Itu 2011 aap

DOI: 10.1542/peds.2011-1330; originally published online August 28, 2011;Pediatrics

COMMITTEE ON QUALITY IMPROVEMENT AND MANAGEMENTSUBCOMMITTEE ON URINARY TRACT INFECTION and STEERING

Management of the Initial UTI in Febrile Infants and Children 2 to 24 MonthsUrinary Tract Infection: Clinical Practice Guideline for the Diagnosis and

http://pediatrics.aappublications.org/content/early/2011/08/24/peds.2011-1330

located on the World Wide Web at: The online version of this article, along with updated information and services, is

of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2011 by the American Academy published, and trademarked by the American Academy of Pediatrics, 141 Northwest Pointpublication, it has been published continuously since 1948. PEDIATRICS is owned, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by guest on January 2, 2012pediatrics.aappublications.orgDownloaded from


http://pediatrics.aappublications.org/

CLINICAL PRACTICE GUIDELINE

Urinary Tract Infection: Clinical Practice Guideline forthe Diagnosis and Management of the Initial UTI inFebrile Infants and Children 2 to 24 Months

abstractOBJECTIVE: To revise the American Academy of Pediatrics practiceparameter regarding the diagnosis and management of initial urinarytract infections (UTIs) in febrile infants and young children.

METHODS: Analysis of the medical literature published since the lastversion of the guidelinewas supplemented by analysis of data providedby authors of recent publications. The strength of evidence supportingeach recommendation and the strength of the recommendation wereassessed and graded.

RESULTS: Diagnosis is made on the basis of the presence of bothpyuria and at least 50 000 colonies per mL of a single uropathogenicorganism in an appropriately collected specimen of urine. After 7 to 14days of antimicrobial treatment, close clinical follow-up monitoringshould be maintained to permit prompt diagnosis and treatment ofrecurrent infections. Ultrasonography of the kidneys and bladdershould be performed to detect anatomic abnormalities. Data from themost recent 6 studies do not support the use of antimicrobial prophy-laxis to prevent febrile recurrent UTI in infants without vesicoureteralreflux (VUR) or with grade I to IV VUR. Therefore, a voiding cystoure-thrography (VCUG) is not recommended routinely after the first UTI;VCUG is indicated if renal and bladder ultrasonography reveals hydro-nephrosis, scarring, or other findings that would suggest either high-grade VUR or obstructive uropathy and in other atypical or complexclinical circumstances. VCUG should also be performed if there is arecurrence of a febrile UTI. The recommendations in this guideline donot indicate an exclusive course of treatment or serve as a standard ofcare; variations may be appropriate. Recommendations about antimi-crobial prophylaxis and implications for performance of VCUG arebased on currently available evidence. As with all American Academy ofPediatrics clinical guidelines, the recommendations will be reviewedroutinely and incorporate new evidence, such as data from the Ran-domized Intervention for Children With Vesicoureteral Reflux (RIVUR)study.

CONCLUSIONS: Changes in this revision include criteria for the diag-nosis of UTI and recommendations for imaging. Pediatrics 2011;128:595–610

SUBCOMMITTEE ON URINARY TRACT INFECTION, STEERINGCOMMITTEE ON QUALITY IMPROVEMENT AND MANAGEMENT

KEY WORDSurinary tract infection, infants, children, vesicoureteral reflux,voiding cystourethrography

ABBREVIATIONSSPA—suprapubic aspirationAAP—American Academy of PediatricsUTI—urinary tract infectionRCT—randomized controlled trialCFU—colony-forming unitVUR—vesicoureteral refluxWBC—white blood cellRBUS—renal and bladder ultrasonographyVCUG—voiding cystourethrography

This document is copyrighted and is property of the AmericanAcademy of Pediatrics and its Board of Directors. All authorshave filed conflict of interest statements with the AmericanAcademy of Pediatrics. Any conflicts have been resolved througha process approved by the Board of Directors. The AmericanAcademy of Pediatrics has neither solicited nor accepted anycommercial involvement in the development of the content ofthis publication.

The recommendations in this report do not indicate an exclusivecourse of treatment or serve as a standard of medical care.Variations, taking into account individual circumstances, may beappropriate.

All clinical practice guidelines from the American Academy ofPediatrics automatically expire 5 years after publication unlessreaffirmed, revised, or retired at or before that time.

www.pediatrics.org/cgi/doi/10.1542/peds.2011-1330

doi:10.1542/peds.2011-1330

PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).

Copyright © 2011 by the American Academy of Pediatrics

COMPANION PAPERS: Companions to this article can be foundon pages 572 and e749, and online at www.pediatrics.org/cgi/doi/10.1542/peds.2011-1818 and www.pediatrics.org/cgi/doi/10.1542/peds.2011-1332.

FROM THE AMERICAN ACADEMY OF PEDIATRICS

PEDIATRICS Volume 128, Number 3, September 2011 595 by guest on January 2, 2012pediatrics.aappublications.orgDownloaded from

www.pediatrics.org/cgi/doi/10.1542/peds.2011-1818 and www.pediatrics.org/cgi/doi/10.1542/peds.2011-1332



pediatrics.aappublications.org/


INTRODUCTIONSince the early 1970s, occult bactere-mia has been the major focus of con-cern for clinicians evaluating febrileinfants who have no recognizablesource of infection. With the introduc-tion of effective conjugate vaccinesagainst Haemophilus influenzae typeb and Streptococcus pneumoniae(which have resulted in dramatic de-creases in bacteremia and meningi-tis), there has been increasing appre-ciation of the urinary tract as the mostfrequent site of occult and serious bac-terial infections. Because the clinicalpresentation tends to be nonspecific ininfants and reliable urine specimensfor culture cannot be obtained withoutinvasive methods (urethral cathe-terization or suprapubic aspiration[SPA]), diagnosis and treatment maybe delayed. Most experimental andclinical data support the concept thatdelays in the institution of appropriatetreatment of pyelonephritis increasethe risk of renal damage.1,2

