It’s safe to assume that if you’re part of the healthcare technology industry, you’ve heard the buzz around FHIR (HL7’s Fast Healthcare Interoperability Resources data standard).
With the swirl around FHIR, what it does, how it operates, and why it matters—it can feel challenging to cut through the noise. When examined, FHIR’s ability to redefine the healthcare space becomes evident.
Understanding interoperability
Healthcare interoperability is hard. It’s a witches’ brew of biologic variability, new tools to measure our unique characteristics, the complexity of available care, and the constant economic pressures shaping electronic data. All these factors combine to make the process of sharing electronic descriptions of patients hard.
Today, we live with high-speed connectivity and supercomputers in our pockets. Most non-healthcare technology is being constantly redefined by high-speed networks as apps talk with back-end servers that power them. The technology delivering U.S. healthcare, however, is still largely defined by non-market mechanisms and a sea of healthcare policy “fixes” dating back to World War II. Past healthcare data interoperability standards such as HL7’s Version 2.x and the Clinical Document Architecture were only optimized for the software tools and transmission technologies of the time. FHIR is the extraordinary data standard that takes the modern concept of an Internet addressable resource and makes those resources available for us to use in our current healthcare system.
Interoperability on FHIR
With personal kudos to Grahame Grieve, the Australian polymath who imagined how this could all work over a decade ago, FHIR sits as a healthcare data model layer on the underlying data standards of JSON (Javascript Object Notation) and XML (extensible markup language). It allows the information needed to provide clinical care to be represented within our software systems. At its heart, FHIR defines roughly 150 “resources,” i.e. core concepts such as a Patient, Observation, or Encounter. The HL7 standards group supports the consistency of this usage with electronic tools such as implementation guides and social tools like “connectathons” which iron out the many details. When combined with the open-source software programs that support the modern Internet, FHIR’s use of JSON as the data format that powers RESTful API (Representational State Transfer Application Programming Interface) provides an extraordinary platform to get the interoperability we need.
So, how do we revalue healthcare?
Historically, healthcare interoperability has been defined as moving pieces of patient data between dedicated healthcare software platforms such as electronic medical record systems (EMRs) and lab or imaging systems. However, thanks to new legislation, today FHIR serves as the core data standard to allow patients to access medical records on their smartphone. In 2016 the 21st Century Cures Act stated that patients have the right to get their medical information on their terms, and also defined APIs(application programming interfaces) required of electronic health records be “without special effort”—in other words, proprietary lockdowns of patient data are not allowed. There has been a lot of discussion about how FHIR APIs will transform today’s patient-provider interactions into the modern service economy we enjoy in the rest of our lives.
FHIR and “Big Data”
While FHIR will certainly fuel a redefinition of healthcare at the individual patient level, some of the most powerful healthcare transformations using FHIR will come at the patient population level. Absent a market economy, the existential question of defining value in healthcare remains unsolved. Today’s quality measures do not scale to the value questions we need to answer as a society. We lack access to the information to combat pandemics. With significant parts of our population underserved, we have not been able to provide the computational services to help those less fortunate. To answer each of these questions, we need to use analytics on a patient population. FHIR and related work from groups such as HL7, the CARIN Alliance, and the DaVinci and Gravity Projects as well as ONC [The Office of the National Coordinator for Health IT] and CMS [The Centers for Medicare and Medicaid Services] offer us the tantalizing opportunity to use the richly computable representations in FHIR to let modern machine learning and artificial intelligence tools inform how we can provide care.
FHIR makes these promises real by providing richly defined and individually computable resources that are more likely to make “big data” approaches informative. The Bulk FHIR API standard required of EHRs at the end of 2022 will provide uniform data sets protected under HIPAA for the providers and payers we have entrusted with our care.
The future of healthcare and FHIR
Searching for value in healthcare is a challenge and while FHIR is central to thinking about healthcare data, many of us are working hard to build the surrounding tools to make the opportunities offered by FHIR a daily reality. FHIR will enable tools to define meaning over vast data sets, to match what we have paid for care with what we have received, to provide rich privacy and security models, and to link all of this data in real time as patients seek care and providers offer that care. With cloud platforms offering near infinitely scalable FHIR-enabled computing we each have the chance to rethink how we provide care.
Donald Rucker, M.D. is the chief strategy officer at 1upHealth. Rucker is a former National Coordinator for Health Information Technology (ONC) in the United States Department of Health and Human Services (HHS).