LAKE BUENA VISTA, FLA. – Renal ultrasound in infants under 2 months of age with febrile urinary tract infection can be used to rule out high-grade vesicoureteral reflux, according to findings from a retrospective cross-sectional study.
This is because renal ultrasound has a high negative predictive value for detecting high-grade vesicoureteral reflux (VUR) in neonates, and although it has poor sensitivity for detecting low-grade VUR, its sensitivity for detecting high-grade VUR is quite good in this population.
The findings could have important implications for the management of neonates with febrile urinary tract infection (UTI), who are not included in American Academy of Pediatrics guidelines for febrile UTI. A 2011 update to those guidelines calls for the use of screening renal and bladder ultrasound, but recommends against routine voiding cystourethrogram (VCUG) after an initial febrile UTI in young children unless indicated by sonographic findings; however, these guidelines apply to children aged 2 months to 2 years only.
Younger infants who present with febrile UTI differ from those over 2 months of age in that they are more likely to be male – a high percentage of whom are uncircumcised and prone to UTI because of colonization, they have a more immature immune system that can also contribute to greater likelihood of UTI, and they can have higher rates of reflux and anatomical anomalies, Dr. Sowdhamini S. Wallace, director of pediatric hospital medicine research at Texas Children’s Hospital, Houston, reported at the Pediatric Hospital Medicine 2014 meeting.
Although many studies have included older children, they have limited applicability to younger infants because of these differences, and those studies that have included infants less than 2 months of age have been of small size or questionable quality, Dr. Wallace noted.
"So the objective of our study was to determine the test properties of renal ultrasound for detecting VUR and high-grade VUR and obstructive uropathies in this [younger] age group," she said at the meeting, which was sponsored by the Society of Hospital Medicine, the American Academy of Pediatrics, the AAP Section on Hospital Medicine, and the Academic Pediatric Association.
Of 200 eligible neonates with a mean age of 33 days, 30% had an abnormal renal ultrasound after presenting with febrile UTI. The most common reason for an abnormal finding was hydronephrosis.
Twenty-six percent of the neonates had reflux and 8% had high-grade reflux; 8% had obstructive uropathies or nonobstructive lower tract anomalies, Dr. Wallace said.
The sensitivity of renal ultrasound for all grades of VUR was 31% overall, but was 87% for high-grade VUR.
"For high-grade VUR, the negative predictive value was 99%, with a 95%-100% confidence interval. There were no obstructive uropathies that were diagnosed by VCUG in patients with a normal renal ultrasound," she said.
The number of patients with a normal ultrasound who would need to undergo VCUG to detect one case of high-grade VUR was 70; the number needed to test to detect one case of low-grade VUR was 4, she said.
The number needed to test "should be very helpful for physicians when they are deciding whether to get a VCUG on a neonate under 2 months if they have a normal renal ultrasound, Dr. Wallace said.
"Overall, I think you can see, with the number needed to test of 70, that you would have to test many babies with normal renal ultrasound to detect one case of high-grade VUR, so you may be able to spare many infants from VCUG," she concluded.
The infants presented to the emergency department during 2008-2011 with culture-proven UTI and fever of at least 100.4° F. They were identified through a microbiology database; those included in the study had urine collected through a catheterization or suprapubic aspiration, and those with a history of abnormal prenatal ultrasound, a previous diagnosis of genitourinary tract anomalies, or greater than 30 days between ultrasound and VCUG were excluded.
Imaging studies were reviewed independently by two radiologists who were blinded to the VCUG findings. Any discrepancies were resolved by a third radiologist.
Renal ultrasound was categorized as abnormal if it showed hydronephrosis and/or caliectasis, or if there was renal size discrepancy greater than 10%, findings of a duplicated collecting system, or urethral thickening, urethral dilatation, or bladder abnormalities.
VUR severity was determined by the standard classification system.
The study is limited by the fact that most renal ultrasounds were performed at the time of UTI diagnosis when inflammation is likely present. Inflammation may also be present in patients with VUR, thus the sensitivity of renal ultrasound in this study may have been higher than with renal ultrasound performed after a UTI has resolved. Also, in the absence of a standard definition for hydronephrosis in infants less than 2 months of age the prenatal parameter of 4 mL was used.