Journal Article

DKFZ-2026-01217

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Accurate Yet Privacy-Preserving Determination of Case Numbers Across German University Hospital Health Data.

Overton, P. ; Kiel, A. ; Semler, S. C. ; Prasser, F. ; Kussel, T.DKFZ* ; Ganslandt, T. ; Gruendner, J.

2026
IOS Press Amsterdam

Abstract: Protecting privacy is essential when handling personal medical records. Even in pseudonymized health data, identifying individual patients remains a persistent risk in large medical datasets. The German Research Data Portal for Health (FDPG) provides access to extensive patient-derived routine data, including diagnoses, procedures, laboratory results, oncological findings, images, and genomic information. These data are stored locally at participating sites and harmonized through a common HL7 FHIR-based model. The portal's feasibility query function allows users to define complex cohorts using diverse search criteria and retrieves real-time patient counts from all German university hospitals and beyond. Despite the high aggregation level, re-identification risks persist due to potential prior knowledge or repeated ('tracker') queries. The FDPG therefore aims to minimize re-identification risk while preserving sufficient data utility for meaningful data discovery. Here, we present the data protection approach implemented in the distributed FDPG infrastructure and share insights from its first two years of operation, offering a blueprint for the upcoming European Health Data Space (EHDS).

Keyword(s): Germany (MeSH) ; Hospitals, University (MeSH) ; Confidentiality (MeSH) ; Computer Security: standards (MeSH) ; Electronic Health Records: organization & administration (MeSH) ; Electronic Health Records: standards (MeSH) ; Humans (MeSH) ; Health Records, Personal (MeSH) ; EHDS ; FDPG ; FHIR ; Medical Informatics Initiative ; data discovery ; differential privacy ; secondary use of health data

Note: ISSN:0926-9630 / #DKTKZFB26#

---

Contributing Institute(s):

1. Verbundinformationssysteme (E260)
2. DKTK HD zentral (HD01)

Research Program(s):

1. 315 - Bildgebung und Radioonkologie (POF4-315) (POF4-315)

Appears in the scientific report 2026

---

Database coverage:
; NCBI Molecular Biology Database ; SCOPUS

---