RÉSUMÉ
BACKGROUND: Onychomycosis affects between 2% to 30% of the world population. Nail biopsy may help in making a diagnosis and can distinguish between invasion and colonisation. AIMS: To evaluate the diagnostic usefulness of nail biopsy with Periodic acid-Schiff (PAS) staining for onychomycosis, compared to direct KOH examination, culture and its combination in a reference laboratory in Colombia. METHODS: The study included 66 patients in whom a blind and independent reading of the three tests was performed. The usefulness was defined based on the validity (sensitivity, specificity, Youden's index, likelihood ratios), performance (predictive values) efficiency (proportion of correctly diagnosed patients), and reproducibility (kappa coefficient). RESULTS: The mean age of the patients was 55±16 years, and included 76% women. The direct tests with KOH were positive in 66.7% (n=44), 62.1% (n=41) were positive with culture, and 56.1% (n=37) with the biopsy. The main causal agents were non-dermatophytes moulds in 36.4% (n=24). The most frequent species were Neoscytalidium dimidiatum (n=11), Trichophyton rubrum (n=11), and Candida parapsilosis (n=13). The sensitivity of nail biopsy, when compared to the standard (KOH and/or culture), was 71%, specificity 83%, Youden's index 0.54, positive likelihood ratio 4.25, negative likelihood ratio 0.35, positive predictive value 92%, negative predictive value 52%, efficiency 74% and kappa coefficient 0.45. When biopsy was evaluated only in patients with onycholysis of the nail plate greater than 50%, all the parameters of diagnostic usefulness increased. CONCLUSIONS: The overall usefulness of the biopsy was moderate for patients with more severe symptomatology, which makes its use advisable in cases of extensive onycholysis, and when discriminating colonisation from invasion is required.