Objective: Clinical evaluation of a rapid screening filter test for urinary tract infection (UTI), FiltraCheck-UTI, comparison to the urine dipstick and conventional urinalysis for test performance and cost effectiveness in children.
Setting: Pediatric emergency department at an urban children's hospital.
Methods: Cross-sectional concordance study of 1298 children age 2 days to 19 years (50% < or = 2 years) for whom a urine culture was ordered; screening tests run by trained laboratory personnel; cost per case detected calculated; retrospective chart review for clinical information.
Results: Prevalence of UTI was 7.1%. Urine obtained from children < or = 2 years by catheter (97%) as part of an evaluation of fever or sepsis (82%). FiltraCheck-UTI was comparable with microscopy for bacteriuria (P = 0.11), sensitivity of 85% (95% confidence interval, 76 to 91) and specificity of 72% (95% confidence interval, 70 to 75%) but it was difficult detect Gram-positive organisms by this method (P < 0.001). Its performance varied by definition of UTI. The urine dipstick had the best specificity and was the most cost-effective rapid test.
Conclusions: FiltraCheck-UTI is more expensive and has more false positives than the urine dipstick in detecting UTI in children. The dipstick continues to be the best inexpensive alternative to microscopy, but it is probably not an adequate screen for when to send a urine culture in young children.