Artigos

Conteúdo (article):

Tendência temporal da mortalidade por homicídios femininos perpetrados por arma de fogo nos estados da Região Nordeste no período 2000-2019
Temporal trends in female firearm homicides across states in the Northeast of Brazil during the period 2000-2019
Autores
Karina Cardoso Meira1*
Stefany Freire Cosme de Oliveira2
Taynãna César Simões3
Carinne Magnago4
Rafael Tavares Jomar5
Pedro Gilson Beserra da Silva6
Eder Samuel Oliveira Dantas7

1Escola de Saúde da Universidade Federal do Rio Grande do Norte, e-mail:ninameira87@gmail.com,orcid: https://orcid.org/ 0000-0002-1722-5703
2Escola de Saúde da Universidade Federal do Rio Grande do Norte,
e-mail:stefanyfreire67@gmail.com,orcid: https://orcid.org/0000-0002-8213-8083
3Instituto de Pesquisa René Rachou, Fundação Oswaldo de Minas Gerais; taynana.simoes@fiocruz.br;orcid https://orcid.org/0000-0002-5849-343X
4Faculdade de Saúde Pública da Universidade de São Paulo,e-mail:carinne@usp.br, orcid:https://orcid.org/0000-0001-8799-3225
5Coordenação de Assistência, Instituto Nacional de Câncer, e-mail: rafaeljomar@yahoo.com.br;orcid: https://orcid.org/0000-0002-4101-7138
6Hospital Universitário Onofre Lopes da Universidade Federal do Rio Grande do Norte,
e-mailpedrogilsonsilva@yahoo.com.br, https://orcid.org/0000-0002-3888-5807
7Hospital Universitário Onofre Lopes da Universidade Federal do Rio Grande do Norte, edersamuel_rn@hotmail.com, orcid:https://orcid.org/0000-0002-6595-6105

Resumo
Objetivo: Analisar a tendência temporal dos homicídios femininos perpetrados por arma de fogo nos estados nordestinos, no período de 2000 a 2019. Métodos: Estudo ecológico, com dados de homicídios por arma de fogo em mulheres com 10 ou mais anos, registrados no Sistema de Informação Sobre Mortalidade. Os dados de mortalidade e os dados populacionais foram obtidos junto ao Departamento de Informática do Sistema Único de Saúde. Após a correção dos registros de óbito para qualidade e cobertura dos óbitos, as taxas de mortalidade foram calculadas Tendências foram avaliadas por regressão binomial negativa, classificadas de acordo com o valor do risco relativo e valor de p. Calculou-se a variação percentual anual média das taxas de mortalidade. Resultados: A região apresentou 4,40 homicídios por arma de fogo por 100 mil mulheres no período do estudo, maiores coeficientes em Alagoas (5,40), na faixa etária de 15-19 anos (5,84), e via pública (1,58). As tendências foram ascendentes, com exceção de Pernambuco em que foi descendente, e estacionárias em Alagoas. Os homicídios por arma de fogo em via pública apresentaram maior percentual de aumento no período estudado. Conclusão: Observou-se tendência ascendente nos homicídios femininos perpetrados por arma de fogo na maioria dos estados nordestinos.
Palavras-chave: Mortalidade; Homicídio; Mulheres; Violência com Arma de Fogo; Violência de Gênero.

Abstract
Objective: To analyze temporal trends in female firearm homicides in the Northeast of Brazil during the period 2000-2019. Methods: We conducted an ecological study using data on firearm homicides of women aged 10 years and over obtained from the Mortality Information System. The population data were taken from the 2010 Census. Homicide rates were calculated after correcting the data to account for differences in the quality and coverage of death records. Trends were assessed using negative binomial regression and described using relative risk and p values. Average annual percentage changes in homicide rates were also calculated. Results: The regional firearm homicide rate during the study period was 4.40 per 100,000 women. Rates were highest in the state of Alagoas (5.40), the 15-19 age group (5.84) and in public thoroughfares (1.58). Trends were upward across all states except Pernambuco, where they were downward, and Alagoas, where rates were stationary. The place of occurrence with the highest percentage increase in firearm homicides over the study period was public thoroughfares. Conclusion: Female firearm homicides showed an upward trend across most northeastern states.
Keywords: Mortality; Homicide; Women; Gun violence; Gender-based violence.

Introduction
Gender violence is recognized worldwide as a serious social problem and grave violation of human rights. It has a major social and economic impact and considerable effect on women’s health. It is estimated that at least one-third of women aged 15 years and over have been subjected to physical and/or sexual violence at least once in their lifetime1.
