ORLANDO – Store-and-forward teledermatology may be useful for grading patch test results.
Erin Warshaw, MD, and Sara Hylwa, MD, both of the University of Minnesota, Minneapolis, sought to compare readings of patch test results both in person and via store-and-forward teledermatology. They patch tested patients at the Hennepin County (Minn.) Medical Center with the North American Contact Dermatitis Group screening series; photos were obtained at the 48-hour reading and the final reading (96-160 hours).
The teledermatology assessment was done by the same physician who assessed the patch test results in person, in order to avoid inter-reader bias. Teledermatology assessments were done 4 weeks and 8 weeks later and the reader was blinded as to the in-person results, Dr. Warshaw said at the annual meeting of the American Contact Dermatitis Society, held just prior to the start of the American Academy of Dermatology’s annual meeting.Almost all (101 of 107) of patients eligible for the trial were enrolled. Patients were overwhelmingly female (72%) with an average age of 50 years in this single-site study. Most screening panels were applied to the back.
Teledermatology assessment was categorized as successful if it matched the in-person assessment and as a failure if it did not; investigators labeled assessed pairs that did not fully match as indeterminate. Successful matches indicated there was no clinically significant difference between teledermatology and in-person assessment, indeterminate matches indicated that there was possible clinically significant difference, and failure to match indicated definite clinically significant difference.
All readings that were negative both in person and via teledermatology were excluded from the analysis.
At 48 hours, 47.2% of 705 reading pairs were labeled successful and 51.3% were labeled indeterminate. Failure, or complete disagreement, occurred in 1.6%, or 11 individual antigen pairs.
More successes – and failures – were seen at the final reading, with 53.8% of 420 final readings labeled successful, 39.8% labeled indeterminate, and 6.4%, or 27 individual antigen pairs, labeled as failures.
In general, teledermatology was more likely to miss or downplay the severity of reactions in the indeterminate pairs, Dr. Warshaw said. “This makes intuitive sense because when you are with a patient live, often the lighting catches an irritant wrinkle reaction or you can feel the lesion and be much more likely to call it irritant or a mild reaction than you would be from a flat photo.”
In the failure group, teledermatology generally overstated reactions, she added.
Dr. Warshaw said that logistical changes would be needed to make teledermatology more effective for reading patch test reactions in her practice. Their method of marking the patch test grid is to use a surgical marker on the corners, but a highlighter to mark the grid between the antigens. The highlighter simply did not show up well in photographs, she noted.
While not perfect, teledermatology does have promise for reading patch test reactions, she added. “I would love to save patients from having to come for their 48-hour reading... In Minnesota we have these horrible snowstorms. Last week there was a blizzard that was predicted. A third of our patients live 2 hours away from the clinic. If they could have taken photographs instead of trying to come through a blizzard for their final reading, that would be helpful.”
Dr. Warshaw noted that their study assessed only the 70 antigens of the North American Contact Dermatitis Research Group series and that it could have been strengthened by using additional series or the patients’ own products.
On Twitter @denisefulton