Three commercially available microbiology identification and susceptibility testing systems were compared with regard to their ability to detect extended-spectrum beta-lactamase (ESBL) production in Enterobacteriaceae, i.e., the Phoenix Automated Microbiology System (BD Diagnostic Systems, Sparks, MD), the VITEK 2 System (bioMérieux, Marcy l'Etoile, France), and the MicroScan WalkAway-96 System (Dade Behring, Inc., West Sacramento, CA), using routine testing panels. One hundred fifty putative ESBL producers were distributed blindly to three participating laboratories. Conventional phenotypic confirmatory tests such as the disk approximation method, the CLSI double-disk synergy test, and the Etest ESBL were also evaluated. Biochemical and molecular characterization of beta-lactamases performed at an independent laboratory was used as the reference method. One hundred forty-seven isolates of Escherichia coli, Klebsiella pneumoniae, Klebsiella oxytoca, Enterobacter cloacae, Enterobacter aerogenes, Citrobacter freundii, Serratia marcescens, Proteus mirabilis, Proteus vulgaris, and Morganella morganii were investigated. Of these isolates, 85 were identified as ESBL producers by the reference method. The remaining isolates were identified as non-ESBL producers; they were either hyperproducers of their chromosomal AmpC, Koxy, or SHV enzymes or lacked any detectable beta-lactamase activity. The system with the highest sensitivity for the detection of ESBLs was the Phoenix (99%), followed by the VITEK 2 (86%) and the MicroScan (84%); however, specificity was more variable, ranging from 52% (Phoenix) to 78% (VITEK 2). The performance of the semiautomated systems differed widely with the species investigated. The sensitivities of the conventional test methods ranged from 93 to 94%. The double-disk synergy test showed the highest specificity and positive predictive value among all test methods, i.e., 97% and 98%, respectively.
