Using Geographically Weighted Linear Regression for County-Level Breast Cancer Modeling in the United States

• Activity Number: 322 - Novel Statistical Methods for Current Health Policy Issues
• Type: Contributed
• Date/Time: Wednesday, August 5, 2020 : 10:00 AM to 2:00 PM
• Sponsor: Health Policy Statistics Section
• Author(s): Srikanta Banerjee* and Matt Jones
• Companies: Walden University and Northwest Emergent Solutions
• Keywords: Geographically Weighted Linear Regression; Cancer; Health Policy; Social Determinants of Health; Inequities

Abstract

Due to the continued disparities in breast cancer, improved models are being needed to inform policy related to existing social disparities related to cancer. First ordinal least squares regression was used to determine the relationship of sociodemographic measures (i.e. poverty rate and social inequity) on breast cancer incidence in the United States. Gini coefficient was used as a measure of income inequality. Next, Geographically Weighted Logistic Regression (GWLR), a local spatial model, was used to explore the impact location has on the relationship between sociodemographic measures and breast cancer. Mappings of the results are presented, which can assist policy makers address inequities and social determinants when funding cancer interventions. The GWLR model is then compared to linear regression models that do not take into consideration location, highlighting the benefits of spatial models in cancer policy research. More studies applying spatial regression techniques are needed in order to accurately inform policy.