FHIR and Health Middleware: Pathways to Real-Time Data Integration in Hospitals
Knowledge database Technology Integration & interoperability A.1: Tech-FoundationThis article builds on the introduction to HL7 FHIR as a standard for interoperable data exchange in healthcare and explores technical and organisational implementation experiences in greater depth.
Problem description, research question and relevance
Despite high-quality medical care, the digital transformation of healthcare lags behind in many countries, including Switzerland (Federal Office of Public Health, 2025).
A core issue is the lack of interoperability: IT systems are proprietary, data formats inconsistent, and data often has to be entered multiple times manually. The reliance on outdated systems further hampers integration efforts. In the Swiss context, this is exemplified by the electronic patient record (EPR), which currently relies on FHIR only in limited areas, such as the electronic vaccination record or medication plan (eHealth Suisse, 2025).
Inadequate standardisation impedes cross-sectoral communication and limits the use of health data for research, planning, and AI-supported applications. Initiatives such as DigiSanté explicitly call for interoperable platforms and legal anchoring of standards such as FHIR (Federal Office of Public Health, 2025). This leads to the central research question: How can FHIR be technically and organisationally integrated into healthcare to promote semantic interoperability and sustainable digitalisation?
Methods and procedures in the project
The article draws on literature analyses from several systematic reviews on FHIR implementation (Ayaz et al., 2021; Nan & Xu, 2023; Vorisek et al., 2022), as well as findings from the Innosuisse flagship SHIFT A.1, particularly in collaboration with practice partners such as TIE AG. In this context, a health middleware solution ("health-engine") was deployed to transform HL7v2 data into FHIR resources. The article’s content is derived from scientific literature, case studies, architectural analyses, and project implementations in Swiss hospital settings (Pimentel & Russ, 2025).
Results and findings
FHIR defines so-called "resources" as the smallest logical units of data exchange and employs technologies such as RESTful APIs, JSON, and XML (Ayaz et al., 2021; HL7 FHIR Foundation, 2025). The literature typically describes the implementation process in three phases (Nan & Xu, 2023):
Data Standardisation: Creating specific FHIR profiles and mapping legacy data (e.g., using TIE’s health-engine®).
Data Management: Storing and providing data via FHIR interfaces.
Data Integration: Linking with applications such as SMART on FHIR or data warehouses.
Compared to earlier standards (HL7v2, CDA), FHIR offers advantages such as granular access, web compatibility, and modular extensibility (Ayaz et al., 2021). The FOXS stack—comprising FHIR, openEHR, IHE XDS, and SNOMED CT—is recommended as a complementary architecture model, not a competing one (Pedrera-Jiménez et al., 2022).
Challenges:
Inadequate infrastructure in many institutions
Resource changes across FHIR versions
Complex transformations from legacy systems
Legal issues regarding e-consent and API security (Nan & Xu, 2023; Vorisek et al., 2022)
Use Cases:
In SHIFT A.1, a Swiss hospital uses FHIR APIs to transmit device data in real time.
In Germany, FHIR is widely used under the ISiK initiative for lab data, medication plans, and master data (gematik GmbH, 2025).
Recommendations for practice
Develop a FHIR-compatible middleware (e.g., with TIE’s health-engine®)
Use CH-Core profiles to ensure standardisation (eHealth Suisse, 2025)
Combine with SNOMED CT and LOINC to enable semantic interoperability
Establish an interoperability strategy and governance structures (Pimentel & Russ, 2025)
Participate in interoperability testing events (e.g., Digital Health Projectathon)
Enable research and AI-based applications (Federal Office of Public Health, 2025)
Conclusion and Outlook
FHIR represents a key building block for digital healthcare delivery—both in primary care and research. Experiences from SHIFT A.1 and the use of the health-engine demonstrate that technical transformation is feasible, provided there are clear frameworks, technical expertise, and coordinated strategies. With increasing standardisation, legal anchoring, and project funding, the use of FHIR in Switzerland is expected to expand further.
Literature and other sources
Ayaz, M. et al. (2021). The Fast Health Interoperability Resources (FHIR) Standard. JMIR Med Inform, 9(7), e21929.
Bundesamt für Gesundheit BAG. (2025). DigiSanté: Förderung der digitalen Transformation. https://www.bag.admin.ch/digisante
eHealth Suisse. (2025). Digital Health Data. https://www.e-health-suisse.ch/
gematik GmbH. (2025). ISiK. https://fachportal.gematik.de/informationen-fuer/isik
HL7 FHIR Foundation. (2025). Welcome to HL7 FHIR. https://fhir.org
Nan, J., & Xu, L.-Q. (2023). Designing Interoperable Health Care Services Based on FHIR. JMIR Med Inform, 11, e44842.
Pedrera-Jiménez, M. et al. (2022). Aligning Semantic Interoperability with the FOXS Stack. JAMIA Open, 5(2), ooac045.
Pimentel, T., & Russ, C. (2025). FHIR im Gesundheitswesen: Grundlagen, Einsatz und Herausforderungen. SHIFT A.1.
Saripalle, R. et al. (2019). Using HL7 FHIR to achieve interoperability. J Biomed Inform, 94, 103188.
Vorisek, C. N. et al. (2022). FHIR for Interoperability in Health Research. JMIR Med Inform, 10(7), e35724.