Resumo

SILVA AYCAGUER, Carlos; BENAVIDES RODRIGUEZ, Alina  e  VIDAL RODEIRO, Carmen Lucía. Análisis espacial de la mortalidad en áreas geográficas pequeñas: El enfoque bayesiano. Rev Cubana Salud Pública [online]. 2003, vol.29, n.4, pp. 0-0. ISSN 1561-3127.

Using maps to represent relevant health data probably provides a more expressive image of their spatial distribution than the simple tables. Through the map it is possible to grasp geographic distribution patterns as well as to suggest possible explanations on identified distributions. For this reason the smoothed construction of disease maps has been the object of recent methodological developments. The traditional approach consists of plotting on a map the standardised mortality ratios. However, when the number of cases in small geographical areas is low, crude mortality rates are very labile, and their representation in maps shows a wide and scarcely informative variation of the existing patterns. Bayesian methods offer the possibility of “correcting” the maps, so that patterns that are not visible when traditional statistical procedures are used can emerge. In order to illustrate and to explain in more detail both methods, mortality data coming from accidents in Cuba for 1998 were used. For this analysis the population was divided in 7 age groups: 0 to 14, 15 to 24, 25 to 34, 35 to 44, 45 to 54, 55 to 64 and 65 and more; as geographic areas we considered the 169 Cuban municipalities. Two maps are shown: one made according to traditional methods, and another one according to the Bayesian ones. In the latter the effect of the smoothing can be clearly observed; the space pattern of mortality, previously diffuse, allows to appreciate that mortality from accidents is specially remarkable in the developed municipalities

Palavras-chave : RESIDENCE CHARACTERISTICS [statistics numerical data]; HEALTH; MAPS; GEOGRAPHIC LOCATIONS; MORTALITY; POPULATION DENSITY; CUBA.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )