Amoxicillinclavulanate and piperacillintazobactam: comparison of different susceptibility testing methods

Abstract number: P582

Objectives: To analyse the susceptibility of clinical Escherichia coli isolates against amoxicillin–clavulanate and piperacillin–tazobactam and compare the performance of three susceptibility testing methods.

Methods: A total of 179 clinical Escherichia coli isolates were included. Species identification was performed with Vitek2, and susceptibility testing was performed with disc diffusion, Etest and Vitek2. Etest-results were considered the gold standard. EUCAST breakpoints were used for both antibiotics when analyzing the MIC and Vitek2 data. Disc diffusion breakpoints for piperacillin–tazobactam were derived from the Swedish Reference Group for Antibiotics. The results from the three methods were compared using Pearson correlation coefficient (r). Discrepancies between the methods were describes as minor, major and very major.

Results: The correlation for MICs for amoxicillin–clavulanate by Etest and Vitek2 was relatively high (r = 0.82). Seventeen major (9.49%) and one very major (0.56%) errors were observed with Vitek2. In contrast there was poor correlation for the MICs for piperacillin/tazobactam by Etest and Vitek2 (r = 0.37). As many as forty-seven (26.26%) major and seven (3.91%) minor discrepancies were observed between the two methods. In all major discrepancies observed in testing both antibiotics, Vitek2 result was resistant whereas the Etest result was susceptible. When disc diffusion method is considered, the breakpoint for amoxicillin–clavulanate was estimated at 18 mm in the wild type E. coli population studied. Results from disc diffusion tests were significantly correlated with Etest for amoxicillin–clavulanate (r = 0.79). There were twenty-three (12.85%) major discrepancies between the two methods. In contrast, relatively poor correlation between Etest and disc diffusion results was observed when piperacillin/tazobactam tested (r = 0.49). There were twenty-eight (15.64%) minor discrepancies between the two methods.

Conclusion: The susceptibility testing results for amoxicillin–clavulanate obtained by three methods were highly correlated. However there was poor correlation in piperacillin/tazobactam testing in the studied material. The discrepancies observed for susceptibility testing of piperacillin/tazobactam might be clinically important. Especially the major errors observed with Vitek2 are of potential great concern.

Session Details

Date: 10/04/2010
Time: 00:00-00:00
Session name: Abstracts 20th European Congress of Clinical Microbiology and Infectious Diseases
Location: Vienna, Austria, 10 - 13 April 2010
Presentation type:
Back to top