This clinical practice guideline is a re-vision of the practice parameter pub-lished by the American Academy ofPediatrics (AAP) in 1999.3 It was devel-oped by a subcommittee of the Steer-ing Committee on Quality Improvementand Management that included physi-cians with expertise in the fields of ac-ademic general pediatrics, epidemiol-ogy and informatics, pediatricinfectious diseases, pediatric nephrol-ogy, pediatric practice, pediatric radi-ology, and pediatric urology. The AAPfunded the development of this guide-line; none of the participants had anyfinancial conflicts of interest. Theguideline was reviewed by multiplegroups within the AAP (7 committees, 1council, and 9 sections) and 5 externalorganizations in the United States andCanada. The guideline will be reviewedand/or revised in 5 years, unless newevidence emerges that warrants revi-sion sooner. The guideline is intended

for use in a variety of clinical settings(eg, office, emergency department, orhospital) by clinicians who treat in-fants and young children. This text is asummary of the analysis. The data onwhich the recommendations arebased are included in a companiontechnical report.4

Like the 1999 practice parameter, thisrevision focuses on the diagnosis andmanagement of initial urinary tract in-fections (UTIs) in febrile infants andyoung children (2–24 months of age)who have no obvious neurologic or an-atomic abnormalities known to be as-sociated with recurrent UTI or renaldamage. (For simplicity, in the remain-der of this guideline the phrase “fe-brile infants” is used to indicate febrileinfants and young children 2–24months of age.) The lower and upperage limits were selected because stud-ies on infants with unexplained fevergenerally have used these age limitsand have documented that the preva-lence of UTI is high (�5%) in this agegroup. In those studies, fever was de-fined as temperature of at least 38.0°C(�100.4°F); accordingly, this definitionof fever is used in this guideline. Ne-onates and infants less than 2months of age are excluded, becausethere are special considerations inthis age group that may limit the ap-plication of evidence derived fromthe studies of 2- to 24-month-old chil-dren. Data are insufficient to deter-mine whether the evidence gener-ated from studies of infants 2 to 24months of age applies to childrenmore than 24 months of age.

METHODS

To provide evidence for the guideline, 2literature searches were conducted,that is, a surveillance of Medline-listedliterature over the past 10 years forsignificant changes since the guidelinewas published and a systematic re-view of the literature on the effective-

ness of prophylactic antimicrobialtherapy to prevent recurrence of fe-brile UTI/pyelonephritis in childrenwith vesicoureteral reflux (VUR). Thelatter was based on the new and grow-ing body of evidence questioning theeffectiveness of antimicrobial prophy-laxis to prevent recurrent febrile UTI inchildren with VUR. To explore this par-ticular issue, the literature search wasexpanded to include trials publishedsince 1993 in which antimicrobial pro-phylaxis was compared with no treat-ment or placebo treatment for chil-dren with VUR. Because all except 1 ofthe recent randomized controlled tri-als (RCTs) of the effectiveness of pro-phylaxis included children more than24 months of age and some did notprovide specific data according tograde of VUR, the authors of the 6 RCTswere contacted; all provided raw datafrom their studies specifically ad-dressing infants 2 to 24 months of age,according to grade of VUR. Meta-analysis of these data was performed.

Results from the literature searchesand meta-analyses were provided tocommittee members. Issues wereraised and discussed until consensuswas reached regarding recommenda-tions. The quality of evidence support-ing each recommendation and thestrength of the recommendation wereassessed by the committee membermost experienced in informatics andepidemiology and were graded ac-cording to AAP policy5 (Fig 1).

The subcommittee formulated 7 rec-ommendations, which are presentedin the text in the order in which a clini-cian would use them when evaluatingand treating a febrile infant, as well asin algorithm form in the Appendix. Thisclinical practice guideline is not in-tended to be a sole source of guidancefor the treatment of febrile infants withUTIs. Rather, it is intended to assist clini-cians in decision-making. It is not in-tended to replace clinical judgment or to

596 FROM THE AMERICAN ACADEMY OF PEDIATRICS by guest on January 2, 2012pediatrics.aappublications.orgDownloaded from


establish an exclusive protocol for thecare of all children with this condition.

DIAGNOSIS

Action Statement 1

If a clinician decides that a febrileinfant with no apparent source forthe fever requires antimicrobialtherapy to be administered be-cause of ill appearance or anotherpressing reason, the clinicianshould ensure that a urine speci-men is obtained for both cultureand urinalysis before an antimicro-bial agent is administered; thespecimen needs to be obtainedthrough catheterization or SPA, be-cause the diagnosis of UTI cannotbe established reliably through cul-ture of urine collected in a bag(evidence quality: A; strongrecommendation).

When evaluating febrile infants, clini-ciansmake a subjective assessment ofthe degree of illness or toxicity, in ad-dition to seeking an explanation for thefever. This clinical assessment deter-mines whether antimicrobial therapyshould be initiated promptly and af-fects the diagnostic process regardingUTI. If the clinician determines that thedegree of illness warrants immediateantimicrobial therapy, then a urinespecimen suitable for culture shouldbe obtained through catheterization orSPA before antimicrobial agents are

administered, because the antimicro-bial agents commonly prescribed insuch situations would almost certainlyobscure the diagnosis of UTI.

