Validity of self-reported oral conditions among Brazilian older women: Do socio-economic factors matter?

Abstract

Objective: To validate the level of agreement between self-report and clinical examination for oral conditions and evaluate the effect of sociodemographic conditions on the validity of self-report among women aged 60 and older. Methods: A cross-sectional study was conducted in a social community center for seniors in Southern Brazil. Sociodemographic data (age, level of education, and income) were measured. Participants were interviewed and clinically examined for the number of teeth (DMF-T index) and the use of dental prostheses. The self-reported number of teeth in each arch and the use of dental prostheses were gathered through interviews. The level of agreement was estimated using the observed agreement, Kappa statistics, sensitivity/specificity (edentulism/prostheses) and Lin's concordance correlation coefficient, and related tests (number of teeth). The validity of the oral conditions was estimated according to sociodemographic information. Results: Ninety-nine women participated in the study. High levels of agreement were observed for edentulism (97.8%; 95%CI 92.8;99.7; Kappa 0.947) and the use of dental prostheses (97.0%; 95%CI 91.3;99.4; Kappa 0.922). In both conditions, despite achieving similar concordance correlation coefficients (ranging from weak to moderate), the mean number of upper teeth was lower in clinical examination (7.1 ± 5.2) compared with self-reported (8.6 ± 3.6), while the opposite was observed for lower teeth (clinical examination: 9.1 ± 3.4; self-reported: 6.6 ± 5.3). Larger differences were found among women of low income and educational levels. Conclusions: Our findings suggest that the participants' socio-economic position might influence their self-reported number of teeth.