456 A prospective cohort study comparing microscopy and culture in the diagnosis of superficial fungal skin infections

Fungal skin infection is highly prevalent worldwide and in rising incidence. It can cause itchy and scaly skin leading to patient distress. Diagnostic tests for fungal infections include microscopy and culture of the skin scrapings which can be done singly or in conjunction, each having its own pros and cons. Direct microscopy yield faster results, is more sensitive and lower in cost but does not report the fungal species involved unlike cultures. To date there are no significant large cohort study studies comparing the agreement between these 2 investigations. In this prospective cohort study, we investigate the agreement between both tests and explore the fungal species isolated from fungal cultures. This epidemiological update could be relevant in emerging resistance patterns when certain fungal species become unresponsive to topical antifungals. 268 skin samples were collected from 149 patients from National Skin Centre Singapore with suspected fungal skin infection and subjected to both microscopy and culture. The most common site sampled was the feet (26.9%). 213 out of 252 (84.5%) samples revealed the same results from both tests - 27 samples (10.7%) were both positive and 186 (73.8%) were both negative, resulting in 84.5% agreement and Cohen’s Kappa κ = 0.487 (p-value < 0.0001), indicating moderate agreement. Out of 210 culture-negative samples, 24 were microscopy-positive, and out of 201 microscopy-negative samples, 15 were culture-positive. Microscopic visualization of fungal morphology showed 46 (68.7%) with mycelium (indicating Dermatophytes), 16 (23.88%) with spores and hyphae (indicating Malassezia species), and 5 (7.46%) budding yeast cells with pseudohyphae (indicating Candida species). The top 3 species isolated in cultures were Trichophyton rubrum (52.4%), Trichophyton mentagrophytes variant interdigitale (28.6%), and Candida albicans (11.9%). Since both tests are in moderate agreement, the authors propose that conducting either test for diagnosis in clinical practice can be sufficient when resources are limited.