Requirements and Design for an Extensible Toolkit for Analyzing EMR Audit Logs

Over the past decade, various regulations have been proposed and promulgated to support the auditing of accesses to Electronic Medical Record (EMRs). Current tools to support this process can improve their use of statistical and machine learning techniques and auditor interfaces. We sketch requirements and design for an Extensible Medical Open Audit Toolkit (EMOAT) to enable progress in these areas. A key objective is to provide interfaces that support three types of stakeholders: (1) expert analysts, (2) privacy and security ocers, and (3) patients. Our system design provides for an application programming interface that enables ocers and patients to access both simple and complex analytic systems. We illustrate how EMOAT has been adapted to support certain audit functionalities with data from the EMR systems of several large hospital systems.