A record linkage and a unique identifier are required in order to merge research data from several projects and studies and assign it to the respective person. In addition to the administration of person-identifying information (PII), it must be possible to assign individual identifiers of the source systems (e.g. laboratories, study centres, etc.) to the correct person. As PII can be incomplete or incorrect, an error-tolerant and reproducible record linkage is required. All these responsibilities are covered by the E-PIX®.
The E-PIX® enables probability-based record linkage and generates a unique identifier (Master Patient Index) for each research project and person. The E-PIX® can manage different versions of a person’s PII and allows possible synonym errors (duplicates) to be detected automatically. This is achieved on the basis of freely definable parameters (e.g. given name, family name, date of birth, KVNR, address information). Several comparison algorithms can be selected for automated matching. Project-specific validations prevent the entry of invalid data and increase data quality right from the data entry stage. Possible synonym errors are protocolled and can be cleared with the help of E-PIX®. Cross-site research projects require special protection of the PII. This is why the E-PIX® also facilitates Privacy-Preserving Record Linkage (PPRL) through the generation and comparison of coded PII. The E-PIX® can be deployed in local, decentralised and centralised infrastructures.
Dashboard
The dashboard provides a quick overview of the number of persons, identities and potential duplicates. All values are presented graphically and can be exported directly from the dashboard if required.
Document persons
People can be registered with demographic characteristics, contact addresses and project-specific parameters. The import and export of Excel lists facilitates the transfer of existing data to E-PIX® and collaboration with other software products.
Detect and resolve duplicates
Customisable algorithms allow multiple entries for the same person to be identified. These duplicates can be resolved automatically or manually with the help of the user interface or can be postponed for subsequent processing.
Log events
The traceability of record linkage events is particularly important when matching individuals. A person’s log provides information about matching results, as well as the creation and updating of the person’s identifying data.
Download
Are you interested in the E-PIX? You can download it here and simply start it using Docker. Alternatively, you can try out the gICS in our live demo.
Interfaces
The E-PIX® is supplied with TTP-FHIR Gateway, which enables documentation of and search for persons in HL7 FHIR® format.
Access to interfaces and interfaces of the E-PIX® can be authorised with Keycloak (OAuth 2.0 + OpenID-Connect).
Integration of E-PIX® into third-party systems or ETL processes using SOAP. Details can be found in the manual.
Notifications about events can be configured and automatically distributed to external systems via JNDI, HTTP and MQTT.
Documentation
User projects
The solutions of the Trusted Third Party of the University Medicine Greifswald are becoming increasingly widespread in the community. More and more consortia, sites and projects have made a conscious decision to use our solutions to realise their individual application scenarios. Users include registries and cancer registries, local trust centres of the MII and the NUM, various study projects and selected DZGs.
We have created the THS Community Dialogue to exchange experiences and share results with each other.
Funding and publications
The E-PIX® is freely available and has been developed by the University Medicine Greifswald (Institute for Community Medicine) since 2009 and was published in 2014 as part of the MOSAIC project (funded by the DFG HO 1937/2-1). Selected functions of the E-PIX® were realised within the context of MIRACUM (funded by the BMBF 01ZZ1801M) and NUM (funded by the BMBF 01KX2021). The E-PIX® is continuously being enhanced on the basis of project-specific requirements and feedback from the user community.
Selected publications
- A workflow-driven approach to integrate generic software modules in a Trusted Third Party
- Assessment of scalability and performance of the record linkage tool E-PIX® in managing multi-million patients in research projects at a large university hospital in Germany
- White Paper – Verbesserung des Record Linkage für die Gesundheitsforschung in Deutschland
- Integration of Trusted Third Party Software into an EDC System for Data Protection – Compliant Identity Management, Consent Management and Pseudonymization in Medical Research Studies