Cariogram-Based Dental Caries Risk Assessment: A Cross-Sectional Statistical Validation of Sector Values Against Caries Severity, Plaque Index, and ICDAS Score

Abstract

Objectives: This study aimed to (1) compare Cariogram sector values between low-risk and very-high-risk individuals, (2) assess the association of plaque index and ICDAS score with caries severity, (3) evaluate the diagnostic accuracy of individual Cariogram sectors in distinguishing carious from non-carious subjects, and (4) identify predictors of caries severity through ordinal logistic regression.
Materials and Methods: Across-sectional sample of 512 subjects was analyzed — 256 with active caries (stratified as Mild, n=96; Moderate, n=60; Severe, n=100) and 256 caries-free controls. Non-parametric tests (Mann-Whitney U; Kruskal-Wallis with D wass-Steel-Critchlow-Fligner post hoccomparisons) were used given non-normal data distributions. Chi-square tests examined associations between categorical clinical variables and severity. Receiver Operating Characteristic (ROC) analysis with Youden's Index determined optimal cut-off values, and ordinal logistic regression identified independent predictors of severity.
Results: All five Cariogram sectors differed significantly between low-risk and very-high-risk groups (all p < .001). Each sector achieved complete separation between the caries and non-caries groups in ROC analysis (AUC = 1.00). Ordinal logistic regression showed that higher darkbluesectorvalues(OR=0.325,95%CI:0.256–0.400)andredsectorvalues(OR=0.117, 95%CI:0.065–0.189)were independently associated with reduced odds of severe caries(both p<.001), while the blue sector approached but did not reach significance(OR=1.123,p=.062). The model explained 76.6% of variance in severity (McFadden's R² = 0.766). Plaque Index and ICDAS score were both significantly associated with caries severity (χ² = 28.647 and 8.927, respectively; p < .05).
Conclusions: Cariogram sector values, particularly the dark blue (dietary) and red (bacterial susceptibility) sectors, were strongly and independently associated with caries severity in this sample, supporting the Cariogram's value as a risk-stratification instrument. Because the caries and non-caries groups in this analysis represented extremes of risk rather than a continuous clinical population, the perfect diagnostic separation observed should be interpreted as internal validation rather than a general is able estimate of real-world diagnostic accuracy.