Creating an Interoperability Data Infrastructure for Research on the Aging Lifecourse

文摘:

Abstract: Digital data that lacks a coherent, stable, discoverable, and reproducible structure is irrelevant to the research process. Improperly or inadequately documented data files represent meaningless numbers, valueless and, ultimately, discardable. One of the significant advances over the past 20 years has been the increased use of descriptive metadata, facilitating the use and value of these data, offering greater discoverability and preliminary exploration. These changes have been part of an evolving process over the past 50 years, roughly categorized into four phases 1) The tabular paper phase, 2) The tape and mainframe phase, 3) The CD and personal computer phase, and 4) The internet distribution phase. This application argues that the broad adoption of best practices for data management, data sharing, and team science remain stalled in Phase 4. The phenomenal growth in data resources has made access to individual research data on aging simpler than ever before. Unfortunately, effective data sharing, even in light of the increased access provided by the Internet, cannot be fully realized without detailed metadata linkages that describe classes of data that share related concepts, constructs, and variables across multiple data waves or multiple related studies. This fifth phase, the interoperability phase, represents the next essential transition to support aging research for multidisciplinary team science. This application's specific goals will advance data interoperability in emerging scientific areas, facilitating team science and multidisciplinary research. By organizing independent but related data collections into a uniform structure, this application's outcomes will accelerate aging research beyond what is achievable using existing collections that treat independent data collections as unique objects. 1) Identify- The universe of aging data needs to be formally cataloged and structured at the metadata level. This process includes variable level information and concepts using a "Common Data Elements" (CDE) approach to create cross-domain XML/DDI templates to unify longitudinal studies across waves. 2) Integrate- Relate these independent longitudinal studies to each other across health thematics, including Alzheimer's disease and related dementias (ADRD), cognitions, risks associated with complications associated with COVID 19, and broader health conditions. 3) Operationalize – Make the data resources and analysis tools available to the research community and provide ongoing support and training for the interoperability portal. A DDI based Cross-Domain Integration (CDI) framework will maintain the integrated data collections, integrate the support services, bibliographic tracking, and social media outreach. The project will provide training and educational services through conference presentations, workshops, and online webinars.