Objectives: a) To describe the methodology used to construct a deprivation index by census tract in cities, to identify the tracts with the least favorable socioeconomic conditions, and b) to analyze the association between this index and overall mortality.
Methods: Several socioeconomic indicators (Census 2001) were defined by the census tracts of the following cities: Barcelona, Bilbao, Madrid, Seville and Valencia. The correlations with the standardized mortality ratio (1996-2003), and the dimensionality of the socioeconomic indicators were studied. Finally, the selected indicators were aggregated in an index, in which the results of the factor loadings from extraction of a factor by principal components were used as weighting values.
Results: The indicators with the strongest correlations with overall mortality were those related to work, education, housing conditions and single parent homes. In the analysis of dimensionality, a first dimension appeared that contained indicators related to work (unemployment, manual and eventual workers) and education (insufficient education overall and in young people). In all the cities studied, the index created with these 5 indicators explained more than 75% of their variability. The correlations between this index and mortality generally showed higher values than those obtained with each indicator separately.
Conclusions: The deprivation index proposed could be a useful instrument for health planning as it detects small areas of large cities with unfavorable socioeconomic characteristics and is associated with mortality. This index could contribute to the study of social inequalities in health in Spain.