摘要：The paper investigates spatio-temporal trends in health disparities through an empir- ical example. We deal with geographical health pattern in Italy from 1991 to 2010, starting from infant mortality data available at the provincial level and assessing the existent disparity among macro-regions (the conventional Northern, Central and Southern macro-regions). After a discussion concerning suitable inequality indices and their decompositions when dealing with small area data, we propose a mo del-based approach that allows to properly tackle sampling variability. Results give evidences of persisting spatial disparity in infant mortality along time.
关键词：health inequality; Gini index; generalised entropy; hierarchical Bayesian modelling; ; small area estimation; spatio-temporal smoothing; sampling variability; INLA
