Antibiotic susceptibility of common pediatric uropathogens in the United States.

TO THE EDITOR—Urinary tract infections (UTIs) remain a common indication for antibiotic therapy in children [1]. They are frequently managed in the outpatient setting where availability of susceptibility results may be delayed, making selection of appropriate initial therapy important. We sought to evaluate susceptibility patterns of antibiotic agents used to treat UTIs in children by developing a pooled antibiogram of urinary isolates from pediatric hospitals across the United States. We obtained 2012 and 2013 antibiotic susceptibility data for pediatric urinary isolates using methods previously described [2]. Susceptibility data were collected separately for institutions that have and have not incorporated the Clinical and Laboratory Standards Institute (CLSI) recommendations to lower cefazolin breakpoints from ≤8 μg/mL to ≤2 μg/mL against Enterobacteriaceae [3]. Comparisons were made using χ analysis. Data were obtained from 43 hospitals (Table 1). The addition of clavulanate to amoxicillin increased activity against Escherichia coli, the most common uropathogen isolated, from 49% to 75%. Ceftriaxone had the highest activity against E. coli at 97%. Cefazolin was active against 90% of E. coli isolates using a breakpoint of 8 μg/mL and 47% of isolates for institutions using a breakpoint of 2 μg/mL (P < .001). Cephalothin, another first-generation cephalosporin, had activity against 56% of E. coli isolates. Third-generation cephalosporins continue to have excellent activity against common gram-negative pediatric uropathogens, but because they can accelerate the development of extended-spectrum β-lactamases [4], other, narrowerspectrum agents should be considered whenever possible. Approximately 90% of E. coli are susceptible to cefazolin (or cephalexin) when a breakpoint of 8 μg/ mL is used, decreasing to <50% with a breakpoint of 2 μg/mL. The cefazolin breakpoint changes were mainly influenced by adult pharmacokinetic/ pharmacodynamic simulation studies, without special consideration for urinary isolates [3]. In the absence of clinical data to support the revised cefazolin breakpoints, our findings underscore the need to reexamine their rationale. Cephalothin activity is currently used as a proxy for susceptibility to other cephalosporins by 21% of hospitals, although recently discouraged by the CLSI [5]. Our results suggest that cephalothin is not a reliable predictor of cephalexin susceptibility. This is important as institutions may consider cephalexin a suboptimal choice based on cephalothin results and resort to prescribing increasingly broad-spectrum antibiotics [5]. Widespread amoxicillin resistance in E. coli, with minimal improvement with the addition of clavulanate, undermines its effectiveness as an empiric agent. Similarly, caution should be used when trimethoprim-sulfamethoxazole is prescribed on an empiric basis as it has activity against 68% of E. coli isolates. Although nitrofurantoin has activity against 95% of E. coli isolates, its poor renal parenchymal penetration precludes its use for pyelonephritis [1]. We were unable to separate urinary isolates obtained from relatively healthy