Routine identification and mixed species detection in 6,192 clinical yeast isolates

International audienceThe clinical laboratory methods used to diagnose yeast infections should be rapid, reliable, and capable of detecting mixed infections with species exhibiting a distinct antifungal susceptibility profile. In this study, we report the performance of a procedure combining the detection of mixed yeast cultures with a chromogenic medium and MALDI-TOF identification of the colonies. We then evaluated the impact that (i) the isolation medium and (ii) lowering the identification log score (LS) threshold value have on yeast identification performance in the routine laboratory. Among 15,661 clinical samples analyzed, 5,671 tested positive and 6,192 yeasts of 42 distinct species were identified. Overall, 6,117 isolates (98.79%) were identified on the first or second MALDI-TOF Mass Spectrometry (MS) attempt, yielding an average yeast species identification turnaround time of 0.346 days (95% CI [0.326 to 0.364]). The 75 remaining isolates were identified via nucleotide sequencing. Mixed infections accounted for 498 (8.78%) of the positive samples. The MALDI-TOF MS identification procedure performed well, regardless of the culture media tested. Lowering the recommended 2.0 LS threshold value to 1.8 would reduce the number of required (i) second MALDI-TOF MS identification attempts (178 vs. 490) and (ii) ITS2 and D1-D2 sequence-based identifications (17 vs. 75), while achieving an adequate identification rate (6,183/6,192, 99.85%). In conclusion, we propose applying a 1.8 LS threshold combined with chromogenic medium subculture to optimize the yeast identification workflow and detect mixed infection in the clinical laboratory