Am Fam Physician. 2005;72(12):2483-2488
A more recent article on UTIs in young children and infants is available.
Author disclosure: Dr. Alper is editor-in-chief of DynaMed.
Up to 7 percent of girls and 2 percent of boys will have a symptomatic, culture-confirmed urinary tract infection by six years of age. Urinary tract infection may be suspected because of urinary symptoms in older children or because of fever, nonspecific symptoms, or failure to thrive in infants. Urine dipstick analysis is useful for ruling out urinary tract infections in cases with low clinical suspicion. However, urine culture is necessary for diagnosis of urinary tract infections in children if there is high clinical suspicion, cloudy urine, or if urine dipstick testing shows positive leukocyte esterase or nitrite activity. Despite current recommendations, routine imaging studies (e.g., renal ultrasonography, voiding cystourethrography, renal scans) do not appear to improve clinical outcomes in uncomplicated urinary tract infections. Oral antibiotics are as effective as parenteral therapy in randomized trials. The optimal duration of antibiotic therapy has not been established, but one-day therapies have been shown to be inferior to longer treatment courses.
Until recently, the management of urinary tract infection (UTI) in children has been controversial and based mainly on opinion, but new evidence regarding imaging studies and treatment prompted this review. Although symptomatic UTI and asymptomatic bacteriuria (i.e., positive urine cultures without symptoms or signs of illness) are discussed in the literature, this review focuses on symptomatic UTI.
|Clinical recommendation||Evidence rating||References|
|Urine culture should be obtained for diagnosis of UTI in children if there is high clinical suspicion, cloudy urine, or positive urine dipstick.||C||13–16|
|Routine imaging studies (e.g., ultrasonography, VCUG, renal scans) do not appear to improve outcomes in children with a first uncomplicated UTI, but the evidence is weak.||B||24, 26, 27|
|Oral antibiotics should be used (when tolerated) instead of parenteral antibiotics to manage UTI in children.||A||35, 36|
|One-day courses of antibiotics should not be used to manage UTI in children.||A||37|
|Consider that short courses of antibiotics (two to five days) may be as effective as longer courses (seven to 14 days).||B||37–39|
|Prophylactic antibiotics may be used to reduce the risk of recurrent UTI.||B||42, 43|
Most UTIs in children result from ascending infections, although hematogenous spread may be more common in the first 12 weeks of life. Most UTIs in children are monomicrobic, often caused by Escherichia coli (60 to 80 percent of cases), Proteus (more common in boys and in children with renal stones), Klebsiella, Enterococcus, and coagulase-negative staphylococci.3
Evidence on risk factors for UTI in children is limited. UTIs were associated with constipation, encopresis, bladder instability, and infrequent voiding in two studies4,5 but not in a cohort of febrile children younger than two years.6 Bathing and back-to-front wiping have not been demonstrated to be risk factors.
Older children with UTI may have dysuria, frequency, urgency, hesitancy, small-volume voids, or lower abdominal pain. Infants with UTI more commonly present with non-specific symptoms such as fever, irritability, jaundice, vomiting, or failure to thrive. Unusual odor of the urine is not helpful in predicting UTI.7
Other conditions may mimic UTI symptoms. Acute urethritis or vulvovaginitis may be caused by various types of irritants, including chemical (e.g., bubble baths, soaps), physical (e.g., self-exploration), and biologic (e.g., pinworms). Self-infusion of water into the urethra was reported in 10 of 31 boys five to 15 years of age presenting with a first UTI.8
For urine collection from infants and young children, suprapubic aspiration or transurethral catheterization generally is recommended. Urethral catheterization is more likely than aspiration to obtain a sufficient sample of urine.10 Collection from bags or pads leads to high contamination rates.11 Clean-catch urine collection from infants requires more patience and effort than the use of pads or bags, but this method is reasonably accurate and rates of contamination are low.11,12
RAPID URINE TESTING
Children with a high likelihood of UTI (e.g., older children with classic urinary symptoms) should have a urine culture; rapid urine testing will not change the treatment of these patients. In children with a low likelihood of UTI (e.g., those with vague urinary complaints and an alternative explanation for fever), urine inspection and dipstick testing may reduce the need for urine culture. Urine culture is indicated if the child's urine is cloudy or if the dipstick test shows positive leukocyte esterase or nitrite activity.13–16 Urine culture also is recommended for children with recurrent symptoms.
