Proposal for interactive anonymization of electronic medical records

One of the most important inputs for medical research is the information registered in electronic medical records. This information typically contains sensitive data that must be preserved in order to be used for research or educational purposes, and protected depending on the regulations of each country and institution. In order to assure confidentiality of data, different techniques can be used to remove basic identifiers (e.g. names, IDs); however, these techniques can be easily bypassed by attackers who know the information that can act as pseudo-identifiers of patients (e.g. birthdates, gender). Although these pseudo-identifiers can also be removed, the information they contain is valuable for medical research. To face this problem, different methods that allow minimizing the risk of sharing confidential information have been proposed. The interactive use of anonymization algorithms for electronic medical records is the main contribution of this chapter, dubbed AnonymousData.co: a proposal for anonymization of electronic health records.