Medidas auto reportadas para predição de periodontite em uma amostra rural de brasileiros

• Autor(a) principal: Reiniger, Ana Paula Pereira
• Data de Publicação: 2017
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Manancial - Repositório Digital da UFSM
• Texto Completo: http://repositorio.ufsm.br/handle/1/13755
• topic: Área rural; Curva ROC; Diagnóstico; Doença periodontal; Epidemiologia; Especificidade; Sensibilidade; Diagnosis; Epidemiology; Periodontal disease; ROC curve; Rural area; Sensitivity; Specificity; Survey; CNPQ::CIENCIAS DA SAUDE::ODONTOLOGIA

Self-reported measures of periodontal disease have shown promising validity in certain populations. There are no studies in the literature evaluating these measures in a representative sample of rural area, low income, low educational level and difficult access to dental care. Our objective was to evaluate the predictive performance of self-reported questions for periodontitis, performed in a representative sample of a rural population in the city of Rosário do Sul, located in southern Brazil. Nine questions were applied, eight of which were subdivided into three domains (self-perception of periodontal disease, self-perception of disease history, and periodontal disease diagnosed by a dentist) and compared with gold-standard (full mouth) clinical exams. Periodontal disease was classified according to two classifications: European and CDC / AAP. Diagnostic tests of sensitivity, specificity, positive predictive value, negative predictive value and ROC curve area were performed for all isolated questions and grouped into models. Binary logistic regression models were used to derive parameter estimates for all variables in given models. The sample consisted of 427 individuals aged 35 years or older who had at least five teeth. Individually, only the measure associated with the self-perception of "loose" teeth was valid to predict severe periodontitis. When self-reported questions were grouped into logistic regression models, the best models combined sociodemographic variables and risk factors with self-reported measures associated with self-perception of gum disease, "loose" teeth, and history of tooth loss. The combination of these variables reached acceptable statistical thresholds, that is, moderate area under the ROC curve (0.71-0.89), and sensitivity and specificity between 60% and 79%, representing moderate validity. The predictive performance of these self-reported questions showed its potential use for surveillance of severe periodontitis in a rural population with high prevalence of periodontitis, low socioeconomic status, and limited access to dental care.