Detection of approximal caries by clinical and radiographic examination in 9-year-old Swedish children.

The aims were to determine caries prevalence in 9-year-olds belonging to a low-caries prevalence population and to assess how accurately some commonly used risk factors/risk markers (predictors) can identify additional approximal caries as judged from bitewing examination (BW). One calibrated dentist examined 117 consecutive 9-year-olds from the inner city of Stockholm. The predictors were the number of clinically detected dfs and DFS, visible plaque, salivary mutans streptococci, tooth brushing habits, consumption of sugary products, parents' education, and (before BW) an overall clinical judgement by the examiner. Analyses of the data included sensitivity and specificity and a multiple logistic regression model. When BW was included, the mean DFS was 0.27 and the mean dfs 1.74. For approximal enamel and dentin lesions, the average gain from BW was 0.20 lesions for permanent first molars and 1.23 lesions for primary molars. Altogether, 48% of the children benefited from BW. The best predictors of additional approximal caries from BW were presence of caries (cavitation/dentin level) on occlusal surfaces of permanent first molars and the dentist's judgement based on clinical findings, both with a combined sensitivity and specificity of 134%. The only statistically significant variables in a logistic regression model were the dentist's judgement and the child's statement about regular tooth brushing. It is concluded that a sizeable proportion of 9-year-olds, representing a low-caries prevalence population, benefited from BW. However, the ability to identify these children from the predictors was limited.