Spatial variation and cause specific infant mortality in Florida

In 2000, the African American infants had twice the infant mortality rate of whites (Hauck et al 2002). This research aims to determine the contribution of individual characteristics and contextual factors that might affect infant mortality by cause in Florida. Individual factors include maternal characteristics such as race, age and socioeconomic status, but also characteristics surrounding the birth and pregnancy process such as gestational age and birth weight. In addition to individual factors surrounding pregnancy and infancy, this research will incorporate contextual variables to assess the impact that residential characteristics have on infant health and survival. Residential characteristics of particular interest are assessed by zip code and include variables such as proportion of households headed by women, proportion of households below poverty line, etc. Further, we seek to map the effects of contextual variables on cause specific infant mortality by employing software using Geographic Information Systems. The data used to support these analyses are individual level birth records linked with subsequent death records for infants born in Florida in 2000. Multinomial logistic regression techniques are used to evaluate the contribution of individual level and contextual characteristics to cause specific infant mortality in Florida. We expect to find poor areas and those areas where the majority of households are headed by women to be associated with mortality from SIDS and external causes. We also expect that affluent areas will be positively associated with risk of death due to congenital anomalies.