VCUG in infants with UTI unnecessary in presence of E. coli infection, normal ultrasound
The presence of Escherichia coli and normal renal ultrasound (US) findings may allow safe avoidance of voiding cystourethrography (VCUG), an invasive procedure associated with significant radiation exposure, in infants aged 0 to 3 months with a first febrile urinary tract infection (UTI), according to a recent study.
“Performing VCUG only in infants with UTI secondary to non-E. coli bacteria and/or abnormal US would save many unnecessary invasive procedures, limit radiation exposure, with a very low risk (<1 percent) of missing a high-grade vesicoureteral reflux (VUR),” said study coauthor Dr Hassib Chehade from the Lausanne University Hospital in Switzerland.
Chehade and colleagues developed a model to identify patients at low risk of high-grade VUR (grade ≥III) using bacterial urine culture results and renal US findings. Urine culture results were categorized as E. coli vs non-E. coli, whereas an abnormal US was defined by renal pelvic anterior–posterior diameter >5 mm and/or any dilation of the ureter or calyces.
In a cohort of 122 infants with a first febrile UTI (mean age at diagnosis 43 days; 79.5 percent male), non-E. coli urinary infection and abnormal renal US were significantly associated with high-grade VUR (p<0.001). [Arch Dis Child 2017;doi:10.1136/archdischild-2016-311587]
The respective sensitivity and specificity for predicting high-grade VUR were 75 and 77 percent in the presence of non-E. coli infection alone, and 58 and 86 percent in the presence of abnormal renal US alone.
Based on the Bayes theorem, the probability of detecting high-grade VUR was 26 percent in the presence of urinary non-E. coli bacteria, and the addition of abnormal US findings further improved the probability to 55 percent, with a diagnostic odds ratio (DOR) of 6.7 (95 percent CI, 1.2 to 33).
Moreover, adding abnormal US findings to E. coli infection improved the probability of detecting high-grade VUR from 3 to 15 percent, with a DOR of 13 (1.1 to 16.1).
On the other hand, the presence of a normal renal US decreases the probability of detecting a high-grade VUR to 1 percent in children with an E. coli UTI and to 15 percent in those with a non-E. coli UTI.
All infants in the cohort underwent VCUG, and none had known urinary tract malformations. A total of 12 had high-grade VUR, and seven of these had abnormal renal US. Majority (72 percent) had urine culture positive for E. coli, which was seen in 97 percent of infants with non-high grade VUR and in 3 percent with high-grade VUR. Non-E. coli was identified in 73 and 26 percent, respectively.
“This study showed that infants younger than 3 months of age with a first febrile UTI secondary to non-E. coli bacteria have a 10 times increased risk of high-grade VUR compared with infants with E. coli infection,” Chehade said.
“Based on these results, we propose a probabilistic diagnostic algorithm … based on the characterisation of bacterial species causing UTI and the renal US findings. In this algorithm, the presence of E. coli UTI associated with a normal renal US reduces the probability of high-grade VUR to <1 percent (1/22) and then allow to safely avoid unnecessary VCUG in this group,” he added.
Professor Peter Hoyer from the University Children’s Hospital Essen in Germany wrote in an accompanying editorial that the study by Chehade et al nudges the healthcare community to go in the direction of more patient-centred algorithms and avoidance of avoidable procedures. [Arch Dis Child 2017;doi:10.1136/archdischild-2017-313011]