Informatics Research Seminar: Geospatially Enabling Duke’s Enterprise Data Warehouse

 February 13 @ 4:00 – 5:00 pm

 

Speaker: Sohayla Pruitt, MA
Presented from Duke University

Broadcast Link: Seminar

 

Abstract:

Duke’s Decision Support Repository (DSR) is the Enterprise Data Warehouse (EDW) for Duke Medicine.  It incorporates more than 26 major clinical and financial systems within the institution and supports clinical research, quality improvement initiatives, financial analysis, and metrics reporting.  To further build upon this infrastructure, geospatial data are being integrated so that automated processes will achieve the following: (a) USPS address standardization, (b) rooftop geocoding, and (c) the pre-calculation of each patient’s address to each geospatial feature (to include neighborhood level socioeconomic and demographic characteristics, proximity to various types of healthcare facilities and other businesses and features in the environment).

There are various tools and services that are already in place that facilitate the use of data in the DSR.  One of which is Duke Enterprise Data Unified Content Explorer (DEDUCE).  DEDUCE is a robust Web application developed for cohort identification and data extraction.  It uses business intelligence to allow investigators the ability to filter millions of administrative and clinical records that are generated during patient care and integrated within Duke Medicine’s EDW.   Through an easy to use interface, investigators have the ability to obtain detailed patient- and observation-level extracts, as well as define cohorts and identify potential research participants without needing to understand structured query language or the underlying EDW model.

To further enhance DEDUCE functionality, the next versions of this application will be focused on the assimilation of geospatial visualization and analytics in an easy-to-use geospatial dashboard.  Investigators will not only be able to visualize the results of their queries on a map, but the integration of both clinical and geospatial data will support various community- and population-based research by allowing the ability to discern environmental and locational patterns present within a queried cohort.

Biosketch:

Sohayla Pruitt is a Senior Geospatial Scientist in the Data Warehouse Group of the Information Management department of Duke Health Technology Solutions (DHTS), the IT group for Duke University Health System (DUHS).

Sohayla is responsible for the design, development, and implementation of geospatial infrastructure to provide reliable and scalable applications and systems to meet objectives and requirements of Duke Medicine’s Enterprise Data Warehouse (EDW).   She works closely with various technical and functional teams to define needs or problems, conduct research, obtain data, and analyze problems to advise on or recommend solutions.

With 14 years of experience with Geospatial Analysis, Sohayla has managed and conducted Geographic Information Systems (GIS), Remote Sensing (RS), and Geospatial Predictive Analytics across multiple disciplines and geographies for Duke, GeoEye, RTI, The Department of Homeland Security, NASA-Goddard Space Flight Center, and The Smithsonian Environmental Research Center.  Sohayla received both her M.A. and B.A. in Geosciences, specializing in GIS and RS from the University of Arkansas.