This study assessed whether a validated, housing-based individual socioeconomic status measure, HOUSES, is associated with adherence to colorectal cancer (CRC) screening in rural communities and whether integrating HOUSES into geospatial analysis can identify geographic targets for intervention. The population comprised Mayo Clinic Midwest patients living in cities without ready access to routine care within the Mayo Clinic Health System in 2019, used to represent rural communities.
The investigators analyzed CRC screening data for a cohort of 34,489 individuals. Key sample descriptors reported were a median age of 64.0 years and 52.4% female. The primary exposure was the individual-level HOUSES index categorized into quartiles; the primary outcome was CRC screening adherence.
A multilevel mixed-effects logistic regression model was used to evaluate the association between HOUSES quartile and CRC screening adherence. Models adjusted for age, sex, race/ethnicity, comorbidity, distance from home address to clinic, and area deprivation index. The analysis focused on individuals residing in areas defined by the study team as lacking ready access to routine care to approximate rural populations within the Mayo Clinic Health System.
In addition to the individual-level regression, the authors performed geospatial analysis to identify geographic hotspots characterized by two features: lower CRC screening rates and concentrations of residents in the lowest HOUSES quartile (HOUSES Q1). The study reported the number of hotspots identified and quantified the correlation between counts of HOUSES Q1 residents and counts of individuals with low CRC screening within those hotspots.
Among the 34,489 individuals analyzed, those in the lowest HOUSES quartile (Q1) had significantly lower odds of being adherent to CRC screening compared with those in the highest quartile (Q4). The adjusted odds ratio for HOUSES Q1 versus Q4 was 0.63 with a 95% confidence interval of 0.58–0.69, corresponding to a 37% lower odds of screening adherence in the lowest SES group after adjustment for covariates.
Descriptive statistics provided by the authors included the cohort size, median age, and sex distribution; other covariates used in adjustment included race/ethnicity, comorbidity, distance to clinic, and area deprivation index, though detailed distributions of these covariates were not reported in the abstract.
The geospatial component identified 14 geographic hotspots that were characterized as HOUSES Q1 hotspots. Within these 14 hotspots, the study found a strong correlation between counts of residents in HOUSES Q1 and counts of low CRC screening, with a reported correlation coefficient of 0.81. This indicates substantial co-location of low individual-level socioeconomic status (SES) and low screening rates in specific geographic areas.
The authors propose that mapping HOUSES alongside screening adherence can highlight target areas where tailored outreach or programmatic interventions might be prioritized to reduce disparities in CRC screening.
The study concluded that lower SES, as measured by the housing-based HOUSES index, was significantly associated with lower CRC screening adherence among the rural populations represented in this Mayo Clinic cohort. The integration of HOUSES into geospatial analysis identified hotspots with both concentrated low SES and low CRC screening, suggesting that HOUSES-enabled mapping could guide targeted interventions to address screening disparities in rural communities.
The authors suggest that HOUSES may serve as a useful digital tool for cancer preventive care and research, particularly for identifying individuals and locations at higher risk of under-screening due to socioeconomic disadvantage.
The source notes that datasets generated and/or analyzed are not publicly available because they include protected health information. The abstract does not provide detailed subgroup distributions (beyond age and sex) or detailed model coefficients beyond the primary adjusted odds ratio and confidence interval for HOUSES Q1 versus Q4. Any additional limitations or sensitivity analyses beyond what is reported in the abstract were not described in the source.
The study obtained Institutional Review Board approval from the Mayo Clinic (IRB #19-009328). The authors declared no competing interests. Funding acknowledged in the source included awards from the National Institute on Aging (R21AG65639), the National Heart, Lung, and Blood Institute Division of Intramural Research (R01HL171508), the National Center for Advancing Translational Sciences (UL1 TR002377), and support from the Mayo Clinic HOUSES Program.