Urinary tract infection (UTI) in children is defined by ≥ 5 × 10 colonies/mL in a catheterized urine specimen or, in older children, by repeated voided specimens with ≥ 10 colonies/mL. In younger children, UTIs are frequently associated with urinary tract anomalies. UTI may cause fever, failure to thrive, flank pain, and signs of sepsis, especially in young children. Treatment is with antibiotics. Follow-up imaging studies of the urinary tract are done if anomalies are suspected.
Urinary tract infection (UTI) may involve the kidneys, bladder, or both.
Mechanisms that maintain the normal sterility of the urinary tract include urine acidity and free flow, a normal emptying mechanism, intact ureterovesical and urethral sphincters, and immunologic and mucosal barriers. Abnormality of any of these mechanisms predisposes to UTI.
By age 7 years, 8% of girls and 2% of boys have had a UTI ( 1 ). The peak age of UTI is bimodal, with one peak in infancy and the other peak between 2 to 4 years of age (at the time of toilet training for many children).
The female:male ratio ranges from 1:1 to 1:4 in the first 2 months of life (estimates vary, likely because of different proportions of uncircumcised males in study groups and the exclusion of infants with urologic anomalies now more commonly diagnosed in utero by prenatal ultrasonography). The female:male ratio quickly rises with age to about 2:1 between 2 months to 1 year, 4:1 during the second year, and > 5:1 after 4 years.
In girls, infections usually are ascending and less often cause bacteremia. The marked female preponderance beyond infancy is attributed both to the shorter female urethra and male circumcision.
Predisposing factors in younger children include
Predisposing factors in older children include
Urinary tract infections in children are a marker of possible urinary tract congenital anomalies or other abnormalities (eg, obstruction, neurogenic bladder , ureteral duplication ); these abnormalities are particularly likely to result in recurrent infection if vesicoureteral reflux (VUR) is present. About 20 to 30% of infants and children age 12 to 36 months with UTI have VUR. The younger the child at the first UTI, the higher the likelihood of VUR. VUR is classified by grade (see table Grades of Vesicoureteral Reflux* ).
Recurrent UTI is clearly associated with VUR, especially VUR of higher grades. This association is likely due to 2 factors—that VUR predisposes to infection and recurrent infections can worsen VUR. The relative contribution of each factor in children with recurrent UTI is unclear. Children with more severe reflux may have higher risk of developing hypertension and renal failure (caused by repeated infection and chronic pyelonephritis), but the evidence is not definitive (see treatment of VUR ).
Grades of Vesicoureteral Reflux*
Only the ureters are involved but not the renal pelvis.
Reflux reaches the renal pelvis, but the calyces are not dilated.
The ureter and renal pelvis are dilated, with minimal or no blunting of calyces.
Dilation increases, and the sharp angle of the calyceal fornices is obliterated.
The ureter, pelvis, and calices are grossly dilated.
Papillary impressions frequently are absent.
* As defined by the International Reflux Study in Children.
Lebowitz RL, Olbing H, Parkkulainen KV, et al : International system of radiographic grading of vesicoureteric reflux. Pediatr Radiol 15(2):105-109, 1985. doi: 10.1007/BF02388714
Many organisms cause urinary tract infection in anatomically abnormal urinary tracts.
In relatively normal urinary tracts, the most common pathogens are
E. coli causes > 85 to 90% of UTIs in all pediatric age groups ( 2 ).
The remaining causes are other gram-negative enterobacteria, especially Klebsiella , Proteus mirabilis , and Pseudomonas aeruginosa . Enterococci and coagulase-negative staphylococci (eg, Staphylococcus saprophyticus ) are the most frequently implicated gram-positive organisms. Organisms such as Staphylococcus aureus and P. aeruginosa are more commonly associated with infections in children with CAKUT.
Fungi and mycobacteria are rare causes, occurring in immunocompromised hosts.
Adenoviruses rarely cause UTIs, and when they do, the disorder is predominantly hemorrhagic cystitis among immunocompromised hosts.
In neonates, symptoms and signs of UTI are nonspecific and include poor feeding, diarrhea, failure to thrive, vomiting, mild jaundice (which is usually direct bilirubin elevation), lethargy, fever, and hypothermia. Neonatal sepsis may develop.
Infants and children 2 years old with UTI may also present with poorly localizing signs, such as fever, gastrointestinal symptoms (eg, vomiting, diarrhea, abdominal pain), or foul-smelling urine. About 4 to 10% of febrile infants without localizing signs have UTI.
In children > 2 years old, the more classic picture of cystitis or pyelonephritis can occur. Symptoms of cystitis include dysuria, frequency, hematuria, urinary retention, suprapubic pain, urgency, pruritus, incontinence, foul-smelling urine, and enuresis. Symptoms of pyelonephritis include high fever, chills, and costovertebral pain and tenderness.
