Purpose

We evaluated a modification of automated antibiograms in urine cultures designed to facilitate the early interpretation of minimum inhibitory concentrations (MICs) and accelerate the targeted treatment of urinary tract infections (UTIs), METHODS: A prospective study was conducted of 309 isolates (219 Enterobacteriaceae, 75 Enterococcus spp., and 15 non-fermenting Gram-negative bacilli (NFGNB), and a retrospective study of 9 carbapenemase-producing clinical isolates from urine cultures. Colonies grown on conventional isolation plates were inoculated in MicroScan Walkaway system panels and incubated for 7 h, using a MicroScan AutoScan-4 plate reader for preliminary MIC determination by turbidimetry. Resulting antibiograms were compared with definitive antibiograms obtained after incubation for 17 h.

Results

Preliminary and definitive readings were concordant for 86.7% of Gram-positive cocci isolates (65/75), 61.6% of Enterobacteriaceae (135/219), and 53.3% of NFGNB. The agreement rate was greater than 90% for most antimicrobials against Gram-positive cocci (94.7% or more) and Enterobacteriaceae, (97.2% or more for 10 of 17 antibiotics) except with nitrofurantoin (89%). The agreement rate was 86.7% or more for most antibiotics against NFGNB apart from piperacillin/tazobactam, aztreonam, amikacin, and ciprofloxacin. Gram-negative bacilli showed the highest differences in MIC values between preliminary and definitive readings.

Conclusions

A preliminary antibiogram reading may be useful in urine cultures to reduce the delay before targeted antibiotherapy, especially against Enterobacteriaceae and Gram-positive cocci, but not in cases of carbapenemase-producing NFGNB. Further local studies are warranted to evaluate the usefulness of this approach in relation to resistance rates.