142nd APHA Annual Meeting and Exposition

Annual Meeting Recordings are now available for purchase

Modeling US county premature mortality

142nd APHA Annual Meeting and Exposition (November 15 - November 19, 2014): http://www.apha.org/events-and-meetings/annual
Monday, November 17, 2014

James L. Wilson, Ph.D. , Department of Geography, Institute for the Study of the Environment, Sustainability, and Energy, Northern Illinois University, DeKalb, IL
Christopher J. Mansfield, PhD , Department of Public Health, Brody School of Medicine, Center for Health Systems Research and Development, East Carolina University, Greenville, NC
In this study, ordinary least square regression and geographically weighted regression models are developed to predict premature mortality at the US county level.  The dependent variable, from among a family of mortality and morbidity metrics, is the years of potential life lost before age 75.  The OLS model is considered a global approach to predicting YPLL75 producing a single R2 value, while the GWR approach produces local R2 values for each observation or county in addition to a global R2. The principal hypothesis of this study is that one or more independent variables will vary in explanatory power across regions of the US and GWR models provide a means to determine which variables are more salient with respect geographic disparities in mortality burden.

Independent variables are taken from the Robert Wood Johnson Foundation’s and University of Wisconsin’s Population Health Institute’s “County Health Rankings National Data” for 2013.  These variables are grouped into two classes for the purpose of this study:  socio-economic determinants and behavioral risk factors.   Using variables from these two classes, OLS regression and GWR models are compared using global R2 values.  Preliminary results show that GWR models produce higher global R2 values for variables selected from either class.  For the OLS regression models, residual maps show a high degree of positive spatial autocorrelation delineating statistically significant county clusters of over and under prediction, while residual maps derived from GWR show close to no spatial autocorrelation.  County or local R2 maps depict county regions of high and low model explanation.

Learning Areas:

Public health or related research
Social and behavioral sciences

Learning Objectives:
Compare the strengths and limitations between OLS and geographically weighted regression in modeling premature mortality. Assess county level residual and local R-squared maps of premature mortality using statistical diagnostic tools. Differentiate the effects of socio-economic determinants and behavioral risk factors on geographic patterns of premature mortality.

Keyword(s): Mortality, Geographic Information Systems (GIS)

Presenting author's disclosure statement:

Qualified on the content I am responsible for because: I have previously published and presented work on premature mortality and health disparities. A major interest of mine is the application of vital statistics data in public health and medical geography research using geographic information systems and statistical analyses.
Any relevant financial relationships? No

I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines, and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed in my presentation.