Toward bidirectional FHIR–OMOP CDM transformations using TermX to support the secondary use of real-world health data within a patient-centered digital health paradigm

The increasing digitization of healthcare has led to vast amounts of clinical data, much of which remains underutilized for research. While Health Level Seven (HL7) Fast Healthcare Interoperability Resources (FHIR) improves interoperability in clinical care, it's primarily designed for real-time data exchange to support diagnosis and treatment, rather than for secondary use of health data. As a result, transforming FHIR data into standardized models such as the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) remains a challenge. This study employs TermX, an open-source terminology and data interoperability platform designed to enhance health data interoperability and support knowledge management. This allowed us to create bidirectional transformation rules between FHIR and OMOP CDM. Using the Design Science methodology, we developed and validated a set of standardized transformation rules that support bidirectional mapping of vital signs data between FHIR and OMOP CDM. In these transformations we used synthetical FHIR JSON data, focusing on five main resources-<i>Observation, Patient, Encounter, Organization</i>, and <i>Practitioner</i>. The focus of this work is primarily on methodological mapping rather than processing real-world datasets; the evaluation concentrates on mapping coverage, i.e., the proportion of FHIR elements that can be reliably transformed into OMOP CDM structures and vice versa. The resulting rules achieved 74% mapping coverage from FHIR to OMOP CDM tables, with unmapped elements primarily related to structural discrepancies. Mapping from OMOP CDM to FHIR reached approximately 23% coverage, capturing mostly values that were previously mapped from FHIR to OMOP CDM. These percentages reflect variations in the standards' structure and granularity. The application of TermX shows the feasibility of reusable, standards-based transformations that support the secondary use of real-world clinical data for medical research and analysis. By addressing key technical and semantic interoperability challenges, this work contributes to advancing digital health interoperability and supports the objectives of the European Health Data Space.