CHICAGO (GenomeWeb) – It is becoming more apparent that the Fast Healthcare Interoperability Resources (FHIR) standard could play a central role in the elusive journey toward a nationwide system of interoperable electronic electronic health records.
Meanwhile, an emerging component called FHIR Genomics is showing promise as a way to bring machine-readable omics data into the equation.
At the annual genomics conference of FHIR creator Health Level Seven International in Washington, DC, last week, one of the lead developers of FHIR Genomics announced the creation of the Consortium for Agile Genomics.
"The idea is to promote implementations of an HL7 genomics platform to provide project support, certification, expertise, and validation tools to help its members to navigate, implement, and to collaborate on challenges of genomic integration," said Gil Alterovitz, director of the Biomedical Cybernetics Laboratory at Harvard Medical School and co-chair of the HL7 Clinical Genomics Work Group.
"Clinicians, they want genomic results," Alterovitz said. But they often still get reports by fax, and those reports have to be scanned into a computer. "If they're accessible [through an EHR], they're certainly not in a format that can be [read] by a computer because they're just digitized images," he noted.
"Reports can be very nice. They can be colorful and pretty, or they can be simple," Alterovitz added. "But they're not actionable in the sense that the computer can't really parse out that information that's in there."
The idea behind FHIR Genomics is to be able to "ask questions and get answers about little bits of information that are in these documents," Alterovitz said. "What's really kind of unique and special about FHIR is that it's really a unified clinical and genomic data standard."
FHIR Genomics started as a means to share data between the EHRs of different organizations and even between departments of the same organization. "But it also integrates the genomic aspects," Alterovitz said.
Getting to this point where a Consortium for Agile Genomics might be necessary took about a year of work, with help from the federally funded Sync for Genes program, which Alterovitz headed up. Sync for Genes is an extension of a data standardization effort called Sync for Science, which itself is part of the US National Institute of Health's All of Us research program.
NIH ran Sync for Genes in concert with the US Food and Drug Administration and the federal Office of the National Coordinator for Health Information Technology (ONC). ONC and the Sync for Genes task force set up a series of pilots in 2017 to test the HL7 Genomics standard and look for ways to overcome barriers to interoperability.
In a November report that featured Alterovitz as lead author, the task force recommended that precision medicine be added to the ONC "Interoperability Roadmap," an evolving document first published in 2015 that outlines a path to achieving secure, nationwide health IT data sharing. "Sync for Genes seeks to fulfill another recommendation — to '[i]dentify opportunities for ONC to collaborate with industry and pilot the use of standards to enable data donation and patient access through application programming interfaces (APIs) using standards'" including FHIR.
Most of the six-month pilots wrapped up by September 2017. One, at Vanderbilt University Medical Center, used a Harvard-developed framework called Substitutable Medical Apps, Reusable Technology (SMART) on FHIR to build applications that advance genomic data sharing in the context of somatic and tumor testing.
At the February HL7 conference, Alterovitz showed an app from Vanderbilt that takes structured data to compare the genome of a patient with lung cancer relative to other genomes. "The computer is labeling and tagging each little piece of data with what it actually represents," he said. "When you're able to do that, then you're able to do some really interesting things" in terms of understanding how genomic and clinical information complement each other.
"Based on the advances made by the Sync for Genes pilot programs, ONC, in partnership with NIH, will continue this important work to help advance genomics standardization in FHIR in a second phase of this project. We expect phase 2 to focus on expanding on the profile work to more fully support the integration of genomics information and clinical information, and to work on additional core use cases," ONC officials said in a blog post that accompanied the November report.
The original set of use cases that HL7 described last year include cancer profiling, specimen identification, family history, drug dosage calculators, sequencing for cytogenic marker identification, public health reporting, pharmacogenomics, and human leukocyte antigen typing. HL7 also recommended various nomenclature, reference databases, and vocabulary constraints for these uses.
Intermountain Health Care in Salt Lake City is working on family history and pharmacogenomics in a Sync for Genes pilot that actually was pushed back to 2018. During the period covered by the ONC report, Intermountain did sort though certain aspects of testing FHIR Genomics, in terms of sharing family histories and pedigree information.
"On the family history side, we were able to work with an online risk assessment service that does risk assessment for breast cancer, and we generated 3,000 different pedigrees," said Grant Wood, a senior IT strategist in the Intermountain Healthcare Clinical Genetics Institute and a member of the HL7 clinical genomics workgroup.
"We actually developed a tool that would create different types of pedigrees and assign people on those pedigrees with different diseases — some without breast cancer, some dying of it, some being diagnosed at an early age — really randomized the pedigrees, and send those in the FHIR format to this web-based risk service," one run by CRA Health, Wood added.
Intermountain took advantage of CRA Health's Risk Engine API to embed risk assessments into health IT systems, according to Wood. He reported that Intermountain was successful in generating and sending 3,000 messages, and then getting results back electronically in FHIR. "We were very pleased with how that worked," Wood said.
There were only minor technical glitches not related to the FHIR standard, he added.
The main Sync for Genes pilot to test FHIR for moving genetic data related to family health history got delayed in no small part because Intermountain just completed an organization-wide migration to a Cerner EHR. That pilot, with genetic testing lab Counsyl, should start by midyear, Wood indicated.
The delay actually is allowing the Utah health system to broaden its use of FHIR in genomic medicine.
"It's lining up with a study that's going on within Intermountain Healthcare," Wood said. Another team there is conducting on heritability of hereditary nonpolyposis colorectal cancer — also known as Lynch syndrome. "With Counsyl, it has that combination of using FHIR for genetic test results and using FHIR to exchange the family history portion of it," Wood said.
Meanwhile, Intermountain is following the recommendations of the ONC report by preparing to incorporate FHIR standards into the capture of genomic results for patient and family history as the Cerner implementation matures. "At the moment, we're keeping it to this small test group, but … it's definitely in our plans," Wood said.
"We already have the FHIR server ready to go that will communicate with the EHR. That FHIR server is ready to go for both pedigrees and genetic test result information. This year, it is our plan to start storing that in Cerner."
Wood said he is involved in other FHIR pilots in addition to the Sync for Genes activity. One is with a National Academy of Medicine workgroup for genomics and EHR integration known as DIGITizE, which stands for Displaying and Integrating Genetic Information Through the EHR. The group is made up of geneticists, representatives of EHR vendors, and standards organizations.
DIGITIzE seeks to transfer pharmacogenomic test results into EHRs.
"All of those different participants try to figure out from that workgroup how to get clinical decision support with genomics into EHRs," Hoda Sayed-Friel, executive vice president of EHR vendor Meditech and a member of the DIGITIzE workgroup, told GenomeWeb recently.