A survey of anonymization algorithms for electronic health records

Abstract

Electronic Health Records (EHRs) contain various types of structured data about patients, such as patients’ diagnoses, laboratory results, active medication, and allergies, which are increasingly shared to support a wide spectrum of medical analyses. To protect patient privacy, EHR data must be anonymized before their sharing. Anonymization ensures that the re-identification of patients and/or the inference of patients’ sensitive information is prevented, and it is possible using several algorithms that have been proposed recently. In this chapter, we survey popular data anonymization algorithms for EHR data and explain their objectives, as well as the main aspects of their operation. After that, we present several promising directions for future research in this area.