Data mining for the unique identification of patients in the national healthcare systems

This paper considers the application of data mining (DM) algorithms as a feasible and necessary strategy for optimal management of databases (DB) in the national healthcare systems. Specifically it deals with the management of multiple DB that consider patient's affiliation information, under the supervision of the authorities in healthcare, an issue that involves not only the issues of every citizen but also its integral right to be treated by any institution. We support the idea that the administrative part of the healthcare system should not obstruct the attention of the patient and a total efficiency must be guaranteed. We believe that DM algorithms are appropriate for this task and human intervention should be minimized. A case study was developed in Colombia that considered the multiple affiliations to DB and its integration to a unique DB managed by the District Health Secretary (DHS, which detected frauds and other type of duplicities. The mechanism used to approach this, indicates not only a significant reduction of manual intervention of the DB, but also allows the extraction of data for future analysis, supporting the patient's need for an efficient and integral health attention, as well as privacy of personal information registered.