**Background**

The trends in mortality inequalities by socio-economic level, both at individual and area levels, have been studied in many countries since the 1990s. The majority of these studies have found an increase in these inequalities over time. However, very few studies have analysed the trends in these inequalities in the context of cities and at the same time employed a high degree of geographical desegregation such as that offered by the census tract.

In order to carry out the study of trends it is necessary to clearly define the information needed to construct the indicators, not only of mortality but also of socio-economic level and population; and to agree on the structure of the data base to be prepared by all groups involved.

This is an ecological study of small areas, therefore the analysis must use a methodology which takes the spatial structure of the data into account. It is proposed to use hierarchical Bayesian modelling, specifically the model of Besag, York and Mollie which includes two random effects: one for spatial dependency, the other for unstructured over-dispersion.

**Objectives**

- To define the causes of death to be included in the analysis.
- To define the data base structure.
- To define the statistical models.
- To elaborate routines for the R statistical package to carry out the statistical analysis.

**Methodology**

A manual will be written which will include all the information necessary to construct the data base and to analyse inequalities in mortality for the two time periods (1996-2001 and 2002-2007).

**References**

Barceló MA, Saez M, Cano-Serral G, Martínez-Beneito MA, Martínez JM, Borrell C et al. Métodos para la suavización de indicadores de mortalidad: aplicación al análisis de desigualdades en mortalidad en ciudades del Estado español (Proyecto MEDEA) [ Methods for the smoothing of mortality indicators: application to analysis of mortality inequalities in cities in Spain (MEDEA project) ]. Gac Sanit. 2008; 22 (6):596-608.