Background: Fast identification of Candida glabrata is important, because empirical antifungal therapy for fungemia with C. glabrata and non-C. glabrata varies. We proposed an algorithm for rapid presumptive diagnosis to identify fungemia with C. glabrata using earlier or only growth from anaerobic bottles and longer time to positivity (TTP) in blood cultures. 

Methods: Positivity and TTP using the BacT/Alert 3D system (bioMerieux Inc, USA) with resin bottles (FA Plus and FN Plus) were analyzed in 215 candidemia patients from June 2014 to June 2016 in a university-affiliated hospital in Korea.

Results: A higher proportion of earlier or only growth from anaerobic bottles was observed in C. glabrata (38.8%, 7/18) than in C. albicans (7.6%, 8/105), C. parapsilosis (10.5%, 4/138), and C. tropicalis (9.2%, 5/54) (P=0.006). The mean (±standard deviation) TTP for C. glabrata was 41.7 h (±16.3 h) compared with 26.7 h (±15.9 h) for C. albicans, 33.4 h (±8.4 h) for C. parapsilosis, and 23.1 h (±17.3 h) for C. tropicalis (P＜0.0001). We could predict fungemia with C. glabrata with a sensitivity of 94.4%, specificity of 63.9%, positive predictive value of 19.3%, and negative predictive value of 99.2% using a two-step algorithm: earlier or only growth from anaerobic bottles and TTP ＞31.4 h.

Conclusion: This two-step algorithm in the BacT/Alert 3D system could be the basis for an initial empirical antifungal therapy for fungemia with C. glabrata prior to final identification. (Ann Clin Microbiol 2018;21:23-27)