Metadata-based assessment of SNOMED CT and LOINC adoption across 800 000 Swiss patients

Abstract Background Semantic interoperability is essential for enabling the secondary use of clinical data in research and healthcare. International standards such as SNOMED CT and LOINC are key enablers, yet empirical evidence on their real-world adoption remains limited. Objective This study presents a nationwide analysis of terminology usage based on metadata collected from hospitals contributing to the SPHN Federated Clinical Routine Dataset. The dataset includes information on approximately 800 000 patients, with a minimum data coverage spanning 2018 to 2025 across six Swiss University Hospitals. The goal is to quantify the adoption of semantic standards, identify strengths and gaps in their usage, and highlight areas requiring further harmonization to support data quality and interoperability. Methods We analyzed the use of SNOMED CT and LOINC international standard terminologies across participating hospitals by querying aggregated metadata from the SPHN Metadata Catalog via its SPARQL endpoint. Results A total of 6 557 distinct SNOMED CT codes and 3 693 distinct LOINC codes were identified, representing approximately 717 million data instances. SNOMED CT and LOINC accounted for 60% and 32% of coded data instances, respectively. The remaining code instances originated from other coding systems (e.g. ATC, ICD-10-GM, CHOP). Despite this adoption, substantial heterogeneity was observed across hospitals, with most codes being institution-specific. A greater diversity of distinct codes was observed in domains such as ‘Substance’ and ‘Organism’, whereas frequently instantiated concepts were concentrated in ‘Clinical Finding’, ‘Procedure’, and ‘Qualifier Value’ hierarchies. LOINC code usage was dominated by routine laboratory tests across all institutions but with little overlap in the coding practice between them. Conclusions While SNOMED CT and LOINC are widely used within the SPHN framework, significant variability in code usage and limited cross-institutional overlap highlight ongoing challenges for semantic harmonization. This study provides the first nationwide empirical baseline of terminology adoption in SPHN and demonstrates the value of metadata-driven approaches for monitoring interoperability readiness. The findings support targeted efforts to improve standardization and enhance the reuse of clinical data for research.

Authors

Institutions

Publication Details

Journal
BMC Medical Informatics and Decision Making
Published
2026-11-17
DOI
https://doi.org/10.1186/s12911-026-03966-4
Citations
1
Primary Topic
Biomedical Text Mining and Ontologies
Type
article
Field-Weighted Citation Impact
1.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Metadata-based assessment of SNOMED CT and LOINC adoption across 800 000 Swiss patients

Harald Witte, Katie Kalt, Gaëlle Vuaridel‐Thurre, Bram Stieltjes et al.
1 citations
BMC Medical Informatics and Decision Making
Biomedical Text Mining and Ontologies
1.00
article

Metadata-based assessment of SNOMED CT and LOINC adoption across 800 000 Swiss patients

Harald Witte, Katie Kalt, Gaëlle Vuaridel‐Thurre, Bram Stieltjes, Karen Triep, Vasundra Touré, Mathias Gassner, Olga Endrich, Seraphina Kissling, Jan Armida, Pero Grgic, Sabine Oesterle, Deepak Unni, Marouan Borja, Amanda Ramirez Ramos
article en
1 citations

Abstract

Abstract Background Semantic interoperability is essential for enabling the secondary use of clinical data in research and healthcare. International standards such as SNOMED CT and LOINC are key enablers, yet empirical evidence on their real-world adoption remains limited. Objective This study presents a nationwide analysis of terminology usage based on metadata collected from hospitals contributing to the SPHN Federated Clinical Routine Dataset. The dataset includes information on approximately 800 000 patients, with a minimum data coverage spanning 2018 to 2025 across six Swiss University Hospitals. The goal is to quantify the adoption of semantic standards, identify strengths and gaps in their usage, and highlight areas requiring further harmonization to support data quality and interoperability. Methods We analyzed the use of SNOMED CT and LOINC international standard terminologies across participating hospitals by querying aggregated metadata from the SPHN Metadata Catalog via its SPARQL endpoint. Results A total of 6 557 distinct SNOMED CT codes and 3 693 distinct LOINC codes were identified, representing approximately 717 million data instances. SNOMED CT and LOINC accounted for 60% and 32% of coded data instances, respectively. The remaining code instances originated from other coding systems (e.g. ATC, ICD-10-GM, CHOP). Despite this adoption, substantial heterogeneity was observed across hospitals, with most codes being institution-specific. A greater diversity of distinct codes was observed in domains such as ‘Substance’ and ‘Organism’, whereas frequently instantiated concepts were concentrated in ‘Clinical Finding’, ‘Procedure’, and ‘Qualifier Value’ hierarchies. LOINC code usage was dominated by routine laboratory tests across all institutions but with little overlap in the coding practice between them. Conclusions While SNOMED CT and LOINC are widely used within the SPHN framework, significant variability in code usage and limited cross-institutional overlap highlight ongoing challenges for semantic harmonization. This study provides the first nationwide empirical baseline of terminology adoption in SPHN and demonstrates the value of metadata-driven approaches for monitoring interoperability readiness. The findings support targeted efforts to improve standardization and enhance the reuse of clinical data for research.

BMC Medical Informatics and Decision Making
SIB Swiss Institute of Bioinformatics (CH), University Hospital of Bern (CH), Centre Hospitalier Universitaire Vaudois (CH), University Hospital of Basel (CH), University Hospital of Zurich (CH), University Hospital of Geneva (CH), University Children's Hospital Zurich (CH), Swiss Personalized Health Network (SPHN) (CH)
Schweizerische Akademie der Medizinischen Wissenschaften, Swiss Institute of Bioinformatics
Good health and well-being
Openalex Percentile: Top 26%
Biomedical Text Mining and Ontologies
1.00
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon