Randomized controlled trial of storytelling compared to a personal risk tool intervention on colorectal cancer screening in low-income patients

Purpose. Screening rates for colorectal cancer (CRC) lag for low-income, minority populations, contributing to poorer survival rates. A model of storytelling as culture-centric health promotion was tested for promoting CRC screening. Design. A two-group parallel randomized controlled trial. Setting. Primary care, safety-net clinics. Subjects. Low-income patients due for CRC screening, ages 50 to 75 years, speaking English or Spanish. Intervention. Patients were exposed to either a video created from personal stories composited into a drama about "Papa" receiving CRC screening, or an instrument estimating level of personal cancer risk. Patients received a health care provider referral for CRC screening and were followed up for 3 months to document adherence. Measures. Behavioral factors related to the narrative model (identification and engagement) and theory of planned behavior. Analysis. Main effects of the interventions on screening were tested, controlling for attrition factors, and demographic factor associations were assessed. Path analysis with model variables was used to test the direct effects and multiple mediator models. Results. Main effects on CRC screening (roughly half stool-based tests, half colonoscopy) did not indicate significant differences (37% and 42% screened for storytelling and risk-based messages, respectively; n 539; 33.6% male; 62% Hispanic). Factors positively associated with CRC screening included being female, Hispanic, married or living with a partner, speaking Spanish, having a primary care provider, lower income, and no health insurance. Engagement, working through positive attitudes toward the behavior, predicted CRC screening. Conclusion. A storytelling and a personalized risk-tool intervention achieved similar levels of screening among unscreened/underscreened, low-income patients. Factors usually associated with lower rates of screening (e.g., no insurance, being Hispanic) were related to more adherence. Both intervention's engagement factor facilitated positive attitudes about CRC screening associated with behavior change.