Analysis of femicides in the state of Pernambuco: gender-based violence and invisibility

• Autor(a) principal: Pimentel, Dayane da Rocha
• Data de Publicação: 2024
• Outros Autores: Souza, Wayner Vieira de, Silva, Amanda Priscila de Santana Cabral
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Revista Baiana de Saúde Pública (Online)
• Texto Completo: https://rbsp.sesab.ba.gov.br/index.php/rbsp/article/view/3973
• topic: Violência de Gênero; Homicídio; Análise espacial; Estatísticas vitais; Sistemas de informação; Violencia de género; Homicidio; Análisis espacial; Estadísticas vitales; Sistemas de información; Gender-based violence; Homicide; Spatial analysis; Vital statistics; Information systems

Feminicide is a phenomenon resulting from the sexist patriarchal heritage and social structures historically permeated by unequal gender relations. The objective of this dissertation was to analyze the epidemiological aspects of feminicides registered in Pernambuco between 2016 and 2019. This is an ecological study. The records of homicides and undetermined causes of death of women in the Mortality Information System were the proxy for feminicide events and accepted as a reference. Search stages were established with probabilistic linkage in health databases, as well as manual reviews on the Pernambuco Court of Justice (Portuguese acronym: TJPE) platform and online news media in a complementary way. Data were described through the distribution of frequencies, measures of central tendency and dispersion. The local empirical Bayesian method and Moran’s autocorrelation index were used in the analysis of the spatial pattern of femicide rates. Hierarchical logistic regression was applied to assess the association between femicide and interpersonal violence. Logistic regression was used to estimate the odds ratio (OR) with 95% confidence intervals (95%CI). A total of 490 feminicides were found, corresponding to a rate of 2.5 per 100,000 women. The highest proportion of feminicides was identified by manual search in online news media and in the database of court documents of the TJPE (n=247; 50.41%). The victims’ characteristics reveal age between 20 and 39 years (n=286; 58.37%), black race/color (n=400; 81.63%), no partner (n=407; 83.06%) and schooling of over eight years (n=303; 61.84%). The spatial autocorrelation of the event was confirmed by the Global Moran Index of smoothed rates (I=0.3; p=0.001); the local autocorrelation reinforces the interiorization of the event by pointing to a critical area (Q1) formed by municipalities located in the macroregion of Vale do São Francisco and Araripe. At the distal determination level, the following variables increase the chances of feminicide: living in a small town (OR=2.10); the unavailability of police stations specialized in assisting women (OR=1.11); and absence of referrals to the assistance and protection network in the event of a previous aggression (OR=1.32). In relation to intermediate determinants, the greater the intensity of the means of action used in the practice of violence, the greater the chance of a fatal outcome, emphasizing the use of sharp objects (OR=3.93) and firearms (OR=11.14). Regarding proximal determinants, victims included in the age group 10-19 years were less likely to suffer a femicide (OR=0.51). The characterization of the study population can expand knowledge about structural violence against women, aspects of notification and the assistance and protection network. The spatial analysis identified the interiorization of the event and transition areas in Pernambuco, which requires prioritization of interventions. The modeling showed that women are exposed to multiple risk factors for the occurrence of feminicide and the discussion at all levels of determination in a particular way is even more necessary.