B0747
Title: Geographical evolution of lung cancer mortality by birth cohorts from 1920-29 to 1960-69 in the Italian provinces
Authors: Annibale Biggeri - University of Florence (Italy) [presenting]
Abstract: Space-time analysis of mortality risk is useful to evaluate the epidemiological transition at the subnational level. We previously analyzed space-time variation at a small geographical scale considering as relevant time axis period or birth cohorts, and, bivariate gender-specific disease mapping. We use almost 20 years of mortality data (1995-2016) for males and females in Italy, and we study the spatio-temporal evolution of lung cancer mortality by province and gender. We define the birth cohort as a relevant time axis. The analysis is performed using a space-time Bayesian model with space-time interaction. We found that the geographical pattern of lung cancer mortality changed during time showing that the well-known Italian north-south gradient has been replaced by a new east-west gradient. We confirmed previous results on men - where lung cancer mortality reached the peak around the birth cohort 1920-29 followed by a strong decline. We documented a different time evolution in women - where a previously undetected decline started in many Italian provinces. Different specifications are possible for the space-time interaction term: in our model interaction terms are structured both in space and time.
