Analise multidimensional de causas múltiplas de óbitos humanos em Belo Horizonte 1998

Abstract

This study of the associations of causes of death was based on th multiple cause aproach, referring to 13,999 deaths certificates ocurred in 1998 in individuals living in Belo Horizonte at the time. Natural deaths and the ones due to external causes such as accidents and violence were evalueted separately. Differently from the classical approach that focuses on one single cause of death - thw underlyng one, this study considers the multiple causes, taking into account all the diagnosis mentioned in the death certificate. The data for this survey were obtained from the review of all the death certificates whose code of basic cause represented the ocurrence of a combination of causes. The affections not considered in the initial phase were individually codified/systematized and included in the survey. All the Deaths Certificates due to external causes were also reviewed in order to include their main consequences in the analysis, like traumatisms, hemorrhages, and fractures, among others not considered in the official statistics of mortality. As a result, a increase in the mean number of diagnosis was observed, even after eliminating the duplications and multiplications of those causes belonging to the same grouping. The ratio between the number of times a certain affection was mentioned as multiple causes and as underlyng one was calculated for the most frequent causes of deaths, as well as the establishment of their positions in both approachs. The association of causes was analyzed according to multiple cause approach, since the objective was to find a broader explanation for the ocurrence of death, trying to fulfill the demographic modifications and the current profile of morbimortality. The Multidimensional Method of Data Analysis through the tecnique of Factorial Analysis of Multiple Correspondence was used to explore the associations of causes, sex and age, both for natural and violent deaths. The results found explained more complex associations by incorporating other causes of death and patterns previously defined, including age and sex. This will certainly define more properly the chain of events that determine death, reinforcing, therefore, the importance of multiple causes of deaths for the studies of mortality and the applicability of the tecnique of Factorial Analysis of Correspondence for this type of surveys.