Exploring Spatial Contrasts of Mortality Patterns in Ukraine

Svitlana Poniakina, Institut National d'Études Démographiques (INED)

The goal is to investigate if there is any systematic pattern in the geographic distribution of death rates, to measure the strength of spatial autocorrelation and to determine whether exploratory cartographic analysis produces consistent with spatial data analysis results. The spatial unit is a district, received after simplifying of the 3rd level of administrative-territorial structure of Ukraine. The strength of the “geographic correlation” is measured by calculating global (Moran’s) and local (LISA) indexes for each cause of death separately. It is notable that that the highest positive spatial correlation is observed for violent deaths and infectious diseases. The weakest autocorrelation and, therefore, more random distribution of death rates in space is observed for respiratory system diseases and cancers. Comparing thematic maps with maps illustrating LISA indexes helps to determine areas of elevated risk and to outline “hot spots” as well as “potential spatial outliers”.