RESUMEN
Background: Homicide rates in Brazil are among the highest worldwide. Although not exclusive to large Brazilian cities, homicides find their most important determinants in cities' slums. In the last decade, an urban renewal process has been initiated in the city of Belo Horizonte, in Brazil. Named Vila Viva project, it includes structuring urban interventions such as urban renewal, social development actions and land regularization in the slums of the city. This study evaluates the project's effect on homicide rates according to time and interventions. Methods: Homicide rates were analyzed comparing five slums with interventions (S1â»S5) to five grouped non-intervened slums (S0), with similar socioeconomic characteristics from 2002 to 2012. Poisson regression model estimates the effect of time of observation and the effect of time of exposure (in years) to a completed intervention, besides the overall risk ratio (RR). Results: Using the time of observation in years, homicide rates decreased in the studied period and even more if considered cumulative time of exposure to a completed intervention for S1, S2, S3 and S4, but not for S5. Conclusions: Although the results of the effect of the interventions are not repeated in all slums, a downward trend in homicide rates has been found, which is connected to the interventions. New approaches could be necessary in order to verify the nexus between slum renewal projects and the reduction of homicide rates.