Can Alteration of Cariogram Risk Categories Improve Caries Prediction in Preschool Children?

Andreas Agouropoulos1, Eleftheria Birpou1, Svante Twetman2, Katerina Kavvadia1
1 Department of Paediatric Dentistry, Dental School, National and Kapodistrian University of Athens, Greece
2 Department of Odontology, Faculty of Health and Medical Sciences University of Copenhagen, Denmark, Denmark

Background: Cariogram is known to have limited performance in predicting caries development in preschool children. This study aimed to explore the possibility of improving the ability of Cariogram to predict caries increment in preschool children, when different cut off points for risk categories were used for Cariograms with and without saliva tests.

Methods: Two to 5-year-old children (N=175) from areas with high caries prevalence in Athens, Greece were clinically examined at baseline and after two years. Mutans streptococci counts (MS) and saliva buffer capacity (SBC) were evaluated with chair-side tests and parents completed a questionnaire on dietary habits and oral health habits and attitudes. Caries increment was expressed as the sum of all new cavitated and non-cavitated caries lesions over the study period. Full Cariogram and reduced Cariogram versions were calculated by extracting the MS, SBC, or both. Statistical methods was used to assess optimal cut off points for low, medium, high risk categories for the different Cariogram versions and the predictive ability was evaluated using Kendall’s tau coefficients and regression models. Statistical significance was set at 5%.

Results: The cut off points for low, medium and high risk categories were: Full Cariogram (C1):0-60, 61-90, 91-100, Cariogram without SBC (C2): 0-80, 80-92, 93-100, Cariogram without MS (C3): 0-78, 79-93, 94-100, Cariogram without MS and SBC (C4): 0-80, 81-91, 92-100. Associations with caries increment (Kendall’s tau) were statistically significant but low; C1: 0.317 (p<0.001), C2: 0.308 (p<0.001), C3: 0.328 (p<0.001) and C4 0.294 (p=0.001). Cariogram 4 showed the best fit in regression models (AIC=180.38, BIC 189.06).

Conclusions: Alteration of the cut off points for the Cariogram risk categories did not improve the ability of the software to predict 2-years caries increment in this preschool children group.

Andreas Agouropoulos