SPA has been considered the standardmethod for obtaining urine that is un-contaminated by perineal flora. Vari-able success rates for obtaining urinehave been reported (23%–90%).6–8

When ultrasonographic guidance isused, success rates improve.9,10 Thetechnique has limited risks, but tech-nical expertise and experience arerequired, and many parents and phy-sicians perceive the procedure asunacceptably invasive, comparedwith catheterization. However, theremay be no acceptable alternative toSPA for boys with moderate or se-vere phimosis or girls with tight la-bial adhesions.

Urine obtained through catheteriza-tion for culture has a sensitivity of 95%and a specificity of 99%, comparedwith that obtained through SPA.7,11,12

The techniques required for catheter-ization and SPA are well described.13

When catheterization or SPA is beingattempted, the clinician should have asterile container ready to collect aurine specimen, because the prepara-tion for the procedure may stimulatethe child to void. Whether the urine isobtained through catheterization or isvoided, the first few drops should beallowed to fall outside the sterile con-

tainer, because they may be contami-nated by bacteria in the distal urethra.

Cultures of urine specimens collectedin a bag applied to the perineum havean unacceptably high false-positiverate and are valid only when they yieldnegative results.6,14–16 With a preva-lence of UTI of 5% and a high rate offalse-positive results (specificity:�63%), a “positive” culture result forurine collected in a bag would be afalse-positive result 88% of the time.For febrile boys, with a prevalence ofUTI of 2%, the rate of false-positive re-sults is 95%; for circumcised boys,with a prevalence of UTI of 0.2%, therate of false-positive results is 99%.Therefore, in cases in which antimicro-bial therapy will be initiated, catheter-ization or SPA is required to establishthe diagnosis of UTI.

● Aggregate quality of evidence: A (diag-nostic studies on relevant populations).

● Benefits: A missed diagnosis of UTIcan lead to renal scarring if left un-treated; overdiagnosis of UTI canlead to overtreatment and unneces-sary and expensive imaging. Once an-timicrobial therapy is initiated, theop-portunity to make a definitivediagnosis is lost; multiple studies ofantimicrobial therapy have shownthat the urine may be rapidlysterilized.

● Harms/risks/costs: Catheterizationis invasive.

● Benefit-harms assessment: Prepon-derance of benefit over harm.

● Value judgments: Once antimicro-bial therapy has begun, the opportu-nity to make a definitive diagnosis islost. Therefore, it is important tohave the most-accurate test for UTIperformed initially.

● Role of patient preferences: There isno evidence regarding patient pref-erences for bag versus catheterizedurine. However, bladder tap has

FIGURE 1AAP evidence strengths.





been shown to be more painful thanurethral catheterization.

● Exclusions: None.

● Intentional vagueness: The basis ofthe determination that antimicro-bial therapy is needed urgently isnot specified, because variability inclinical judgment is expected; con-siderations for individual patients,such as availability of follow-upcare, may enter into the decision,and the literature provides only gen-eral guidance.

● Policy level: Strong recommendation.

Action Statement 2

If a clinician assesses a febrile infantwith no apparent source for the feveras not being so ill as to require imme-diate antimicrobial therapy, then theclinicianshouldassessthelikelihoodofUTI (see below for how to assesslikelihood).

Action Statement 2a

If the clinician determines the febrileinfant to have a low likelihood of UTI(see text), then clinical follow-upmonitoring without testing is suffi-cient (evidence quality: A; strongrecommendation).

Action Statement 2b

If the clinician determines that thefebrile infant is not in a low-riskgroup (see below), then there are 2choices (evidence quality: A; strongrecommendation). Option 1 is to ob-tain a urine specimen through cath-eterization or SPA for culture andurinalysis. Option 2 is to obtain aurine specimen through the mostconvenient means and to perform aurinalysis. If the urinalysis resultssuggest a UTI (positive leukocyteesterase test results or nitrite testor microscopic analysis resultspositive for leukocytes or bacte-ria), then a urine specimen should

be obtained through catheteriza-tion or SPA and cultured; if urinaly-sis of fresh (<1 hour since void)urine yields negative leukocyte es-terase and nitrite test results, thenit is reasonable to monitor the clin-ical course without initiating anti-microbial therapy, recognizing thatnegative urinalysis results do notrule out a UTI with certainty.

If the clinician determines that the de-gree of illness does not require imme-diate antimicrobial therapy, then thelikelihood of UTI should be assessed.As noted previously, the overall preva-lence of UTI in febrile infants who haveno source for their fever evident on thebasis of history or physical examina-tion results is approximately 5%,17,18

but it is possible to identify groupswith higher-than-average likelihoodand some with lower-than-averagelikelihood. The prevalence of UTIamong febrile infant girls is more thantwice that among febrile infant boys(relative risk: 2.27). The rate for uncir-cumcised boys is 4 to 20 times higherthan that for circumcised boys, whoserate of UTI is only 0.2% to 0.4%.19–24 Thepresence of another, clinically obvioussource of infection reduces the likeli-hood of UTI by one-half.25

In a survey asking, “What yield is re-quired to warrant urine culture in fe-brile infants?,” the threshold was less

than 1% for 10.4% of academicians and11.7% for practitioners26; when thethreshold was increased to 1% to 3%,67.5% of academicians and 45.7% ofpractitioners considered the yield suf-ficiently high to warrant urine culture.Therefore, attempting to operational-ize “low likelihood” (ie, below a thresh-old that warrants a urine culture) doesnot produce an absolute percentage;clinicians will choose a threshold de-pending on factors such as their confi-dence that contact will be maintainedthrough the illness (so that a specimencanbeobtainedat a later time) andcom-fort with diagnostic uncertainty. Fig 2 in-dicates the number of risk factors as-sociated with threshold probabilitiesof UTI of at least 1% and at least 2%.

