METHODS: From 35 patients, 121 and 63 interproximal surfaces in, respectively, primary and permanent teeth were included. NILT images were obtained using DIAGNOcam™ (KaVo) and scored by two calibrated raters. A consensus diagnosis was reached for BW radiography; whereas, the ICDAS scores were obtained by one calibrated rater. Weighted Kappa (wκ) was used to evaluate inter- and intra-rater reliability of NILT and to evaluate the correlation between NILT, ICDAS and BW radiography.
RESULTS: The correlation between NILT and BW radiography was moderate to substantial for primary teeth [Rater 1: wκ = 0.61 (95% CI = 0.49-0.75), Rater 2: wκ = 0.55 (95% CI = 0.41-0.69)] and fair for permanent teeth [Rater 1: wκ = 0.34 (95% CI = 0.15-0.53), Rater 2: wκ = 0.33 (95% CI = 0.08-0.58)]. The correlation between ICDAS and BW radiography was moderate for primary teeth [wκ = 0.49 (95% CI = 0.35-0.63)] and substantial for permanent teeth [wκ = 0.62 (95% CI = 0.32-0.92)]. No significant differences were found between primary and permanent teeth.
CONCLUSIONS: NILT cannot be recommended as a single diagnostic tool for interproximal caries detection in primary teeth. The number of false negatives for dentine caries, especially in first primary molars, was too high. For the use in permanent teeth, NILT could be more accurate than BW radiography.