While men are more likely to be murdered than women, this pattern is different when it comes to intimate partner homicide. The proportion of female victims of this type of homicide is six times greater than that of men2. This data exposes the extremely grave consequences of gender violence and the most brutal manifestation of patriarchal power: femicide, the intentional murder of women because they are women3.
In Brazil, rates of violence against women are alarming. In 2020, 1 in every 4 women aged 16 years and over was subjected to some type of violence4. More than 260.000 cases of bodily harm due to domestic violence, 54,453 cases of rape and 4,631 cases of attempted murder were reported. There were also 3,913 female homicides, corresponding to 3.6 deaths per 100,000 women. Of this total, 34.5% were considered femicides, and more than half were committed with a firearm5. In recent years, the Northeast has been the region that has reported the highest absolute number of intentional violent deaths of women, most of which involving a firearm6.
At national level, data from the Mortality Information System (SIM, acronym in Portuguese) do not always make it clear whether female homicides are gender-related. This is partially due to the limitations of the SIM, which does not always show the relationship between the victim and the perpetrator6-10, and the fact that the classification of deaths is the responsibility of the police, who are not always prepared to identify gender violence, especially because the legislation on femicide in Brazil is recent5.
To contribute to the classification of gender violence, researchers have created indirect evaluation indicators1,8,11-16. In addition, the Institute for Applied Economic Research (IPEA) uses female homicides committed at home and female homicides perpetrated with a firearm as a proxy for femicide14-15.
The use of this indicator is justified by the fact that the majority of female homicides4-6,8-15 and homicides of women with prior notification of violence7,16 in Brazil are committed with firearms. In addition, research indicates that previous history of intimate partner violence, carrying of a firearm by an intimate partner and living with an intimate partner are leading risk factors for femicide1-2,17-18.
The present study therefore uses female firearm homicides as a proxy for femicides. The aim of this study was to analyze temporal trends in female firearm homicides in states in the Northeast of Brazil during the period 2000-2019 and calculate the proportion of homicides by race/skin color, place of occurrence and marital status.

Methods
Study design, data source and variables
We conducted an ecological study of temporal trends in female firearm homicides in states in the Northeast of Brazil during the period 2000-2019. We used data from the SIM, which is run by the national health system’s Department of Informatics (SIM/DATASUS). The de-identified data are available in the public domain at the following website: https://datasus.saude.gov.br/mortalidade-desde-1996-pela-cid-1019. The Northeast is made up of nine states: Alagoas (AL), Bahia (BA), Ceará (CE), Maranhão (MA), Paraíba (PB), Pernambuco (PE), Piauí (PI), Rio Grande do Norte (RN) and Sergipe (SE).
We analyzed all records of deaths of women aged 10 years and over classified as female firearm homicides. We included the 10-14 year age group because childbearing age is assumed to start at age 10 in Brazil and sexual violence against girls, teenage pregnancy and teen marriage are common, increasing the risk of domestic violence and femicide20. In addition, homicide rates begin to rise from this age group6-13.
The study variables were year of occurrence (2000-2019), state (AL, BA, CE, MA, PB, PE, PI, RN and SE), age group (10-14, 15-19, 20-39, 40-59 and 60 years and over), place of occurrence (hospital and other health facilities, at home, public thoroughfare, other and ignored), race/skin color (white, black [black and brown], yellow, indigenous and ignored), and marital status (single, married, widow, separated/other and ignored).
Deaths classified in the SIM as assault by firearm (X93 to X95) and events of undetermined intent (EUI) based on the 10th Revision of the International Statistical Classification of Diseases (Unspecified firearm discharge, undetermined intent, Y22, Y23, Y24; and Unspecified event, undetermined intent, Y34)) were included.
The population data were taken from the 2010 Census, conducted by the Brazilian Institute of Geography and Statistics, and intercensal estimates of population for the period 2000-2019 obtained from DATASUS.
Major differences in the quality and coverage of death records across states were observed during the study period8-13, meaning that the data required correction.
Record correction and descriptive analysis
First, the death record data were corrected to improve information quality by performing a proportional redistribution according to year, age group and state involving the following stages proposed by Garcia et al8: (1) calculation of the number of deaths classified as firearm homicides (X93 to X95) as a proportion of total deaths due to accidental and intentional external causes (self-harm, assault with a firearm, accidental trauma and legal intervention); (2) multiplication of the result of the first stage by total number of deaths classified as EUI under the codes Y22, Y23, Y24 and Y34; (3) the result of the second stage was then added to the total number of deaths originally classified as firearm homicides to give the corrected number of deaths; 4) We then applied the correction factors published by Queiroz et al. (2017)21, which were obtained by using the adjusted Synthetic Extinct Generations method (SEG-adj)22. To maximize the reliability and trustworthiness of the correction process, the analyses were performed independently by two researchers. The correction process was performed for all states for firearm homicides according to race/skin color, marital status and place of occurrence.
