The prevalence and molecular types of extended-spectrum beta-lactamases (ESBLs) were determined during a 1-year period in unselected clinical nonduplicate isolates of Escherichia coli (n = 1,738), Klebsiella pneumoniae (n = 436), and Klebsiella oxytoca (n = 208), cultured at the University Medical Centre Nijmegen, The Netherlands. Isolates identified as ESBL producer by the Phoenix automated system were collected prospectively and subjected to molecular analysis for the most common ESBLs TEM, SHV, and CTX-M, as well as OXA and GES. Both the Etest ESBL and double-disk synergy test were performed as confirmatory tests. The estimated prevalence of ESBLs was 2.1% in E. coli, 5.2% in K. pneumoniae, and 2.4% in K. oxytoca. TEM-12 and -26, SHV-5 and -12, and CTX-M groups 1 and 9 were the most frequent ESBLs found. Isolates identified as ESBLs by the Phoenix were confirmed by polymerase chain reaction (PCR) in only 42%. In ESBL PCR-positive E. coli and K. pneumoniae, both confirmatory tests were positive in 95% of the isolates. In 28% of the Etest and 13% of the double-disk synergy test-positive isolates, PCR could not detect any ESBL gene. In these cases, other resistance mechanisms may play a role. Confirmatory tests were unreliable for K. oxytoca. A previously described mutation in the K1 enzyme was detected in one ceftazidime-resistant K. oxytoca. The prevalence of ESBLs in The Netherlands is increasing. The predominant molecular types of ESBLs detected were comparable to other studies. Phoenix ESBL results need to be confirmed as advocated by ESBL detection guidelines.

A rectal culture-guided antimicrobial prophylaxis strategy may prevent infections after transrectal ultrasound-guided prostate biopsy (TRUSP). The use of selective culture media could assist the choice of appropriate antibiotic prophylaxis. The objective of our study was to evaluate the performance of four selective media used for guidance of oral antibiotic prophylaxis in TRUSP. In this prospective validation study, we used MacConkey media with vancomycin plus one of the following antibiotics: ciprofloxacin (McC3+CIP/V), trimethoprim (McC3+TMP/V), fosfomycin (McC3+FOF/V), and amdinocillin-amoxicillin-clavulanic acid (McC3+MEC/V). First, clinical strains of Gram-negative bacilli (GNB) ( = 33) were evaluated for growth on the selective media. Thereafter, rectal swabs ( = 97) were randomly collected from residual material of fresh stool samples and plated on a growth control and the four selective media. Levels of recovery of GNB on the growth control and selective media were compared, and the MICs of the antibiotics used in this study were determined. The sensitivity and specificity of the four selective media amounted, respectively, to 90.0% (55.5 to 99.8%) and 98.7% (93.1 to 100.0%) for McC3+CIP/V, 95.7% (85.2 to 99.5%) and 100.0% (91.6 to 100.0%) for McC3+TMP/V, 95.5% (84.5 to 99.4%) and 97.8% (88.2 to 99.9%) for McC3+FOF/V, and 100.0% (76.8 to 100.0%) and 97.6% (87.4 to 99.9%) for McC3+MEC/V. In conclusion, the four selective media were sufficiently sensitive and specific for the identification of rectal GNB resistant to ciprofloxacin, trimethoprim, fosfomycin, or amdinocillin-amoxicillin-clavulanic acid. These media can have added value in streamlining the optimal culture based antibiotic prophylaxis in TRUSP in a non-labor-intensive manner.