Abstract B23: Socioeconomic factors and correlates of prostatespecific antigen testing in urban African-American Medicare beneficiaries

Abstract

Abstract Background: Prostate specific antigen (PSA) testing is frequently used for early detection of prostate cancer, including among older adults. Since both race and socioeconomic status are associated with prostate cancer incidence and outcomes, it is often difficult to separate race from socioeconomic status as a determinant of prostate cancer screening. We sought to describe the relationship between socioeconomic status and PSA testing in a sample of urban, African American Medicare beneficiaries, and to examine factors that may contribute to this relationship. Methods: We examined baseline questionnaire data collected from the Cancer Prevention and Treatment Demonstration, a community-based randomized, controlled trial in Baltimore, MD. The study population consisted of 511 community-dwelling African-American men, age 65–75, without a history of prostate cancer. Our main outcome was PSA testing within the past year. Sequential multivariate logistic regression models were performed to study factors that may mediate the relationship between socioeconomic status and PSA screening. Results: Eighty percent of participants reported having had at least one PSA test in their lifetime, and approximately 50% reported having one within the past year. About half of the sample population reported a household income of less than $30,000 a year and one-third reported education past high school, but only 14% reported a bachelor's degree or higher. In bivariate analyses both higher income ($30,000 vs. < $10,000) (Odds Ratio [OR] 3.14, 95% Confidence Interval [CI] 1.79, 5.52) and higher levels of education (Bachelor's degree vs. less than high school) (OR 3.19, CI 1.72, 5.71) were associated with screening in the past year. In a multivariable regression model, which controlled for marital status, family history, and self-reported health status, income was no longer a significantly associated with PSA testing while the association between educational attainment and PSA testing was reduced (OR 2.13, 95% CI 1.07–4.22). The addition of increased healthcare access and patient-provider relationship measures in sequential models did not alter the relationship between educational attainment and screening. However, upon inclusion of cancer knowledge and belief measures the association between education and PSA testing was no longer significant (OR 1.84, 95% CI 0.90–3.75). Conclusions: In this urban, African American Medicare population, socioeconomic status was associated with PSA testing. This relationship may be explained, in part, by demographic factors and cancer knowledge and beliefs. Citation Information: Cancer Epidemiol Biomarkers Prev 2011;20(10 Suppl):B23.