105 - Accuracy of urine WBC count compared to urine dipstick for the diagnosis of urinary tract infections in febrile infants 2 to 6 months of age

Poster Number: 105
Publication Number: 105.1506

Background: Urinary tract infection (UTI) is the most common invasive bacterial infection of infancy. Urine dipsticks can be performed as point-of-care while the urinalysis requires formal laboratory testing. Clinicians make initial treatment decisions based on results of urine screening tests, as cultures take several days to yield results.

Objective: To compare the accuracy of urine white blood cell (WBC) count to urine dipstick for the diagnosis of UTI in febrile infants 2 to 6 months of age.

Design/Methods: We assembled a multi-center retrospective cohort of infants 2 to 6 months of age who presented to one of five emergency departments with a recorded temperature of ≥ 38.0°C and had a urine culture obtained. Our primary outcome was a UTI defined by a urine culture with ≥ 50,000 colony-forming units/mL of a single bacterial pathogen. We used receiver operator curve analysis with the Liu method to select the optimal urine WBC cut-point to identify UTI in infants with dilute [specific gravity (SG) < 1.015] and concentrated (SG ≥ 1.015) urine. We then compared the sensitivity and specificity of positive urine dipstick (≥ 1+ leukocyte esterase or positive nitrite) to positive urine WBC count for UTI.

Results: Of 9,387 febrile infants with urine culture, 1,034 (11.0%) had a UTI. Escherichia coli was the most common pathogen identified (916, 88.6%). Overall, 9,104 (97.0%) had a urine dipstick, 5,111 (54.4%) urine WBC count, and 5,106 (54.4%) both. The optimal urine WBC cut-point was ≥ 7 cells/high powered field (hpf) in dilute urine and ≥ 12 cells/hpf in concentrated urine. The dipstick had a higher sensitivity (813/877, 92.7%) than urine WBC count [727/877, 82.9%; difference 9.8%, 95% confidence interval (CI) 6.78 to 12.9%], but similar specificity (dipstick: 6231/6865, 90.7% vs. urine WBC count: 3804/4234, 89.8%; difference 0.9%, 95% CI -0.2 to 2.1%). Of the 64 infants with UTI and false negative dipstick (Negative Predictive Value 99.0%, 95% CI 98.7 to 99.2%), 58 (92%) also had normal urine WBC count (Table).

Conclusion(s): Urine dipstick is an accurate screening test for identifying UTI in febrile infants 2 to 6 months of age. Urine WBC count may not be required to make empiric antibiotic treatment decisions.