IADR Abstract Archives
In-vivo Performance of Different Methods for Occlusal Caries Detection
Objectives: The purpose of this in vivo study was to evaluate the performance of caries detection techniques based on alternating current impedance spectroscopy, laser fluorescence and bitewings radiographs (ICDAS Radiographic scoring system) and compare it to ICDAS-visual scores.
Methods: In 60 patients, one occlusal surface was examined by two examiners using ICDAS-visual (ICDAS), CarieScan PRO(ACIS), DIAGNOdent Pen(LF Pen) and bitewing radiographs (BW). One-four weeks later repeated measurements were performed. Intra- and inter-examiner reproducibility was assessed with intra-class correlation coefficients (ICC) for numerical results and kappa values for categorical findings. Spearman correlation coefficients between ICDAS and the remaining methods were calculated. Sensitivity (SEN), specificity (SPE), positive predicted values (PPV), negative predicted values (NPV) and diagnostic odds-ratio (DOR) were found with ICDAS (dichotomized) as reference. McNemar’s test was used to evaluate the outcome of ACIS, LF and BW compared to ICDAS.
Results: ICDAS dichotomized, sound/initial lesions constituted (n=33) 55% versus moderate/extensive lesions constituted (n=27) 45% of the lesion sample. The inter- and intra-examiner reproducibility ranged between ICC 0.65-0.87 for ACIS and ICC 0.86-0.93 for LF Pen. For ICDAS the weighted kappa values were 0.81-0.91 and for BW 0.90-0.92. Spearman correlation coefficients were 0.65, 0.60 and 0.71 for ICDAS- vs ACIS, LF Pen and BW, respectively. ACIS demonstrated high specificity but low sensitivity, whereas moderate sensitivity and high specificity was found for LF Pen. DOR was highest for BW. McNemar’s test showed significant differences between caries detection with ICDAS/ACIS (p<0.05) and ICDAS/bitewings (p<0.05), but not between ICDAS/LF Pen (p=0.25).
Conclusions: Reproducibility was high for LF pen, BW and ICDAS and moderate to high for ACIS. BW and ACIS differed significantly from ICDAS in their abilities to detect occlusal caries and no significant difference was found between ICDAS and LF.
Methods: In 60 patients, one occlusal surface was examined by two examiners using ICDAS-visual (ICDAS), CarieScan PRO(ACIS), DIAGNOdent Pen(LF Pen) and bitewing radiographs (BW). One-four weeks later repeated measurements were performed. Intra- and inter-examiner reproducibility was assessed with intra-class correlation coefficients (ICC) for numerical results and kappa values for categorical findings. Spearman correlation coefficients between ICDAS and the remaining methods were calculated. Sensitivity (SEN), specificity (SPE), positive predicted values (PPV), negative predicted values (NPV) and diagnostic odds-ratio (DOR) were found with ICDAS (dichotomized) as reference. McNemar’s test was used to evaluate the outcome of ACIS, LF and BW compared to ICDAS.
Results: ICDAS dichotomized, sound/initial lesions constituted (n=33) 55% versus moderate/extensive lesions constituted (n=27) 45% of the lesion sample. The inter- and intra-examiner reproducibility ranged between ICC 0.65-0.87 for ACIS and ICC 0.86-0.93 for LF Pen. For ICDAS the weighted kappa values were 0.81-0.91 and for BW 0.90-0.92. Spearman correlation coefficients were 0.65, 0.60 and 0.71 for ICDAS- vs ACIS, LF Pen and BW, respectively. ACIS demonstrated high specificity but low sensitivity, whereas moderate sensitivity and high specificity was found for LF Pen. DOR was highest for BW. McNemar’s test showed significant differences between caries detection with ICDAS/ACIS (p<0.05) and ICDAS/bitewings (p<0.05), but not between ICDAS/LF Pen (p=0.25).
Conclusions: Reproducibility was high for LF pen, BW and ICDAS and moderate to high for ACIS. BW and ACIS differed significantly from ICDAS in their abilities to detect occlusal caries and no significant difference was found between ICDAS and LF.