Physical findings suggesting associated urinary tract congenital anomalies include abdominal masses, enlarged kidneys, abnormal urethral orifice, and signs of lower spinal malformations. Diminished force of the urinary stream may be the only clue to obstruction or neurogenic bladder.
A reliable diagnosis of UTI requires the presence of pyuria on urinalysis and positive bacterial culture in properly collected urine, before an antimicrobial is given ( 1 ). A diagnosis of probable UTI may be made by the presence of pyuria on urinalysis, while culture results are pending. Most clinicians obtain urine by transurethral catheterization in infants and young children, reserving suprapubic aspiration of the bladder for boys with moderate to severe phimosis. Both procedures require technical expertise, but catheterization is less invasive, slightly safer, and has sensitivity of 95% and specificity of 99% compared with suprapubic aspiration. Bagged specimens are unreliable and should not be used for diagnosis.
Urine culture results are interpreted based on colony counts. If urine is obtained by catheterization, UTI commonly is defined by a count ≥ 1 × 10 4 colonies/mL (see also How To Catheterize the Bladder in a Female Child and How To Catheterize the Bladder in a Male Child ). If urine is obtained by suprapubic aspiration, a count 1 × 10 3 colonies/mL defines a UTI (see also How To Do Suprapubic Aspiration of the Bladder in a Child ). Clean-catch, midstream-voided specimens are significant when colony counts of a single pathogen (ie, not the total count of mixed flora) are ≥ 1 × 10 5 colonies/mL. However, at times symptomatic children may have UTI despite lower colony counts on urine cultures.
Urine should be examined by urinalysis and cultured as soon as possible or stored at 4 ° C if a delay of > 10 minutes is expected. Occasionally, UTI may be present despite colony counts lower than the described guidelines, possibly because of prior antibiotic therapy, very dilute urine (specific gravity < 1.003), or obstruction to the flow of grossly infected urine. Sterile cultures generally rule out UTI unless the child is receiving antibiotics or the urine is contaminated with antibacterial skin-cleaning agents.
Microscopic examination of urine is very useful but not definitive. Pyuria (defined as > 5 white blood cells (WBCs)/high-power field in centrifuged or spun urine sediment) is about 96% sensitive for UTI and 91% specific. Currently, many laboratories use automated instruments that report pyuria in WBC/high-power field using unspun urine, which seems to approximate counts based on centrifuged urine examined microscopically. A WBC count (using a hemocytometer) > 10/mcL (0.01 × 10 9 /L) in unspun urine has greater sensitivity (90%) but is not used by many laboratories. Presence of bacteria on urinalysis of spun or unspun fresh urine is about 80 to 90% sensitive but only 66% specific; Gram stain of the urine to detect the presence of bacteria is about 80% sensitive and 80% specific.
How to Catheterize the Urethra.Dipstick tests on urine to detect gram-negative bacteria (nitrite test) or WBC (leukocyte esterase test) are typically done together; if both are positive, the diagnostic sensitivity for UTI is about 93 to 97% and the specificity is about 72 to 93%. Sensitivity is lower for each individual test, especially for the nitrite test (about 50% sensitivity), because it may take several hours for bacterial metabolism to produce nitrites, and frequent voiding by children may preclude nitrite detection. The specificity of the nitrite test is quite high (about 98%); a positive result on a freshly voided specimen is highly predictive of UTI. Sensitivity of the leukocyte esterase test is 83 to 96% and specificity is 78 to 90%.
In a recent multicenter analysis of infants with fever, the aggregate urinalysis findings of pyuria, positive leukocyte esterase, or presence of nitrites was 94% sensitive and 91% specific for UTI; in the study population, this yielded a positive predictive value of 43% and a negative predictive value of 100% ( 2 ).
Differentiating an upper UTI from a lower UTI can be difficult. High fever, costovertebral angle tenderness, and gross pyuria with casts indicate pyelonephritis; an elevated C-reactive protein level or procalcitonin level also tends to be associated with pyelonephritis. However, many children without these symptoms and signs have an upper UTI. Tests to distinguish upper infection from lower infection are not indicated in most clinical settings because treatment is not altered.
A complete blood count and tests for inflammation (eg, erythrocyte sedimentation rate, C-reactive protein) may help diagnose infection in children with borderline urine findings.
Blood cultures are appropriate for infants with UTIs and for children > 1 to 2 years old who appear toxic.
Many major renal or urologic anomalies now are diagnosed in utero by routine prenatal ultrasonography, but a normal result does not completely exclude the possibility of anatomic anomalies. Thus, renal and bladder ultrasound imaging is typically done in children 3 ). Some clinicians do imaging on children up to 7 years of age or older.
