Skip to content Skip to navigation

FHIR’s Key Role in Precision Medicine Initiative

June 28, 2016
| Reprints
From Sync for Science to family history tools, FHIR is central to PMI efforts

With the National Institutes for Health preparing to announce the grant award for the Precision Medicine Initiative coordinating center, Health Level Seven International held a meeting June 28 to discuss the increasing role HL7’s Fast Healthcare Interoperability Resources (FHIR) standard is playing in genomics and precision medicine.

William Riley, Ph.D., director of the Office of Behavioral and Social Sciences Research at NIH, started out by mentioning that Eric Dishman, the new PMI cohort director, is now on board and overseeing activities. NIH has awarded $142 million over five years to the Mayo Clinic to establish the world’s largest research-cohort biobank. Riley also mentioned the Sync for Science (S4S) pilot project. Because individual data donation will be a key component of the PMI Cohort Program, ONC, NIH, and the Harvard Medical School Department of Biomedical Informatics are coordinating the implementation of the S4S pilot in collaboration with EHR developers Allscripts, athenahealth, Cerner, drchrono, Epic, and McKesson.

S4S pilot developers will implement a consistent, standards-based workflow, building on open specifications including FHIR and OAuth. Once developed and implemented, this functionality will allow individuals to connect a research app to their electronic health data, facilitating individual data donation for research.

Grant Wood, senior IT strategist at Intermountain Health Care’s Clinical Genetics Institute and co-chair of HL7 Clinical Genomics Work Group, described work being done to make it easier for clinicians to include family health history pedigree in routine care.

There are standards to do pedigrees in HL7 version 3, he said, and a work group of the  Global Alliance for Genomics and Health is now translating that effort into FHIR resources and profiles that can be used by family health history app developers. Most healthcare organizations, however, do not have a system-wide family health history program, he said. “That is another area we need to move forward on.”

There also was work done in HL7 version 2 on genetic variation implementation guides for transmission to EHRs. Although that version 2 guide was never fully adopted, that work’s data models and structures are being put into FHIR resources, he said. Once you collect the data and want to put it in an “omic” repository and link it to an EHR, you may want to utilize online services that perform clinical genomic interpretation. SMART on FHIR developers are creating apps that can run in any EHR and access that information. Wood also mentioned that the Global Alliance for Genomics and Health is deploying FHIR to address computable consent for research as it applies to family health history.

Gil Alterovitz, Ph.D., assistant professor in the Division of Medical Sciences/Computational Health Informatics at Harvard Medical School/Boston Children's Hospital and co-chair of HL7’s Clinical Genomics Work Group, led off his presentation by showing how physicians currently receive results of clinical genomics lab tests: on a paper report with boxes checked by hand, and no structural data to share. “This is not the way we want to do clinical genomics,” he said. “What we want is to have apps. In a few years, we will have apps to do cutting edge clinical genomic medicine at the point of care.”

Developers are working on Smart on FHIR apps that launch from the medical record and allow clinicians to make orders and receive results from genomic tests and visualize and present that information, he said. The apps can be customizable for different types of needs.

One example is the SMART Precision Cancer Medicine app developed at Vanderbilt-Ingram Cancer Center, designed to aid cancer care by enabling clinicians to pull up contextual information about a patient’s cancer genome in the clinic setting.

Alterovitz said advances in clinical genomics would lead to a new category of computer system. Much like a PACS was needed to store large amounts of imaging data, health systems will need a GACS, or Genomics Archive Computer System, that is separate from the EHR. “There is so much data and you need a lot of access to it. A GACS integrates with the EHR, yet stores the data.” It is the server with all the raw data, and the clinically relevant data is attached on a FHIR server.

The thing that is fascinating about this work, Alterovitz said, is that it is all free and open source: You can check it out at