Clustering EU Countries by Causes of Death

Countries with different mortality patterns face different health and demographic challenges. Knowing a country’s position relative to other countries with respect to sex-age- and cause-specific mortality can be very helpful when deciding policy and trying to manage costs efficiently. We argue that supplementing ‘quantitative’ mortality information (i.e., mortality rates) with the ‘qualitative’ information (i.e., structure by causes of death) provides additional insight that can inform health and social policy. This new knowledge, which cannot be obtained simply by considering data about life expectancy at birth, could be used to enhance the transfer of good practices between countries. Using the Eurostat 2015 mortality data, which are grouped by sex, age and cause of death, we analysed similarities amongst the 28 EU countries using classical clustering methods, applying five-year age groups and organising causes of death into three main groups, namely neoplasms, diseases of the circulatory system and diseases of the respiratory system. To demonstrate the advantages of including additional information on mortality, we compared clustering based on the above data with the clustering based on life expectancy at birth, one of the most common demographic indicators. The clusters obtained using a sex-age-cause approach were more geographically coherent. We also identified the sex-age-cause combinations that discriminate best between the clusters of countries. The factor that discriminated best between the clusters was not overall mortality rate, but mortality from diseases of the circulatory system in people aged over 80 years.