Purpose: To evaluate if high-powered magnification could improve the performance of visual inspection and compare with other methods to detect occlusal caries lesions in primary teeth.
Methods: 110 suspected sites were analyzed by two examiners with four methods: visual inspection, visual inspection aided with high-powered magnification (x20), laser fluorescence (DIAGNOdent) and radiographs. The first examiner performed the examinations in a second session to assess the intra-examiner reproducibility. Then, the teeth were cut and the sections were evaluated in stereomicroscope. Sensitivity, specificity and accuracy were calculated at enamel and dentin caries thresholds, and compared using McNemar change test. The ROC analysis was performed and the mean of areas under ROC curves were compared using ANOVA and Student-Newman-Keuls test. Inter- and intra-examiner reproducibility was assessed by calculating Cohen's Kappa.
Results: The visual inspection with magnification did not change the accuracy and reproducibility of the visual inspection alone. The best caries detection method was laser fluorescence, and the least accurate was with radiographs.