Geographic and Temporal Patterns of Screening for Breast, Cervical, and Colorectal Cancer in the US, 1997-2019

Importance Despite guidelines and recommendations, cancer screening remains low and variable across geographic and sociodemographic groups, contributing to persisting disparities. Objective To analyze geographic and temporal patterns of county-level cancer screening prevalence for breast, cervical, and colorectal cancer over a 22-year period in the US and examine sociodemographic factors associated with screening clusters. Design, Setting, and Participants This cross-sectional study used an ecological panel design using county-level screening prevalence as the unit of analysis. Data were collected from US mainland counties (ie, excluding Alaska, Hawaii, and Puerto Rico) from 1997 to 2019, with prevalence estimated over 3- to 5-year periods. Data were analyzed from 2024 to 2025. Exposures County-level screening prevalence was estimated from the Behavioral Risk Factor Surveillance System and the National Health Interview Survey from 1997 to 2019. Socioeconomic and demographic characteristics were estimated from the 2000 US Census and linked with county and geographic clusters of screening. Main Outcomes and Measures Spatial autocorrelation analyses, including Global Moran I and bivariate local indicator of spatial autocorrelation, were conducted to assess geographic clustering of county-level cancer screening prevalence over time. A queen contiguity matrix defined neighboring counties, and permutation tests evaluated statistical significance. Results Data from 3142 counties were assessed. Spatial analyses of mammography, Papanicolaou test, and colorectal cancer screening in the contiguous US from 1997 to 2019 revealed consistent geographic clustering, with high screening prevalence in the Northeast and lower prevalence in the Southwest. Spatial autocorrelation, measured by Global Moran I, declined over time. For example, the distribution of mammography screening became 83% more uniform in more recent years (eg, Moran I = 0.57 in 1997-1999 vs Moran I = 0.10 in 2017-2019). Similarly, Papanicolaou test screening became more uniform in more recent years (eg, Moran I = 0.44 in 1997-1999 vs Moran I = 0.07 in 2017-2019). Bivariate local indicator of spatial autocorrelation analyses identified clusters of persistently high and low screening, with sociodemographic characteristics associated with clusters that changed from low to high. For example, for both Papanicolaou test and mammography, areas that changed from low to high were more likely to be low socioeconomic status, suggesting improvements in disadvantaged areas in 2017 to 2019 vs 1997 to 1999. Disadvantaged areas were most likely to experienced persistently low screening, particularly for breast or colorectal cancer. Conclusion and Relevance These findings suggest that despite increasing screening overall, which led to reduced geographic variation in screening, local clusters of high and low screening persisted in the Northeast and Southwest US, respectively. Future studies could incorporate county-level health care access characteristics to explain why areas of low screening did not catch up to optimize cancer screening practices.