Predictors of untreated dental decay among 15-34-year-old Australians

© 2009 The Authors. Journal Blackwell MunksgaardObjectives: To determine predictors of untreated dental decay among 15–34-year-olds in Australia. Methods: Data were from Australia's National Survey of Adult Oral Health, a representative survey that utilized a three-stage, stratified clustered sampling design. Models representing demographic, socioeconomic, dental service utilization and oral health perception variables were tested using multivariable logistic regression to produce odds ratios. Results: An estimated 25.8% (95% CI 22.4–29.5) of 15–34-year-old Australians had untreated dental decay. After controlling for other covariates, those who lived in a location other than a capital city had 2.0 times the odds of having untreated dental decay than their capital city-dwelling counterparts (95% CI 1.29–3.06). Similarly, those whose highest level of education was not a university degree had 2.1 times the odds of experiencing untreated dental decay (95% CI 1.35–3.31). Perceived need of extractions or restorations predicted untreated coronal decay, with 2.9 times the odds for those who perceived a treatment need over those with no such treatment need perception (95% CI 1.84–4.53). Participants who experienced dental fear had 2.2 times the odds of having untreated dental decay (95% CI 1.38–3.41), while those who reported experiencing toothache, orofacial pain or food avoidance in the last 12 months had 1.9 times the odds of having untreated dental decay than their counterparts with no such oral health-related quality-of-life impact (95% CI 1.20–2.92). The multivariate model achieved a 'useful' level of accuracy in predicting untreated decay (area under the ROC curve = 0.74; sensitivity = 0.63; specificity = 0.73). Conclusions: In the Australian young adult population, residential location, education level, perceived need for dental care, dental fear, toothache, orofacial pain or food avoidance together were predictors of untreated dental decay. The prediction model had acceptable specificity, indicating that it may be useful as part of a triage system for health departments wishing to screen by means of a questionnaire for apparently-dentally healthy 15–34-year-olds.Lisa M. Jamieson, Gloria C. Mejía, Gary D. Slade and Kaye F. Roberts-Thomso