In order to be able to combine research data from several projects or studies and assign them to a single person, so-called record linkage is necessary. It requires both a duplicate procedure and a unique system-wide identifier, to which both the person-identifying data and individual local identifiers of the source system (e.g. laboratories, study centres, etc.) are assigned. Since this must be done in a fault-tolerant and traceable manner even in the case of incomplete or incorrect person-identifying data, a sustainable Record Linkage and Identity Management is required.
E-PIX enables probabilistic record linkage and implements the concept of a Master Patient Index. The integrated identity concept allows potential synonym errors to be detected automatically and supports their resolution using a user-friendly graphicalinterface. The recognition of duplicates is based on freely definable parameters and the Levensthein distance. Consequently, possible synonym errors are logged automatically and can be resolved using the corresponding functions.
Who uses the E-PIX?
The diagram shows productive (red) and planned projects (light red).
Date: October 2022
You are interested in using E-PIX? You can either try it online with our demo version or install it via Docker.