Complementary results on sex-, age-, and cause-specific mortality in EU countries obtained by SYR symbolic data analysis software

Abstract

Different mortality patterns across countries require different health and demographic policies. Positioning of the countries according to their characteristic mortality pattern can help allocate scarce resources appropriately. We use symbolic data analysis within SYR software to analyse 28 European Union countries' sex-, age-, and cause-specific mortality in 2015. There are two main advantages for using symbolic analysis: (i) it permits more transparent and informative data descriptions along with contextual relations, and (ii) advanced methods adapted for complex data representations can be employed to analyse such data, taking contextual relations into account. Clustering results based on symbolic data analysis show that groups of countries are strongly related to the geographical position of countries, with a clear east–west cut on the first-level partition and with an even more geographically consistent lower-level partition compared to the classical clustering result. Relations between the obtained clusters of countries and their external social and health indicators are well pronounced. We also identify the mortality rates as symbolic variables that discriminate the most between individual countries as well as between the resulting clusters. Knowledge of a country's mortality pattern and its position among comparable countries is valuable information for health and demographic policymakers and can be exploited to exchange good practices.