Clinical performance of clinical-visual examination, digital bitewing radiography, laser fluorescence, and near-infrared light transillumination for detection of non-cavitated proximal enamel and dentin caries

The aim of this study was to compare the clinical performance of clinical-visual examination using the International Caries Detection and Assessment System (ICDAS) II, digital bitewing radiography, near-infrared light transillumination (NIR-LT), and laser fluorescence (LF) for the detection of non-cavitated proximal enamel and dentin caries. The study included 335 patients, aged 12–18 years, with no cavities in the posterior teeth. Clinical-visual inspections of 335 non-cavitated proximal caries were performed by two examiners. For enamel caries, clinical validation included a combination of clinical-visual and digital bitewing radiography assessments. For dentin caries, the clinical validation was opening the cavity. The accuracy rate, sensitivity, specificity, predictive values, and areas under receiver operating characteristic curves were determined. The agreement between the examiners’ measurements was calculated using the kappa coefficient. The sensitivity, specificity, and accuracy of the methods were compared using the McNemar test. The significance level was set at p < 0.05. Digital bitewing radiography had the highest sensitivity (0.96) and accuracy (0.96), and LF had the lowest sensitivity (0.38) and accuracy (0.39). After separation of the lesions into enamel and dentin caries, clinical-visual examination had the highest sensitivity (0.98) and accuracy (0.98) for enamel caries, while digital bitewing radiography had the highest sensitivity (0.97) and accuracy (0.97) for dentin caries. The NIR-LT method had a higher sensitivity for enamel caries (0.86). Each method also differed significantly from the others (p < 0.001). Digital bitewing radiography gave the best prediction of proximal enamel and dentin caries. NIR-LT showed good potential for detection of proximal caries.