Analysis of prostate cancer incidence using geographic information system and multilevel modeling.

This study investigated racial differences in prostate cancer incidence, stage and grade in Florida using individual, community and environmental data from three data sets. Prostate cancer data were obtained from the Florida Department of Health. Census-tract-level characteristics were extracted from census data. County-level environmental data were obtained from the Department of Environmental Protection. Geographic Information Systems (GIS) was used to show racial and geographic disparities. Multilevel modeling was applied to examine the relationship of prostate cancer stage and grade to factors at the aforementioned levels. The results indicated that at the individual patient level, advanced/late cancer stage was significantly associated with older age, being black, being unmarried, tobacco use and being diagnosed in early years. At the census-tract level, late cancer stage was related to low median income and low percentage of people with some college education. No significant association was found for environmental factors. Similar results were found for tumor grade. These findings are consistent with national data demonstrating striking racial/ethnic disparities, improved stage and grade over time, and the importance of socioeconomic status. The GIS results also add local community perspectives important for planning community education and outreach to reduce racial disparities in low-income neighborhoods and low-literacy populations.