Hierarchical Bayesian Modeling of Spatio-temporal Patterns of Lung Cancer Incidence Risk in Georgia, USA: 2000–2007

UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
Ping Yin, Assistant Professor (Creator)
The University of North Carolina at Greensboro (UNCG )
Web Site: http://library.uncg.edu/

Abstract: Lung cancer is the second most commonly diagnosed cancer in both men and women in Georgia, USA. However, the spatio-temporal patterns of lung cancer risk in Georgia have not been fully studied. Hierarchical Bayesian models are used here to explore the spatio-temporal patterns of lung cancer incidence risk by race and gender in Georgia for the period of 2000–2007. With the census tract level as the spatial scale and the 2-year period aggregation as the temporal scale, we compare a total of seven Bayesian spatio-temporal models including two under a separate modeling framework and five under a joint modeling framework. One joint model outperforms others based on the deviance information criterion. Results show that the northwest region of Georgia has consistently high lung cancer incidence risk for all population groups during the study period. In addition, there are inverse relationships between the socioeconomic status and the lung cancer incidence risk among all Georgian population groups, and the relationships in males are stronger than those in females. By mapping more reliable variations in lung cancer incidence risk at a relatively fine spatio-temporal scale for different Georgian population groups, our study aims to better support healthcare performance assessment, etiological hypothesis generation, and health policy making.

Additional Information

Journal of Geographical Systems
Language: English
Date: 2014
Hierarchical Bayesian model, Spatio-temporal pattern, Lung cancer, Socioeconomic status, Georgia, GIS

Email this document to