Validation of a Risk Assessment Form

Objectives: This IRB-approved retrospective study evaluated the effectiveness of the IUSD caries risk management (CRM) tool to predict future caries risk.  Methods:   A convenience sample of 303 children (5-13 yrs old) was examined at baseline, 12- and 48-months using the International Caries Detection and Assessment System (ICDAS). Caregivers (79) completed a 25-item questionnaire including socio-demographic, dietary, protective factors, disease experience, and access-to-care at baseline. The CRM was completed using ICDAS scores, including lesion activity at baseline and questionnaire data. A flow chart was used as a guide to assess the risk for caries as low, moderate or high independently by a senior and a junior examiner.  One-way analysis of variance (ANOVA) was used to analyze the separation of subjects into three risk levels and the significance of specific risk factors.  Results: Inter-evaluator agreement of risk level was moderate (weighted kappa = 0.49). Mean incidence of caries (ICDAS ≥3) at 12- and 48-months was 4.4 and 9.3, respectively.  Percentage of subjects according to significant factors: cavitated lesions and lesions radiographically into dentin 54%; lesions radiographically in inner enamel 14%; in outer enamel 17%, no radiographically evident lesions 14%; active incipient lesions 75%. Separation into high, moderate, and low caries risk was significant at all thresholds (ICDAS ≥1, ≥3, ≥5) and years (p<0.05).  At 12-months, cavitated lesions, severity of lesions, and clinically active incipient lesions (ICDAS 1 and 2) were significant predictors of caries incidence (p<0.04) at thresholds of ICDAS 1 and 3.  At 48-months the severity of radiographically evident lesions was significant at ICDAS ≥1 and ≥3 (p<0.01), with cavitated lesions and clinically active incipient lesions significant at the ICDAS ≥3 (p≤0.03).  Conclusion: In this population the CRM was effective in discriminating the risk levels, even when using incipient lesions. Past caries experience was a significant factor at all ICDAS thresholds.