By Abby S. Van Voorhees, MD, March 01, 2013
In this month’s Acta Eruditorum column, Physician Editor Abby S. Van Voorhees, MD, talks with April W. Armstrong, MD, MPH, about her recent Journal of the American Academy of Dermatology article, “Application of mobile teledermatology for skin cancer screening.”
Dr. Van Voorhees: We’ve all been hearing about teledermatology, we know it’s coming of age. Tell us why you thought about studying the use of cellular phone-based store-and-forward teledermatology.
Dr. Armstrong: Teledermatology has been around for decades. But one of the trends that we’re seeing in the last five years is the use of mobile teledermatology. In North America, over 90 percent of Americans own a mobile phone and over 50 percent of them own a smartphone, a phone with advanced telecommunication features. With their availability and portability, smartphones have become a useful platform for medical communication.
I was interested in studying how mobile teledermatology could work due to the changing technology landscape in medicine and thanks to a larger volunteer teledermatology initiative organized by the AAD. In 2011, Drs. William James and Carrie Kovarik started organizing a national effort in providing teledermatology services to underserved communities as a part of the AAD’s Telemedicine Task Force. Dermatologists around the country started using a mobile platform to perform volunteer services for clinics that otherwise do not have dermatology access. This captures one part of the research aspect of that project. [pagebreak]
Dr. Van Voorhees: What do we know about the feasibility of using smartphones to make clinical diagnoses? How accurate can one expect to be?
Dr. Armstrong: There are advantages and limitations associated with smartphone use. With regard to the advantages: Current smartphones are extremely capable. They’re portable and they don’t have to be connected to a computer — so the image transmission can be instantaneous. The limitations are often dependent on two things. Some challenges are due to limitations in technology capabilities — with certain smartphones, their ability to capture high-resolution images may be suboptimal compared to a digital camera. Also, the small screen size of the mobile phone may limit the amount of information that can be displayed at once, both in terms of the way people enter information into the mobile phones and the way information is presented to users. Finally, regarding transmission, often the speed at which images can be transferred to the consultant depends on the bandwidth of the cellular network. Limited network bandwidth could limit the speed of the data transmission. Fortunately, these technology-related issues decrease with every year that goes by.
I also think the skill of the person taking the photograph is really important. How well the users can take quality photographs with the smartphones can have a big impact on how images will appear.
Accuracy will largely depend on the phone technology itself, including image resolution and lighting adjustments, and the skill of the person taking the photo. In our study, the staff were trained to take the photos — they had a training session ahead of time — and the phone used was an earlier Google Android phone. We found a concordance of 82 percent for diagnostic accuracy. When you look at other studies, Dr. Kovarik has done a lot of work with international teledermatology and she found similar concordance with Egyptian patients using mobile teledermatology in the developing world. So I think the accuracy depends on the technology itself as well as the skill of the person taking the photograph. [pagebreak]
Dr. Van Voorhees: Tell us about your study.
Dr. Armstrong: We organized a skin cancer screening event, and as part of it, patients who came were offered the opportunity to participate in the study. If they agreed, we had one dermatologist, myself, on-site at the skin cancer screening event. And then we had another dermatologist, Dr. Haines Ely, an experienced teledermatologist who was available remotely to view images. For each patient in the study, I evaluated them in person, and our study staff took photos of their skin lesions and then transmitted them to Dr. Ely off-site. He evaluated those images independently and was blinded to what I had diagnosed and my management plan and came up with his own diagnosis and management plan. We then compared our diagnostic concordance and our management concordance.
I should add that before our study we did a pre-study diagnostic concordance and management concordance between the two dermatologists because we all have different practice patterns and can come to different conclusions even looking at the same patient in person. So Dr. Ely and I did a clinic together where we assessed the same patients in person; he went in first, evaluated the patients, and came up with his diagnosis and management plan, and then I did the same. We were above 90 percent in both diagnostic and management concordance. We also did that with the teledermatology photos and we found that we were pretty high on our concordance for that as well.
