Poster, Podium & Video Sessions
Presentation Authors: Daniel Au*, Johar Syed, Sameer Siddiqui, Saint Louis, MO
Introduction: A decline in prostate cancer incidence was reported after USPSTF recommendations in 2008 and 2012. It is unknown if this decline is uniformly present among various education and socioeconomic levels within the US population.
Methods: Age-adjusted prostate cancer incidence from Surveillance Epidemiology and End Results (SEER) data were studied from 2008-2013. Utilizing American Community Survey (ACS) 2009-2013 and US Census 2010, 3142 counties were stratified by education (percent county pop. >= 25 years with >= high school degree or equivalent), poverty (percent county pop. >= 200% federal poverty line), and urbanization (percent county pop. living in an urban area) categories to set national quintile cut points. SEER county incidence data were matched with their corresponding national education, poverty, and urbanization quintiles. The highest, middle, and lowest quintiles were compared over time using incidence rate ratio (IRR) and between quintiles by absolute disparity (AD; highest and lowest quintile range difference) and relative disparity (RD; highest and lowest quintile range ratio). Analysis was performed to 95% confidence intervals (CI) using the Tiwari et al. method.
Results: Counties with highest education, economic, and urbanization levels had the highest prostate cancer incidence, 112.7, 108.7, and 108.1 per 100,000 respectively in 2013. Counties with the lowest education, economic, and urbanization levels had the lowest prostate cancer incidence, 97.9, 104.0, 97.2 per 100,000 respectively in 2013. The AD demonstrated a convergence in incidence rate between highest and lowest quintiles. From 2008-2013 the percentage change in absolute disparity declined -39.2%, -34.3%, -46.2%, for education, economic, and urbanization categories respectively. IRR declined equally across all quintiles within each category ranging 0.69-0.71. Equivalent IRRs among quintiles indicates decline in incidence was homogenous across counties and not weighted to county educational, poverty, or urbanization level.
Conclusions: Incidence rate is positively correlated with higher county education, wealth, and urbanization levels. The incidence absolutely converged between highest, middle, and lowest education, poverty and urbanization quintiles in all categories after USPSTF recommendations, relative disparity remained unchanged. Statistically equivalent IRRs suggest prostate cancer incidence has declined uniformly among counties of varying education, poverty, and urbanization levels between 2008 and 2013.
Source Of Funding: None