In a series of studies, Gorelick, Shaw,and colleagues27–29 derived and vali-dated a prediction rule for febrile in-fant girls on the basis of 5 risk factors,namely, white race, age less than 12months, temperature of at least 39°C,fever for at least 2 days, and absenceof another source of infection. Thisprediction rule, with sensitivity of88% and specificity of 30%, permitssome infant girls to be considered ina low-likelihood group (Fig 2). For ex-ample, of girls with no identifiablesource of infection, those who are non-white and more than 12 months of agewith a recent onset (�2 days) of low-

FIGURE 2Probability of UTI Among Febrile Infant Girls28 and Infant Boys30 According to Number of FindingsPresent. aProbability of UTI exceeds 1% even with no risk factors other than being uncircumcised.



grade fever (�39°C) have less than a1% probability of UTI; each additionalrisk factor increases the probability. Itshould be noted, however, that someof the factors (eg, duration of fever)may change during the course of theillness, excluding the infant from alow-likelihood designation andprompting testing as described inaction statement 2a.

As demonstrated in Fig 2, the majorrisk factor for febrile infant boys iswhether they are circumcised. The prob-ability of UTI can be estimated on the ba-sis of 4 risk factors, namely, nonblackrace, temperature of at least 39°C, feverfor more than 24 hours, and absence ofanother source of infection.4,30

If the clinician determines that the in-fant does not require immediate anti-microbial therapy and a urine speci-men is desired, then often a urinecollection bag affixed to the perineumis used. Many clinicians think that thiscollection technique has a low contam-ination rate under the following cir-cumstances: the patient’s perineum isproperly cleansed and rinsed beforeapplication of the collection bag, theurine bag is removed promptly afterurine is voided into the bag, and thespecimen is refrigerated or processedimmediately. Even if contaminationfrom the perineal skin is minimized,however, there may be significant con-tamination from the vagina in girls orthe prepuce in uncircumcised boys,the 2 groups at highest risk of UTI. A“positive” culture result from a speci-men collected in a bag cannot be usedto document a UTI; confirmation re-quires culture of a specimen collectedthrough catheterization or SPA. Be-cause there may be substantial delaywaiting for the infant to void and a sec-ond specimen, obtained through cath-eterization, may be necessary if theurinalysis suggests the possibility ofUTI, many clinicians prefer to obtain a

definitive urine specimen throughcatheterization initially.

● Aggregate quality of evidence: A (diag-nostic studies on relevant populations).

● Benefits: Accurate diagnosis of UTIcan prevent the spread of infectionand renal scarring; avoiding overdi-agnosis of UTI can prevent over-treatment and unnecessary and ex-pensive imaging.

● Harms/risks/costs: A small propor-tion of febrile infants, considered atlow likelihood of UTI, will not receivetimely identification and treatmentof their UTIs.


● Value judgments: There is a risk ofUTI sufficiently low to forestall fur-ther evaluation.

● Role of patient preferences: Thechoice of option 1 or option 2 andthe threshold risk of UTI warrantingobtaining a urine specimen may beinfluenced by parents’ preferenceto avoid urethral catheterization (ifa bag urine sample yields negativeurinalysis results) versus timelyevaluation (obtaining a definitivespecimen through catheterization).

● Exclusions: Because it depends on arange of patient- and physician-specific considerations, the precisethreshold risk of UTI warranting ob-taining a urine specimen is left tothe clinician but is below 3%.

● Intentional vagueness: None.

● Policy level: Strong recommendation.

Action Statement 3

To establish the diagnosis of UTI,clinicians should require both uri-nalysis results that suggest infec-tion (pyuria and/or bacteriuria)and the presence of at least 50 000colony-forming units (CFUs) per mLof a uropathogen cultured from aurine specimen obtained throughcatheterization or SPA (evidencequality: C; recommendation).

Urinalysis

General Considerations

Urinalysis cannot substitute for urineculture to document the presence ofUTI but needs to be used in conjunctionwith culture. Because urine culture re-sults are not available for at least 24hours, there is considerable interestin tests that may predict the results ofthe urine culture and enable presump-tive therapy to be initiated at the firstencounter. Urinalysis can be per-formed on any specimen, includingone collected from a bag applied to theperineum. However, the specimenmust be fresh (�1 hour after voidingwith maintenance at room tempera-ture or�4 hours after voiding with re-frigeration), to ensure sensitivity andspecificity of the urinalysis. The teststhat have received the most atten-tion are biochemical analyses of leu-kocyte esterase and nitrite through arapid dipstick method and urinemicroscopic examination for whiteblood cells (WBCs) and bacteria(Table 1).

TABLE 1 Sensitivity and Specificity of Components of Urinalysis, Alone and in Combination

Test Sensitivity (Range), % Specificity (Range), %

Leukocyte esterase test 83 (67–94) 78 (64–92)Nitrite test 53 (15–82) 98 (90–100)Leukocyte esterase ornitrite test positive

93 (90–100) 72 (58–91)

Microscopy, WBCs 73 (32–100) 81 (45–98)Microscopy, bacteria 81 (16–99) 83 (11–100)Leukocyte esterase test,nitrite test, ormicroscopy positive

99.8 (99–100) 70 (60–92)





Urine dipsticks are appealing, becausethey provide rapid results, do not re-quire microscopy, and are eligible fora waiver under the Clinical LaboratoryImprovement Amendments. They indi-cate the presence of leukocyte es-terase (as a surrogate marker forpyuria) and urinary nitrite (which isconverted from dietary nitrates in thepresence of most Gram-negative entericbacteria in the urine). The conversion ofdietary nitrates to nitrites by bacteria re-quires approximately 4 hours in thebladder.31 The performance characteris-tics of both leukocyte esterase and ni-trite tests vary according to the defini-tion used for positive urine cultureresults, the age and symptoms of thepopulation being studied, and themethod of urine collection.

