Melanoma is a completely curable cancer, if diagnosed early before the tumor has penetrated into the dermis. The challenge for physicians is to determine which of the many pigmented lesions a patient may have is a melanoma that needs to be removed.
Dermoscopy has been advocated as a tool that can help physicians make this decision. Its usefulness is usually answered by determining how well it permits a lesion to be identified correctly as melanoma. Based on a study by Ferrara et al.1 that was published in Cancer, the answer is not very well. Of 107 pigmented lesions examined by 8 experts in the field of dermoscopy, the number of lesions diagnosed as melanomas ranged from 26–60, a variation of > 100%. Diagnostic accuracy should be even less in the day-to-day use of this procedure by physicians who are not experts.
However, the real decision a physician has to make is not whether a pigmented lesion is a melanoma. That determination is best left to the pathologist. The real decision to be made by the physician is whether a lesion is sufficiently suspicious for melanoma to warrant a biopsy.
How good is dermoscopy? The real question that needs to be answered is whether it improves the ability of physicians to biopsy lesions that are melanomas or precursors to melanoma but that might otherwise have been missed. That question remains to be answered.