This in vivo study aimed to evaluate the performance of methods of approximal caries detection in primary molars and to assess the influence of the discomfort caused by these methods on their performance. Two examiners evaluated 76 children (4-12 years old) using visual inspection (ICDAS), radiography and a laser fluorescence device (DIAGNOdent pen, LFpen). The reference standard was visual inspection after temporary separation with orthodontic rubbers. Surfaces were classified as sound, noncavitated (NC) or cavitated (Cav), and performance was assessed at both NC and Cav thresholds. Wong-Baker faces scale was employed to assess the discomfort. Multilevel analysis was performed to verify the influence of discomfort on performance, considering the number of false-positives and false-negatives as outcome. At NC threshold, visual inspection achieved better performance (sensitivities and accuracies around 0.67) than other methods (sensitivities around 0.25 and accuracies around 0.35). At Cav threshold, visual inspection presented lower sensitivity (0.23 and 0.19), and LFpen (0.52 and 0.42) and radiography (0.52) presented similar sensitivities. Concerning the influence of the discomfort, at NC threshold, when discomfort was present, the number of false-negative results was lower with LFpen and the number of false-positive results was higher with visual inspection. At Cav threshold, the number of false-positive results was higher with LFpen. In conclusion, radiography and LFpen achieved similar performance in detecting approximal caries lesions in primary teeth and the discomfort caused by visual inspection and LFpen can influence the performance of these methods, since a higher number of false-positive or false-negative results occurred in children who reported discomfort.
Copyright © 2010 S. Karger AG, Basel.