Nitrite Test

A nitrite test is not a sensitive markerfor children, particularly infants, whoempty their bladders frequently.Therefore, negative nitrite test resultshave little value in ruling out UTI. More-over, not all urinary pathogens reducenitrate to nitrite. The test is helpfulwhen the result is positive, however,because it is highly specific (ie, thereare few false-positive results).32

Leukocyte Esterase Test

The sensitivity of the leukocyte es-terase test is 94% when it used in thecontext of clinically suspected UTI.Overall, the reported sensitivity in var-ious studies is lower (83%), becausethe results of leukocyte esterase testswere related to culture results withoutexclusion of individuals with asymp-tomatic bacteriuria. The absence ofleukocyte esterase in the urine of indi-viduals with asymptomatic bacteriuriais an advantage of the test, rather thana limitation, because it distinguishesindividuals with asymptomatic bacte-riuria from those with true UTI.

The specificity of the leukocyte es-terase test (average: 72% [range:

64%–92%]) generally is not as good asthe sensitivity, which reflects the non-specificity of pyuria in general. Accord-ingly, positive leukocyte esterase testresults should be interpretedwith cau-tion, because false-positive results arecommon. With numerous conditionsother than UTI, including fever result-ing from other conditions (eg, strepto-coccal infections or Kawasaki dis-ease), and after vigorous exercise,WBCsmay be found in the urine. There-fore, a finding of pyuria by no meansconfirms that an infection of the uri-nary tract is present.

The absence of pyuria in children withtrue UTIs is rare, however. It is theoret-ically possible if a febrile child is as-sessed before the inflammatory re-sponse has developed, but theinflammatory response to a UTI pro-duces both fever and pyuria; therefore,children who are being evaluated be-cause of fever should already haveWBCs in their urine. More likely expla-nations for significant bacteriuria inculture in the absence of pyuria in-clude contaminated specimens, insen-sitive criteria for pyuria, and asymp-tomatic bacteriuria. In most cases,when true UTI has been reported to oc-cur in the absence of pyuria, the defi-nition of pyuria has been at fault. Thestandard method of assessing pyuriahas been centrifugation of the urineand microscopic analysis, with athreshold of 5 WBCs per high-powerfield (�25 WBCs per �L). If a countingchamber is used, however, the findingof at least 10 WBCs per �L in uncentri-fuged urine has been demonstrated tobe more sensitive33 and performs wellin clinical situations in which the stan-dard method does not, such as withvery young infants.34

An important cause of bacteriuria inthe absence of pyuria is asymptomaticbacteriuria. Asymptomatic bacteriuriaoften is associated with school-agedand older girls,35 but it can be present

during infancy. In a study of infants 2 to24 months of age, 0.7% of afebrile girlshad 3 successive urine cultures with105 CFUs per mL of a single uropatho-gen.26 Asymptomatic bacteriuria canbe easily confused with true UTI in afebrile infant but needs to be distin-guished, because studies suggest thatantimicrobial treatment may do moreharm than good.36 The key to distin-guishing true UTI from asymptomaticbacteriuria is the presence of pyuria.

Microscopic Analysis for Bacteriuria

The presence of bacteria in a fresh,Gram-stained specimen of uncentri-fuged urine correlates with 105 CFUsper mL in culture.37 An “enhanced uri-nalysis,” combining the countingchamber assessment of pyuria notedpreviously with Gram staining of dropsof uncentrifuged urine, with a thresh-old of at least 1 Gram-negative rod in10 oil immersion fields, has greater sen-sitivity, specificity, and positive predic-tive value than does the standard urinal-ysis33 and is the preferred method ofurinalysis when appropriate equipmentand personnel are available.

Automated Urinalysis

Automated methods to perform uri-nalysis are now being used in manyhospitals and laboratories. Image-based systems use flow imaginganalysis technology and software toclassify particles in uncentrifugedurine specimens rapidly.38 Resultscorrelate well with manual methods,especially for red blood cells, WBCs,and squamous epithelial cells. In thefuture, this may be the most commonmethod by which urinalysis is per-formed in laboratories.

Culture

The diagnosis of UTI is made on the ba-sis of quantitative urine culture re-sults in addition to evidence of pyuriaand/or bacteriuria. Urine specimensshould be processed as expediently as



possible. If the specimen is not pro-cessed promptly, then it should be re-frigerated to prevent the growth of or-ganisms that can occur in urine atroom temperature; for the same rea-son, specimens that require transpor-tation to another site for processingshould be transported on ice. A prop-erly collected urine specimen shouldbe inoculated on culture medium thatwill allow identification of urinary tractpathogens.

