Evaluation of performance of pulsed field gel electrophoresis in subtyping of Klebsiella pneumoniae

Objective To evaluate the performance of pulsed field gel electrophoresis (PFGE) in the subtyping of Klebsiella pneumoniae. Methods PFGE was used to detect 220 K. pneumoniae strains to evaluate its subtyping capability and reproducibility; three groups of epidemiologic unlinked strains of carbapenemase-producing clinical strains, diarrhea pathogen strains and foodborne strains were used to evaluate PFGE discriminatory power, 34 strains isolated during an outbreak were used to evaluate epidemiologic concordance and the subtyping consistency between PFGE and multilocus sequence typing (MLST) and 64 strains from different sources were used to evaluate the subtyping consistency of epidemiologic unlinked strains between PFGE and MLST. BioNumerics was used to develop the molecular typing database based on PFGE patterns. Results The PFGE asasay used in this study had high subtyping capability (98.18%), reproducibility (100%) and discriminatory power (D=0.9910, 0.9964, 1.0000 for three groups of epidemiologic unlinked strains respectively). For 34 strains isolated during an outbreak, PFGE identified the strains causing the outbreak and unlinked strains, and had good subtyping consistency with MLST. However, PFGE could not subtype and cluster the strains with same MLST types. Conclusion PFGE has good subtyping capability, reproducibility, discriminatory power, epidemiologic consistency in detecting K. pneumoniae strains; PFGE had good subtyping consistency with MLST for outbreak strains, but the consistency was poor between PFGE and MLST for epidemiologic unlinked strains.