Since 2016, the Office of the National Coordinator for Health Information Technology has worked with the National Institutes of Health (NIH) on the Sync for Genes project to make genomic data available at the point-of-care and for research. A new report describes some lessons learned in Phase 2 of the project involving pilots at the Utah Newborn Screening Program, Weill Cornell Medicine, Lehigh Valley Health Network, and the National Marrow Donor Program.
Phase 2 continued the goal of Sync for Genes Phase 1 by further exploring approaches of exchanging and integrating genomic data into healthcare systems and for research. It tested and refined the FHIR Clinical Genomics specification by pilot testing the standard, providing feedback for the refinement of the specification directly to the HL7 Clinical Genomics Work Group, and participating in a FHIR Connectathon.
The pilot use cases were based on genomic results obtained through next-generation sequencing (NGS) techniques and represented a variety of scenarios pertaining to newborn screening, supporting the availability of point-of-care knowledge resources, cancer pharmacogenomic testing, and donor matching. Each organization mapped workflows illustrating their unique genomic use case to FHIR and, based on their experience, provided feedback for the refinement of the FHIR Clinical Genomics specification.
At the January 2019 HL7 FHIR Connectathon, three of the four pilot sites successfully demonstrated connectivity and exchange of clinical genomic data using the FHIR Clinical Genomics specification and their own genomic diagnostic reports.
In its report on Phase 2, ONC identified challenges that pilot projects identified. Regarding the implementation of the currently available standards, challenges included missing or mis-aligned semantics, the need for diverse community representation in standard development organizations, and the need for an understanding of the complexities of the genomic field among developers and implementers.
Regarding the genomics industry more broadly, pilot sites noted issues with needed alignment among legislation or policies that address privacy issues; security; data provenance; data storage and management; educational support for providers and patients; clinical and laboratory information systems that are not designed to accommodate the complexities of genomic use cases; health IT developer readiness to implement FHIR; and cost and business drivers.
ONC said the work conducted under Phase 2 provided valuable lessons regarding the importance of appropriately scaling and scoping projects related to the exchange of genomic data; having working knowledge of FHIR and genomics; understanding the current health IT developer support of FHIR; and the need for additional guidance and documentation to support the interoperable implementation of standards.
In one of the pilot use cases, a Weill Cornell Medicine provider could request an EXaCT1 test for a patient who is diagnosed with metastatic bladder cancer. This test can help the provider determine if the patient’s tumor has any genetic variants that would be responsive to certain chemotherapeutic or immunotherapeutic agents. (EXaCT1 is a New York State Clinical Laboratory Improvement Amendments-approved laboratory-developed test for whole-exome sequencing.)
The request for this test is entered into the EHR system and sent to a molecular pathology laboratory. The laboratory generates variant results as pathogenic, likely pathogenic, and those of unknown significance, along with several discrete components pertaining to each of the altered genes. The provider is able to see the pathogenic and likely pathogenic variant results in the EHR. Variants of unknown significance are stored in an external repository because of architectural constraints against diminishing returns of storing the data in the EHR. The provider is able to query external genomic knowledge resources against results stored in the EHR, as well as those stored in the external repository, directly from their EHR using a connection via a FHIR API. This point-of-care knowledge support can present provider and patient educational materials that include information on companion therapeutics.
