Design: Systematic review and diagnostic accuracy meta-analysis
Population: 90 studies of children after first or recurrent urinary infection
Primary outcome: Sensitivity and specificity of renal and bladder ultrasound for reflux
Effect: First UTI: sensitivity 50% (42-58%), specificity 71% (62-78%); recurrent UTI sensitivity 26%

This systematic review and meta-analysis was one of the technical reports behind the new AAP guideline. It pooled 90 studies of renal and bladder ultrasound against an accepted reference standard for vesicoureteral reflux in children after a urinary infection.

After a first infection, ultrasound had a sensitivity of 50% (95% CI 42% to 58%) and specificity of 71% (62% to 78%). For high-grade reflux (grades III to V), sensitivity was 66% and specificity 77%. After recurrent infection, sensitivity fell to 26%. Certainty was low to moderate.

Ultrasound still has a role in finding structural abnormalities such as hydronephrosis, but it cannot be used to exclude reflux. A child with a normal scan and recurrent febrile infections may still have reflux.

Do not use a normal kidney and bladder ultrasound to reassure parents that reflux is absent.
Use ultrasound to look for structural abnormalities, not as a test for reflux.
Consider further evaluation for reflux after recurrent febrile urinary infections even when the ultrasound was normal.
Record ultrasound findings clearly as structural, so later clinicians do not read them as excluding reflux.

Why it matters

Ultrasound misses about half of reflux after a first infection and three-quarters after recurrent infection, so a normal scan can falsely reassure.

The statistics, in plain English

Sensitivity of 50% means ultrasound detected only half of children who had reflux. Specificity of 71% means about three in ten children without reflux had a scan suggesting it. Neither figure is good enough for a test used to rule reflux in or out.