Abstract B99: A community-based randomized trial to increase colorectal cancer screening: Approaches to assessing intervention effectiveness, subgroup analyses and behavioral mediators of screening

Abstract

Abstract Background: We recently conducted a randomized colorectal cancer (CRC) screening intervention trial among Filipino Americans (N=548). The trial used multiple intervention components with different participants receiving different combinations of components (small group education in community settings, print materials, free FOBT kits, reminder letter and letter to physicians to encourage them to recommend CRC screening to their patient). We observed a significant increase in CRC screening by 6 month follow-up (25-30% in the intervention arms versus 9% in the control arm), based on self-report and an intent-to-treat analysis that assumed a not screened outcome for participants lost to follow-up (Maxwell et al, in press). The purpose of this paper is to report additional analyses to gain a more detailed understanding of this overall intervention effect. Methods: The multi-component intervention was based on the Health Behavior Framework and targeted such constructs as knowledge and awareness of CRC screening, communication with health care provider about CRC screening, health beliefs, social support and barriers to screening as potential mediators of behavior change. Telephone interviews 6 months after baseline assessed receipt of any CRC screening during follow-up, the primary outcome of the trial. Self-reported screening was verified by providers in a subsample. The effect of the intervention on CRC screening receipt during follow-up was estimated using analytical approaches that adjusted for potential biases due to missing data and self-report, and for different combinations of intervention components. Subgroup analyses were also conducted. In addition, we conducted mediation analyses to test whether variables of the Health Behavior Framework were associated with intervention arm assignment and the outcome in a manner consistent with mediation. Results: Multiple analytical approaches accounting for self-report bias and missing data confirmed the robust effect of the intervention. The intervention was also broadly effective across various demographic characteristics of the sample. Estimates of the intervention effect were highest among subjects whose providers received a letter as part of the intervention. Consistent with the Health Behavior Framework, knowledge/ awareness of CRC screening and patient-provider communication about CRC screening mediated receipt of CRC screening. Increase in knowledge/awareness of CRC screening tests accounted for 13% (95% confidence interval 1%-25%) of the total effect size while patient-provider communication accounted for 20% (95% confidence interval 6%-34%). Conclusions: The findings support a robust effect which increases confidence that the intervention can be implemented broadly. Subgroup analyses and attempts to deconstruct multi-component interventions can provide important information for future intervention development and implementation. The field could benefit from similar approaches employed in other arenas. Examining potential mediators in intervention trials may help identify constructs to target in order to enhance the effectiveness of interventions to increase CRC screening. Citation Information: Cancer Epidemiol Biomarkers Prev 2010;19(10 Suppl):B99.