Pregnancy Outcomes and Segregation Based on Race, Ethnicity, and Built Environment​

Objectives: This study investigates the impact of individual, community and provider characteristics on pregnancy outcomes in New Jersey and develops risk-adjustment tools to facilitate evaluation.   

Methods: This study uses individual-level birth outcomes data linked with socioeconomic characteristics of the parents’ residential ZIP code   Multivariate regression is used to examine the association between pregnancy outcomes and individual, family, and  community characteristics.  To assess differential patterns across populations, we examine heterogeneity by race and ethnicity, focusing on whether neighborhood characteristics, including segregation, are differentially associated with outcomes. We estimate a series of nested regression models, starting with a baseline model that includes maternal post-conception characteristics.  Subsequent models incorporate pre-conception factors such as body mass index (BMI), education, and family characteristics, while final specifications additionally include neighborhood-level variables, including poverty and segregation.  Throughout the analysis process, findings are shared with maternal and child health communities to inform model building and recognize qualitative factors. 

Results: This analysis identifies how individual, family, and community characteristics are associated with pregnancy outcomes and evaluates alternative model specifications for risk adjustment. By comparing nested models, the analysis identifies the relative contribution of different risk factors to observed disparities. Findings are communicated through peer-reviewed publication and presentation. By comparing alternative model specifications, the study provides insights into effective approaches for risk adjustment. These are subsequently developed into interactive graphics and maps to support enable risk-adjusted measures of provider and policy effectiveness that allow community organizations, researchers and policy analysts to look at pregnancy outcomes through a variety of filters.