IGF::OT::IGF Cancer Data Ecosystem

Leidos Biomedical Research, Inc., Frederick MD

Investigators

Abstract

As called for by the Blue Ribbon Panel Enhanced Data Sharing Working Group, the National Cancer Data Ecosystem was envisioned to consist of multiple components: ? Fundamental infrastructure to connect the components and ensure interoperability; this includes centralized APIs, data schemas, and a common data dictionary ? Components such as repositories, analytic services, and interactive portals for data analysis and visualization. Initial fundamental activities to support the National Cancer Data Ecosystem include: ? Establishing the ?blue print? to define the requirements of the fundamental infrastructure and components of the National Cancer Data Ecosystem and serve as a reference to the wider community that will contribute to and guide future activities of the NCI ? Identifying key data contributors, key data elements, and structured mechanisms for sharing these data elements to support various components of that National Cancer Data Ecosystem ? Building the infrastructure to enable data linkages, which is crucial to the interoperability of the National Cancer Data Ecosystem The first two of these activities will require input and expertise from the community, which NCI will achieve by convening workshops to bring together critical stakeholders, including bioinformatics experts, IT experts, patient advocates, electronic medical record vendors, and cancer registries. The third activity focuses on the ability to connect is the ability to connect disparate data sources at the patient-level, making it available to researchers, but without compromising patient privacy or confidentiality. This connectivity should include data generated from a variety of data sources and representing different aspects of each cancer patient's experience over time. Linking data at the patient level poses a number of challenges, including: protecting patient confidentiality, data security, scalability and accuracy in linking the same patient assuring consistency of patient identifying information (PII) across a variety of sources that may hold different components of PII (e.g. name and date of birth versus SSN and date of birth). This work will support the exploration of existing solutions to address these issues through the generation of encrypted unique patient identifiers to link individual patient data from disparate sources.