Answers are grounded in this paper and cite their sources.
Retrospective cross-sectional study compares bacterial identification methods, suggesting high agreement for routine use.
To compare bacterial identification using the VITEK 2 Compact (bioMérieux) and Bruker Microflex LT matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) systems in routine clinical microbiology. This retrospective cross-sectional study included consecutive clinical isolates recovered between February and April 2022 from a 1355-bed academic center. Identification was performed using VITEK 2 GN/GP cards and the Bruker MALDI Biotyper workflow using a Microflex LT instrument. Identifications were compared at species and genus levels; concordance was defined as identical taxon names. Agreement was summarized as percent agreement (Wilson 95% confidence intervals (CI)) and Cohen’s kappa (κ); discordance was interpreted as between-system disagreement because no molecular adjudication was performed. A total of 1394 isolates were analyzed. Most isolates were obtained from urine-related specimens (n=563, 40.4%) and blood cultures (n=446, 32.0%). Species-level concordance between systems was 93.1% (95% CI, 91.7–94.3%; κ=0.923 (95% CI, 0.908–0.937)), and genus-level concordance was 98.3% (95% CI, 97.5–98.8%; κ=0.979 (95% CI, 0.970–0.987)). Among the 96 discordant species identifications, 72 (75.0%) occurred within the same genus, indicating disagreements concentrated in closely related taxa. The most frequent discordant pairs involved the Klebsiella genus and coagulase-negative Staphylococcus species; the single most common pair was Klebsiella pneumoniae versus Klebsiella variicola (n=12). VITEK 2 and Bruker Microflex LT MALDI-TOF MS demonstrated very high agreement for routine bacterial identification, with most discrepancies occurring among closely related taxa. When species-level precision is clinically critical (e.g., outbreaks or invasive infections), confirmatory approaches such as 16S rRNA sequencing and up-to-date databases may be required.
No takes yet. Share an insight, caveat, or question.
Karabıyık et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: