Interval likelihood ratios of urinalysis components can help predict the probability of a true urinary tract infection (UTI) in young children, new research suggests.
“Our findings can help clinicians use the levels of inflammation or bacteria of each urinalysis component to more accurately predict a positive urine culture in young children,” first author Dr. Tian Liang of NYC Health + Hospitals/Kings County, in Brooklyn, told Reuters Health by email.
“More accurately diagnosing UTIs can improve appropriate antibiotic use, and potentially lead to decreased medication side effects, antibiotic resistance, and ultimately healthcare costs,” he added.
“Although urinalysis components often contain multiple result values (negative, trace, 1+, 2+, 3+), their likelihood ratios in predicting UTI were historically calculated as dichotomized variables, only indicating the presence or absence of each component,” Dr. Liang and colleagues explain in Pediatrics.
They calculated the interval likelihood ratio (ILR) of urinalysis components to estimate the probability of UTI in children younger than age 2 years presenting to their pediatric emergency department. They had 2,144 urinalysis and urine culture samples from 1,954 individual patients.
The UTI prevalence was 9.2 percent and the most common pathogen was Escherichia coli (75.2 percent).
Using this prevalence as the pretest probability, the researchers calculated the ILR of leukocyte esterase, hemoglobin, protein and nitrite in the urinalysis dipstick.
They found that the probability of UTI significantly increases with 3+ leukocyte esterase, positive nitrite results, 3+ protein, 20 white blood cells (WBCs) per high-power field (HPF) or greater, or “many” bacteria in the urinalysis.
The probability of UTI only “marginally” increases with trace or 1+ leukocyte esterase or 5 to 20 WBCs per HPF.
“Our findings can be used in conjunction with existing tools for evaluating UTIs in young children to more accurately diagnose a true infection,” Dr. Liang told Reuters Health.
“There is minimal time and cost added of using our calculations to help diagnose UTIs and can be used in conjunction with existing tools,” he added.
The research had no specific funding and the authors report no relevant conflicts of interest.