After data correction, we calculated the uncorrected homicide rates, rates of EUI (Y22-Y24, and Y34) and corrected firearm homicide rates (all places of occurrence, public thoroughfare and at home) by age group and period.
After correcting the data by race/skin color, marital status and place of occurrence, we calculated the percentage of deaths for each variable and state by dividing the number of deaths for each variable category by total deaths during the period 2000-2019.
Mortality rates for the study period (2000-2019) were calculated by dividing the total number of deaths during the period by the total population during the period and multiplying by 100,000 women. The rates were standardized using the direct method to control for age, using the female population in Brazil according to the 2010 census as the standard population.
Temporal trends in corrected firearm homicide rates (all places of occurrence, public thoroughfare and at home) were assessed by calculating mortality rates smoothed by three-year moving averages to correct random fluctuations in annual mortality rates and create line charts23.

Statistical methods
The analysis of the time series was performed using negative binomial regression and the Durbin-Watson statistic to check for the presence of serial autocorrelation9,10. The dependent variable for each state was number of deaths in each year (y_t) and the independent variable was the t-centered calendar year. Negative Binomial regression models were run, adding and an offset term as a natural logarithm of the at risk population in each state in the respective calendar year, represented by the following equation: y_t~Negative Binomial (μ_t,θ)
□log (μ_t ) =log (N_t )+β_0+β_1 (t-E(t))+ϵ_t
ϵ_t~Gama(θ,θ)
Where µ is the number of expected deaths in year t, Ɵ is the dispersion parameter, N_t is the at risk population in year i , β_0 is the intercept, β_1 is the time-centered variable (t-E(t)) and ϵ_i is the random error with Gamma(θ,θ) distribution. Overdispersion is considered in the estimate of variance VAR(Y_t )=μ_t+(μ_t^2)/θ.
Trends were classified as stationary, downward or upward, based on the relative risk (RR) value, calculated by exponentiating the regression coefficient (e^(β_1 (t-E(t)) )) and respective 95% confidence intervals (95%CI) and p-values24.
We then calculated the average annual percentage change in rates to determine whether the percentage increase or decrease was statistically different from zero using the two-tailed t-test with two degrees of freedom (p < 0.05).
Ethical aspects
The study did not require ethical approval as it was conducted using exclusively secondary data available in the public domain. The study was conducted in accordance with the norms and standards set out in National Health Council Resolution 466 (12 December 2012).

Results
There were 13,733 female firearm homicides in the Northeast during the period 2000-2019 (3.06 homicides per 100,000 women) and 8,571 deaths classified as EUIs (1.91 deaths per 100,000 women). The state with the highest rate of EUIs was BA (2.98), followed by PE (2.82) and RN (2.14). The state with the lowest rate was AL (0.07), followed by PB (0.19) and MA (0.39) (Table 1).
After data correction, the homicide rate increased by 32.03%, from 3.06 to 4.04. The highest percentage increases were observed in BA (2.85 vs 4.03; +41.40%) and RN (2.96 vs 4.05; +36.82%) and the lowest increases were found in PB (3.31 vs 3.78;+14.20) and AL (4.68 vs 5.40;+15.38%) (Table 1).
The states in which corrected rates were higher than the regional rate were PE, AL, CE, PI and RN. Rates were higher in women aged 15-19 years and 20-29 years across all states, with rates declining steadily with age. In contrast, the EUI rate showed a positive age gradient (Table 1). Firearm homicide rates were higher in public thoroughfares than at home across all states except PI and MA, where the rates for these places of occurrence were similar (Table 1).
The most common place of occurrence of firearm homicides was public thoroughfare across all states except MA and PE, where most deaths occurred in a hospital. Hospitals and other health facilities were also common places of occurrence of homicides. Black and single women were the most victimized groups, accounting for between 61.57% and 80.56% and 49.22% and 74.52% of firearm homicides, respectively (Table 2).
The exploratory analysis of temporal trends in overall rates of firearm homicide and by place of occurrence revealed three patterns: (1) a reduction in rates from 2010 in AL, PB and PE; (2) an increase in rates in BA, CE and RN; and (3) an increase in rates with a reduction at the end of the time series in MA, SE and PI (Figures 1 and 2).
