The lack of a universal terminology standard is a major barrier to communication between different electronic health record systems and the ability to derive clinical meaning from the shared EHR data.

While industry experts are working on ways to overcome this challenge, they disagree on the extent to which stakeholders should try to solve the problem of semantic interoperability.

Disparate EHR systems have their own inherent clinical terminologies. By leveraging common data models and vocabularies, these systems can “talk” to each other in the same “language,” providing the ability to share clinical information and use it meaningfully.

Stan Huff, MD, chief medical informatics officer at Intermountain Healthcare, would like to see the healthcare industry make EHR data consistent through standardization of information models that are mapped to clinical terminology standards such as LOINC and SNOMED.

Also See: AMA and IHTSDO set collaborative health terminology agreement

Health Level Seven International’s Fast Healthcare Interoperability Resources (FHIR) is gaining momentum as an open health data standard that has potential for meeting the need for semantic interoperability and minimizing the need for metadata translation services. FHIR is supported by a rich information model to achieve semantic interoperability of clinical data.

The Argonaut Project, an industry-wide effort to accelerate the development and adoption of FHIR, has created some profiles with detailed specifications for physical exam measurements such as heart rate, blood pressure and respiratory rate, Huff says.

Nonetheless, he contends that the Argonaut Project “has not really specified the thousands of things you would need for lab data in a consistent way.” As a result, Huff insists that, “Now’s the time to have the conversation about how we truly standardize those representations of data.

“Even though (FHIR) is very good, there’s a lot of work that remains to make it truly interoperable” through the standard use of terminology and codes in transactions between systems, he adds. Even when FHIR is a normative standard, it will not be truly interoperable without additional information.

Also See: 2017 emerges as pivotal year for FHIR interoperability standard

“Just saying that you’re going to use LOINC codes or SNOMED codes doesn’t mean that you’ll end up with interoperable software,” says Huff. “FHIR services from one vendor are not identical to the same services from another vendor. Without some further understanding created, vendors will do it differently because the codes exist and they will just choose different codes. The good news for the vendors is that they don’t have to do as much work to retrieve the data if they have choices in the codes.”

For its part, Cerner is committed to implementing FHIR, which is already a part of its product offering, with the emerging standard serving as the basis for the company’s open application programming interface, according to David McCallie, MD, Cerner’s senior vice president of medical informatics.

David McCallie, MD

“There are parts of FHIR that we’ve already implemented and that our clients are using, and app developers are using and have deployed in production systems today,” McCallie observes. “But it’s not so simple as to say, ‘Go use FHIR and we’re done,’ ” he adds. “FHIR out of the box isn’t ready to use—you have to put profiles in place to constrain it to particular use cases, and which vocabularies and nomenclatures should be used.”

He says that the higher the level of detail in the FHIR profiles for things such as conditions and vital signs, the greater the likelihood that the meaning will be exactly understood between sender and receiver systems. Nonetheless, McCallie notes that more detailed profiles are also more difficult to implement.

“There is a tradeoff you have to make depending on how deep you go into the profile of these FHIR resources. And Stan wants to go very deep, and some of us in the vendor community are concerned that that tradeoff might be too expensive for where we are today,” he contends.

While there are many ways to describe certain lab tests in LOINC, McCallie contends that Huff’s standpoint is “there should be one right way to do it, and the other ways should be deprecated.” For now, EHR vendors are not willing to make a commitment to that level of standardization, he says.

“Why should they have to change just because some profile said there’s a better way to do it?” he asks. “That’s the debate that we have around how deep to go into the semantic interoperability. On the vendor side, we want to be good enough to be useful and good enough to be implementable—without a lot of delay and cost—and then fix it and go deeper, as time allows in the future.

“There’s a tradeoff point that’s good enough and it’s good enough for where we are in 2017,” he concludes.

For his part, Micky Tripathi, president and CEO of the Massachusetts eHealth Collaborative and manager of the Argonaut Project, diplomatically notes that Huff and McCallie are both on the Argonaut Project team and offer valuable perspectives on FHIR.

Ultimately, however, he believes McCallie’s point is correct that “you can’t let the perfect be the enemy of the good as it relates to getting FHIR out into the field so people can start to use it—I tend to be very practical in that sense.” At the same time, Tripathi offers that “you don’t want to stop doing the work that Stan’s doing.”

Still, Huff laments that some EHR vendors “think it’s just too complex (and) too hard to do, and they compensate for that basically by putting more of the work on the people who are trying to use the data and who are trying to create applications that could be used across different vendor systems.”

In the end, he predicts that “what we’re going to see is people implementing FHIR and getting real value from it and then recognizing that we need to do more work and even more standardization to get to the level of what I would call being truly semantically interoperable.”

Register or login for access to this item and much more

All Health Data Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access