Evaluation of the NCCLS extended-spectrum beta-lactamase confirmation methods for Escherichia coli with isolates collected during Project ICARE

To determine whether confirmatory tests for extended-spectrum beta-lactamase (ESBL) production in Escherichia coli are necessary, we selected 131 E. coli isolates that met the National Committee for Clinical Laboratory Standards (NCCLS) screening criteria for potential ESBL production from the Project ICARE (Intensive Care Antimicrobial Resistance Epidemiology) strain collection. For all 131 isolates, the broth microdilution (BMD) MIC of at least one extended-spectrum cephalosporin was >/=2 micro g/ml. For 21 of 131 (16%) isolates, the ESBL confirmatory test was positive; i.e., the BMD MICs of ceftazidime or cefotaxime decreased by >/=3 doubling dilutions in the presence of clavulanic acid (CA) or the disk diffusion zone diameters increased by >/=5 mm around ceftazidime or cefotaxime disks in the presence of CA. All 21 isolates were shown by PCR to contain at least one of the genes bla(TEM), bla(SHV), and bla(OXA), and in isoelectric focusing (IEF) tests, all isolates demonstrated at least one beta-lactamase band consistent with a TEM, SHV, or OXA enzyme. Of the 21 isolates, 3 showed a CA effect for cefotaxime by BMD but not by disk diffusion testing. A total of 59 (45%) of the 131 isolates demonstrated decreased susceptibility to cefpodoxime alone (MIC = 2 to 4 micro g/ml), and none had a positive ESBL confirmatory test result. These were classified as false positives according to ESBL screen test results. For the remaining 51 (39%) isolates, the cefpodoxime MICs ranged from 16 to >128 micro g/ml and the MICs for the other extended-spectrum cephalosporins were highly variable. All 51 isolates gave negative ESBL confirmatory test results. Most showed IEF profiles consistent with production of both a TEM and an AmpC beta-lactamase, and representative isolates of several phenotypic groups showed changes in porin profiles; these 51 isolates were considered true negatives. In all, only 16% of 131 E. coli isolates identified as potential ESBL producers by the current NCCLS screening criteria were confirmed as ESBL producers. Thus, changing the interpretation of extended-spectrum cephalosporins and aztreonam results from the susceptible to the resistant category without confirming the presence of an ESBL phenotype would lead to a large percentage of false resistance results and is not recommended. However, by increasing the cefpodoxime MIC screening breakpoint to >/=8 micro g/ml, 45% of the false-positive results could be eliminated. NCCLS has incorporated this change in the cefpodoxime screening breakpoint in its recent documents.