Abstract 2354: Multilevel analysis of colorectal cancer screening in southeastern Pennsylvania

Abstract

Listen

Translate article

Abstract Background: Colorectal cancer (CRC) is a leading cause of mortality in the US. There are racial and socioeconomic disparities in CRC incidence and mortality, in part due to underuse of CRC screening. The goal of this cross-sectional study was to examine multilevel factors associated with CRC screening in Southeastern PA. Methods: We conducted a cross-sectional study using the Public Health Management Corporation's Community Health Database household survey (2010-2015) and American Community Survey data (2010-2014) to examine multilevel factors related to CRC screening via sigmoidoscopy and colonoscopy. We compared adults age 50+ that received CRC screening within 10 years (N=12963) to non-compliant individuals (N=4585), according to American Cancer Society guidelines. We conducted descriptive analyses to compare individual and residential census tract characteristics of the compliant and non-compliant groups. Multilevel regression models including random intercepts for census tracts (CTs) and CT-level socioeconomic status (SES) were used to examine independent associations with CRC screening compliance. P-values <0.05 were considered statistically significant. Results: In bivariate analysis, we observed that individuals that did not comply with CRC screening recommendations were more likely to be younger, male, non-black, unmarried, less educated, uninsured, and current smokers than those in compliance. Non-compliant individuals were also more likely to live in areas with lower SES, shorter distance to public transportation, lower educational attainment, lower median income, and higher percentages of uninsured, unemployed, poverty and households on public assistance compared to those in compliance. In multilevel regression analysis, being unmarried (OR=1.49, 95% CI=1.34-1.65), uninsured (OR=1.65, 95% CI=1.30-2.10), or living in lower SES census tracts (below the first SES quartile: OR=1.48, 95% CI=1.30-1.69 and between the 2nd and 3rd quartile: OR=1.51, 95% CI=1.29-1.75 vs. above the highest quartile) were associated with non-compliance in CRC screening recommendations. There were no interactions detected between race and census tract SES in these models. Conclusions: We identified multilevel factors (marital status, health insurance status, and residential SES) significantly associated with non-compliance to CRC screening recommendations. Individuals and CTs with low levels of compliance can be identified and targeted for tailored interventions, such as peer support and community education about CRC in future studies. Future studies might examine how the availability and location of screening facilities may relate to CRC screening recommendation compliance rates in different neighborhoods. Citation Format: Charnita M. Zeigler-Johnson, Scott Keith, Kari Moore, Russell McIntire, Preethi Selvan, Steven Melly, Ana Diez-Roux. Multilevel analysis of colorectal cancer screening in southeastern Pennsylvania [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 2354.