Atlas of Mortality in Spanish Municipalities and Census Tracts (1984-2004)

In order to be able to compare the small areas we need suitable estimates of the relative mortality risk and of their time trends in the study period. Two important analytical issues needed to be considered towards this goal. First, the large heterogeneity in population size across areas leads to various levels of precision of areaspecific mortality risk estimates and its temporal trend. Second, age-distribution differences across areas result in mortality risk estimates that are influenced by age effects, a major confounding factor in the use of atlas. To reduce the impact of these problems, we applied the next methods: Empirical Bayes estimation; age-adjustment procedure and maps of high and low mortality risk. Also is the approach used to study evolution over time of relative risk (See Appendix II in the Book for technical details).