Cancer Control Research5R01CA115983-05
Rawl, Susan M.
PROMOTING COLON CANCER SCREENING AMONG AFRICAN AMERCIANS
Colorectal cancer is the third leading cause of cancer deaths in the United States, affecting both men and women of all racial and ethnic groups. The majority of these deaths could be prevented since CRC, when discovered early, is highly treatable and screening that leads to removal of adenomatous polyps decreases CRC incidence by 75% to 90%. Nationally, in 2002, only 50% of adults aged 50 or older reported having had any screening test in the recommended intervals. Both CRC incidence and mortality rates are higher among African Americans than any other racial group. These health disparities are due, in part, to delay in diagnosis resulting in advanced stage disease at presentation. Lower rates of CRC screening participation among African Americans contribute to this burden. Interactive computer-based interventions effectively promote health behavior change by delivering individualized, or tailored, health information and counseling. In pilot testing, a tailored, interactive computer intervention designed to promote CRC screening among African Americans resulted in increased CRC knowledge and higher rates of fecal occult blood testing. The proposed study is a randomized controlled trial to determine the efficacy of a tailored, interactive computer intervention to promote CRC screening by comparing it to, and combining it with, a physician recommendation, which has been found to be a significant predictor of CRC screening. A sample of 1248 African American men and women who are being seen in two Midwestern primary care networks and who are non-adherent to colon cancer screening guidelines will be randomly assigned to receive either: 1) a CRC screening brochure; 2) an interactive computer intervention; or 3) the interactive computer intervention plus the brochure. Both process and outcome variables will be examined. Rates of adherence to fecal occult blood testing, sigmoidoscopy and colonoscopy are primary outcomes examined at 6 and 15 months. Logistic regression models will be used to test intervention effects separately for two outcomes, fecal occult blood testing (initial and repeat) and endoscopic screening, while adjusting for potentially confounding covariates. Mediators and moderators of intervention effects and changes in health beliefs related to the interventions also will be determined.