Utilização de banco de dados policiais como perspectiva para redução do sub-registro da violência contra mulheres

Abstract

Introduction: Violence against women is considered a public Health problem for bringing uncountable issues. To inscrease the problem’s visibility, was created in 2006 the Violence and Accident Survaillance System, wich has in its tools the Notifiable Disease System – SINAN. However, even though there were many efforts to improve the system, underreporting is a reality that demands strategies to solve the problem. Objective: Evaluate whether police data can help reduce the underreporting of violence against women in the SINAN analyze the socio-demographic profile of the women, as well as the profile of the aggressions and aggressors; and compare the agreement between these two data sources for common cases. Methods: cross-sectional study with data on interpersonal violence (physical and sexual) against women, occurred between 2015 and 2016 in Ribeirão das Neves (RN) - Minas Gerais (MG) - and reported in SINAN and/or identified at the Legal Medical Office (PML - Civil Police unit) and in police reports. Information was collected from these databases and a consolidated database (n=1,185) was created containing all cases, where descriptive analyses (frequency and central tendency measures) were performed. A paired database (n=56) containing common cases from both sources was also created to verify the agreement of records, using the Fleiss' Kappa test. Results: There was an under-registration of 83.54% in the SINAN data, in addition to incomplete variables. The police database showed about eight times greater capture than SINAN. For 56 cases present in the two databases, the information agreement was high / moderate for seven of the 11 characteristics evaluated. In the consolidated database, the victims were predominantly black, single or widowed, aged between 18 and 39 years old. Men, especially (ex) partners and family members, were the main aggressors. There was a predominance of physical violence, especially that involving blunt trauma. Conclusions: The underreporting of violence against women in SINAN can be reduced through cross-referencing with police data sources, both of which show good agreement for common data. Domestic physical violence against women remains the most prevalent form of violence.