UTI is common among patients at Veterans Affairs Medical Centers (VAMCs), many of whom are elderly men with underlying medical or urological problems. Most UTI treatment guidelines address uncomplicated UTI in women and presume knowledge of local uropathogen susceptibility patterns, which often are unknown or are inferred from E. coli. To inform selection of empiric therapy for UTI at our VAMC, we compiled susceptibility data for one year's urine isolates.
Susceptibility results (from a bioMerieux VITEK® instrument) were compiled for the 2,109 significant urine isolates in the Minneapolis VAMC clinical microbiology laboratory from June 1, 2013 to May 31, 2014. For "drug-bug" combinations that were not tested directly we imputed results based on local or published data, and/or expert opinion (40% of 52,725 data points). Cumulative % susceptible was then calculated for 25 relevant agents, both overall and stratified by Gram stain group. Given the uncertainty, susceptibility of Enterococcus spp. (ENTC) to trimethoprim/sulfamethoxazole (SXT) was analyzed as both 0% and 100%.
Of 2,109 urine isolates, 561 were Gram-positive and 1,548 Gram-negative. E. coli predominated (32%), followed by ENTC (18%), Klebsiella (15%), Staphylococcus (9%), and Pseudomonas (8%). The most active oral agents (as % susceptible) were SXT (79%, if ENTC susceptible; 60%, if resistant), fluoroquinolones (FQs, 68%), and nitrofurantoin (NF, 64%). The most active intravenous agents were imipenem (93%), piperacillin/tazobactam (91%), aminoglycosides (77-81%), cefepime (74%), ertapenem (69%), and 3rd-gen. cephalosporins (65-69%). Antibiotic rank order by % susceptible differed greatly for Gram-positives vs. -negatives, and for E. coli vs. other Gram-negatives (not shown).
Among urine isolates from veterans, no oral agent provided ≥ 80% overall susceptibility; SXT either led or trailed FQs and NF, depending on its presumed activity against ENTC; and FQs and NF were closely matched. E. coli was a minor contributor and represented poorly the total population. Data such as these could improve selection of empirical UTI therapy for veterans, as could urine Gram stains.
K. J. Marsh,
L. Mundy, None
J. R. Johnson, Crucell: Consultant , Consulting fee
ICET: Consultant , Consulting fee
Merk: Investigator , Grant recipient
Tetraphase: Investigator , Grant recipient