Background: Little is known about the occurrence of false negatives in the diagnosis of melanoma using dermoscopy in clinical practice: in the literature dermoscopy only increased the accuracy of diagnosis of equivocal lesions that were to undergo biopsy anyway.
Aim and methods: We collected the 81 lesions clinically diagnosed as probable melanomas by experienced specialists (from a series of 256 pigmented skin lesions submitted to excisional biopsy) and reviewed them for possible false negative results in the diagnosis of melanoma using in vivo dermoscopy and dermoscopy performed on slide images. Both procedures were carried out by the same experienced dermatologists who classified the lesions clinically.
Results: Dermoscopy made on slide images (observers blinded for clinical features of the lesions) yielded three false negative results (91% sensitivity) in a group of 35 histologically confirmed melanomas. In vivo dermoscopy showed better results, with no melanomas missed (100% sensitivity). The frequency of false positive results in the diagnosis of melanoma was 13.5% (11 of 81) with dermoscopy on slide images and 2.5% (two of 81) with in vivo dermoscopy.
Conclusions: Only in vivo dermoscopy by fully trained dermatologists with both clinical and dermoscopic experience avoids the risk of misclassification of melanomas otherwise correctly classified on clinical grounds.