OBJECTIVES: Recently, a device for detecting occlusal caries lesions based on laser fluorescence measurements (DIAGNOdent) has been introduced. The reliability (reproducibility) and validity of this laser-fluorescence device were investigated. METHODS: In the in vitro part of the study, 49 permanent molars were measured by two observers using two DIAGNOdent devices. In addition, visual inspection was performed. The teeth were sectioned to measure the actual depth and area of the caries lesions. In the in vivo part of the study, 45 sites at the occlusal surfaces of permanent molar teeth in 13 patients were measured by two observers using two DIAGNOdent devices, manufactured in two successive years. RESULTS: The reliability of both DIAGNOdent devices expressed in the intraclass correlation coefficient (ICC=0.90 for observer 1, and 0.88 for observer 2) was high, and so was the interobserver reliability (ICC=0.85 for the 1998-device, ICC=0.90 for the 1999-device). The correlation between DIAGNOdent measurements and the actual depth of the caries lesions was lower than that of visual inspection by one observer, yet higher than that of a second observer. The specificities of visual inspection (0.94 and 0.88) were higher than those of the DIAGNOdent devices, whereas the sensitivities of the DIAGNOdent devices (0.93-1.00) exceeded those of visual inspection The validity of the DIAGNOdent, expressed as the area under the receiver operating characteristic curve, was not statistically significantly different from that of visual inspection (P>0.05). The DIAGNOdent measurements showed a higher association with the enamel part of the caries lesion than with the dentinal part. CONCLUSIONS: The DIAGNOdent is suitable for monitoring small caries lesions. Because of the high reproducibility of DIAGNOdent devices produced in successive years, a DIAGNOdent device may be replaced by a new one. Due to the higher specificity, visual inspection should be preferred to diagnose occlusal caries lesions in populations with a low caries prevalence.
OBJECTIVES: Recently, a device for detecting occlusal caries lesions based on laser fluorescence measurements (DIAGNOdent) has been introduced. The reliability (reproducibility) and validity of this laser-fluorescence device were investigated. METHODS: In the in vitro part of the study, 49 permanent molars were measured by two observers using two DIAGNOdent devices. In addition, visual inspection was performed. The teeth were sectioned to measure the actual depth and area of the caries lesions. In the in vivo part of the study, 45 sites at the occlusal surfaces of permanent molar teeth in 13 patients were measured by two observers using two DIAGNOdent devices, manufactured in two successive years. RESULTS: The reliability of both DIAGNOdent devices expressed in the intraclass correlation coefficient (ICC=0.90 for observer 1, and 0.88 for observer 2) was high, and so was the interobserver reliability (ICC=0.85 for the 1998-device, ICC=0.90 for the 1999-device). The correlation between DIAGNOdent measurements and the actual depth of the caries lesions was lower than that of visual inspection by one observer, yet higher than that of a second observer. The specificities of visual inspection (0.94 and 0.88) were higher than those of the DIAGNOdent devices, whereas the sensitivities of the DIAGNOdent devices (0.93-1.00) exceeded those of visual inspection The validity of the DIAGNOdent, expressed as the area under the receiver operating characteristic curve, was not statistically significantly different from that of visual inspection (P>0.05). The DIAGNOdent measurements showed a higher association with the enamel part of the caries lesion than with the dentinal part. CONCLUSIONS: The DIAGNOdent is suitable for monitoring small caries lesions. Because of the high reproducibility of DIAGNOdent devices produced in successive years, a DIAGNOdent device may be replaced by a new one. Due to the higher specificity, visual inspection should be preferred to diagnose occlusal caries lesions in populations with a low caries prevalence.