Herting R L, Barnes M R
Department of Health Management and Informatics, University of Missouri-Columbia School of Medicine, USA.
Proc AMIA Symp. 1998:508-12.
Now that case managers, quality improvement teams, and researchers use medical databases extensively, the ability to share and disseminate such databases while maintaining patient confidentiality is paramount. A process called scrubbing addresses this problem by removing personally identifying information while keeping the integrity of the medical information intact. Scrubbing entire databases, containing multiple tables, requires that the implicit relationships between data elements in different tables of the database be maintained. To address this issue we developed DBScrub, a Java program that interfaces with any JDBC compliant database and scrubs the database while maintaining the implicit relationships within it. DBScrub uses a small number of highly configurable object-oriented software components to carry out the scrubbing. We describe the structure of these software components and how they maintain the implicit relationships within the database.
鉴于病例管理员、质量改进团队和研究人员广泛使用医学数据库，在维护患者隐私的同时共享和传播此类数据库的能力至关重要。一种称为清理的过程通过去除个人识别信息同时保持医学信息的完整性来解决这个问题。清理包含多个表的整个数据库要求维护数据库不同表中数据元素之间的隐式关系。为了解决这个问题，我们开发了DBScrub，一个与任何符合JDBC的数据库接口并在维护其中隐式关系的同时清理数据库的Java程序。DBScrub使用少量高度可配置的面向对象软件组件来执行清理。我们描述了这些软件组件的结构以及它们如何维护数据库内的隐式关系。
