Methods: A total of 94 sound permanent third molars had the occlusal enamel sectioned perpendicular to the long axis of the tooth to expose adjacent dentin. Flat dentin surfaces were obtained after grinding/polishing until all enamel had been removed. Samples were submitted to an in vitro cariogenic bacterial model to create caries-like lesions. The specimens were randomly divided into two groups (n = 47) according to the caries removal method: (A) bur excavation and (B) hand excavation. After dentinal caries removal, all specimens were assessed by two independent examiners, using different fluorescence-based methods: DIAGNOdent (LF), DIAGNOdent pen (LFpen), Vista Proof fluorescence camera (VP) and Quantitative Light-induced Fluorescence (QLF). The teeth were sectioned through the excavated cavities and the sections were analyzed under a stereomicroscope as the gold standard.
Results: Of 94 teeth, 59 had residual dentinal caries, being 29 from group A and 30 from group B. Inter-examiner reproducibility values (ICC) varied from 0.557 (VP) to 0.978 (LFpen). For groups A and B, QLF presented statistically higher sensitivity values (0.73 and 0.75, respectively). Specificity values for groups A/B were, respectively, 0.94/0.94 for LF, 1.00/1.00 for LFpen, 0.35/0.33 for QLF and 1.00/1.00 for VP. Accuracy values were statistically similar and ranged from 0.38 (VP) to 0.58 (QLF) and from 0.38 (VP) to 0.59 (QLF) for groups A and B, respectively. The excavation methods did not affect the performance of fluorescence-based methods.
Conclusion: QLF presented the greatest sensitivity and accuracy values in detecting residual dentinal caries of any of the fluorescence-based methods tested. Consequently, it could be an effective method in checking dentinal caries removal by bur or hand excavation.