Urine dipstick analysis can rule out UTI if the result is negative (in nondilute urine), but a positive result on dip-stick testing is insufficient to diagnose UTI because false-positive results are common. In a cohort study13 with 18 percent prevalence of UTI, a negative result on urine dipstick analysis (i.e., negative for leukocyte esterase and nitrites) had a negative predictive value of 96 percent. Urine dipstick analysis appears to be more accurate than microscopic analysis for the detection of pyuria in children,13,14 although this has been debated.17 Urine dipstick analysis may produce false-negative results with dilute urine (i.e., specific gravity less than 1.005). Clear urine (defined as the ability to read text through the urine in a test tube as easily as through water) had 96 to 100 percent negative predictive values in cohort studies,13,15,16 but these values may be unreliable in dilute urine.
Some authors recommend urine culture testing for infants with suspected UTI because of lower sensitivities for urine dipstick analysis in infants.18,19 However, investigators in several large cohort studies2,20 of infants younger than two years with undiagnosed fevers or UTI symptoms have found negative predictive values greater than 98 percent despite sensitivities of 79 percent. Urine dipstick testing achieved these high negative predictive values because of the low likelihood of positive urine cultures (2.8 to 8 percent) in these studies.2,20
Previously, imaging studies were recommended for any child with a UTI. The 1999 clinical practice guideline23 from the American Academy of Pediatrics (AAP) recommends imaging for children two months to two years of age, noting that the strength of evidence for the recommendation is fair. Recent evidence has suggested that imaging work-ups for children with a first uncomplicated UTI may not improve patient care. In one prospective study,24 renal ultrasonography did not change the management of UTI in any of 255 children younger than five years who were admitted with a first uncomplicated febrile UTI. An earlier systematic review25 of 63 descriptive studies found no evidence of the impact of routine imaging on clinical outcomes in children with a first UTI.
In a randomized trial,26 150 children two to 10 years of age with a first UTI were assigned to routine imaging (i.e., ultrasonography and voiding cystourethrography [VCUG]) or to selective imaging (only if recurrent UTI or persistent problems were present). Approximately one in five children (21 percent) in the selective imaging group had imaging performed. Routine imaging increased the use of prophylactic antibiotics (28 versus 5 percent) but did not reduce the rate of recurrent UTIs (26 versus 21 percent) or renal scarring at two years (9 versus 9 percent). These results suggest that imaging is not useful in determining which children might benefit from prophylactic antibiotics.
Extensive imaging work-up (i.e., ultrasonography and renal scan within 72 hours, VCUG at one month, and repeat renal scan at six months) was evaluated in a prospective study27 of 309 children one to 24 months of age with a first febrile UTI. Results of ultrasonography were normal in 88 percent, and abnormalities did not change management (i.e., no obstructive lesions were found). Initial renal scans suggested acute pyelonephritis in 61 percent of patients, but febrile UTI in infants is managed with the presumption of pyelonephritis. VCUG and repeat renal scans showed vesicoureteral reflux (VUR) and renal scarring in 39 and 9.5 percent of infants, respectively. However, the manner in which VUR or renal scarring should alter the course of management is controversial. In some reviews,28,29 authors have suggested that VUR predisposes children to recurrent pyelonephritis and renal scarring. Although some children with renal scarring develop hypertension and renal insufficiency in adulthood, this association appears limited to those with extensive renal scarring or hypertension in childhood.28 Recommendations have been made to identify VUR early to stop this cascade of events, despite a lack of studies showing that interventions along this pathway prevent renal scarring, hypertension, and renal insufficiency. Severe VUR has been associated with recurrent UTI30 and recurrent pyelonephritis,31 but the association of VUR with renal scarring has been inconsistent.32,33 Renal scarring is associated with recurrent UTI32 and recurrent pyelonephritis,34 but a causal relationship has not been established.25
Although imaging traditionally has been recommended in children with recurrent UTI, no randomized trials are available to support or refute these recommendations. Bladder ultrasonography to determine postvoid residual volume may identify urinary obstruction, and children with renal scarring may benefit from follow-up for early detection of hypertension or renal insufficiency, but data supporting these approaches are sparse.