Urine culture results are consideredpositive or negative on the basis of thenumber of CFUs that grow on the cul-ture medium.36 Definition of significantcolony counts with regard to themethod of collection considers thatthe distal urethra and periurethralarea are commonly colonized by thesame bacteria that may cause UTI;therefore, a low colony count may bepresent in a specimen obtainedthrough voiding or catheterizationwhen bacteria are not present in blad-der urine. Definitions of positive andnegative culture results are opera-tional and not absolute. The time theurine resides in the bladder (bladderincubation time) is an important deter-minant of the magnitude of the colonycount. The concept that more than100 000 CFUs per mL indicates a UTIwas based on morning collections ofurine from adult women, with compar-ison of specimens from women with-out symptoms and women consideredclinically to have pyelonephritis; thetransition range, in which the propor-tion of women with pyelonephritis ex-ceeded the proportion of women with-out symptoms, was 10 000 to 100 000CFUs per mL.39 In most instances, anappropriate threshold to considerbacteriuria “significant” in infants andchildren is the presence of at least50 000 CFUs per mL of a single urinarypathogen.40 (Organisms such asLactobacillus spp, coagulase-negativestaphylococci, and Corynebacterium

spp are not considered clinically rele-vant urine isolates for otherwisehealthy, 2- to 24-month-old children.)Reducing the threshold from 100 000CFUs per mL to 50 000 CFUs per mLwould seem to increase the sensitivityof culture at the expense of decreasedspecificity; however, because the pro-posed criteria for UTI now include evi-dence of pyuria in addition to positiveculture results, infants with “positive”culture results alone will be recog-nized as having asymptomatic bacteri-uria rather than a true UTI. Some labo-ratories report growth only in thefollowing categories: 0 to 1000, 1000 to10 000, 10 000 to 100 000, and morethan 100 000 CFUs per mL. In suchcases, results in the 10 000 to 100 000CFUs per mL range need to be evalu-ated in context, such as whether theurinalysis findings support the diagno-sis of UTI and whether the organism isa recognized uropathogen.

Alternative culture methods, such asdipslides, may have a place in the of-fice setting; sensitivity is reported tobe in the range of 87% to 100%, andspecificity is reported to be 92% to98%, but dipslides cannot specify theorganism or antimicrobial sensitivi-ties.41 Practices that use dipslidesshould do so in collaboration with acertified laboratory for identificationand sensitivity testing or, in the ab-sence of such results, may need to per-form “test of cure” cultures after 24hours of treatment.

● Aggregate quality of evidence: C (ob-servational studies).

● Benefits: Accurate diagnosis of UTIcan prevent the spread of infectionand renal scarring; avoiding overdi-agnosis of UTI can prevent over-treatment and unnecessary and ex-pensive imaging. These criteriareduce the likelihood of overdiagno-sis of UTI in infants with asymptom-atic bacteriuria or contaminatedspecimens.

● Harms/risks/costs: Stringent diag-nostic criteria may miss a smallnumber of UTIs.


● Value judgments: Treatment ofasymptomatic bacteriuria may beharmful.

● Role of patient preferences: We as-sume that parents prefer no actionin the absence of a UTI (avoidingfalse-positive results) over a verysmall chance of missing a UTI.



● Policy level: Recommendation.

MANAGEMENT

Action Statement 4

Action Statement 4a

When initiating treatment, the clini-cian should base the choice ofroute of administration on practi-cal considerations. Initiating treat-ment orally or parenterally isequally efficacious. The clinicianshould base the choice of agent onlocal antimicrobial sensitivity pat-terns (if available) and should ad-just the choice according to sensi-tivity testing of the isolateduropathogen (evidence quality: A;strong recommendation).

Action Statement 4b

The clinician should choose 7 to 14days as the duration of antimicrobialtherapy (evidence quality: B;recommendation).

The goals of treatment of acute UTI areto eliminate the acute infection, to pre-vent complications, and to reduce thelikelihood of renal damage. Most chil-dren can be treated orally.42–44 Patientswhom clinicians judge to be “toxic” orwho are unable to retain oral intake(including medications) should re-ceive an antimicrobial agent parenter-





ally (Table 2) until they exhibit clinicalimprovement, generally within 24 to 48hours, and are able to retain orally ad-ministered fluids andmedications. In astudy of 309 febrile infants with UTIs,only 3 (1%) were deemed too ill to beassigned randomly to either paren-teral or oral treatment.42 Parenteraladministration of an antimicrobialagent also should be considered whencompliance with obtaining an antimi-crobial agent and/or administering itorally is uncertain. The usual choicesfor oral treatment of UTIs includea cephalosporin, amoxicillin plusclavulanic acid, or trimethoprim-sulfamethoxazole (Table 3). It is essen-tial to know local patterns of suscepti-bility of coliforms to antimicrobialagents, particularly trimethoprim-sulfamethoxazole and cephalexin, be-cause there is substantial geographicvariability that needs to be taken intoaccount during selection of an antimi-crobial agent before sensitivity resultsare available. Agents that are excretedin the urine but do not achieve thera-peutic concentrations in the blood-stream, such as nitrofurantoin, shouldnot be used to treat febrile infants withUTIs, because parenchymal and serumantimicrobial concentrations may beinsufficient to treat pyelonephritis orurosepsis.

Whether the initial route of administra-tion of the antimicrobial agent is oralor parenteral (then changed to oral),

the total course of therapy should be 7to 14 days. The committee attempted toidentify a single, preferred, evidence-based duration, rather than a range, butdata comparing 7, 10, and 14 days di-rectly were not found. There is evidencethat 1- to 3-day courses for febrile UTIsare inferior to courses in the recom-mended range; therefore, the minimalduration selected should be 7 days.

● Aggregate quality of evidence: A/B(RCTs).

● Benefits: Adequate treatment of UTIcan prevent the spread of infectionand renal scarring. Outcomes ofshort courses (1–3 d) are inferior tothose of 7- to 14-d courses.

● Harms/risks/costs: There are mini-mal harm and minor cost effects ofantimicrobial choice and durationof therapy.


● Value judgments: Adjusting antimi-crobial choice on the basis of avail-able data and treating according tobest evidencewill minimize cost andconsequences of failed or unneces-sary treatment.

