Evaluation of an Automated Yeasts Identification System for Identification of Yeast Isolates.

ABSTRACT

BACKGROUND: Invasive candidiasis is the most important health-care-associated fungal infection worldwide. In the last two decades, the cause of the increase of fungal infections is immunosuppression or serious underlying diseases. Additionally, Rhodotorula species, Blastoschizomyces capitatus, and Trichosporon species are emerging as important human pathogens in immunocompromised patients with hematological malignancy. METHODS: Between January 2012 and January 2018, a total of 603 fungal organisms were isolated from blood culture samples and included in the study. All of the isolates were identified by using standard mycological methods, MALDI TOF MS system, and the Phoenix system. Sequence analysis was performed on yeasts that could not be definitively identified by using SMM and incompatible according to the results with Phoenix and MALDI-TOF MS analysis. RESULTS: 603 fungal isolates including 594 Candida spp. and 9 other yeasts like species were analyzed. C. albicans was the most frequently isolated species. The results of identification by conventional methods and MALDI TOF MS were compared to the results of the Phoenix system. The observed concordance was 99.2%. The compatibility with other systems of the Phoenix system was 100%, 100%, 97.3%, 100%, and 96.9% for C. albicans, C. parapsilosis, C. tropicalis, C. glabrata, and C. krusei, respectively. CONCLUSIONS: The BD Phoenix system was found to be a simple, reliable, and effective method to identify the main species of the genus Candida in our study.