Background: Most dermoscopic algorithms to diagnose melanoma were established more than 10 years ago and have been tested primarily on clear-cut melanomas and excised melanocytic naevi. Objectives: To assess the diagnostic performance of pattern analysis and seven-point checklist on lesions that reflect the current clinical setting, compared with a revised seven-point checklist with a lower threshold for excision. Methods: Eight experienced dermatologists viewed dermoscopic images of 100 excised melanomas, 100 excised naevi and 100 monitored naevi. Each lesion was evaluated by pattern analysis and scored as naevus, melanoma or lesion to be excised. Images were then evaluated using the seven-point criteria, with both standard and revised thresholds for excision. Results: Pooled data using the pattern analysis algorithm showed that 82% of melanomas and 87.5% of monitored naevi were correctly scored as lesion to be excised and benign naevus, respectively. Using the standard and revised thresholds for the seven-point checklist, excision was recommended for 77.9% and 87.8% of the lesions in the melanoma set, respectively. The standard threshold produced 'no excision' recommendations for 85.6% of the monitored naevi, compared with 74.5% using the revised threshold. Pattern analysis, standard seven-point and revised seven-point algorithms resulted in recommendations of 'excision' for 63.6%, 60.3% and 72.0% of the excised naevi, respectively. Conclusions: The diagnostic approach to naevi and melanoma should be adapted to the current clinical setting, in which patients may present with early-stage melanomas and multiple atypical naevi. To increase sensitivity, a revised seven-point checklist with a lower threshold for excision should be used.
ASJC Scopus subject areas