IADR Abstract Archives
Modeling of Children’s Caries Risk in Medical Settings
Objectives: Develop and validate an easy to score caries risk tool for use in pediatric medical settings to target prevention and referral strategies.
Methods: 982 primary caregiver (PCG)-child pairs (of 1326 enrolled) were followed for 36 months to assess the caries predictive validity of a 52-item questionnaire. Children received caries examinations using the ICDAS criteria at 12±3 months (baseline), 30±3 months (80% retention) and 48±3 months of age (74% retention). Cavitated caries lesion (dmft; d=ICDAS >or=3) experience at 4 years-of-age was tested for associations with questionnaire items using generalized estimating equation models applied to logistic regression. Multivariable analysis used backwards model selection, with a limit of 10 items. The coefficients from the model were used to create scores that might be implemented more easily for predicting caries risk.
Results: At the 4 years-of-age exam, children were 49% female, 14% Hispanic/41% White/33% Black/2% Other/10% Multi-racial, 58% Medicaid-enrolled, 95% living in urban communities, and 24% had cavitated caries lesions. The prediction model using baseline responses (c-statistic=0.73) included the following significant (p<0.001) variables (odds ratio/scoring points): child participating in public assistance programs-Medicaid (1.74/1), being non-White (1.80-1.96/1), born premature (1.48/1), not born by C-section (1.28/1), and snacking on sugary snacks (3 or more/day 2.22/2; 1-2/day or weekly 1.55/1); and PCG cleaning the pacifier with juice/soda/honey or sweet drink (2.17/2), PCG daily sharing/tasting food with child using same spoon/fork/glass (1.32/1), PCG brushing their teeth less than daily (2.72/2), PCG’s gums bleeding daily when brushing or PCG having no teeth (1.83-2.00/1), and PCG having cavities/fillings/extractions in last 2 years (1.55/1). The resulting total score after completing the questionnaire is associated with the accuracy of the prediction, with scores ranging from 0 (100%Sensitivity/0%Specificity/29%PositivePredictiveValue) to 11 points (100%Specificity/0%Sensitivity/100%PositivePredictiveValue/71%NegativePredictiveValue).
Conclusions: A 10-item caries prediction model for use in medical settings shows good agreement with cavitated caries lesion development.
Methods: 982 primary caregiver (PCG)-child pairs (of 1326 enrolled) were followed for 36 months to assess the caries predictive validity of a 52-item questionnaire. Children received caries examinations using the ICDAS criteria at 12±3 months (baseline), 30±3 months (80% retention) and 48±3 months of age (74% retention). Cavitated caries lesion (dmft; d=ICDAS >or=3) experience at 4 years-of-age was tested for associations with questionnaire items using generalized estimating equation models applied to logistic regression. Multivariable analysis used backwards model selection, with a limit of 10 items. The coefficients from the model were used to create scores that might be implemented more easily for predicting caries risk.
Results: At the 4 years-of-age exam, children were 49% female, 14% Hispanic/41% White/33% Black/2% Other/10% Multi-racial, 58% Medicaid-enrolled, 95% living in urban communities, and 24% had cavitated caries lesions. The prediction model using baseline responses (c-statistic=0.73) included the following significant (p<0.001) variables (odds ratio/scoring points): child participating in public assistance programs-Medicaid (1.74/1), being non-White (1.80-1.96/1), born premature (1.48/1), not born by C-section (1.28/1), and snacking on sugary snacks (3 or more/day 2.22/2; 1-2/day or weekly 1.55/1); and PCG cleaning the pacifier with juice/soda/honey or sweet drink (2.17/2), PCG daily sharing/tasting food with child using same spoon/fork/glass (1.32/1), PCG brushing their teeth less than daily (2.72/2), PCG’s gums bleeding daily when brushing or PCG having no teeth (1.83-2.00/1), and PCG having cavities/fillings/extractions in last 2 years (1.55/1). The resulting total score after completing the questionnaire is associated with the accuracy of the prediction, with scores ranging from 0 (100%Sensitivity/0%Specificity/29%PositivePredictiveValue) to 11 points (100%Specificity/0%Sensitivity/100%PositivePredictiveValue/71%NegativePredictiveValue).
Conclusions: A 10-item caries prediction model for use in medical settings shows good agreement with cavitated caries lesion development.