● Role of patient preferences: It is as-sumed that parents prefer themost-effective treatment and theleast amount of medication that en-sures effective treatment.


● Intentional vagueness: No evidence

distinguishes the benefit of treating7 vs 10 vs 14 days, and the range isallowable.

● Policy level: Strong recommendation/recommendation.

Action Statement 5

Febrile infants with UTIs shouldundergo renal and bladder ultra-sonography (RBUS) (evidencequality: C; recommendation).

The purpose of RBUS is to detect ana-tomic abnormalities that require fur-ther evaluation, such as additional im-aging or urologic consultation. RBUSalso provides an evaluation of the re-nal parenchyma and an assessment ofrenal size that can be used to monitorrenal growth. The yield of actionablefindings is relatively low.45,46 Wide-spread application of prenatal ultra-sonography clearly has reduced theprevalence of previously unsuspectedobstructive uropathy in infants, but theconsequences of prenatal screeningwith respect to the risk of renal abnor-malities in infants with UTIs have notyet been well defined. There is consid-erable variability in the timing andquality of prenatal ultrasonograms,and the report of “normal” ultrasono-graphic results cannot necessarily berelied on to dismiss completely thepossibility of a structural abnormalityunless the study was a detailed ana-tomic survey (with measurements),was performed during the third tri-

TABLE 2 Some Empiric Antimicrobial Agentsfor Parenteral Treatment of UTI

AntimicrobialAgent

Dosage

Ceftriaxone 75 mg/kg, every 24 hCefotaxime 150 mg/kg per d,

divided every 6–8 hCeftazidime 100–150 mg/kg per d,

divided every 8 hGentamicin 7.5 mg/kg per d,

divided every 8 hTobramycin 5 mg/kg per d,

divided every 8 hPiperacillin 300 mg/kg per d,

divided every 6–8 h

TABLE 3 Some Empiric Antimicrobial Agents for Oral Treatment of UTI

Antimicrobial Agent Dosage

Amoxicillin-clavulanate 20–40 mg/kg per d in 3 dosesSulfonamideTrimethoprim-sulfamethoxazole 6–12 mg/kg trimethoprim and 30-60 mg/kg sulfamethoxazole

per d in 2 dosesSulfisoxazole 120–150 mg/kg per d in 4 doses

CephalosporinCefixime 8 mg/kg per d in 1 doseCefpodoxime 10 mg/kg per d in 2 dosesCefprozil 30 mg/kg per d in 2 dosesCefuroxime axetil 20–30 mg/kg per d in 2 dosesCephalexin 50–100 mg/kg per d in 4 doses



mester, and was performed and inter-preted by qualified individuals.47

The timing of RBUS depends on theclinical situation. RBUS is recom-mended during the first 2 days of treat-ment to identify serious complications,such as renal or perirenal abscessesor pyonephrosis associated with ob-structive uropathy when the clinical ill-ness is unusually severe or substantialclinical improvement is not occurring.For febrile infants with UTIs who dem-onstrate substantial clinical improve-ment, however, imaging does not needto occur early during the acute infec-tion and can even be misleading; ani-mal studies demonstrate that Esche-richia coli endotoxin can producedilation during acute infection, whichcould be confused with hydronephro-sis, pyonephrosis, or obstruction.48

Changes in the size and shape of thekidneys and the echogenicity of renalparenchyma attributable to edemaalso are common during acute infec-tion. The presence of these abnormal-ities makes it inappropriate to con-sider RBUS performed early duringacute infection to be a true baselinestudy for later comparisons in the as-sessment of renal growth.

Nuclear scanning with technetium-labeled dimercaptosuccinic acid hasgreater sensitivity for detection ofacute pyelonephritis and later scar-ring than does either RBUS or voidingcystourethrography (VCUG). The scan-ning is useful in research, because itensures that all subjects in a studyhave pyelonephritis to start with and itpermits assessment of later renalscarring as an outcome measure. Thefindings on nuclear scans rarely affectacute clinical management, however,and are not recommended as part ofroutine evaluation of infants with theirfirst febrile UTI. The radiation dose tothe patient during dimercaptosuccinicacid scanning is generally low (�1mSv),49 although it may be increased in

children with reduced renal function.The radiation dose from dimercapto-succinic acid is additive with that ofVCUG when both studies are per-formed.50 The radiation dose fromVCUG depends on the equipment thatis used (conventional versus pulseddigital fluoroscopy) and is related di-rectly to the total fluoroscopy time.Moreover, the total exposure for thechild will be increased when bothacute and follow-up studies are ob-tained. The lack of exposure to radi-ation is a major advantage of RBUS,even with recognition of the limita-tions of this modality that were de-scribed previously.


● Benefits: RBUS in this populationwill yield abnormal results in�15%of cases, and 1% to 2% will have ab-normalities that would lead to ac-tion (eg, additional evaluation, re-ferral, or surgery).

● Harms/risks/costs: Between 2%and 3%will be false-positive results,leading to unnecessary and invasiveevaluations.


● Value judgments: The seriousnessof the potentially correctable abnor-malities in 1% to 2%, coupled withthe absence of physical harm, wasjudged sufficiently important to tipthe scales in favor of testing.

● Role of patient preferences: Be-cause ultrasonography is noninva-sive and poses minimal risk, we as-sume that parents will prefer RBUSover taking even a small risk ofmissing a serious and correctablecondition.




Action Statement 6

Action Statement 6a

VCUG should not be performedroutinely after the first febrileUTI; VCUG is indicated if RBUS re-veals hydronephrosis, scarring,or other findings that would sug-gest either high-grade VUR or ob-structive uropathy, as well as inother atypical or complex clinicalcircumstances (evidence qualityB; recommendation).

