MALDI-TOF mass spectrometry to identify clinical bacterial isolates: Evaluation in a teaching hospital

Purpose. - The aim of our study was to evaluate the capacity of MALDI-TOF mass spectrometry to identify clinical bacterial isolates, as compared to the automated identification system Vitek 2 (bioMerieux) used routinely in a teaching hospital. Methods. - Three hundred and sixty-two strains representing 178 species from the laboratory collection were analysed by a Microflex spectrometer (Bruker Daltonics) and Vitek 2. Discrepancies between MALDI-TOF and Vitek 2 identifications were investigated by genetic identification (rrS, sodA, rpoB), considered as a reference. Results. - Among the 362 isolates, 264 (73%) were consistently identified by Vitek 2 and Microflex. Taking into account genetic identification, we found that 44 (44.9%) of the 98 remaining isolates were correctly identified by mass spectrometry but not by Vitek 2. Conversely, 33 isolates (33.7%) were correctly identified by Vitek 2, but not by Microflex. The genetic identification of the 21 remaining isolates (21,4%) did not match either Vitek 2 or Microflex results. Conclusion. - The performances of MALDI-TOF mass spectrometry for bacterial identification correspond to those of a reference automated identification system. (C) 2009 Elsevier Masson SAS. All rights reserved.