Mainzelliste

Mainzelliste provides pseudonymization and record linkage for medical research networks, generating non-speaking pseudonyms from identifying attributes (IDAT) to enable privacy-preserving data integration, informational separation of powers, and real-time re-identification by authorized users.

Key Features:

• RESTful Web Interface: Implements a RESTful, resource-oriented architecture using standard HTTP verbs over HTTP.
• Pseudonymization and Record Linkage: Generates personal identifiers (PID) from identifying attributes (IDAT), supports suboptimal IDAT quality, and uses the EpiLink algorithm for record linkage across datasets.
• Session Management and Tokenization: Uses session management to preserve statelessness while retaining client-specific data across requests and employs tokens to authorize actions and delegate authentication.
• Firewall-Friendly Design: Communicates solely over HTTP to facilitate integration through standard network channels.
• Real-Time Re-Identification: Supports re-identification by authorized users via temporary identifiers.
• Open-Source Reference Implementation: Provides a Java Servlet reference implementation that implements generation of non-speaking pseudonyms and the web interface and is production stable.

Scientific Applications:

• Medical Research Networks: Enables secure linkage of patient data across institutions while maintaining informational separation of powers.
• Epidemiological Studies: Supports large-scale epidemiological studies by linking records even with poor-quality IDAT.
• Multi-Center Clinical Trials: Facilitates identity management and linkage across centers with controlled re-identification.
• Dynamic Research Environments: Provides timely data access and re-identification via temporary identifiers for time-sensitive research workflows.

Methodology:

Implements a RESTful HTTP API; generates PIDs from IDAT; applies the EpiLink algorithm for record linkage; manages sessions and token-based authorization; issues temporary identifiers for re-identification; reference implementation realized as a Java Servlet that generates non-speaking pseudonyms.