Empiric antibiotic therapy (Table 223 ) is reasonable while awaiting culture results if there is a high clinical suspicion of UTI.23 Subsequently changing antibiotics based on reported sensitivities may be unnecessary if clinical resolution occurs. According to the AAP, the child should be reevaluated with a repeat urine culture and renal/bladder ultrasonography if clinical improvement does not occur within two days.23
The optimal duration of antibiotic therapy has not been established. One-day courses of antibiotics are less effective than longer courses.37 Systematic reviews37–39 of randomized trials comparing shorter and longer courses of antibiotics have had varying results, based in part on the inclusion of one-day courses in the analysis. Short courses (three to five days) of antibiotics may be as effective as longer courses (seven to 14 days), but this has not been clearly proved.37–39
No evidence was found to support a recommendation for routine posttreatment urine cultures in children with UTIs. In one study40 of 364 children hospitalized with UTI, 291 (80 percent) had follow-up urine cultures, none of which was positive.
In a one-year longitudinal study,41 recurrent symptomatic UTI occurred in 12 percent of children younger than five years presenting to an emergency department with a first UTI. Children younger than six months had a higher risk of recurrence (18.6 percent).
Prophylactic antibiotics may reduce the risk of recurrent UTI. Two systematic reviews42,43 found weak evidence of benefit with poor quality trials. Table 323 describes potential prophylactic regimens in children younger than two years. Only one comparison trial of antibiotic selection was found. In this six-month trial,44 nitrofurantoin (Furadantin) was more effective than trimethoprim (Primsol) in preventing UTIs (number needed to treat [NNT] = 5.4) but was more likely to be discontinued because of side effects (number needed to treat to harm = 5.3), which were mainly gastrointestinal.
Because VUR has been considered a predisposing factor to UTI and its complications, treatments for children with VUR have been promoted. Medical management often has used prophylactic antibiotics, but the systematic reviews43 suggesting that prophylactic antibiotics prevent recurrent UTI did not include studies of children with VUR. Nonetheless, the AAP recommends prophylaxis until imaging studies are completed and until additional treatment based on the imaging findings is instituted.23 Two systematic reviews45,46 of randomized trials comparing surgical and medical management of VUR found no significant differences in UTI recurrence rates, renal function, renal scarring, hypertension, or end-stage renal disease. In two randomized trials47,48 involving more than 400 children younger than 11 years who had grade III or IV VUR and documented UTI, surgery reduced the rate of acute pyelonephritis (21 to 22 percent incidence with surgery versus 8 to 10 percent with antibiotic prophylaxis alone; NNT = 7 to 9), but overall rates of UTI did not differ. A new treatment for children with VUR, dextranomer/hyaluronic acid copolymer (Deflux), is a bulking agent placed endoscopically near the distal ureteral orifice. Deflux was more effective than antibiotic prophylaxis in resolving childhood VUR in one randomized trial49 of 61 children but did not prevent UTI. Nine UTIs were recorded in six of the 40 children in the Deflux group, whereas none occurred in the 21 children receiving antibiotics.49
Cranberry juice and related products have not been shown to prevent or reduce UTI in children; in a systematic review50 of seven trials, the evidence base was too limited to support such conclusions.
Circumcision prevented recurrent symptomatic UTI in one randomized trial.51 Seventy uncircumcised boys three months to 10 years of age with a first symptomatic culture-confirmed UTI were randomized to immediate circumcision or circumcision after six months. Recurrent symptomatic UTIs developed in three of 35 boys during the six-month delay (8.6 percent) but not in the boys who received immediate circumcision (NNT=12). Circumcision may provide primary prevention of UTI in boys, based on prospective follow-up of 69,100 boys for two to three years.52 However, circumcision may not be warranted for this specific purpose because 195 circumcisions would be needed to prevent one hospital admission for UTI in the first year of life.
Data sources: Online sources including PubMed, The Cochrane Library, Clinical Evidence, DynaMed, and the National Guideline Clearinghouse were searched in April 2003 using “urinary tract infection” and age limitations (younger than 18 years) where applicable. Reference lists of retrieved articles and reviews also were used. The highest level of evidence found for each specific topic was included. This review is based on a more thorough version of this topic that is available in DynaMed (http://www.DynamicMedical.com) and is updated as additional relevant information is identified.53