BackgroundMatrix-assisted laser desorption-ionization time-of-flight mass spectrometry (MALDI-TOF MS) recently became available for the identification of bacteria in routine diagnostic laboratories. It is rapid and cost-effective and likely to replace phenotypic identification. This study was undertaken to compare two MALDI-TOF MS-based, Bruker Microflex MS (BMS) and VITEK MS (VMS) systems, for identification (ID) of clinically significant bacterial isolates. Clinically relevant broad diversity of bacterial isolates obtained during a 6-consecutive months of routine laboratory processing of clinical specimens were subjected to ID by the BMS and VMS in parallel with Vitek 2, a conventional phenotypic system (CPS). For the BMS, the isolates were tested in duplicates directly and after pretreatment. Identification was provided with accompanying scores according to manufacturers¿ instructions. With VMS, single deposits of the same sets of isolates were tested in duplicates directly on MALDI-plate. Results were interpreted according to the manufacturer¿s protocols. Discrepant results were resolved by 16S rRNA gene amplification and sequencing.ResultsA total of 806 pathogens comprising 507 Gram-negative bacilli (GNB), 16 Gram-negative cocci (GNC), 267 Gram-positive cocci (GPC), and 16 Gram-positive bacilli (GPB) were tested. BMS and VMS correctly identified isolates to genus and species levels (ID 97.3% and 93.2%, and 99.8% and 99.0%, respectively). Both systems as well as the CPS correctly identified the majority of the species in the family Enterobacteriaceae, Pseudomonas spp., and Acinetobacter baumannii. Turnaround time for identification by BMS and VMS was <20 min compared with 24-48 h by the CPS.ConclusionsVMS performed slightly better than BMS with GPC ID, especially the Streptococcus spp. Some S. mitis isolates were identified as S. pneumoniae by BMS. BMS and VMS were rapid and proved to be consistently accurate for producing bacterial identification in a fraction of time it takes for identification by CPS.