ᰔAmpC ␤-lactamases can interfere with extended-spectrum-␤-lactamase (ESBL) confirmatory tests. Resulting failures to detect ESBLs can endanger patients because false susceptibility to cephalosporins may be reported (5). This problem occurs with CLSI and some other ESBL confirmatory tests, but there are tests that can be used to provide more accurate detection of ESBLs in AmpC-producing isolates (4, 5). While inaccurate susceptibility reports are never acceptable, it is of interest to know whether it is common for a lab to encounter ESBL-screen-positive, AmpC-producing isolates that yield negative CLSI confirmatory tests. That is, can it be common to encounter negative ESBL confirmatory test results of unknown accuracy? Apart from a small Escherichia coli study in which 20 of 26 cefpodoxime ESBLscreen-positive, cefoxitin-resistant E. coli isolates produced a plasmid-mediated AmpC ␤-lactamase (4), little is known about how often AmpC production creates uncertainty about the accuracy of CLSI ESBL confirmatory tests. Therefore, a study was conducted at a Louisville, KY, teaching hospital to determine how many ESBL-screen-positive isolates of E. coli, Klebsiella pneumoniae, Klebsiella oxytoca, and Proteus mirabilis were AmpC positive and ESBL confirmatory test negative by CLSI methodology.During the period of 1 May 2006 to 28 February 2007, 952 isolates from patients at the University of Louisville Hospital, Louisville, KY, were analyzed. They comprised 682 E. coli, 152 K. pneumoniae, 44 K. oxytoca, and 70 P. mirabilis isolates. ESBL screening was performed by using CLSI microdilution methodology using cefotaxime, ceftriaxone, ceftazidime, cefpodoxime, and aztreonam with positive isolates then tested by the CLSI ESBL disk confirmatory method (2). Screen-positive isolates with negative ESBL confirmatory tests were investigated for AmpC production using a Tris-EDTA-based disk test (1) to determine if they were AmpC producers. If this test was positive, it meant that the negative ESBL confirmatory result was of unknown accuracy. Representative AmpC-positive isolates of K. pneumoniae were investigated for gene identification by PCR using primers specific for bla DHA , bla FOX , bla CMY , bla ENT , bla ACT , and bla ACC (3).Only 13 of the 57 screen-positive isolates (22.8%) were ESBL producers by CLSI methodology, while 40 were AmpC producers (70.2%) ( Table 1). Other resistance mechanisms were assumed to account for the other four positive screens. The screening-positive isolates comprised 37 E. coli, 16 K. pneumoniae, 3 Klebsiella oxytoca, and 1 P. mirabilis. Of these, 8 E. coli isolates (22% of screen-positive E. coli), 4 K. pneumoniae isolates (25% of screen-positive K. pneumoniae), and 1 P. mirabilis isolate were confirmed as ESBL producers. Of the isolates that gave an ESBL-negative result, 28 E. coli (76% of screen-positive isolates) and 12 K. pneumoniae (75% of screen-positive isolates) were AmpC positive. No K. oxytoca or P. mirabilis isolates were AmpC positive. The AmpC-positive K. pneumoniae isolates were assu...

No abstract