Reproduzierbarkeit, Sensitivität und Spezifität des "International Caries Detection and Assessment System (ICDAS-II)" zur Diagnose der Okklusalkaries -Eine in-vitro Studie

The International Caries Detection and Assessment System (ICDAS-II), introduced in 2005, is a visual caries detection system which detects and differentiates both initial and advanced carious dental lesions. Until now, only few studies investigated sensitivity, specificity and reproducibility of examiners with different clinical experience as well as the correlation of ICDAS-II with radiographical and histological findings. Aim: The aim of this study was to compare reproducibility, sensitivity and specificity of the ICDAS-II applied by an experienced and an inexperienced examiner. Furthermore, the correlations between visual findings versus analog and digital radiographic findings as well as versus histological results were assessed. Material and methods: 100 human extracted teeth (85 molars and 15 premolars) were available for this study. The teeth were cleaned, numbered and stored in water. Digital photographs were made of each occlusal surface and 1-4 examination points were chosen (n=181). Prior to the visual examination the experienced examiner trained the inexperienced examiner in the ICDAS-II system. Afterwards the teeth were examined visually by means of ICDAS-II by both investigators blind to each other. After three weeks, the visual examination was repeated in order to analyze the intraexaminer reproducibility. Moreover, analog and digital X-ray images were made of all teeth and separately investigated by the two examiners. Thereafter, the teeth were serially sectioned and the lesion depth was assessed microscopically by the Downer-criteria. Statistics: The data were analyzed with SPSS (14.0). To determine reproducibility, Cohen´s kappa (κ) was calculated. Spearman’s rank correlation coefficient (rs) was used to illustrate the correlation. Sensitivity and specificity were measured for D1 and D3 diagnostic threshold. Results: The unweighted and weighted interexaminer reproducibility of the visual examination by the means of ICDAS-II were 0.36 and 0.63, respectively. For the experienced examiner the unweighted and weighted intraexaminer kappa values were 0.67 and 0.83. The unweighted and weighted intraexaminer kappa values of the inexperienced examiner were 0.53 and 0.72. The assessed unweighted and weighted interexaminer kappa values were 0.36 and 0.43 for the analog radiography and 0.28 and 0.33 for the digital radiography. The Spearman rank correlation coefficient for the correlations between the visual findings and the analog and digital radiography of both the experienced (rs = 0.59 resp. rs = 0.56) and inexperienced examiner (rs = 0.53 resp. rs = 0.42) were moderate. The correlation of the visual and histological findings was rs = 0.54 for the experienced and rs= 0.50 for the inexperienced examiner. Sensitivity and specificity of the visual examination (ICDAS-II) for the experienced examiner were 92.2% resp. 38.8% at the D1 diagnostic threshold and 78.7% resp. 81.3% at the D3 diagnostic threshold. For the inexperienced examiner the values for sensitivity and specificity were 98.4% resp. 30.0% at the D1 diagnostic threshold and 62.1% resp. 79,6% at the D3 diagnostic threshold. Conclusion: This comparative study shows that even an inexperienced examiner can achieve good sensitivity, specificity and reproducibility values after a short introduction. Therefore, ICDAS-II can be used in daily practice and in education. The differences in the correlation between the visual findings and both radiographic and histologic results might depend on the clinical experience of the examiner.