Ana Paula Pereira Reiniger1, Ananda Barrachini Londero1, Ticiane de Góes Mário Ferreira1, José Mariano da Rocha2, Carlos Heitor Cunha Moreira3, Karla Zanini Kantorski3. 1. Graduate Program in Oral Science, Periodontology Unit, School of Dentistry, Federal University of Santa Maria, Santa Maria, Rio Grande do Sul, Brazil. 2. Graduate Program in Oral Science, Periodontology Unit, School of Dentistry, Federal University of Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil. 3. Graduate Program in Oral Science, Department of Stomatology, School of Dentistry, Federal University of Santa Maria, Santa Maria, Rio Grande do Sul, Brazil.
Abstract
BACKGROUND: To evaluate the predictive performance of self-reported questions for periodontitis screening in a representative sample of a rural population. METHODS: Nine questions were compared with gold standard clinical examinations (probing six sites/tooth, full-mouth). Case definition for severe periodontitis was defined according to World Workshop (2017-WW) and Centers for Disease Control and Prevention/American Academy of Periodontology (CDC/AAP). Diagnostic tests such as sensitivity (SN), specificity (SP), positive and negative predictive values were performed for all questions alone and grouped into models. Binary logistic regression modeling was used to derive parameter estimates for all variables in a given model and the area under ROC curve was calculated. RESULTS: Clinical examinations showed a prevalence of periodontitis in the sample (n = 585) of 99.4% and 86.3%, being 40.3% and 33.8% of severe disease according to 2017-WW and CDC/AAP case definitions, respectively. Individually, only the questions regarding the self-perception of teeth/gum health and loose and lost teeth were valid to predict severe periodontitis. The best logistic regression models combined sociodemographic variables and risk-factors with the self-reported measures of self-perception of gum disease, teeth/gum health, loose teeth and history of tooth loss. CONCLUSION: Predictive performance of these self-reported questions presented herein support its potential use for surveillance of severe periodontitis in rural populations with high periodontitis prevalence.
BACKGROUND: To evaluate the predictive performance of self-reported questions for periodontitis screening in a representative sample of a rural population. METHODS: Nine questions were compared with gold standard clinical examinations (probing six sites/tooth, full-mouth). Case definition for severe periodontitis was defined according to World Workshop (2017-WW) and Centers for Disease Control and Prevention/American Academy of Periodontology (CDC/AAP). Diagnostic tests such as sensitivity (SN), specificity (SP), positive and negative predictive values were performed for all questions alone and grouped into models. Binary logistic regression modeling was used to derive parameter estimates for all variables in a given model and the area under ROC curve was calculated. RESULTS: Clinical examinations showed a prevalence of periodontitis in the sample (n = 585) of 99.4% and 86.3%, being 40.3% and 33.8% of severe disease according to 2017-WW and CDC/AAP case definitions, respectively. Individually, only the questions regarding the self-perception of teeth/gum health and loose and lost teeth were valid to predict severe periodontitis. The best logistic regression models combined sociodemographic variables and risk-factors with the self-reported measures of self-perception of gum disease, teeth/gum health, loose teeth and history of tooth loss. CONCLUSION: Predictive performance of these self-reported questions presented herein support its potential use for surveillance of severe periodontitis in rural populations with high periodontitis prevalence.