Reliability of Diagnostic Modalities for Dental Caries

Objectives: The aim of this study was to compare three clinical diagnostic modalities; Visual-
Tactile, Radiographic and laser fluorescence (DIAGNOdent) device; to detect dental caries, by skilled clinicians.

Methods: This in-vitro study utilized 30 extracted premolar and molar teeth. The teeth were washed with water using a tooth brush to remove debris and mounted in wax for evaluation. Nine clinicians examined each tooth, corresponding radiographic films and separately utilized laser fluorescence device to decide if the tooth was carious. All the evaluators followed the same dental caries assessment criteria for visual-tactile,radiographic and laser fluorescence device manufacturer’s caries detection criteria. For each tooth thus diagnosed with caries the evaluators then decided if they required receiving restoration. The data received was analyzed to yield Cohen’s Kappa and interclass correlation coefficients (ICC).
Results: Across evaluators 76% of teeth were considered decayed based on visual and tactile examination. For ratings based on such exploration, kappa ranged from 0.20 to 0.87(Med= 0.49) and the ICC was 0.54. Amid evaluators, 40.6% of teeth were considered decayed upon examination of the corresponding X-ray films. While the data based on the results showed that kappa ranged from 0- 0.81 (Median= 0.24) and ICC was 0.24. Finally based on laser fluorescence device readings 91.1% were considered caries and hence the kappa measured between 0.14 -0.78(Median= 0.52) and ICC was 0.49.
Conclusions: In the present study laser fluorescence device detected more carious lesions than the visual-tactile and radiographic methods. While clinical and laser fluorescence device assessments yielded moderate ICCs of about 0.5, the X-ray data were much less reliable. All three modalities of caries detection are not singularly reliable and reproducible by the nine examiners.