Abstract A036: Associations between area-level social determinants and late-stage diagnosis of colorectal cancers

Manami Bhattacharya,Lindsey Enewold,Amy Davidoff

DOI: https://doi.org/10.1158/1538-7755.disp22-a036

2023-01-10

Abstract:Introduction Systemic policies and disinvestment in some communities have created deep inequities in the social and economic conditions in which some communities live. There are documented disparities in the detection and diagnosis of colorectal cancer (CRC) and stage-appropriate treatment, which drives cancer outcomes, such as survival. Using the Andersen-Aday model of access to healthcare, prior researchers have identified person-specific factors associated with late-stage diagnosis. In this study, we investigated whether area-level measures that may related to CRC screening barriers were associated with late-stage CRC diagnoses. Methods We used data from the Surveillance, Epidemiology, and End Results (SEER) and selected cases of adult (18+ years) colorectal cancer diagnosed from 2006-2013 (most recent years available; n=234,232). This data was linked to Medicaid enrollment information. We merged census-tract level data from the American Community Survey (2010 midpoint) capturing transportation access (% utilizing public transit for commuting), housing quality and cost (more than 1 person occupying per room and rent > 30% of income), language proficiency (% speaking English less than well), receipt of SNAP benefits, as well as poverty rate and educational attainment (% in tract attaining a high school education or less). We tested these variables as separate multivariable logistic regression models using late-stage CRC diagnosis (AJCC Stage IV vs Stages 0-III) as the dependent variable to test the association between each area factor, controlling for age, sex, race, marital status, and others. Results In all, 24% of CRC cases were diagnosed in Stage IV. High poverty areas (20+% poverty compared vs. <5% poverty OR = 1.10, CI: 1.07-1.15) and areas with lower educational attainment (lowest educational quartile vs. highest OR = 1.11 CI: 1.07-1.15) are associated with higher odds of Stage IV CRC diagnosis. Areas utilizing more public transit (most use quartile vs. least use quartile, OR = 1.04, CI: 1.01-1.07) and more SNAP benefits (most use vs. least use quartile, OR = 1.13, CI = 1.09-1.17), are also associated with higher odds of Stage IV diagnosis. High occupancy housing areas were associated with higher odds of Stage IV CRC diagnosis (highest occupancy quartile vs. lowest OR = 1.04, CI = 1.01-1.07), while affordability of rent was not. Area-level English proficiency was also not associated with later diagnosis. Conclusions Area-level social determinants are important factors in CRC detection and control. Early diagnosis is a key factor in better survival after CRC. It is important to understand specifically which area-level factors are most important in early detection to inform interventions to reduce disparities in late-stage CRC diagnosis. Citation Format: Manami Bhattacharya, Lindsey Enewold, Amy Davidoff. Associations between area-level social determinants and late-stage diagnosis of colorectal cancers [abstract]. In: Proceedings of the 15th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2022 Sep 16-19; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2022;31(1 Suppl) nr A036.

oncology,public, environmental & occupational health