The trend analysis revealed an upward trend in firearm homicide rates in the region and across all states (RR > 1; p < 0.05), except AL (RR = 1.01; 95%CI 0.98-1.04) and PB (RR = 1.03; 95%CI 0.99-1.07), where the trend was stationary, and PE, which showed a downward trend (RR = 0.98; 95%CI 0.96-0.99). The average annual percentage increase in the regional firearm homicide rate was 3.58%. The states with the highest percentage increase in rates were CE (11.39%), RN (7.43%) and MA (6.32%). Average annual percentage reduction was 2.32% (95%CI -3,35-1.29%) in PE (Table 3). Similar rates were found for firearm homicides committed at home (Table 3).
The rate of firearm homicides in public thoroughfares showed an upward trend in the region and across all states except AL (RR = 1.013; 95%CI 0.97-1.05) and PE (RR = 0.97; 95%CI 0.96-0.99). Firearm homicides committed in a public thoroughfare showed the highest average percentage increases, with the highest rate being found in CE (15.75%), followed by RN (12.14%) and PI (8.56%) (Table 3).

Discussion
Death record quality and coverage were similar across states during the study period. The highest rates of deaths classified as EUI (Y22-Y24 and Y34) were found in BA, PE and RN. The states with the lowest rates of deaths classified as EUI also had the poorest coverage by the SIM. After data correction to account for death record quality and coverage, the states with the highest female firearm homicide rates were PE, AL, CE, PI and RN. All states showed an upward trend in firearm homicide rates (all places of occurrence, at home and public thoroughfare), except AL and PB, where rates (overall and at home) were stationary, and PE, where the rate showed a downward trend.
Data correction is required in time trend analysis studies of areas with differing levels of death record quality and coverage, as changes in information system quality and coverage can lead to period effects in data, distorting trends and over or underestimating mortality rates6,8,13,25-26.
Despite improvements in the quality of death certification, coverage and data completeness across all states, the proportion of deaths in the SIM where the underlying cause is a garbage code (undetermined intent) remains high9-10. While the proportion of deaths with ill-defined external causes decreased by 17.2% between 2009 and 20139, this rate increased by 25.6% between 2017 and 2018. Furthermore, five of the ten states with the lowest levels of quality of information on external causes were in the Northeast (BA, PE, RN, CE and SE), with rates of deaths classified as EUI exceeding 5.0 per 100,000 population15.
Studies have shown increasing rates of female homicides with decreasing rates of rates of deaths classified as EUI and vice versa6,12-13. Similar results were reported with male homicides in PE between 1980 and 2014, with the reduced risk of homicide in the 2000s being accompanied by an increase in the classification of deaths as EUI26. These findings point to the need to correct death record data before analyzing temporal trends in rates of death due to external causes, which was done by the present study, resulting in an increase of more than 30% in the firearm homicide rate in Northeast as a whole and in the states of BA and RN.
Six countries (Brazil, United States, Mexico, Colombia, Venezuela and Guatemala) account for over half of worlwide deaths from firearm injuries27. In 2016, Brazil was the country with the highest number of firearm deaths (43,200), accounting for one-sixth of all fatalities worlwide27. In 2020, the country reported just over 50,000 intentional homicides, 78.0% of which were committed using a firearm5. Around 4,000 women are murdered each year in Brazil, corresponding to 5% of female victims of homicide worldwide, reiterating that violence against women is a serious public health problem4-5.
Brazil’s most violent region is the Northeast, accounting for 44.0% of total intentional violent deaths in 2020 and harboring the three states with the highest homicide rates (CE, BA and SE)5. Between 1980 and 2019, the female homicide rate grew faster than the male homicide rate, from 1.64 to 4.38 per 100,000 women (+167%)6. Different studies show that the most commonly used weapon was a firearm4-15.
In the present study, the overall corrected average regional homicide rate was 4.04 per 100,000 women. The state with the highest rate was AL (5.40 per 100,000 women, which is 3.6 times higher than that of PI, with 1.50 deaths per 100,000 women). In concordance with the results of other studies, our findings show that rates were higher among young, single black women, and that the most common place of occurrence among this group was a public thoroughfare8,13-16.
Studies in Brazil reveal different trends in homicide rates between white and black women in the 2000s, with a downward trend among the former and upward trend in the latter. The difference in homicide rates between these groups increased from 48.5% in 2009 to 65.8% in 201915. Racial disparities were greatest in states in the Northeast, with homicide rates among black women being 5.2 and 4.4 times higher than in white women in RN and SE, respectively. In AL, in 2019, all female victims of homicides were black15.
