Validation of a self-report questionnaire for periodontitis in a Japanese population

We aimed to assess the validity of the self-report questionnaire for periodontitis in a Japanese population. A Japanese 9-item self-report questionnaire, developed by translating English-version questions that were used to detect periodontitis, was validated against full-mouth clinically-assessed periodontitis in 949 Japanese adults (average age = 43.2 years). Multivariable logistic regression modeling was used to calculate the area under the receiver operating characteristic curve (AUC), wherein the periodontitis case definition of the Centers for Disease Control and Prevention/American Academy of Periodontology was considered the gold standard. Severe, moderate, and mild periodontitis were identified in 6.2%, 30.0%, and 6.7% of the study population, respectively. Self-reported oral health questions combined with socio-demographic and health-related variables had an AUC > 0.70 (range, 0.71-0.87) for any periodontitis category. Four oral health questions ("have gum disease," "loose tooth," "lost bone," and "bleeding gums") were selected in the parsimonious model for severe periodontitis. The periodontitis screening score generated by the responses to these four questions had an AUC, sensitivity, and specificity of 0.82, 73.1%, and 74.3%, respectively, where the cut-off was set at 2 points. In conclusion, a locally adapted version of the self-report questionnaire had an acceptable diagnostic capacity for the detection of periodontitis in this study population.