written to be easily understood by implementers, there's still much for beginners to learn, especially those who do not write code. clinFHIR, developed David Hay, is an educational tool that allows people to create or search for FHIR-based resources and link them to tell a clinical story.
clinFHIR is intended to help those not currently familiar with FHIR to understand what it is and how it can be used. It is a front-end tool that can be used to connect to any number of FHIR servers to explore and store new data.
During the Datathon, we used clinFHIR connected to the SyntheticMass FHIR server and enabled participants to start exploring the data with a point and click interface.
Those just starting to learn how to explore and build resources in FHIR should consider checking out clinFHIR. You can learn more about it on David Hay's blog Hay on FHIR.
Apache Drill - Direct Clinical BI on FHIR
The common FHIR API query capabilities are meant to address data access at the resource level. It does not address more complex query logic, such as joins, unions, and subqueries that may be needed for clinical informatics.
One approach that was taken at the Datathon to address the challenge of complex query logic was to employ more traditional analytics tools to analyze a native FHIR resource data store. Using the NoSQL to SQL engine Apache Drill to directly query FHIR formatted resources stored in NoSQL data stores, participants were able to perform complex queries. Learn more about this approach from Chris Grenz here.
This approach is quite powerful and allows one to perform common SQL-based queries directly on stored FHIR resources. Using this technique, an individual could either analyze the data directly in their own FHIR server or analyze a different data store where a synchronization approach, such as the history method, is employed.
Watch for More!
The HL7 FHIR standard has great potential to improve the accessibility and use of clinical information. Events like the HL7-AMIA FHIR Datathon introduced FHIR to a new community of users who will now be able to harness the power of FHIR for clinical informatics.
Look for HL7 to have more collaborative events like this in the future to further build the FHIR community.
