Concept and feasibility of privacy-preserving record linkage of cancer registry data and claims data in Germany: results from the DigiNet study on stage IV non-small cell lung cancer

Objective While cancer registry and health insurance data are valuable resources for oncological health services research, these are rarely linked at the individual level due to data protection concerns and technical limitations. The prospective, controlled cohort study DigiNet aims to optimize personalized care for patients with stage IV non-small cell lung cancer (NSCLC) in the German study regions Berlin, Saxony and North Rhine-Westphalia. The population-based control group (pCG) was identified through cohort matching within the participating cancer registries. For health economic analyses, case-specific linkage of cancer registry data with claims data without informed consent was required. Methods A privacy-preserving record linkage (PPRL) concept was developed, ensuring that no conclusions about individual identities can be drawn. The approach relied on irreversible encryption of the statutory health insurance number (KVNR) within the data-holding institutions, using a study-specific configuration of a publicly available software. Results Following cohort matching in the cancer registries, N = 9,597 pCG cases with stage IV NSCLC diagnosis between June 2022 and March 2024 were identified. Of these, n = 1,437 (15.0%) had insurance coverage with one of three participating statutory health insurance funds and were eligible for PPRL. Among those, 94.2% ( N = 1,354) were successfully linked with claims data. A trusted third party performed the linkage based on encrypted identifiers, removed the linkage keys, and provided the data to the evaluating parties. Conclusions This study demonstrates the feasibility of PPRL of cancer registry and claims data in a real-world oncological research setting. The concept is transferable to other research contexts requiring secure, identifier-based linkage without disclosure of personal identifiers.