By Charlotte LoBuono
Smartphone apps that analyze user photographs of suspicious lesions for the likelihood of skin cancer seem like the perfect screening tool, but they can actually be life-threatening. Their variable and inaccurate results can delay the correct diagnosis of melanoma, the most deadly form of skin cancer, and potentially lifesaving treatment.
Three out of the four apps tested in a recent study incorrectly diagnosed at least 30 percent of the submitted images of melanomas as "unconcerning." The only app with a high rate of accuracy was one in which the digital image was read by a dermatologist.
The applications usually come with a disclaimer that they are intended for educational purposes, not to be used to make a definitive diagnosis, and that they should not substitute for standard medical care. However, the authors of the study are concerned that such a tool may be misused by the general public, particularly in a tough economy, where many people are either uninsured or reluctant to pay a co-pay for an office visit.
Researchers at the University of Pittsburgh uploaded 188 images of skin lesions (60 melanomas and 128 benign control lesions) to each of four applications available on the two most popular smartphone platforms. The apps were not identified by name. Of these four apps, three read the images using an automated analysis algorithm. The fourth app used a network of anonymous, board-certified dermatologists to evaluate the image.
"Smartphone usage is rapidly increasing, and the applications available to consumers have moved beyond communication and entertainment to everything under the sun, including health care," Laura Ferris, an assistant professor in the department of dermatology at the University of Pittsburgh School of Medicine, said in a statement. "These tools may help patients be more mindful about their health care and improve communication between themselves and their physicians, but it's important that users don't allow their 'apps' to take the place of medical advice and physician diagnosis."
The study was published online in JAMA Dermatology.
March 6, 2013