Magnitude and trend of the AIDS epidemic in Brazilian cities, from 2002 to 2006.

OBJECTIVE To analyze different patterns of occurrence of AIDS in Brazilian cities between 2002 and 2006, associating trend and magnitude with socio-demographic indicators and local characteristics of the epidemic. METHODS This was an ecological study that categorized cities according to magnitude and trend of the epidemic and subsequently analyzed, considering social indicators, types of HIV transmission and year of first case reported. Data came from the Brazilian Epidemiological Surveillance System, the Brazilian Institute of Geography and Statistics and the United Nations Development Program for Brazil. Linear regression was used to estimate trend and chi-square statistics and ANOVA to analyze indicators. RESULTS A total of 4,190 cities (75.3%) reported AIDS cases between 2002 and 2006. Of these, 3,403 (81.2%) had an occurrence of "small magnitude" (mean=4.7 cases), 367 (8.8%) of "average magnitude" (mean=30.3 cases) and 420 (10.0%) of "great magnitude" (mean=378.7 cases). Cases of "small magnitude" were associated with lower incidence; beginning of the epidemic after 1991; presence of one or two types of transmission; especially heterosexual contact; with occurrences of cases in one or two years of the period; and lower human development index (HDI). Those of a "great magnitude" were associated with larger cities and higher HDI; presence of all types of transmission; beginning of the epidemic between 1980/1991; and trend towards reduction/stabilization, especially due to a decrease in transmission among injecting drug users. Growth of the epidemic was concentrated in "small magnitude" cities, although without significance to the point of changing proportional participation (8.7%) of these cities in the group of cases in Brazil. CONCLUSIONS The AIDS epidemic remains concentrated in urban centers and the spread of cases to the countryside is characterized by irregular occurrence and small magnitude. Cities with low HDI and exclusive transmission through heterosexual contact showed low capacity of increase and the reduction of the epidemic is especially associated with the decrease in transmission among injecting drug users.