Objectives: To assess the performance of the agar disc diffusion method for the detection of carbapenemase-producing Enterobacteriaceae (CPE) referred to the national reference laboratories (NRLs) in Belgium and France.
Methods: All Enterobacteriaceae isolates referred to the NRLs for the confirmation of CPE in 2012 were included. The inhibition zone diameters of meropenem, piperacillin/tazobactam and temocillin using CLSI disc diffusion methodology were recorded. Phenotypic and molecular detection of carbapenemases was performed on all isolates.
Results: A total of 1354 Enterobacteriaceae isolates, including 435 (32.1%) confirmed CPE isolates [OXA-48 (n = 323), KPC (n = 60), VIM (n = 32) and NDM (n = 20)] and 919 carbapenemase-negative isolates, were tested. Using recommended interpretative criteria, non-susceptibility to meropenem had poor sensitivity (52.0% by CLSI susceptibility breakpoint and 80.0% by EUCAST screening breakpoints), while non-susceptibility to piperacillin/tazobactam (according to CLSI breakpoint) or to temocillin (according to Fuchs, Barry, Thornsberry et al. Eur J Clin Microbiol 1985; 4: 30-3) was highly sensitive (99.8% and 98.2%, respectively) but poorly specific (29.4% and 42.9%, respectively) for the detection of CPE. Temocillin diameters <12 mm alone had high specificity (90.0%) and the combination of temocillin diameters ≥12 mm with piperacillin/tazobactam diameters ≥16 mm observed in 40% of all referred isolates displayed excellent negative predictive value (99.2%).
Conclusions: In geographical areas with a high prevalence of OXA-48 producers, recommended meropenem susceptibility or screening breakpoints failed to detect CPE in a large proportion of isolates. The combination of modified zone diameter cut-offs for piperacillin/tazobactam (≥16 mm) and temocillin (≥12 mm) can be used to rule out the presence of carbapenemase and avoid unnecessary additional testing for confirmation of CPE.
Keywords: OXA-48; breakpoints; meropenem.