Skip to main content

Re-Identification Risk Modeling

Meaning

Re-Identification Risk Modeling is the formal, quantitative process of using statistical and computational techniques to estimate the probability that an individual patient, whose data has been de-identified, can be successfully identified by an external party. This modeling involves simulating potential linkage attacks using known external data sources and calculating the expected number of successful re-identifications based on the uniqueness of the remaining quasi-identifiers. This critical process provides a verifiable, objective metric of privacy protection, moving beyond subjective assessment to ensure that sensitive hormonal and clinical data can be shared for research with a demonstrably low risk of patient harm. The resulting risk score dictates the appropriate level of data sharing.