Examining geographic patterns of mortality

Background: Small-area mortality atlases have been demonstrated to be a useful tool for both showing general geographical patterns in mortality data and identifying specific high-risk locations. In Spain no study has so far systematically examined geographic patterns of small-area mortality for the main causes of death. This paper presents the main features, contents and potential uses of the Spanish Atlas of Mortality in small areas (1987–1995). Methods: Population data for 2,218 small areas were drawn from the 1991 Census. Aggregated mortality data for 14 specific causes of death for the period 1987–1995 were obtained for each small area. Empirical Bayes-modelbased estimates of age-adjusted relative risk were displayed in small-area maps for each cause/gender/age group (0–64 or 65 and over) combination using the same range of values (i.e. septiles) and colour schemes. Results: The ‘Spanish Atlas of Mortality’ includes multiple choropleth (area-shaded) small-area maps and graphs to answer different questions about the data. The atlas is divided into three main sections. Section 1 includes the methods and comments on the main maps. Section 2 presents a two-page layout for each leading cause of death by gender including 1) a large map with relative risk estimates, 2) a map that indicates highand low-risk small areas, 3) a graph with median and interquartile range of relative risk estimates for 17 large regions of Spain, and 4) relative-risk maps for two age groups. Section 3 provides specific information on the geographical units of analysis, statistical methods and other supplemental maps. Conclusion: The ‘Spanish Atlas of Mortality’ is a useful tool for examining geographical patterns of mortality risk and identifying specific high-risk areas. Mortality patterns displayed in the atlas may have important implications for research and social/health policy planning purposes.