SPHN Releases Clinical Data From 800,000 Patients

The Swiss Personalised Health Network has released the metadata of a uniquely rich national clinical dataset, spanning 800,000+ patients and billions of semantically harmonized data points across diagnoses, procedures, labs, medications, vital signs, oncology, and more. Researchers can now easily discover what data exists and how to access it.

After eight years of coordinated national effort, the Swiss Personalized Health Network (SPHN) has reached a major milestone: The SPHN Federated Clinical Routine Dataset with FAIR data from 800,000+ patients was prepared by university hospitals and its metadata recently published. Through the SPHN Metadata Catalog, researchers can now explore what clinical health data is available across Switzerland, how it is structured, and under which conditions it can be accessed.  

Additional healthcare institutions can also contribute to the SPHN Federated Clinical Routine Dataset by processing their data in the SPHN format.  

A national dataset of remarkable scale 

The SPHN Federated Clinical Routine Dataset is one of the most comprehensive harmonized clinical datasets ever created in Switzerland. As of December 2025, it included: 

• 800,000+ patients who have consented to the reuse of their data for research 
• 6 university hospitals: Basel, Bern, Geneva, Lausanne, Zurich (adult and children’s hospitals)  
• Clinical data from 2018 to today 
• 12.5 billion RDF triples mapped to 125 SPHN semantic concepts, covering:
  • Demographics 
  • Diagnoses and procedures 
  • Medications 
  • Laboratory tests, including microbiology 
  • Vital signs and clinical scores 
  • Allergies 
  • Intensive care and oncology data 
  • Biological samples.

All data and metadata fully adhere to the FAIR principles. This achievement reflects years of close collaboration within SPHN, notably of the university hospitals and SPHN’s FAIR data specialists.  

Information security, data protection and semantic interoperability at the core 

Since 2018, SPHN has built a national data infrastructure that enables healthcare institutions to harmonize clinical routine data without centralizing sensitive patient information. A key accomplishment is the large‑scale semantic standardization of clinical data across institutions.  

Participating hospitals map their data to internationally recognized terminologies, using:

• 6,500+ distinct SNOMED CT codes 
• 3,700+ distinct LOINC codes 

Each institution performs semantic annotation locally and stores SPHN‑compliant RDF data in dedicated triple stores. To support data discovery, hospitals contribute descriptive metadata and summary statistics to the central SPHN Metadata Catalog, which follows standards aligned with the European Health Data Space (EHDS).

Fully FAIR, fully federated 

In this federated architecture, patient-level data securely remain within each hospital, ensuring strict privacy protection. At the same time, both data and metadata fully adhere to the FAIR principles – Findable, Accessible, Interoperable, and Reusable for humans and machines.  

Enabling data-driven health research in Switzerland 

The publication of the SPHN Federated Clinical Routine Dataset marks a significant step toward a more connected and interoperable Swiss health data ecosystem. It provides a foundational resource for researchers, clinicians, and innovators seeking to plan studies, and design data-driven solutions and ultimately advance personalized health and patient care in Switzerland.  

➡️ Source: Press Release | 📸 ©TamimT9B/Pixabay