Longer incubation times for yeast fungemia: importance for presumptive treatment

Highlights

• To determine factors that influence incubation time using the BacT/ALERT system, we analyzed 1741 episodes of yeasts fungemia from 12 French hospitals participating in a prospective surveillance network (Paris area, 2002–2009). The overall mean incubation time was 2.17 ± 1.10 days and 2.97 ± 1.18 days for the 289 Candida glabrata episodes. The time of recovery was ≤2 days for 1221 (70%) and >2 days for 520 (30%) of the 1741 episodes. The species frequency changed according to the time of incubation with C. glabrata isolation rate increasing from 8.9% (108/1221) to 34.8% (181/520) ≤2 days and >2 days, respectively. Similarly, Cryptococcus neoformans fungemia represented only 0.9% (11/1221) of the fungemia ≤2 days and 8.6% (45/520) afterwards. The proportion of recovery for non-Candida and non-Cryptococcus fungi was not significantly different ≤2 days or >2 days of incubation. In parallel, a multivariate analysis showed an incubation time >2 days significantly associated with C. glabrata isolation, solid organ transplantation, intensive care units, and fluconazole within the past 30 days before fungemia. Although an incubation time over 2 days is associated with C. glabrata isolation and previous fluconazole therapy, supporting candins as the first antifungal treatment, this attitude must be challenged with the risk of C. neoformans–related fungemia.

• This underlines the limits of presumptive treatment for fungemia and stresses the necessity to introduce rapid identification methods for yeast species identification.

Abstract

Isolation rates of Candida glabrata at ≤2 days were 8.9% and 34.8% at >2 days; for Cryptococcus neoformans, they were 0.9% and 8.6%, respectively (1741 fungemia analyzed). An incubation time >2 days supports candins as presumptive treatment for C. glabrata, keeping in mind the risk of Cryptococcus fungemia.