Validation of self-reported measures of periodontitis in a Spanish Population

Background and Objective Use of self-reported questionnaires in Dentistry may be useful to estimate the prevalence of periodontitis in epidemiological studies. This study aims to assess the accuracy of self-reporting for predicting the prevalence of periodontitis in a Spanish population participating in a diabetes incidence study. Materials and Methods Data were collected from 231 patients participating in the Di@bet.es study. Eight questions about periodontal health were included in a health patient-reported questionnaire. The outcomes from self-reporting were validated against a full-mouth periodontal examination. Multivariable logistic regression predictive modeling was used to determine the sensitivity, specificity, and area under the receiver operator characteristic curve (AUROCC). Results Self-reported gum health, loose teeth, tooth appearance, and use of dental floss were associated with different definitions of severe periodontitis. Correlations between responses to the questions were weak. The question Do you think you might have gum disease? combined with demographic and well-established risk factors resulted in an AUC value of 0.75, sensitivity of 75.2%, and specificity of 60.6% for severe periodontitis. The answer to 4 questions combined with age, educational level, smoking status, and tooth loss was 76.4% sensitive and 63.5% specific, with an AUC of 0.75 in predicting prevalence of >= 25% of teeth with probing pocket depth (PPD) >= 6 mm. Conclusion Predictive models, combining self-reporting on oral health status with demographic and risk factors, were useful for estimating the prevalence of severe periodontitis in the Spanish population.