180 likes | 432 Views
The Role of Poverty in Prostate Cancer in African-Americans. M. Norman Oliver, M.D., M.A., Assistant Professor, University of Virginia Departments of Family Medicine, Health Evaluation Sciences, and Anthropology; Eric Smith, M.S., University of Virginia Department of Family Medicine;
E N D
The Role of Poverty in Prostate Cancer in African-Americans M. Norman Oliver, M.D., M.A., Assistant Professor, University of Virginia Departments of Family Medicine, Health Evaluation Sciences, and Anthropology; Eric Smith, M.S., University of Virginia Department of Family Medicine; Mir Siadaty, M.S., M.D., University of Virginia Department of Health Evaluation Sciences
Acknowledgements • Research supported by grants from the National Cancer Institute (1 K07 CA099983-01) and the University of Virginia Paul Mellon Prostate Cancer Research Institute.
Prostate cancer scourge • Second leading cause of cancer death among men • Unequal burden: African-Americans have both prostate cancer incidence and mortality rates twice that of whites • Reasons for disparities unclear. Diet? Environmental exposures? Access to health care? Genetic susceptibility?
Data description • VCR: Incident prostate cancer cases 1990-99 • Geocoded 78% (nearly 31,000 cases) • Cases aggregated to census block group level for comparison to sociodemographic variables in the 1990 U.S. Census • Census variables: low education (<12yrs), high education (≥16yrs), income, % rural – all stratified by racial category
0.00000-72.38900 72.38901-221.70465 221.70466-375.20333 375.20334-579.86368 579.86369-990.50718
LOW HIGH
0.00000-61.71469 61.71470-162.93925 162.93926-293.66711 293.66712-508.78444 508.76445-969.08000
LOW HIGH
LOW HIGH
0-6 7-17 18-26 27-47 48-100
LOW HIGH
LOW HIGH
Limitations of study • VCR a poor-quality cancer registry; problems with data collection • Only used 78% of VCR data, with large fall out of rural data • U.S. Census suppresses some data at block group level • Incidence rates calculated using average 1990 and 2000 populations • Spatial and temporal lag: a general challenge for GIScience in public health
Next steps • GIS in cancer control research:1) statistical modeling of incidence and mortality to identify predictors for intervention2) surveillance – identify clusters for further investigation • Develop statistical models of VCR data • Diet surveys/ethnographic studies in areas of low and high incidence • Repeat these steps on data from southeastern U.S.
The Role of Poverty in Prostate Cancer in African-Americans M. Norman Oliver, M.D., M.A., Assistant Professor, University of Virginia Departments of Family Medicine, Health Evaluation Sciences, and Anthropology; Eric Smith, M.S., University of Virginia Department of Family Medicine; Mir Siadaty, M.S., M.D., University of Virginia Department of Health Evaluation Sciences