If you’re concerned about skin cancer and you notice a suspicious mole, what do you do?
Chances are you show it to a loved one, who either brushes it off or voices alarm and nags you to go to the doctor to get it checked out.
Imagine, though, if you could get that mole vetted by a crowd of hundreds? New research out of the University of Utah and Texas Tech University suggests it would significantly improve your odds of an accurate diagnosis.
Melanoma is the most deadly type of skin cancer and rates of the disease have increased every year for the past 30 years. Utah has higher than average rates of melanoma deaths.
“It’s one of the handful of cancers where we are not winning the war,” said one of the study’s authors, Jakob Jensen, an assistant communications professor at the U.
The problem stems not necessarily from lack of awareness but from a shortage of dermatologists, especially in rural parts of America, and the failure of self-examinations, he said.
Studies have shown self-examinations using the old “ABCDE” method of skin cancer detection — looking for asymmetry, a jagged border, multiple shades of color, a larger diameter and an evolving shape — aren’t effective.
“It takes quite a bit of skill to look at a lesion and determine whether it is cancerous or not,” said Jensen.
But it turns out that groups of people — or “crowdsourcing,” better known by communication theorists as “collective effort” — are very good at detecting melanoma, he explained.
Jensen and colleagues, including oncology and dermatology experts, showed 500 adults high-resolution images of 40 moles, including nine previously diagnosed as melanomas, and asked them to circle those they found suspicious.
The average individual was only able to identify about half of the melanomas, researchers found. But 19 percent of the participants were able to correctly identify 90 percent of the melanomas, the study showed.
In other words, if 19 people out of 100 think your mole is a concern, you should get to a doctor.
It’s the sheer numbers that give the herd this diagnostic edge, not group decision-making, said Jensen, noting that each mole analyzer performed the task independently.
The study was funded by the National Institutes of Health and the National Cancer Institute and published in the December issue of Cancer Epidemiology.
“There’s still a lot of work to be done” to understand how to apply the findings, said Jensen.
The information could be leveraged by health centers looking to expand their reach into isolated communities via telemedicine, he said.
“Dermatologists will tell you that most of the patients they see are those who tend to worry, but who aren’t high risk. This could be a way to identify and reach those who are at high risk,” Jensen said.
His team is working on a smart phone app that will allow people to take a photo of their mole and have the image evaluated by hundreds, if not thousands, of users.
“Another line of research is looking at how to improve self skin examinations through educational campaigns by understanding why the crowd is better,” he said.
Melanoma detection apps already exist. Most ask users to upload photos and use algorithms to produce a risk score. But software on the market today isn’t evidenced-based and users have filed lawsuits over improper diagnoses, said Jensen.
“We want to put this finding out to the community,” he said. “We want researchers around the world who are interested in this to play with it and find new opportunities.”