Our primary aim was looking at management concordance; we wanted to determine if we would come to similar management decisions regardless of whether the patient was assessed in person or through mobile teledermatology. Because these were actionable items, that was our primary endpoint. Our secondary endpoint was looking at the diagnostic concordance. Our management concordance was 81 percent. We separated our diagnostic concordance; patients came in with their primary lesion of concern but they could also have a number of other lesions of concern. So we evaluated the primary categorical diagnostic concordance and also the categorical diagnostic concurrence for additional lesions, and then had an aggregated diagnostic concordance. For the primary lesion of concern, Dr. Ely and I had a diagnostic concordance of about 82 percent. [pagebreak]
Dr. Van Voorhees: Categorical diagnostic concordance implies that you’re in the same family of diagnosis — so you could have both decided it was a benign pigmented lesion, but you didn’t have to agree about whether it was dysplastic or more typical?
Dr. Armstrong: In this case, we would have the same primary categorical concordance but would differ in management concordance — observance versus biopsy. The categories were defined very narrowly. If he thought it was a basal cell and I thought it was a squamous cell, those were in different categories. However, if he thought it was a squamous cell carcinoma in situ and I thought it was a squamous cell carcinoma, we would consider that concurrence for the primary categorical concordance.
Dr. Van Voorhees: What were the strengths and limitations of your study?
Dr. Armstrong: The strength of the study is that it was the first study in North America to look at the use of mobile technology with regard to management and diagnostic concordance. It allows clinicians to think about how to potentially use that technology in their own practices in the future. It applies to situations where the dermatologist is located in a different location from the patient or in volunteer situations, and there are other applications that I’ll expand on later.
The limitations of the study include the fact that we didn’t include dermoscopy. There are technologies now that allow you to attach the dermatoscope to a mobile phone and transmit dermoscopic images. It would be interesting to do that and see what the outcomes turned out to be. Also, this study was done using one of the earlier mobile platforms designed for this purpose, so there were a few images that couldn’t be transmitted at the time of the skin cancer screening and had to be transmitted at a later time; that was not the instantaneous transmission we were looking for.
We found that atypical nevi were difficult for our teledermatologist to assess. He felt like some of the images were suboptimal. For the future, incorporating dermoscopy would be helpful. We also found that advanced patient age was associated with diagnostic disagreement. These patients are more likely to have pre-cancerous and cancerous lesions and their overall skin tone may be more mottled as well. [pagebreak]
Dr. Van Voorhees: Do you see smartphones as the future for data transmission from patients to their physicians? What do you see as the practical applications for the practicing dermatologist?
Dr. Armstrong: I think smartphones will be one form of data transmission from patients to physicians. Future health care consumers, our current young generation of patients, are extremely comfortable with mobile technology. I see smartphones being a large part of their communication to their physicians.
With regard to if I see this as a practical application for a practicing dermatologist, I think part of this will be consumer-driven in the future. We’re at a time when we can get extremely disparate opinions on mobile teledermatology depending on the level of comfort that the dermatologist has with communicating directly with their patients via this technology. Mobile teledermatology can be helpful, for example, for things like patient consultations where they otherwise wouldn’t have a dermatologist available in an inpatient setting; in those situations I can see it being quite helpful if the team in an inpatient setting can send images to a dermatologist who has an affiliation with the hospital who can view the images from anywhere conveniently on their phone and relay their recommendation. (Whichever mobile teledermatology application a practicing dermatologist chooses, it is important to make sure that the application is HIPAA-compliant and there are various information-security systems in place that allow for secure transmission.)
I think a direct-to-consumer trend in teledermatology is emerging where patients can send their photos directly to their dermatologists; the mobile platform will probably grow in that direction in the next five or 10 years. Patient demand will push in that direction because they want the convenience of taking a photo of their skin lesion, sending it instantaneously, and having it evaluated by their dermatologist quickly. Using mobile phones for medicine will likely be a growing trend as technology continues to improve and our health care consumer demographics evolve.
AAD telederm program
Interested in trying out teledermatology? AccessDerm is an Academy-sponsored teledermatology program that allows volunteer AAD member dermatologists to provide care to underserved populations in the United States. By volunteering in the program, members can consult remotely on dermatology cases using mobile devices and the Internet. To learn more or volunteer, visit www.aad.org/member-tools-and-benefits/volunteer-and-mentor-opportunities/accessderm-teledermatology-program.
Dr. Armstrong is director of the Dermatology Clinical Research Unit, director of teledermatology, and an assistant professor of dermatology at the University of California Davis. Her article was published in the Journal of the American Academy of Dermatology, 2012 (October); 67(4):576-581. doi: 10.1016/j.jaad.2011.11.957.