Objective: To assess, by means of meta-analysis techniques for diagnostic tests, the accuracy of dermoscopic (also known as dermatoscopy and epiluminescence microscopy) diagnosis of melanoma performed by experienced observers vs. naked-eye clinical examination.
Data sources: MEDLINE, EMBASE, PASCAL-BIOMED, and BIUM databases were screened through May 31, 2000, without any language restrictions.
Study selection: Original studies were selected when the following criteria were met: spectrum of lesions well described, histologic findings as standard criterion, and calculated or calculable sensitivity and specificity. Eight of 672 retrieved references were retained.
Data extraction: Three investigators extracted data. In case of disagreement, consensus was obtained. Summary receiver operating characteristic curve analysis was used to describe the central tendency of the studies, and to compare dermoscopy and clinical examination.
Data synthesis: Selected studies represented 328 melanomas, mostly less than 0.76 mm thick, and 1865 mostly melanocytic benign pigmented skin lesions. For dermoscopic diagnosis of melanoma, the sensitivity and specificity ranges were 0.75 to 0.96 and 0.79 to 0.98, respectively. Dermoscopy had significantly higher discriminating power than clinical examination, with respective estimated odds ratios of 76 (95% confidence interval, 25-223) and 16 (95% confidence interval, 9-31) (P =.008), and respective estimated positive likelihood ratios of 9 (95% confidence interval, 5.6-19.0) and 3.7 (95% confidence interval, 2.8-5.3). The roles of the number of lesions analyzed, the percentage of melanoma lesions, the instrument used, and dermoscopic criteria used in each study could not be proved.
Conclusion: For experienced users, dermoscopy is more accurate than clinical examination for the diagnosis of melanoma in a pigmented skin lesion.