Racial disparity in birth outcomes: Diversity between local communities

Background: Despite evidence regarding the influence of community-level infrastructure and culture upon race-related health disparity, programming at the local, state and federal levels rarely takes community differences into account. The study goal was to examine variation within a single county regarding racial disparity in birth-outcomes, and to examine the community-level features associated with the variation.

Methods: This was a cross-sectional study linking two secondary datasets using geographic information systems (GIS): 1) 2008 birth certificate record data, generated by the state vital records department (N=2,873), and 2) Y2000 census tract data exported from the Interuniversity-Consortium-for-Political-and-Social-Research data warehouse (N=60). Within each tract, disparity was measured by subtracting the percent of White mothers with a poor birth outcome (premature or low birthweight infant) from the percent of Black mothers with a poor birth outcome. Covariates included socio-economic, residential segregation, environmental, and maternal-health variables.

Results: Census tracts varied widely on the study disparity measure, from -28% (Whites worse) to +93% (Blacks much worse). Communities had distinct socio-economic (median income, vacant housing), demographic (prevalence of education, foreign-born), maternal (prevalence of adolescent pregnancy, prenatal smoking) and healthcare (prenatal care onset) features associated with disparity. In fact, among communities with poorer Black outcomes, those with moderately poorer outcomes were very different from those with substantially poorer outcomes.

Conclusions: Even within a single county, there is wide variation in disparity. Mapping at the community level is a critical step in identifying communities at risk and in targeting the resources to combat that community's particular risk profile.