In this study, we aimed to detect the proportion of Candida dubliniensis among yeast strains previously identified as C. albicans by using several phenotypic methods and PCR.
For this purpose, we screened 300 strains by using phenotypic tests suggested for the identification of C. dubliniensis in the literature, but we detected high proportion of false-positive reactions. Only two strains (0.6%) were detected as true C. dubliniensis by PCR and API ID 32C methods. Moreover, these two strains gave the expected results with all the phenotypic tests, including modified salt tolerance test for C. dubliniensis.
In conclusion, none of the phenotypic methods, except for the modified salt tolerance test, revealed 100% successful results in discrimination of C. albicans and C. dubliniensis species. However, in the tobacco agar test, the rate of false positivity was as low as 0.6%. We suggest that in the case of absence of PCR and other automatized identification systems, these two phenotypic tests can be used in routine laboratories to obtain a presumptive result.