Identifying Communities at Risk for Sudden and Unexpected Infant Deaths Using ArcGIS®

Abstract

In 2004, approximately 4,500 cases of sudden, unexpected infant death occurred in the United States and between 1994 and 2004 there were 1,007 SUID deaths in Kentucky. Linking morbidity and mortality rates to geographic areas is a fundamental epidemiological tool, which can be applied to preventing infant death. In 2006 the Centers for Disease Control and Prevention funded seven states to record and collect statewide data to clarify certification practices, and identify if state performances fall short of national expectations. Statewide SUID data were retrospectively collected in Kentucky for the years 1999-2005. There were 575 evaluable SUID cases during the study period. For visual analysis of the data, cases were geocoded then spatially joined to the county GIS data layer in a combined data set in order to create maps. Standardized mortality ratio and probability maps were generated and areas with unexpectedly high or low SUIDs were identified. Of Kentucky's 120 counties, 42 were found to have SUIDs higher than expected (including 20 counties, 48% considered Appalachian and 86% non-metropolitan). The remaining 78 counties were considered average with an expected number of SUID cases or a lower number of SUIDs than would be expected. Identifying regions with higher than expected SUID rates allows specific communities and regions to be targeted. Understanding geographically based risk factors allows for more effective and focused prevention strategies. Similar analyses in other states could target needy areas with limited resources to optimize risk reduction and promote more effective pregnancy planning. Keywords: Sudden infant death syndrome, sudden unexplained infant death, mapping analysis, Appalachian, rural