Evaluation of the VITEK 2 Advanced Expert System performance for predicting resistance mechanisms in Enterobacterales acquired from a hospital-based screening program

Pathology. 2021 Oct;53(6):763-767. doi: 10.1016/j.pathol.2021.01.009. Epub 2021 May 3.

Abstract

There is limited literature examining the accuracy of the VITEK 2 Advanced Expert System (AES) in characterisation of β-lactamase resistance patterns. We present a prospective single centre study to better ascertain the performance characteristics of this program. The VITEK 2 AES interpretation was compared to established laboratory phenotypic methods. The overall sensitivity for detection of broad-spectrum β-lactamase by the AES was 95%, with a specificity of 78%. One or more discrepancies were noted in 36% of samples, with the majority of these (87/100) due to incorrect 'overcall' of a resistance mechanism. AES characterisation of AmpC resistance mechanisms was excellent. In contrast, the AES had poor specificity in classifying extended spectrum β-lactamases (ESBLs). As a screening aid, the AES can be a valuable tool. However, optimal use requires an adequate working knowledge of resistance mechanisms in order to correctly interpret and accept the result output.

Keywords: Advanced expert system; AmpC; ESBL; VITEK; beta-lactamase resistance.

MeSH terms

  • Anti-Bacterial Agents / pharmacology*
  • Bacterial Proteins / metabolism
  • Enterobacteriaceae / pathogenicity*
  • Escherichia coli / drug effects
  • Expert Systems*
  • Hospitals
  • Humans
  • beta-Lactam Resistance / drug effects
  • beta-Lactamases / metabolism*

Substances

  • Anti-Bacterial Agents
  • Bacterial Proteins
  • beta-Lactamases