The objective of this study was to measure validity and interexaminer reliability of current methods for diagnosis of approximal caries.
Eighty extracted premolars and molars without fillings or marginal defects were selected. Firstly, selected teeth were examined using direct visual inspection by one dentist, and ten models of mandibular or maxillary arch each containing 12 contacting approximal surfaces were constructed. 15 dentists were asked to examine for approximal caries under standard condition in a dental practice table unit. The methods employed were visual inspection (VI), visual inspection combined with probing (VI+P), visual inspection combined with fiber-optic transillumination (VI+FOTI), conventional bitewing radiography (BW) and visual inspection combined with conventional bitewing radiography (VI+BW). Histological sections of the teeth were examined to give the valid state of disease in each surface. The results were as follows;
1. In case of histological dentin caries, the mean sensitivity and specificity values were VI = 0.35 and 0.91, VI+P = 0.29 and 0.94, VI+FOTI = 0.34 and 0.98, BW = 0.40 and 0.98, and VI+BW = 0.48 and 0.99, respectively. The positive predictive values of VI+FOTI, BW and VI+BW were markedly higher than those of VI and VI+P. The negative predictive values of five methods were similar.
2. In case of histological dentin caries, interexaminer reliability of VI and VI+P were poor (0 (kappa (0.4), but that of VI+FOTI, BW, and VI+BW were good (0.4 ( kappa ( 0.75).
3. Sensitivity and specificity values of direct visual inspection were 0.75 and 0.98 for histological dentin caries, and the values for histological enamel caries were 0.95 and 0.94, respectively.
According to the results, the validity of current methods for the diagnosis of approximal caries is acceptable for specificity but unsatisfactory for sensitivity. Interexaminer reliability of clinical examination using diagnostic aids such as FOTI, BW was generally good. But results of unaided clinical
examination was very variable between examiners like many other previous studies. From this study, it is implied that the development of diagnostic aids with improved sensitivity is necessary and that furtherstudy is needed about how to reduce interexaminer variation of unaided clinical examination to appropriate level.
|