Mainzelliste SecureEpiLinker (MainSEL): privacy-preserving record linkage using secure multi-party computation

Abstract

Motivation Record Linkage has versatile applications in real-world data analysis contexts, where several data sets need to be linked on the record level in the absence of any exact identifier connecting related records. An example are medical databases of patients, spread across institutions, that have to be linked on personally identifiable entries like name, date of birth or ZIP code. At the same time, privacy laws may prohibit the exchange of this personally identifiable information (PII) across institutional boundaries, ruling out the outsourcing of the record linkage task to a trusted third party. We propose to employ privacy-preserving record linkage (PPRL) techniques that prevent, to various degrees, the leakage of PII while still allowing for the linkage of related records. Results We develop a framework for fault-tolerant PPRL using secure multi-party computation with the medical record keeping software Mainzelliste as the data source. Our solution does not rely on any trusted third party and all PII is guaranteed to not leak under common cryptographic security assumptions. Benchmarks show the feasibility of our approach in realistic networking settings: linkage of a patient record against a database of 10.000 records can be done in 48s over a heavily delayed (100ms) network connection, or 3.9s with a low-latency connection. Availability and implementation The source code of the sMPC node is freely available on Github at https://github.com/medicalinformatics/SecureEpilinker subject to the AGPLv3 license. The source code of the modified Mainzelliste is available at https://github.com/medicalinformatics/MainzellisteSEL.

Record Linkage based Patient Intersection Cardinality for Rare Disease Studies using Mainzelliste and Secure Multi-Party Computation

“…The methods performing this "fuzzy" matching between patient records are called Record Linkage methods (RL) [14], and specifically in the context of this work Privacy-Preserving Record Linkage (PPRL) [15,35]. For a more detailed discussion of related works in the field of Record linkage, please confer [18,48], and for an overview of PPRL methods, including a discussion w.r.t. cryptographic privacy and statistical perturbation, please see [24].…”

“…For the detailed security guarantees of ABY in general and MainSEL in particular, please see [11,48].…”

“…In a previous work we modularly extended Mainzelliste with a MPC based trans-institutional PPRL system, Secure EpiLinker [48]. The secure EpiLinker is connected to Mainzelliste via its RESTful API, following Mainzelliste's general design, thus forming a holistic pseudonymization, record linkage and ID management system: MainSEL (freely available under https: //github.com/medicalinformatics/MainSEL).…”

“…Even though recent versions claim protection against those known attacks [46], new exploits are expected to be found [55]. Alternatives to centralized PPRL are Record Linkage systems based on secure Multi-Party Computation (MPC) [18,35,48] which allow participating parties to perform a joined linkage calculation over distributed input data without the need of any central linkage component. In particular, by ensuring that the data is not transferred, thus never leaving its owner's institutional boundaries, the attack surface of the identifying data is reduced, increasing the patients' data protection level and prohibiting potential re-identification.…”

Privacy-preserving data sharing infrastructures for medical research: systematization and comparison

“…by enriching data with secure record linkage tokens [85]. Also, secure multi-party computation environments could be extended with libraries including different record-linkage algorithms (see [86] for a recent example). Moreover, future work could explore ways to provide strong protection guarantees for inherently flexible approaches, such as the Personal Health Train.…”