Open standards for semantic interoperability of clinical and genomic data: a proof-of-concept bridging openEHR, Phenopacket and FHIR

At present, research activities and clinical practice make typically use of two completely separate classes of models, methodologies and tools to support data management. From the research side, for example, the OMOP vocabulary and the MIABIS standard are used to characterise samples within the biobank context and beyond. In clinical practice, openEHR and HL7 (v2, CDA, FHIR) are widely used for data capture and data exchange, together with a high number of Clinical Document Repository systems, proprietary and open-source. The Phenopacket-schema is a part of the GA4GH efforts to help to bridge this gap, enabling clinicians and researchers to share detailed phenotype descriptions connected to genetic information, but data acquisition and management remain the foundation of the quality of health information. In fact, a major issue is that medical data depend strongly on the intended semantics at the moment of recording and their value is limited, if sufficient contextual information is not captured and maintained simultaneously. openEHR addresses this issue with multi-level modelling, providing a way to capture clinical information with a computable and ontological semantic approach, independent of specific software implementations. The openEHR Clinical Knowledge Manager contains a broad library of clinical knowledge artefacts - openEHR archetypes and templates - allowing to model clinical, professional, cultural and regulatory variance, using standardised atomic data patterns. This work is a proof-of-concept to explore the feasibility of combining openEHR, Phenopacket and HL7-FHIR for the management and interexchange of biomedical data. We developed the reference implementation for an openEHR system, based on the Code4health open tools, suitable to store genomic information and the related clinical data. We also represented the Phenopacket data schema as openEHR archetypes for the purposes of integration, to enable us to engage with the Phenopacket platform and related HL7-FHIR resources. At the moment, we are testing the system with a synthetic dataset. Future works include the use of the platform with real dataset coming from two specific use cases: the German HiGHmed Consortium cancer data and the rare disease data deriving from the Next Generation Sequencing Facility at the CRS4 in Italy.

@Misc{FMULDZLBMGB19,
  author       = {Frexia, F. and Mascia, C. and Uva, P. and Lianas, L. and Delussu, G. and Zanetti, G. and Leslie, H. and Bakke, S. and McNicoll, I. and Giacomelli, G. and Boscà, D.},
  title        = {Open standards for semantic interoperability of clinical and genomic data: a proof-of-concept bridging openEHR, Phenopacket and FHIR},
  month        = {october},
  year         = {2019},
  note         = {Poster presented at GA4GH 7th Plenary Meeting},
  type         = {Poster},
  keywords     = {GA4GH, openEHR, Phenopacket, FHIR, genomics, computable semantic interoperability},
  url          = {https://publications.crs4.it/pubdocs/2019/FMULDZLBMGB19},
}