Clinical validation of near-infrared light transillumination for early proximal caries detection using a composite reference standard

Objectives The aim of this in vivo prospective clinical study was to validate a near-infrared light transillumination (NILT) device for the detection of proximal caries lesions and compare it to the established diagnostic method of bitewing radiography (BWR). Materials and methods A total of 116 interproximal posterior teeth/surfaces without gross cavitation were included and assessed with BWR and NILT. After clinical examination, the teeth with sound surfaces or non-cavitated caries lesions were separated with the use of an orthodontic separation rubber for 24h. Upon removal of the separator a silicon impression was obtained and the teeth surfaces were directly investigated. Intra- and inter-examiner reproducibility, sensitivities (SE) and specificities (SP) as well as the area under ROC curves (AUC) were calculated. Results The SE of NILT was 0.92 in D1 lesions and 1.00 in D2 and D3 lesions, respectively. BWR showed SE values of 0.81 in D1 lesions, 0.74 in D2 and 1.00 in D3 lesions, respectively. However, NILT showed lower values of SP (D1 0.38, D2 0.42 and D3 0.99) compared to BWR (D1 0.65, D2 0.88, D3 1.00). Intra-examiner reproducibility increased for both examiners from the initial examination (0.68 and 0.62) to the recall appointment (0.89 and 0.76) respectively. Conclusion The performance of NILT in detection of enamel caries was comparable to BWR. NILT could be considered as a useful and reliable tool in clinical practice. However, monitoring of enamel lesions might not be accurate, since a low specificity of NILT at D1,2 level and a tendency to false positive ratings have been shown. Therefore, it is recommended that treatment decisions should not be based on NILT alone. Clinical relevance This study might indicate that NILT can be used for routine patient screening but the use of BWR is still important for treatment decisions.