[Temporal effects of corrected female homicide mortality estimates in Northeast Brazil]

This study aimed to analyze the temporal effects (age, period, and cohort) on female homicide mortality in the states of Northeast Brazil from 1980 to 2017. This ecological time trend study used APC with a Bayesian approach and the deterministic method Integrated Nested Laplace Approximations (INLA) in the parameters' inference. The female homicide rates for each state of the Northeast were standardized by the direct method after correction of the death records for quality of information and underreporting. Data were also obtained on race/color, place of death, and means of perpetration. During the period analyzed, after correcting the death records, the Northeast region showed a mean rate of 5.40 female homicide deaths per 100,000 women, with a significant increase in all the states in the 2000s. In all the states, there was an increase in relative risk (RR) of homicide death in the second and third decades of life and a protective effect in older women. Except for the state of Sergipe, there was an increase in the risk of death in all five-year periods in the 2000s. The Northeast region as a whole and the states of Paraíba, Pernambuco, and Piauí showed a protective effect for women from older generations. There were also higher proportions of deaths in black women, homicides committed at home, and those perpetrated with firearms. The current study's findings may correlate with the spread of violence in Brazil in the 2000s and the Brazilian State's failure to protect women from violence.