RESUMEN
OBJECTIVES: The objective of this study was to estimate mortality risk among women exposed to violence in Brazil using population-based data. STUDY DESIGN: This study used a linked database containing nearly 800,000 violence (against women) notifications and 16,500 associated deaths over the period 2011-2016. METHODS: Aggregate age-standardized population-based rates of mortality were built to estimate risk ratios (RRs) at the national and state level, and for different forms of violence and causes of death, as well as type of offender involved, and across various characteristics of the women. RRs compared the rate of mortality among women exposed to violence with that in the general population of women - excess mortality due to violence was also derived from this comparison. The analysis was divided into two time periods (2011-13 and 2014-16). RESULTS: During 2014-16, women exposed to violence had an estimated mortality risk that was 8.3 [95% confidence interval (CI): 8.2-8.5] times higher than that of the general woman population, and an estimated 100 women died on a weekly basis as a direct or indirect consequence of exposure to violence. Higher (all-cause) mortality risk was associated with physical violence and violence that involved repetition and that was self-inflicted. The risk of mortality increased when the cause of death involved external causes (RR: 51.2, 95% CI: 49.6-52.8). When death was attributable to (i) non-communicable diseases and (ii) communicable, maternal, neonatal, and nutritional diseases, the risk was 5.4 [95% CI: 5.3-5.6] and 6.7 [95% CI: 6.1-7.2] times, respectively. Women at greatest (all-cause) mortality risk include white and multiracial (parda) and single women in the age group 10-29 years, who live in the northeast part of the country. When the offender was a partner/ex., women aged 10-19 years showed the greatest (all-cause) mortality risk at 16.9 [95% CI: 13.9-19.8] times. Higher risk was also observed within the age group 30-59 years when death was attributable to external causes (RR: 74.6, 95% CI: 71.3-77.9). For younger women and girls, there was a clear gradient in (all-cause) mortality risk, with those living in the poorest municipalities at greater risk. Age-specific mortality risk also showed significant variation within and across states. CONCLUSIONS: This analysis suggests that most women exposed to violence will likely experience an increased risk of mortality, regardless of her place of residence, age group, racial/ethnic background, marital status situation, and socio-economic status. The estimated RRs are only an approximation given the design of this analysis and should be interpreted with caution.