Assessment of Self-report Measures for Predicting Prevalence of Periodontitis

Objectives: The Centers for Disease Control and Prevention (CDC) in collaboration with the American Academy of Periodontology (AAP) has been investigating the potential use of self-report measures for predicting prevalence of periodontitis in the U.S population. Findings from a number of analyses, including data from an Australian National Survey, have identified 6 questions as possible predictors for periodontitis. This study attempts to validate these self-report measures against clinically assessed periodontitis in a convenience sample of U.S adults.

Methods: A sample of 456 adults, aged 35 years and older was recruited among non-Hispanic whites and blacks, and Hispanics. Participants were interviewed (by phone and in person) using six oral health questions along with demographic questions. Clinical periodontal disease status was assessed from a full mouth periodontal examination for measurements of pocket depth and attachment loss at six sites per tooth for all teeth (except the third molars) with cases defined by the CDC-AAP definition for periodontitis. Multivariable modeling was used to assess the collective predictiveness of these oral health questions in predicting prevalence of severe and total (severe and moderate combined) periodontitis.

Results: Clinical examinations identified 102 (22.4%) cases of periodontitis consisting of 80 (17.5%) moderate periodontitis and 22 (4.8%) severe periodontitis cases. Response rates to all questions were high for all three race/ethnic. Combined measures from questions about scaling and root planing, loose teeth and use of mouth wash were most useful in predicting prevalence of total periodontitis, while questions on gum disease, loose teeth and tooth appearance were most useful for severe periodontitis reaching a combined sensitivity and specificity of 162 and AUC of 9.0.

Conclusion: Results from this study further suggest that measures obtained using self-report oral health questions, in combination with demographic information, are promising for predicting the prevalence of periodontitis in the adult U.S. population.