UZ Leuven (3 of 3)

Data Capabilities: SNOMED CT and FHIR as a foundation for an interoperable standardised EPD

EMEA

Collaboration, Data quality, Global/International, Implementation, Mapping, Patient safety

In this project, we aimed to enhance the structural and semantic interoperability and secondary usability of clinical data by mapping structured elements from our Electronic Health Record (KWS) to SNOMED CT and FHIR (Fast Healthcare Interoperability Resources) standards. The overarching goal was to prepare our EHR for meaningful data exchange with other systems, while unlocking the potential for reusing structured clinical data for research, quality monitoring, and organizational analytics.

We focused on data elements already captured in a structured way within the EHR and aligned them with appropriate FHIR resources. These included:

* Medical procedures (both surgical and non-surgical)
* Imaging procedures
* Laboratory
* Nursing observations and procedures
* Diagnoses (goal= SNOMED CT as primary language)

We identified the relevant data elements within the FHIR resource that contribute to the term's semantic expression and are widely applicable. This enabled us to define FHIR-based templates tailored to our context‚Äîserving both as a mapping guide and as a foundation for generating valid FHIR resources.

FHIR data elements were then semantically enriched using SNOMED CT codes. This ensured that clinical meaning was preserved, even in scenarios where precoordinated SNOMED CT concepts were unavailable. Much of the semantic mapping occurs in the background, ensuring minimal impact on the end-user workflow.

This initiative demonstrated that thoughtful, context-aware mapping of EPD data to international standards is both feasible and valuable. It provides a strong foundation for improved data exchange, semantic interoperability, and meaningful reuse of clinical data across healthcare and research ecosystems.

Description

This project focused on enabling semantic interoperability and secondary data usability of clinical data within our hospital's Electronic Health Record (EHR) system, KWS. We mapped existing structured clinical content to SNOMED CT and aligned relevant elements with FHIR (Fast Healthcare Interoperability Resources) standards. The project covered various clinical domains, including diagnoses, medical procedures, imaging, laboratory, and nursing observations and procedures. Our aim was twofold: to prepare the EHR for interoperable, standards-based data exchange and to enable meaningful reuse of clinical data for research, quality improvement, and organizational analytics. We developed context-aware FHIR templates that served as both a technical guide and a tool for generating structured FHIR resources enriched with SNOMED CT concepts.

Scope

SNOMED CT was selected for its comprehensive clinical scope, standardized structure, and robust international interoperability, making it the most widely adopted clinical terminology system globally. As the mandated national standard in Belgium, its adoption aligns with strategic guidance from national healthcare bodies such as the eHealth Platform, while also ensuring seamless alignment with international standards such as HL7, ICD, and LOINC. This dual compliance guarantees consistency with both national policy and global healthcare interoperability frameworks, fostering cross-border data exchange and collaboration.

SNOMED CT provides a semantically rich and precise foundation for representing clinical data, making it ideal for structured health information exchange across diverse healthcare systems worldwide. Its capacity for both pre- and post-coordination enables flexibility in mapping a wide range of clinical scenarios, supporting multilingual and multicultural healthcare environments.

Furthermore, its deep integration with FHIR (Fast Healthcare Interoperability Resources) ensures future-proof, standards-based infrastructure, facilitating compliance with evolving digital health regulations. This interoperability enhances secondary uses of data, such as (international) research collaborations, global health surveillance, quality benchmarking, and advanced analytics across the healthcare ecosystem.

By adopting SNOMED CT, Belgium not only strengthens its national health data infrastructure but also positions itself as an active participant in the global digital health landscape, enabling cross-border patient care, interoperable EHR systems, and worldwide clinical research initiatives.

How SNOMED CT will be used

SNOMED CT is used to semantically enrich the structured data elements identified within the FHIR resource that contribute to the term's semantic expression and are widely applicable. Each FHIR template was designed to incorporate relevant SNOMED CT codes at appropriate data points, ensuring that the clinical intent and meaning are preserved during data exchange. For domains where precoordinated SNOMED CT concepts were insufficient, post-coordination was explored and selectively applied. The use of SNOMED CT occurs primarily in the background (for medical procedures, imaging, laboratory results, and nursing observations and procedures), allowing clinical workflows to remain unaffected.

Our hospital has adopted SNOMED CT as the default and standard terminology for diagnosis coding in the problem list. While some medical disciplines are already successfully using SNOMED CT, we are taking an incremental approach to implementation across all departments, with full hospital-wide adoption targeted for completion by the end of 2026.