Validation of caries risk assessment programs in Hong Kong children

Abstract

Identifying caries-prone children through caries risk assessment (CRA) is important for cost-effective prevention and evidence-based treatment planning. OBJECTIVE: to evaluate/compare validity of various CRA programs (NUS-CRA, Cariogram, CAT, and CAMBRA) in predicting caries in Hong Kong preschoolers. METHODS: With parental consent, 544 children aged 3-4 years were recruited from four kindergartens located in different districts. Parents completed a self-administered questionnaire on their socio-demographic background, dental knowledge/attitude and child's oral health habits. Children received oral examinations and biological tests on the saliva buffering capacity and levels of mutants Streptococci (MS) and Lactobacilli (LB). Children's caries risk was predicted through the abovementioned CRA programs. After 12 months, caries increment among 485 (89.2%) children was followed and compared with the predicted risks generated by the CRA programs. The validity of CRA programs was evaluated through ROC analysis. RESULTS: Within 12 months, 178 (36.7%) children developed new caries. The mean (SD) increment in dmft was 0.78 (1.36). The sensitivity/specificity of the NUS-CRA screening model (without biological tests) was 72%/82%; the full-blown model (with MS and LB tests) reached a sensitivity/specificity of 85%/81%. The sensitivity/specificity of Cariogram was 76%/64%. Both CAMBRA and CAT reached a high sensitivity (100% for CAT; 92.7% for CAMBRA); however, their specificity was low (3% for CAT; 46% for CAMBRA). Validity of the NUS-CRA models was higher than those of other programs (p<0.05). CONCLUSION: NUS-CRA program outperformed other programs and appeared to be a more suitable CRA tool for predicting early childhood caries in Hong Kong preschoolers