Modelo preditor de risco para violência doméstica contra a mulher

• Ano de defesa: 2020
• Autor(a) principal: Deininger, Layza de Souza Chaves
• Tipo de documento: Tese
• Tipo de acesso: Acesso embargado
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Ciências Exatas e da Natureza; Programa de Pós-Graduação em Modelos de Decisão e Saúde; UFPB
• Palavras-chave em Português: Saúde pública; Violência contra a mulher; Tomada de decisão; Public health; Violence against women; Decision making; CNPQ::CIENCIAS DA SAUDE::SAUDE COLETIVA
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/123456789/20226

Objective: To validate a risk predictor model to support specific care decision-making for women in situations of domestic violence. Method: This is a population-based applied methodological research. The research consisted of two distinct phases: the elaboration of a predictor model based on the Neural Network model having as input variables, the data extracted from the database of the study by Lucena (2015). The database contains variables related to the sociodemographic, epidemiological profile and the quality of life of women over 18 years of age, in addition to data on the measurement of the types of violence perpetrated by the intimate partner. For the operationalization of the second phase, mobile devices (tablet / cell phones) were used, where, through the sampling plan, 58 primary care professionals from the city of João Pessoa-PB were selected and instructed to participate in a pilot project, applying the predictor model in women seen at the service. The study met all the requirements of the ethics and research committee of the Paraíba State Health Department. Results: using the questionnaire that infers quality of life (WHOQOL-BREF) and the questionnaire that evaluates violence against women (WHO VAW STUDY), the significant variables were obtained through the neural network model and multiple logistic regression. Thus, the likelihood of a woman experiencing domestic violence was calculated. This numerical expression was transcribed to the VCMulher application software. The application was created to be used by primary care professionals, who are closer to women, in order to predict and identify victims of domestic violence. The VCMulher application achieved 83% approval by professionals. Regarding the risk of suffering violence, the data collected pointed out that of the 165 women participants, 98.8% have a medium to high risk of suffering violence. About 19% had a more than 90% chance of suffering domestic violence. Conclusion: The construction of the risk predictor model met the objective proposed in the study, presenting itself as a powerful instrument to identify risks / cases of domestic violence against women in the scope of primary care. It is necessary to invest in the qualification of health professionals to provide comprehensive care to women victims of violence.