• Kenneth Frumkin.
• Emergency Medicine Department, Naval Medical Center Portsmouth, Portsmouth, Virginia.
• J Emerg Med. 2015 Apr 1;48(4):405-15.

BackgroundBecause urinary tract infection (UTI) is the most frequent source of serious bacterial infections in young children, we studied the bacteriology of such infections in our institution.ObjectivesPediatric urine cultures were reviewed for age- and sex-specific differences in testing, prevalence, causative organisms, and antibiotic sensitivities.MethodsA retrospective 5-year observational study of all urine cultures from Emergency Department patients aged 0-36 months found 4403 cultures. Primary outcomes were numbers of cultures obtained, infections found, the organisms isolated, and their antibiotic sensitivities.ResultsBoys were cultured much less frequently (overall 4.7% vs. 19.4% of girls, p < 0.001). Three hundred ninety-six cultures yielded infection episodes. Although far fewer boys were cultured, their overall positive culture rate (10.9%) was significantly higher than for girls (8.4%; p = 0.01), with more Gram-positive organisms (52.3%; vs. 18.6% Gram-positives in girls; p < 0.001). The rate of positive cultures in boys and girls remained 8.4% and 10.3%, respectively, even in 2-year-olds. Trimethoprim-sulfamethoxazole and amoxicillin/clavulanic acid were effective against all isolates.ConclusionsAge- and sex-based assumptions guiding evaluation for and treatment of UTIs in young children should be reevaluated. We may not be culturing enough young boys, risking missed UTIs with potential for renal injury in this vulnerable group. Based on their significant rate of Gram-positive infections, those boys we treat empirically might benefit more from trimethoprim/sulfamethoxazole or amoxicillin/clavulanic acid than from third-generation cephalosporins. The persistence of positive cultures in 2-year-olds suggests we should be culturing beyond 24 months in both sexes. Urine Gram stains should be more frequently considered.Published by Elsevier Inc.

30 keywords...

hide…