Feasibility and diagnostic accuracy of teledermatology in Swiss primary care: process analysis of a randomized...

To test feasibility and diagnostic accuracy of dermatologist's feedback based on digital images of skin lesions collected in Swiss primary care.

This was a process analysis of a randomized controlled trial, conducted in 2011/2012. 30 of 78 general practitioners (GPs) were randomized to an intervention, which included dermatologist's feedback on digital images of skin lesions. Feedback was categorized into four categories: (1) no further investigation; (2) clinical observation; (3) biopsy; and (4) other. Histologic findings were allocated to the same categories. Feasibility was measured in the perspective of GPs concerning technical handling and of dermatologists as proportion of images usable for feedback. Diagnostic accuracy was measured as proportion of malignant histology of the first three feedback groups. No long-term data was collected.

981 images of skin lesions were collected, two were not eligible due to low quality of images. The majority of GPs (77.8%) reported no problem with the procedure. 207 images were in feedback category 1, 353 in 2, 360 in 3 and 59 in 4. A total of 236 histologic tests were collected. Three cases in category 1 indicated malignancy (1 melanoma). 201 of category 3 received a biopsy, where in 91 (45.3%) malignancy was confirmed.

Teledermatology with digital images taken in primary care was feasible from a GP and from a specialist perspective. However, diagnostic appropriateness regarding avoidance of specialist care and possible missed skin cancer raises concerns. These results therefore question a promotion of teledermatology in clinical routine.