[Relation between the level of the development of the care system and the level of "preventable" mortality according to department in France].

"Avoidable" mortality may be defined as a selection of causes of death whose occurrence is closely related to medical intervention. We have attempted to relate the variations in mortality from these selected causes to health service input among the 95 French departments. This analysis was controlled for social factors. Firstly we performed independently two factor analysis of the departmental variations in health care variables and in social factors. These studies were carried out with the aim of creating new discriminant and independent variables. Multiple stepwise regressions were then used to analyse the independent and joint associations of these new variables with mortality. The data show principally the considerable weight of social factors in explaining mortality differences. On the contrary, we found very weak relationships between health service input and mortality. However in view of the analysis method, it would appear unjustified to interpret this result in terms of a criticism of the "avoidable" mortality concept.