Abstract
Bloodstream infections due to yeast are associated with a high mortality rate. There is a lack of data that evaluate the real-world sensitivity of a rapid detection system for bloodstream infections due to yeast or the impact of these results on antimicrobial stewardship. The aim of this study was to evaluate the sensitivity of an ePlex panel (BCID-FP) for rapid detection of yeast from a positive blood culture bottle and to evaluate the impact of these rapid results on antifungal escalation or de-escalation. We evaluated 63 episodes of fungemia and found a sensitivity of 94%, lower than the 99%-100% stated in the package insert. Most common pathogens were Candida glabrata (36%), Candida albicans (24%), Candida parapsilosis (10%) and Candida krusei (10%). Only 57.1% of BCID-FP results lead to a change in antifungal therapy, most of which was initiation of an echinocandin. The real-world sensitivity of the BCID-FP panel was lower than anticipated and rarely led to de-escalation of antifungal therapy.