Evaluation of record linkage methods for iterative insertions

This paper evaluates record-linkage methods in the realistic setting where records are inserted iteratively over time, as happens in disease and patient registries that grow continuously. Rather than linking a fixed pair of datasets once, the study considers how linkage quality behaves as new records are added and matched against an evolving database. It compares approaches on their ability to maintain accurate matching under these dynamics. Published in Methods of Information in Medicine, the work informs the design of record-linkage workflows for longitudinally maintained medical registries.