Initial in vitro evaluation of DIAGNOdent for detecting secondary carious lesions associated with resin composite restorations

Objective: To assess the potential of DIAGNOdent (DD) readings for detecting secondary enamel and dentin carious lesions associated with resin composite restorations in a set of preserved extracted adult teeth. Method and materials: Thirty test sites adjacent to resin composite restorations in 15 extracted teeth were evaluated visually and with DD readings for enamel and dentin caries. The results were compared to gold standard diagnosis determined by sectioning through each site and subsequent microscopic observation. Sensitivity, specificity, accuracy, and likelihood ratios at optimum threshold values for enamel and dentin caries were determined. Receiver operator characteristic (ROC) analysis was performed to compare visual and DD diagnostic methods. Results: Histologic incidence of enamel caries was 20% and dentin caries was 36.67%. All composite restorations in the sample teeth fluoresced (range 5 to 22). For enamel caries detection with DD, the optimum threshold of >/= 26 produced a sensitivity of .67 and a specificity of .79. For dentin caries detection with DD, the optimum threshold of >/= 22 produced a sensitivity of .73 and a specificity of .84. ROC analysis showed that the DD readings do contain diagnostic information. DD readings for dentin caries diagnosis produced better diagnostic results compared to visual readings for the 30 test sites in the study, but these results were not significantly different. Conclusion: DIAGNOdent readings may have potential for resin composite secondary enamel and dentin carious lesion detection. DIAGNOdent instrument design factors and technique should be optimized to improve performance prior to clinical trial for this application.