Limited evidence for benefits of computer-assisted techniques for diagnosing skin cancer in adults

Limited evidence for benefits of computer-assisted techniques for diagnosing skin cancer in adults

Brian McAvoy
PEARLS No.
636
Clinical question

Compared with dermoscopy, how effective are computer-assisted diagnosis techniques (dermoscopy and spectroscopy-based) for diagnosing skin cancer in adults?

Bottom line

Meta-analysis found CAD systems had high sensitivity for correct identification of invasive cutaneous melanoma and atypical intraepidermal melanocytic variants in highly selected populations, but with low and very variable specificity, particularly for spectroscopy-based CAD systems. Due to the paucity of studies, it was not possible to make summary statements about the use of CAD in unreferred populations, its accuracy in detecting keratinocyte cancers or its use in any setting as a diagnostic aid. Twenty-four studies evaluated dermoscopy-based CAD, and 18 evaluated spectroscopy-based CAD.

Caveat

The evidence base was too limited to draw conclusions on which individual systems might be preferred for practice. It was not always possible to tell whether suspicion of skin cancer in study participants was based on clinical examination alone, or both clinical and dermoscopic examinations.

Context

Early accurate detection of all skin cancer types is essential, to guide appropriate management and to improve morbidity and survival. When used in unreferred settings (primary care), CAD may assist GPs or other clinicians to more appropriately triage high‐risk lesions to secondary care. Used alongside clinical and dermoscopic suspicion of malignancy, CAD may reduce unnecessary excisions without missing melanoma cases.

Cochrane Systematic Review

Ferrante di Ruffano L et al. Computer-assisted diagnosis techniques (dermoscopy and spectroscopy-based) for diagnosing skin cancer in adults. Cochrane Reviews, 2018, Issue 12. Art. No.: CD013186.DOI: 10.1002/14651858. CD013186. This review contains 42 studies involving 15,938 lesions.