Using GIS to Model Utilization of Breast Cancer Screening Services in Minnesota, 2010-2015

Authors: David Haynes*, University of Minnesota
Topics: Medical and Health Geography, Geographic Information Science and Systems, Population Geography
Keywords: GIS, health disparities, cancer screening, breast cancer
Session Type: Paper
Day: 4/5/2019
Start / End Time: 8:00 AM / 9:40 AM
Room: Marshall South, Marriott, Mezzanine Level
Presentation File: No File Uploaded


Minnesota’s National Breast and Cervical Cancer Early Detection Program, Sage, is a cancer screening enrollment program whose mission is to reduce breast and cervical cancer morbidities for the uninsured and underinsured populations. A primary activity of Sage is to recruit and connect women to breast cancer screening services. Sage connects approximately 15,000 women to screening each year; however, the utilization of Sage by specific communities has not been adequately quantified. Sage seeks to create a better systematic approach to understand which communities are maximizing and under-utilizing the program. We use innovative spatial analysis techniques to create a utilization map of Sage screening services by Minnesotan women. The big data geospatial platform Spark is employed to create adaptive spatial filters that stabilize rate calculation for small or sparsely populated areas. Integrating a novel population dataset allows for the calculation of utilization at the neighborhood level. After creating the utilization dataset, it is validated with small area estimates of percent uninsured. While the overall utilization level of Sage screening services is low in the state (13%), we identified areas of high (104.1%) and low (0.1%) utilization. Screening utilization varied greatly within the state, and we determined that uninsurance rates significantly predict Sage utilization rates. We successfully built a novel geospatial model of Sage services utilization. This work will enable Sage to target specific areas they have yet to address, especially where the community need is the highest. Similar programs could use such models to improve programmatic activities.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login