The State of São Paulo accounts for approximately 40% of the AIDS cases notified in Brazil and provides a suitable opportunity for space-time analysis aimed at better understanding of the dissemination of HIV/AIDS. Using the AIDS cases notified to the Ministry of Health between 1990 and 2004, among individuals aged 15 years or over, and the Ministry of Health's information system for disease notification (Sistema de Informação de Agravos e Notificação, SINAN) as the information source, the relative risks of AIDS over three-year periods were estimated using full Bayesian models, for each gender. The models used were shown to be adequate for explaining the process of AIDS dissemination in the State of São Paulo and demonstrated the growth among females and in small-sized municipalities. They also suggested that the municipalities currently most affected are in regions of economic growth and have populations of less than 50,000 inhabitants.
Space-time; Bayesian models; HIV/AIDS; São Paulo
