Mammography facilities are accessible, so why is utilization so low?

Background: A lot of attention has been paid to whether mammography facilities are conveniently placed. Apart from service proximity, there are many other factors that might impact one's decision to use mammography.

Objective(s): This paper studies a region with good density of mammography facilities but poor mammography use rates, to better understand other factors that drive mammography use decisions. Over half the women in the sample lived in the same ZIP code as a facility, but less than half received a mammogram, which was fully covered by Medicare for all.

Methods: Our conceptual model is a spatial-interactions approach that recognizes the interplay between personal, social, and physical environments along the pathways to healthcare utilization. We study about 70,000 elderly California women from SEER-Medicare data in 2002-2003, located in all regions of the state. California has widely variable mammography use over geographic regions and considerable heterogeneity in socio-economic, socio-cultural, and geo-spatial factors. We estimate a multilevel probit model of the mammography use decision. We include person-level and community-level covariates to assess independent effects of these different levels of influence, and assess the relative importance of person versus environmental factors, and whether there are cross-level interactions between them.

Results and Discussion: Women in rural communities were more likely, while those living in poor or commuter communities were less likely to use services. Racial segregation was positively associated with use for whites only, and a person's race or ethnicity had an independent effect from community segregation.