Renal and bladder ultrasonography helps exclude obstruction, hydronephrosis, and CAKUT in children with febrile UTIs and is typically done within a week of diagnosing UTI in infants. Ultrasonography is done within 48 hours if infants do not respond quickly to antimicrobials or if their illness is unusually severe. Beyond infancy, ultrasonography may be done in the few weeks after the UTI diagnosis.
Voiding cystourethrography (VCUG) and radionuclide cystography (RNC) are better than ultrasonography for detecting VUR and anatomic anomalies and previously were recommended for most children after a first UTI. However, VCUG and RNC both involve use of radiation and are more uncomfortable than ultrasonography. Also, the role VUR plays in the development of chronic renal disease is undergoing reevaluation, making the immediate diagnosis of VUR less urgent. Thus, VCUG is no longer routinely recommended after the first UTI in children, especially if ultrasonography is normal and if children respond quickly to antibiotic therapy. VCUG is reserved for children with the following:
If VCUG is to be done, it is done at the earliest convenient time after clinical response, typically toward the end of therapy, when bladder reactivity has resolved and urine sterility has been regained. If imaging is not scheduled until after therapy is due to be completed, children should continue antibiotics at prophylactic doses until VUR is excluded.
Radionuclide scanning is now used mainly to detect renal scarring. It is done using technetium-99m-labeled dimercaptosuccinic acid (DMSA), which produces images of the renal parenchyma. DMSA scanning is not a routine test, but it may be done if children have risk factors, such as abnormal ultrasound results, high fever, and organisms other than E. coli , or recurrent febrile UTIs ( 1 , 3 ).
Treatment of UTI is aimed at eliminating the acute infection, preventing urosepsis, and preserving renal parenchymal function. Antibiotics are begun presumptively in all toxic-appearing children and in nontoxic children with a probable UTI (positive leukocyte esterase, nitrites, or pyuria). Others can await the results of the urine culture, which are important for both confirming the diagnosis of UTI and yielding antimicrobial susceptibility results ( 1 ).
Pseudomonas .
> 1 year with complicated UTI caused by multidrug-resistant E. coli , P. aeruginosa , or other gram-negative bacteria, a fluoroquinolone) selected on the basis of antimicrobial sensitivities can be used to complete a 7- to 10-day course. A poor clinical response (eg, fever persisting > 72 hours) suggests a resistant organism or an obstructive lesion and warrants urgent evaluation with ultrasonography and repeat urine culture.
It has long been thought that antibiotic prophylaxis reduces UTI recurrences and prevents kidney damage and should be started after a first or second febrile UTI in children with VUR. However, this conclusion was not based on long-term, placebo-controlled trials (important because it has been observed that much VUR abates with time as children mature). A large, controlled trial, the Randomized Intervention for Children with Vesicoureteral Reflux (RIVUR) trial ( 2 ), did show that antibiotic prophylaxis using SMX-TMP reduced UTI recurrences by 50% (from about 25% to 13%) compared to placebo but did not show a difference in the rate of renal scarring at 2 years (8% in each group). Also, the children in the RIVUR trial who did develop UTI while taking prophylactic antibiotics were more likely to be infected with resistant organisms, but this likelihood decreased over time. Additional studies may show that antibiotic prophylaxis does provide some renal protection but at the risk of more antibiotic-resistant infections. Thus, the optimal strategy remains somewhat uncertain. Prophylaxis seems to be of greatest benefit to children with VUR and bowel-bladder dysfunction and perhaps also to those with the highest risk of scarring (eg, those with recurrent febrile UTIs).
For children with grade IV or V VUR, open repair or endoscopic injection of polymeric bulking agents is usually recommended, often along with antibiotic prophylaxis until repair is completed. For children with lesser grades of VUR, further research is required. Because renal complications are probably unlikely after only 1 or 2 UTIs, pending further research one acceptable strategy may be to closely monitor children for UTIs, treat them as they occur, and then reconsider antimicrobial prophylaxis in those children with recurrent infections (especially for those with recurrent febrile UTIs, more severe VUR, or bowel or bladder dysfunction).
3 ), usually given at bedtime.
Properly managed children rarely progress to renal failure unless they have uncorrectable urinary tract abnormalities. However, repeated infection, particularly in the presence of VUR, is thought (but not proved) to cause renal scarring, which may lead to hypertension and end-stage renal disease. In children with high-grade VUR, long-term scarring is detected at a 4- to 6-fold greater rate than in children with low-grade VUR and at an 8- to 10-fold greater rate than in children without VUR. The risk of scarring after recurrent UTI (≥ 2 febrile episodes) is as high as 25%, or 10- to 15-fold greater than that in children with only 1 febrile UTI; however, few children will have recurrent febrile UTI.