Simulating Realistic Enough Patient Records

Information systems for storing, managing and manipulating electronic medical records must place an emphasis on maintaining the privacy and security of those records. Though the design, development and testing of such systems also requires the use of data, the developers of these systems, rarely also their final end users, are unlikely to have ethical or governance approval to use real data. Alternative test data is commonly either randomly produced or taken from carefully anonymised subsets of records. In both cases there are potential shortcomings that can impact on the quality of the product being developed. We have addressed these shortcomings with a tool and methodology for efficiently simulating large amounts of realistic enough electronic patient records which can underpin the development of data-centric electronic healthcare systems.