Laypersons' sensitivity for melanoma identification is higher with dermoscopy images than clinical photographs

Abstract

Background: Most melanomas are first recognized by patients themselves or by their friends and family.

Objectives: To assess the ability of laypersons to identify melanomas using dermoscopy images.

Methods: This is an image-based study using laptop computers in the community. Seventeen laypersons were given a one-page educational brochure on the AC Rule for melanoma (asymmetry, colour variation). These laypersons and three expert dermoscopists completed two image sets, each containing a series of 100 pigmented skin lesions. Set 1 contained five melanomas, while set 2 contained 20 melanomas. Participants viewed a clinical image followed by a dermoscopy image for each lesion. For each image a score of 0-10 was assigned for asymmetry and colour, and then an overall assessment was made for suspicion of melanoma. Mean estimates have been calculated for sensitivity and specificity.

Results: Laypersons achieved a clinical sensitivity of 91·2% and a significantly higher dermoscopy sensitivity of 94·0%, P = 0·013. This improvement was not associated with a significant change in overall specificity, which for the clinical image was 64·2% and with dermoscopy was 62·0%, P = 0·97.

Conclusions: These results indicate that laypersons may be able to use dermoscopy to identify more melanomas than naked eye examination alone. Further study into the practice of dermoscopy by laypersons is warranted.