Comparison of three test methods used for the diagnosis of candidiasis.

A total of 266 specimens was taken from oral mucosa or dentures of 88 patients with suspected Candida-infected oral mucosa. One-third of the debris from each sample was cultured on Microstix-Candida (C), strips, one-third was cultured on Oricult-N-plates, and the rest was collected on glass plates and stained by the periodic acid-Schiff (PAS) method. The PAS-stained samples were analyzed under the light microscope for the presence of Candida hyphae. The other samples (Oricult-N or Microstix-C) were studied visually, according to the manufacturer's recommendation. PAS-stained specimens showed significantly less often positive results than those of the two culture methods. These data suggest that oral candidiasis may be incorrectly diagnosed if based upon results obtained with the culture methods. It is also possible that the PAS-staining method yielded false negative findings. This possibility should always be considered, especially if other findings and symptoms disagree with the test results obtained.