Paper 840-2017: Healthcare Data Sharing and Innovative Analytic Development: Lessons Learned from Distributed Data Networks

Secondary use of administrative claims data, electronic health and medical record (EHR/EMR) data, registry data, and other sources within the health data ecosystem provide rich opportunity and potential to study topics ranging from public health surveillance to comparative effectiveness research to identifying patients who may be eligible for inclusion in a specific study or for a specific intervention. Data sourced from individual sites can be limited in their scope, coverage, and statistical power. Sharing and pooling data from multiple sites and sources, however, present administrative, governance, analytic, and patientprivacy challenges. Distributed data networks represent a paradigm shift in health-care data sharing. They have evolved at a critical time when big data and patient privacy often have competing priorities. A distributed data network is one for which no central repository of data exists. Data reside behind the firewalls of each data-contributing partner in a network. Each partner transforms its source data in accordance with a common data model and allows indirect access to those data through a standard query approach using flexibly designed informatics tools. This paper discusses how distributed data networks have matured to make important contributions to the health-care data ecosystem and the evolving Learning Healthcare System.