For the first time, biomedical researchers can assess the availability of consented datasets over all five university hospitals in Switzerland at once. This is a key feasibility step for large-scale biomedical projects, which rely on the integration of a diversity of patient data, scattered across institutions. So far, such queries had been limited to individual clinics or single hospitals. The query system relies on Clinerion’s Patient Network Explorer. Its adoption results from a nationwide coordination among clinical partners, piloted by SIB’s Personalized Health Informatics Group, as part of the Swiss Personalized Health Network (SPHN) initiative.
A key step in assessing the feasibility of biomedical research studies
Finding out if, where and how many patients present the condition of interest for a specific study (e.g. predicting individual sepsis onset risk in ICU patients) is an essential preliminary step in the context of biomedical research. With the new Federated Query System (FQS), established in the realm of the Swiss Personalized Health Network (SPHN), it is now possible for researchers to search for fully anonymized clinical information across all five university hospitals (see figure below), while allowing the hospitals to retain full control over their data. “For the first time, it is now possible to run feasibility queries for data-driven research projects over all five university hospitals in Switzerland. This is a major step for SPHN,” says Katrin Crameri, Director, SIB Personalized Health Informatics.
Over 70 million data elements from over 450’000 patients
The available information in the system, i.e. currently more than 70 million data elements from over 450’000 patients who have consented to the further use of their data for research purposes, is a subset of the clinical data of all five university hospitals. It includes demographic data (age and gender), diagnosis (coded in ICD-10), procedures (coded in CHOP), medication (coded in ATC), laboratory tests (coded in LOINC) and laboratory test results with units (according to UCUM). Under the project lead of the SPHN Data Coordination Center (DCC) managed by SIB, and with support of IT-Logix, the clinical partners harmonized and standardized the anonymized clinical information, making it available for federated queries via a common front-end.
A federated cloud infrastructure
The Federated Query System (FQS) of SPHN uses a customized version of Clinerion’s Patient Network Explorer, due to the unique nature of its patented federated cloud infrastructure. It enables researchers to search for fully anonymized clinical information across all five university hospitals, while allowing the hospitals to retain full control over their data which remain inside their IT infrastructure. It fully complies with privacy requirements, since only aggregated search results are shared with users and small patient numbers are obfuscated to prevent potential re-identification.