From the Journals

Teledermoscopy using smartphones fails to boost skin cancer detection



Skin cancer detection does not improve when patients use a dermatoscope for smartphones to capture and send photos of lesions for diagnosis, according to the Australian SKIN Project trial.

“Mobile health applications are increasingly used in cancer prevention and early detection but rarely tested stringently for their value with regard to patient care,” noted trial investigators, who were led by Monika Janda, PhD, of the University of Queensland in Brisbane, Australia. The trial was published in Lancet Digital Health.

The investigators studied 234 adults at high risk for skin cancer, asking them to perform whole-body skin self-exams in their homes at baseline, 1 month, and 2 months.

Half of patients were randomized to perform standard naked-eye exams, note suspicious lesions on a body chart, and submit the chart by email. The other half were randomized to supplement their exams with mobile dermoscopy, and they were provided with a dermatoscope (FotoFinder) that interfaces with smartphones to capture and submit photos of suspicious lesions to a dermatologist for telediagnosis.

Both groups received Web-based instructions on how to complete whole-body skin self-exams and were examined in person within 3 months of their last self-exam to provide a reference standard.

In comparing the two approaches for early detection of skin cancer, the investigators determined that teledermoscopy-enhanced exams would have to show 20% greater sensitivity to establish their superiority.

The median number of lesions submitted was six per person in both the group using adjunctive teledermoscopy and the group performing only naked-eye exams.

At the lesion level, teledermoscopy-enhanced exams had lower sensitivity than naked-eye exams in detecting suspected skin cancers or precursor lesions (75% vs. 88%; P = .04) and similar specificity (87% vs. 89%; P = .42). At the patient level, the two approaches had statistically indistinguishable sensitivity (87% vs. 97%; P = .26) and specificity (95% vs. 96%; P = .96).

At the same time, telediagnosis showed good overall diagnostic concordance with in-person clinical skin examination (88%).

“For the early detection of skin cancer, naked-eye skin self-examination should continue to be recommended by cancer agencies,” Dr. Janda and colleagues concluded. “Further improvements to the instructions for participants on the relevance of nonpigmented skin lesions, training for partners, and the integration of automatic algorithms that rule out clearly benign skin lesions at the time of photographing might increase sensitivity of teledermoscopy in the future.”

This trial was funded by the National Health and Medical Research Council. The authors disclosed relationships with e-derm-consult, SciBase, Canfield Scientific, and other companies.

SOURCE: Janda M et al. Lancet Digit Health. 2020 March; 2(3):e129-e137.

Next Article: