José L Moreno-Camacho1,2, Diana Y Calva-Espinosa1, Yoseli Y Leal-Leyva1,2, Dolores C Elizalde-Olivas1,2, Abraham Campos-Romero3, Jonathan Alcántar-Fernández3.
Abstract
BACKGROUND: To validate the performance, reproducibility, and reliability of BD automated instruments in order to establish a fully automated clinical microbiology laboratory.
METHODS: We used control strains and clinical samples to assess the accuracy, reproducibility, and reliability of the BD Kiestra WCA, the BD Phoenix, and BD Bruker MALDI-Biotyper instruments and compared them to previously established conventional methods. The following processes were evaluated: sample inoculation and spreading, colony counts, sorting of cultures, antibiotic susceptibility test, and microbial identification.
RESULTS: The BD Kiestra recovered single colonies in less time than conventional methods (e.g. E. coli, 7h vs 10h, respectively) and agreement between both methodologies was excellent for colony counts (κ=0.824) and sorting cultures (κ=0.821). Antibiotic susceptibility tests performed with BD Phoenix and disk diffusion demonstrated 96.3% agreement with both methods. Finally, we compared microbial identification in BD Phoenix and Bruker MALDI-Biotyper and observed perfect agreement (κ=1) and identification at a species level for control strains. Together these instruments allow us to process clinical urine samples in 36h (effective time).
CONCLUSION: The BD automated technologies have improved performance compared with conventional methods, and are suitable for its implementation in very busy microbiology laboratories. © American Society for Clinical Pathology 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
BACKGROUND: To validate the performance, reproducibility, and reliability of BD automated instruments in order to establish a fully automated clinical microbiology laboratory.
METHODS: We used control strains and clinical samples to assess the accuracy, reproducibility, and reliability of the BD Kiestra WCA, the BD Phoenix, and BD Bruker MALDI-Biotyper instruments and compared them to previously established conventional methods. The following processes were evaluated: sample inoculation and spreading, colony counts, sorting of cultures, antibiotic susceptibility test, and microbial identification.
RESULTS: The BD Kiestra recovered single colonies in less time than conventional methods (e.g. E. coli, 7h vs 10h, respectively) and agreement between both methodologies was excellent for colony counts (κ=0.824) and sorting cultures (κ=0.821). Antibiotic susceptibility tests performed with BD Phoenix and disk diffusion demonstrated 96.3% agreement with both methods. Finally, we compared microbial identification in BD Phoenix and Bruker MALDI-Biotyper and observed perfect agreement (κ=1) and identification at a species level for control strains. Together these instruments allow us to process clinical urine samples in 36h (effective time).
CONCLUSION: The BD automated technologies have improved performance compared with conventional methods, and are suitable for its implementation in very busy microbiology laboratories. © American Society for Clinical Pathology 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
Entities:
Keywords:
BD Kiestra; BD Phoenix; CLSI; InoqulA; MALDI Biotyper; full automation; microbiology laboratory; urine
Mesh：
Year: 2017
PMID： 29253199 DOI： 10.1093/labmed/lmx079
Source DB: PubMed Journal: Lab Med ISSN： 0007-5027