Validity of a self-reported questionnaire for periodontitis in Spanish population

Abstract

Background: clinical examination is the gold-standard approach for surveillance of periodontitis; however, it requires large resources. Several self-reported measures have been developed and tested in diverse scenarios with results suggesting that it may be a useful tool for screening periodontal disease in different populations; yet they have not being tested in Spanish population. We aimed to assess the validity of a self-reported questionnaire for periodontitis in a Spanish population from Barcelona during 2018. Methods: one hundred and twelve participants were enrolled in the study and, in one appointment; a dentist performed the self-reported questionnaire and a full-mouth periodontal examination. Periodontitis was defined as at least mild periodontitis according to three criterion of classification. Receiver-operating characteristics curve analyses were used to test the discriminatory capability, sensitivity and specificity of the self-reported questionnaire and logistic regression models were adjusted to estimate the minimal-set of questions associated with periodontitis. Results: the self-reported questionnaire had a useful discriminative capability for detecting individuals with periodontitis (area under the curve [AUC] = 0.85 95% CI 0.78-0.92) and its moderate/severe form (AUC = 0.86 95% CI 0.79-0.04) with sensitivity and specificity of 77% and 74% and 73% and 87%, respectively; representing moderate validity. The combination of four-specific questions had high accuracy (AUC = 0.88 95% CI 0.81-0.94) and validity (sensitivity = 92.2%) and was strongly associated with moderate/severe periodontitis after adjusting by socio-demographic factors. Conclusions: the use of several self-reported questions proved to have a good performance for screening periodontitis in the population under study; specifically, those related with tooth mobility and gum migration. Large community-based studies are needed to test its validity and predictive capability.