Factors influencing violence are directly associated with unequal gender relations characterized by female submission. While all women are vulnerable to violence in a patriarchal, racist and capitalist society, factors such as skin color/race/ethnicity and material conditions can exacerbate the risk of violence28. Historically, due slavery and structural racism, black people are more likely to experience violence and die violently. In addition, poor black women are at the bottom of the social pyramid and therefore encounter greater hardship in coping with situations of violence27,29.
The firearm homicide rate was highest in single women across all northeastern states. This may be explained by the age effect, given that the risk of death was higher among younger women, most of whom are single6,8,13,30-34. However, the fact that victims were classified as single on the death certificate does not necessarily mean they were not in a non-formal relationship16 or single women who were killed by their ex-partner.
Young women (15-39 years) are exposed to greater risk of gender violence and homicides because this group is exposed to multiple types of male violence, such as family violence, intimate partner and ex-partner violence and is more exposed to risk factors linked to machismo and sexism in public thoroughfares, especially at night35-40. Studies29,32 show that women are more likely to be victims of intimate partner and ex-partner violence and gender violence, although rates vary across states, countries and continentes33.
It is estimated that 87,000 women worldwide are victims of intentional homicide each year and 137 women are killed every day by a family member33, which makes the home the most likely place of occurrence. In the present study, the main place of occurrence of homicides at regional level was a public thoroughfare (36.73%), followed by a hospital (27.75%) and at home (21.56%).
Place of occurrence is not a determinant of gender violence. This type of violence can take place in a range of different settings. Public thoroughfares would appear to be an opportunistic setting for homicides committed by ex-partners and partners who do not live with the victims14,15. A previous study reported that 25.2% of female homicides with previous reports of domestic violence between 2011 and 2016 were committed in a public thoroughfare7.
Our findings reveal an upward trend in rates of firearm homicide at home and in public thoroughfares at regional level and across all states except AL (stationary) and PE (downward). Similar trends were reported in small, medium and large cities in northeastern states during the period 2000-201540. The states with the highest annual percentage increase in rates of firearm homicide (all places of occurrence, public thoroughfare and at home) were CE, MA and RN.
Similar results have been observed by other studies showing an increase in the rate of female homicides across all regions during the period 1980-2014, except the Southeast, where there was a downward trend6, in most northeastern states13, and in health regions in RN during the period 2000-201630. In all northeastern states except SE, there was an increase in the risk of death in the 2000s after data correction when compared to the baseline period 1995-199913. In addition, there was an increase in the likelihood of death among women born after the 1960s.
This situation may be linked to the conservative culture of this region and the fact that younger generations of women went through significant sociocultural changes, questioning the role assigned to women in a patriarchal society6,13,37,39. Daring not to play traditional gender roles increases the likelihood of exposure to gender violence due to the reaction of the patriarchy, which uses violence, including lethal violence, to force women to reoccupy their original position36-39. In addition, in the 2000s the Northeast witnessed the spread and interiorization of violence, with a significant increase in rates of homicide in small, medium and large towns and cities (2000-2015)40. Studies have reported a correlation between male8-10,14-15 and female homicides in cities and regions marked by organized crime and drug disputes, resulting in armed conflict, disrespect for human rights and gender violence11,31,36-37.
The results of the present study are worrying because of the high rates of firearm homicide among women living in the Northeast, both at home and in public thoroughfares, showing that these settings are unsafe for women in most states with upward trends in mortality. During the 2000s, the government implemented policies to curb violence against women, reducing the circulation of firearms through the Disarmament Statute. Other legislation includes Law 11340/2006, known as the Maria da Penha Law, which is the country’s most important legal milestone on the road to tackling violence against women. Creating mechanisms to curb domestic and family violence, the law establishes more severe punishments for perpetrators and introduces measures to protect women exposed to violence41-42. A reduction in rates of homicide due to this group of causes was therefore expected14-15,41-42.
However, these policies have not been sufficient to contain the increase in violence against women. Twelve years after the creation of the Maria da Penha Law, only 137 (2.4%) of Brazil’s 5,570 municipalities have shelters for women exposed to domestic violence. These shelters are concentrated in large cities in the South and Southeast and less than 10% of the country’s municipalities offered specialist services for women suffering sexual violence, and only 8.3% of cities had police stations that deal specifically with crimes involving female victims43-44. During the period 2017-2019, there was a 75% reduction in funding for the “2016 Program: Policies for Women: Promoting Autonomy and Combating Violence”45. Underfunding of social services and women’s and child protection services is a reflection of government disinterest in the life of women45-46.
