LINKJA - A Privacy-Preserving Record Linkage Software System

About the Program

Category: Information Technology (Best in Category)

Year: 2023

Linking datasets across agencies provides opportunities to inform and evaluate government-directed programs. Despite being critically important, such linkages often are not realized due to requirements to protect personal-identifying and protected health information. To share elements needed for linking person-level data between datasets (e.g., name, date-of-birth, Social Security Number) appropriate protections and agreements are needed; sometimes, despite much discussion, data are determined to be too high-risk, and trust cannot be utilized for data sharing. Software has been developed by vendors that creates encrypted combinations of record-linkage fields that are one-way (hashed values can’t be reversed). Vendor systems are expensive to use, often come with license restrictions that require additional fees for partners, can be complicated to execute, and linkages among individuals of low socio-economic status (SES) or with cultural diversity are less accurate. At Cook County Health, we developed an open source, state-of-the art privacy preserving record (PPRL) software system, Linkja, with novel rules that improve the accuracy of record linkages for low SES populations. We have used Linkja to join health system, public health, homeless management information system, adult probation and other health system datasets. Our system is available for other jurisdictions, government organizations, and health systems to use without license fees.