FHIR expected to supplement, not replace, other standards
While Health Level Seven International’s Fast Healthcare Interoperability Resources (FHIR) application programming interface holds great potential for the interoperable exchange of health information, the emerging standard will not replace older HL7 standards.
That’s the contention of Russell Leftwich, MD, senior clinical advisor for interoperability at software vendor InterSystems and assistant professor of biomedical informatics at Vanderbilt University School of Medicine.
Leftwich believes that by leveraging RESTful application programming interfaces, the latest web standards that are the basis of Google and Twitter, FHIR has tremendous potential to serve as the core functionality to support data access in healthcare enabling health information exchange.
At the same time, he points out that “interoperability is not in the FHIR base specification itself, but comes from using the same profile as another system uses for a particular use case—so, shared profiles create interoperability.” Leftwich defines a FHIR profile as the “equivalent of an implementation guide for a specific use case” such as blood pressure or a discharge summary. “The profiles are not part of FHIR. The profiles are what the implementers do with FHIR,” he adds.
One of the advantages of FHIR over older HL7 standards is that “once you have expressed the data in FHIR, it can be moved from one interoperability paradigm to another, so the data that comes in as a message that could be placed in a document without any transformation,” said Leftwich during a January 25 eHealth Initiative webinar.
“That’s not true of the existing HL7 standards like Version 2 and CDA. It would take a great deal of effort to move data from one of those interoperability paradigms to another,” he adds.
Ultimately, however, Leftwich does not see FHIR replacing other standards but supplementing them.
“Many standards will be in existence for the foreseeable future, probably for our careers at least,” he added. “HL7 Version 2 is the standard used by literally hundreds of thousands of systems around the world. It’s going to take years to replace those systems, and that standard, for example, does what is does very well. There’s no reason or economic justification for replacing those systems just because we have something new in FHIR.”
HL7 CEO Chuck Jaffe, MD, noted that half of all healthcare data worldwide is exchanged on some flavor of Version 2, adding that “in my perspective, they’re not going to take out systems that work well for their use—people aren’t going to convert technology that works for their needs.
“Every week, the FHIR community meets with the developers of the legacy specifications,” said Jaffe. “I think one of the more successful ones is CDA on FHIR. But, there are also opportunities to integrate Version 2 and other data streams into FHIR. I’ll highlight only a few of them, but they’re enabling legacy systems for clinical decision support, for public health, and even for research.”
Nonetheless, according to Leftwich, HL7’s existing standards were designed for the use case of connecting two systems, not the digital health ecosystem that is emerging.
“Thirty years ago, interoperability meant connecting two systems together. But, interoperability is starting to mean being able to access data in multiple systems in real time in order to see a complete and longitudinal view of an individual’s data across many different systems and environments—including their own personal devices that are now becoming data sources,” said Leftwich.
“HL7 Version 2, the messaging standard, is almost 30 years old,” he noted. “HL7 CDA, the basis of C-CDA, which is the current standard in use for document exchange, is over 10 years old.”
At the same time, Leftwich acknowledged that FHIR is relatively young, early in its development and continues to undergo testing.
“Much of the use of FHIR is within organizations to access their own data in ways that are very valuable to them,” he concluded. “The thing that’s holding up the potential use of FHIR across organizations is really not the technology but the policy and governance around security and consent of accessing data in different systems from outside.”
Still, Jaffe contends that “as opportunities arise for accountable care we’ll see broader FHIR implementation.”