In addition to the above, there has been a reduction in funding for tackling violence against women. Since 2019, the federal government has introduced 11 decrees, one law and 15 army regulations weakening firearms and munitions control and enforcement14-15, exacerbating the risk of domestic violence, femicide, suicide and accidents involving children47-50. In Brazil, as in other countries, policies restricting access to guns have resulted in a reduction in homicides7-18,47-52. A recent decree issued by Brazil’s new government is therefore an important step forward. Decree 11366, introduced earlier this year, suspends the issuing of new firearms licenses for hunters, sport shooters, collectors and private owners, reduces the maximum limit for purchases of permitted guns and ammunition, suspends the issuing of new firearms licenses for shooting clubs and schools, and sets up a working group to propose new regulations for the 2003 Disarmament Statute53.
This study has some limitations. First, the use of secondary data with incomplete entries and inconsistencies may have influenced the results, especially given the number of records where death was classified as a EUI (Y22-Y24 and Y34) and changes in coverage of records over the study period. However, this limitation was partially addressed by data correction.
Another limitation was the use of an indirect indicator of femicides due to the lack of a specific database with this information. However, it is important to highlight that we adopted a proxy used by the IPEA14-15. The use of all female homicides as an indirect indicator of femicide can lead to the overestimation of gender-related killings. In this respect, public security data indicate that 34.5% of female homicides are femicides5.
Despite these limitations, our findings provide important insights for the planning and evaluation of violence prevention policies in the Northeast, as the generation of information about this grave public health problem can help push measures to tackle violence onto health and development agendas and contribute to the implementation of policies targeting specific groups.

Conclusion
Most of the states in the Northeast showed an upward trend in female firearm homicides, both at home and in public thoroughfares. The findings signal the need for health services to report cases of violence against women and promote health education, especially in primary care settings, focusing on prevention and lines of care for women exposed to violence. It is therefore important that health services promote coordinated actions with security and social protection services to ensure the provision of assistance and care to women in situations of violence. Finally, there is an urgent need to increase funding for violence prevention and protection measures.

References
1. World Health Organization. Violence against women prevalence estimates, 2018: global, regional and national prevalence estimates for intimate partner violence against women and global and regional prevalence estimates for non-partner sexual violence against women. Switzerland: WHO; 2021.
2. Stöckl H, Devries K, Rotstein A, Abrahams N, Campbell J, Watts C, Moreno CG. The global prevalence of intimate partner homicide: a systematic review. Lancet 2013; Sep 7;382(9895):859-65.
3. Pasinato W. “Feminicídios” e as mortes de mulheres no Brasil. Cad. Pagu 2011; 37:219-246.
4. Fórum Brasileiro de Segurança Pública; DataFolha Instituto de Pesquisas. Visível e invisível: a vitimização de mulheres no Brasil. 3ª ed. São Paulo: FBSP; DataFolha; 2021.
5. Fórum Brasileiro de Segurança Pública. 15º Anuário Brasileiro de Segurança Pública. São Paulo: FBSP; 2021.
6. Souza ER, Meira KC, Ribeiro AP, Santos J, Guimarães RM, Borges LF, et al. Homicídios de mulheres nas distintas regiões brasileiras nos últimos 35 anos: análise do efeito da idade-período e coorte de nascimento. Cien Saude Colet. 2017;22(9):2949-62.
7. Barufaldi LA, Souto RMCV, Correia RSB, Montenegro MMS, Pinto IV, Silva MMA, et al. Gender violence: a comparison of mortality from aggression against women who have and have not previously reported violence. Cien Saúde Colet. 2017;22(9):2929-38.
8. Garcia LP, Freitas LRS, Silva GDM, Höfelmann DA. Estimativas corrigidas de feminicídios no Brasil, 2009 a 2011. Rev Panam Salud Publica 2015;37(4-5):251-7.
9. Brasil. Ministério da Saúde. Secretaria de Vigilância em Saúde. Departamento de Vigilância de Doenças e Agravos Não Transmissíveis e Promoção da Saúde. Saúde Brasil 2014: uma análise da situação de saúde e das causas externas. Ministério da Saúde, 2015. 462 p.
10. Brasil. Ministério da Saúde. Secretaria de Vigilância em Saúde. Departamento de Vigilância de Doenças e Agravos não Transmissíveis e Promoção da Saúde. Saúde Brasil 2017: uma análise da situação de saúde e os desafios para o alcance dos objetivos de desenvolvimento sustentável [recurso eletrônico]. Ministério da Saúde, 2018. 426 p
11. Meneghel SN,Hirata VN.Feminicídios: Homicídios femininos no Brasil. Rev. Saúde Pública 2011; 45(3):564-74.
12. Meira KC, Costa MAR, Honório ACC, Simões TC, Camargo MP, Silva GWS. Tendência temporal da mortalidade por homicídio de mulheres em região brasileira. Rev Rene. 2019;20:e39864.
