Periorbital lupus profundus successfully treated with oral tacrolimus plus low-dose prednisolone.

analyzed data, while nurses took photographs of the nails, assisted in collection of specimens, and stored them. In total, 113 images were prepared which included 58 (51.3%) nails with tinea unguium. Diseases other than tinea unguium included onychogryphosis, pachyonychia, onycholysis, leukonychia and subungual hematoma. The dermatologists judged whether the cases in the images were tinea unguium and the correct diagnostic rate was calculated. Among the 31 dermatologists, the average, minimum, maximum and median years of experience in dermatology were 6.5, 0.6, 35.0 and 4.0, respectively. The dermatologists identified tinea unguium by visual inspection at an overall average correct rate of 67.7 7.2% (mean SD), with minimum, maximum and median rates of 47.8%, 79.6% and 68.1%, respectively. There was no correlation (r = 0.162, P = 0.385) between experience in dermatology and the correct diagnostic rate (Fig. 1a) although the average correct rate was 62.3% for dermatologists with 1 year experience or less, which was relatively lower than the rates of 68.4%, 69.2% and 68.8% for dermatologists with experience of 2–3 years, 4–5 years and more than 5 years (Fig. 1b). The overall rate of correct diagnosis was 67.7%; thus, diagnosis by inspection resulted in misdiagnosis in more than 30% of cases, in comparison with the diagnostic sensitivity of KOH direct microscopy of 80–90%. Dermatologists with 1 year or less of experience had a lower diagnostic accuracy based on inspection, but greater experience in dermatology beyond 2 years did not improve the correct diagnostic rate. These results demonstrate the difficulty of clinical diagnosis of tinea unguium based only on visual inspection and emphasize the importance of mycological examination.