1.1.5. Analytical Strategy

In order to identify small geographic areas that may reflect gaps in screening efforts. We first performed a spatial analysis using the crude model to determine the unadjusted geographic variation in the prevalence of non-participation in the NMSP. Then covariates with a geographic component were added to the adjusted model and the maps were visually inspected to see if the variable changed the appearance of the map surface, decreasing or not the size of the cluster. Three adjusted analyses were conducted: (1) the deprivation index alone, (2) the deprivation index and proportion of women with non-programme related screenings and (3) the deprivation index and the proportion of women in the census block aged 50 or more as a proxy of social cohesion. Census blocks without any women invited to participate to organized breast cancer screening (for example, an industrial census block or a park) were excluded from the analysis (n = 13, 2.5%).