The availability of geographic data provides public health officials with the capability to perform two unique types of analysis: 1) finding areas of high or low incidence that can be labelled as statistically significant and worthy of further investigation, and 2) examining the spatial relationship between health outcomes and population and contextual factors that vary across space. Ideally, statistical analyses should simultaneously examine individual and area-level risk factors that lead to health outcomes and the spatial relationship between these factors that give rise to these spatial patterns. Even more complicated is to monitor how outcomes, and the relationship between outcomes and risk factors, vary across space and time. However, a tension exists between geographic variation analysis, which identifies the location and nature of the variation, and non-spatial analysis (e.g., predictive modeling), which may identify characteristics of environments or individuals associated with variation, but does so without spatially specific models. In this project we combine spatially (and temporally) explicit modeling with geographic variation analysis to identify spatial patterns of infant mortality and other poor birth outcomes and the individual- and area-level risk factors associated with these outcomes. Ultimately, a spatial understanding of this problem can contribute to the implementation and evaluation of population-level public health programs which target a variety of birth outcomes.
The project will create a spatial database which includes individual-level data from linked Medicaid and Vital Statistics data sets, area-level risk factors (e.g., poverty, racial segregation, housing quality, etc), and health resources. After investigating risk factors associated with infant mortality (IM) using multilevel spatial modelling, we map our results to investigate how risk factors and infant mortality rates vary across space. The project will also create an online mapping tool to disseminate and update spatial IM data and analyses to stakeholders.
This project seeks to identify high risk communities in Ohio that can be targeted for intervention or allocation of resources and provide a deeper understanding of why these communities are high risk. Mapping of risk factors, model results, and spatial clusters of IM will provide visual representation of the spatial inequalities in both risk factors and IM across the state. Practitioners in high risk communities may integrate assessment of key social, economic and health systems factors when working with women of childbearing age. The spatial analytic methods and the online mapping system will support continued monitoring of IM and other birth outcomes in Ohio – including changes in spatial patterns of IM. Such a tool is valuable for evaluating the impact of population-level health interventions, changes in access to care, and shifts in the socioeconomic climate.