13. Meira KC, Jomar RT, Santos J, Silva GWS, Dantas ESO, Resende EB, Rodrigues WTS, Silva CMFP, Simões TC. Efeitos temporais das estimativas de mortalidade corrigidas de homicídios femininos na Região Nordeste do Brasil. Cad. Saúde Pública 2021; 37(2):e00238319.
14. Instituto de Pesquisa Econômica Aplicada; Fórum Brasileiro de Segurança Pública. Atlas da violência [Internet]. https://www.ipea.gov.br/atlasviolencia/ (19 fev. 2022)
15. Cerqueira D, Ferreira H, Bueno S. (Coord.) Atlas da violência 2021. São Paulo: Ipea; FBSP; IJSN; 2021.
16. Barros SC, Oliveira CM, Silva APS, Melo MFO, Pimentel DR, Bonfim CV. Spatial analysis of female intentional homicides. Rev Esc Enferm USP. 2021;55:e03770.
17. Matzopoulos RG, Thompson ML, Myers JE. Firearm and nonfirearm homicide in 5 South African cities: a retrospective population-based study. Am J Public Health. 2014;104(3):455–60.
18. Malta DC, Soares Filho AM, Pinto IV, Souza Minayo MC, Lima CM, Machado, ÍE et al. Association between firearms and mortality in Brazil, 1990 to 2017: a global burden of disease Brazil study. Population health metrics2020;18(Suppl 1):19.
19. Brasil, Ministério da Saúde. Banco de dados do Sistema Único de Saúde - DATASUS. Informações de Saúde, Sistema de Informações sobre Mortalidade (SIM). Disponível em http://www.datasus.gov.br/catalogo/sim.htm
20. Goes EF, Ferreira AJF,Meira KC, Myrrha LJD, Reis AP, Nunes VGA et al.Desigualdades raciais nas tendências da maternidade adolescente e no acesso ao pré-natal no Brasil, 2008-2019. Research, Society and Development 2023; 12 (1): e8312139404.
21. Queiroz BL, Freire FHMA, Gonzaga MR, Li - ma EEC. Completeness of death-count cover - age and adult mortality (45q15) for Brazilian states from 1980 to 2010. Rev Bras Epidemiol 2017; 20 Suppl 1:21-33.
22. Hill K, You DZ, Choi YJ. Death distribution methods for estimating adult mortality: sensitivity analysis with simulated data errors. Demogr Res 2009; 21: 235-5.
23. Morettin PA, Toloi CMC. Análise de séries temporais. 3ª ed. São Paulo: Blücher; 2018.
24. Antunes JLF, Cardoso MRA. Using time series analysis in epidemiological studies. Epidemiol Serv Saúde. 2015; 24(3):565-76. https://doi: http://dx.doi. org/10.5123/S1679-49742015000300024
25. Yang Y, Land KC. Age-period-cohort analysis. New models, methods, and empirical applica - tions. London: Chapman & Hall/CRC Press; 2013. (Interdisciplinary Statistics Series)
26. Borges LF, Souza ER, Ribeiro AP, Silva GWS, Silva CMFP, Santos J, Meira KC. Homicídios masculinos em duas regiões brasileiras: análise do efeito da idade, período e coorte. Cad. Saúde Pública 2019; 35(12):e00008719.
27. Naghavi M, Marczak LB, Kutz M, et al. Global mortality from firearms, 1990- 2016. JAMA. 2018;320(8):792–814.
28. Cisne M. Direitos humanos e violência contra as mulheres: uma luta contra a sociedade patriarcal-racista-capitalista. Serv Soc Rev. 2015;18(1):138-54.
29. Petrosky E, Blair JM, Betz CJ, Fowler KA, Jack SP, Lyons BH. Racial and Ethnic differences in homicides of adult women and the role of intimate partner violence - United States, 2003-2014. MMWR Morb Mortal Wkly Rep. 2017;66(28):741-6.
30. Ferreira IRS, Simões TC, Resende EB, Rodrigues WTS, Silva PEF, Santos J et al. Homicídios femininos no estado do Rio Grande do Norte e suas regiões de saúde no período de 2000 a 2016. Cad Saúde Coletiva 2021;29(spe):92-102.
