How to Access Synthetic Data in the VA

“Emerging opportunities from this wealth of digital data come at the steep cost of individual privacy. With the breadth of data sharing amongst consumers, data brokers, technological platforms, etc., the risk to patient privacy has increased.Synthetic data is data that mimics realistic patterns but does not correspond to real data records. If data can be generated that reflects the patterns in clinical EHR data without any identifiable correlation to the actual data records used in their generation, this would have significant benefits for data privacy in the research and clinical contexts.”

海量数字数据所催生的新兴应用机遇，往往以牺牲个人隐私为高昂代价。伴随着消费者、数据经纪商、技术平台等主体间的数据共享范围持续扩大，患者隐私面临的风险也随之攀升。合成数据（Synthetic data）是一类模仿真实数据分布模式，但并不对应真实数据记录的数据集。若能够生成既贴合临床电子健康记录（Electronic Health Record，EHR）的数据分布模式，又与生成过程中所使用的真实数据记录不存在任何可识别关联的合成数据，那么该技术将为科研与临床场景下的数据隐私保护带来显著益处。