31. Meneghel SN, Rosa BAR, Ceccon RF, Hirakata VN, Danielevictz IM. Feminicídios: estudo em capitais e municípios brasileiros de grande porte populacional. Cien Saúde Colet. 2017;22(9):2963-70.
32. Margarites AF, Meneghel SN, Ceccon RF. Feminicides in Porto Alegre: How many? Who are they?. Rev Bras Epidemiol. 2017;20(2):225-36.
33. United Nations Office on Drugs and Crime. Global study on homicide: gender-related killing of women and girls. Vienna: UNODC; 2019.
34. Portella AP. Como morre uma mulher? Editora UFPE,2019. https://editora.ufpe.br/books/catalog/book/49
35. Safiotti HI. Gênero, patriarcado, violência. Fundação Perseu Abramo,2004.
https://fpabramo.org.br/publicacoes/wpcontent/uploads/sites/5/2021/10/genero_web.pdf
36. Segato RL. Gênero e colonialidade: em busca de chaves de leitura e de um vocabulário estratégico descolonial. e-Cadernos CES 2012;18,106-131.
37. Meneghel SN, Ceccon RF, Hersler LZ, Marga - rites AF, Silva SR, Vasconcelos VD. Femicídios: narrativas de crimes de gênero. Interface (Botucatu) 2013; 17:523-33.
38. Faludi S. Backlash: o contra-ataque na guerra não declarada contra as mulheres. Rio de Janeiro: Rocco; 2001.
39. Aguirre, K. K. D. (2020). Masculinidades colonizadas e feminicídios na América Latina. Crítica Histórica 2020;11(22), 38-67.
40. Soares Filho AM, Duarte EC, Merchan-Hamann E. Tendência e distribuição da taxa de mortalidade por homicídios segundo porte populacional dos municípios do Brasil, 2000 e 2015 Ciência & Saúde Coletiva 2020; 25(3):1147-1156.
41. Garcia LP, Freitas LRS, Höfelmann DA. Avaliação do impacto da Lei Maria da Penha sobre a mortalidade de mulheres por agressões no Brasil, 2001-2011. Epidemiol Serv Saude. 2013;22(3):383-94.
42. Gattegno MV, Wilkins JD, Evans DP. The relationship between the Maria da Penha Law and intimate partner violence in two Brazilian states. Int J Equity Health. 2016;15(1):138.
43. Instituto Brasileiro de Geografia e Estatística (IBGE). (2019b). Perfil dos municípios brasileiros:2019. Rio de Janeiro: IBGE.
44. Instituto Brasileiro de Geografia e Estatística (IBGE). (2021). Estatísticas de gênero: indicadores sociais das mulheres no Brasil. 2. ed. Rio de Janeiro: IBGE.
45. Costa-Jr ESD. A Pandemia Frente à Constituição Fragilizada: Impactos da Emenda nº 95. Revista Direito Público 2020;17(96),9-36.
46. Barbosa JPM, Lima RCD, Santos GBM, Lanna SD,Andrade MAC Interseccionalidade e violência contra as mulheres em tempos de pandemia de covid-19: diálogos e possibilidades. Saúde & Sociedade.2021;30(2): e200367.
47. Haviland M, Rahbar AR, Rivara FP.Age, period and cohort effects in firearm homicide and suicide in the USA, 1983–2017. Injury prevention.2021; 27: 344-348, 2021.
48. Langmann C. Effect of firearms legislation on suicide and homicide in Canada from 1981 to 2016. West JC, editor. PLOS ONE. 2020 Jun 18;15(6):e0234457.
49. Dahlberg LL,Ikeda RM,Resnow MJ. Guns in the home and risk of a violent death in the home: findings from a national study. American Journal of Epidemiology2004; 160(10),929–936.
50. Cerqueira D, Mello JMP. Evaluating a national anti-firearm law and estimating the causal effect of guns on crime. Rio de Janeiro: PUC, 2013. http://twixar.me/JBpn>. (Texto para Discussão, n. 607).
51. Santaella-Tenorio J, Cerdá M, Villaveces A, Galea S. What Do We Know About the Association Between Firearm Legislation and Firearm-Related Injuries? Epidemiol Rev 2016;38(1):140-57.
52. Chapman S, Alpers P, Jones M. Association between gun law reforms and intentional firearm deaths in Australia, 1979-2013. J Am Med Assoc. 2016; 316(3):291–9.
53. Brasil. Presidência da República Secretaria-Geral Subchefia para Assuntos Jurídicos Decreto nº 11.366, de 1º de janeiro de 2023. https://www.planalto.gov.br/ccivil_03/_ato2023-2026/2023/decreto/D11366.html