Action Statement 6b

Further evaluation should be con-ducted if there is a recurrence of fe-brile UTI (evidence quality: X;recommendation).

For the past 4 decades, the strategy toprotect the kidneys from further dam-age after an initial UTI has been to de-tect childhood genitourinary abnor-malities in which recurrent UTI couldincrease renal damage. The most com-mon of these is VUR, and VCUG is usedto detect this. Management includedcontinuous antimicrobial administra-tion as prophylaxis and surgical inter-vention if VUR was persistent or recur-rences of infection were not preventedwith an antimicrobial prophylaxis reg-imen; some have advocated surgicalintervention to correct high-grade re-flux even when infection has not re-curred. However, it is clear that thereare a significant number of infantswho develop pyelonephritis in whomVUR cannot be demonstrated, and theeffectiveness of antimicrobial prophy-laxis for patients who have VUR hasbeen challenged in the past decade.Several studies have suggested thatprophylaxis does not confer the de-sired benefit of preventing recurrentfebrile UTI.51–55 If prophylaxis is, in fact,not beneficial and VUR is not requiredfor development of pyelonephritis,then the rationale for performingVCUG routinely after an initial febrileUTI must be questioned.





RCTs of the effectiveness of prophy-laxis performed to date generally in-cluded children more than 24 monthsof age, and some did not provide com-plete data according to grade of VUR.These 2 factors have compromisedmeta-analyses. To ensure direct com-parisons, the committee contacted the6 researchers who had conducted themost recent RCTs and requested rawdata from their studies.51–56 All com-plied, which permitted the creation ofa data set with data for 1091 infants 2to 24months of age according to gradeof VUR. A �2 analysis (2-tailed) and aformal meta-analysis did not detect astatistically significant benefit of pro-phylaxis in preventing recurrence offebrile UTI/pyelonephritis in infantswithout reflux or those with grades I, II,III, or IV VUR (Table 4 and Fig 3). Only 5infants with grade V VUR were in-cluded in the RCTs; therefore, data forthose infants are not included in Table4 or Fig 3.

The proportion of infants with high-grade VUR among all infants with fe-brile UTIs is small. Data adapted fromcurrent studies (Table 5) indicate that,of a hypothetical cohort of 100 infantswith febrile UTIs, only 1 has grade VVUR; 99 do not. With a practice of wait-ing for a second UTI to perform VCUG,only 10 of the 100 would need to un-dergo the procedure and the 1 withgrade V VUR would be identified. (Italso is possible that the 1 infant withgrade V VURmight have been identifiedafter the first UTI on the basis of abnor-mal RBUS results that prompted VCUGto be performed.) Data to quantify ad-ditional potential harm to an infantwho is not revealed to have high-gradeVUR until a second UTI are not precisebut suggest that the increment is in-sufficient to justify routinely subject-ing all infants with an initial febrile UTIto VCUG (Fig 4). To minimize any harmincurred by that infant, attempts havebeen made to identify, at the time of

the initial UTI, those who have thegreatest likelihood of having high-grade VUR. Unfortunately, there are noclinical or laboratory indicators thathave been demonstrated to identify in-fants with high-grade VUR. Indicationsfor VCUG have been proposed on thebasis of consensus in the absence ofdata57; the predictive value of any of theindications for VCUG proposed in thismanner is not known.

The level of evidence supporting rou-tine imaging with VCUG was deemedinsufficient at the time of the 1999practice parameter to receive a rec-ommendation, but the consensus ofthe subcommittee was to “strongly en-courage” imaging studies. The positionof the current subcommittee reflectsthe new evidence demonstrating anti-microbial prophylaxis not to be effec-tive as presumed previously. More-over, prompt diagnosis and effectivetreatment of a febrile UTI recurrencemay be of greater importance regard-less of whether VUR is present or thechild is receiving antimicrobial pro-phylaxis. A national study (the Ran-domized Intervention for Children WithVesicoureteral Reflux study) is cur-rently in progress to identify the ef-fects of a prophylactic antimicrobialregimen for children 2 months to 6years of age who have experienced aUTI, and it is anticipated to provide ad-ditional important data58 (see Areasfor Research).

Action Statement 6a

● Aggregate quality of evidence: B(RCTs).

● Benefits: This avoids, for the vastmajority of febrile infants with UTIs,radiation exposure (of particularconcern near the ovaries in girls),expense, and discomfort.

● Harms/risks/costs: Detection of asmall number of cases of high-grade reflux and correctable abnor-malities is delayed.


● Value judgments: The risks associ-ated with radiation (plus the ex-pense and discomfort of the proce-dure) for the vastmajority of infantsoutweigh the risk of delaying the de-tection of the few with correctableabnormalities until their second UTI.

● Role of patient preferences: Thejudgment of parents may come intoplay, because VCUG is an uncomfort-able procedure involving radiationexposure. In some cases, parentsmay prefer to subject their childrento the procedure even when thechance of benefit is both small anduncertain. Antimicrobial prophy-laxis seems to be ineffective in pre-venting recurrence of febrile UTI/py-elonephritis for the vast majority ofinfants. Some parents may want toavoid VCUG even after the secondUTI. Because the benefit of identify-ing high-grade reflux is still in somedoubt, these preferences should beconsidered. It is the judgment of thecommittee that VCUG is indicated af-ter the second UTI.


TABLE 4 Recurrences of Febrile UTI/Pyelonephritis in Infants 2 to 24 Months of Age With andWithout Antimicrobial Prophylaxis, According to Grade of VUR

RefluxGrade

Prophylaxis No Prophylaxis P

No. ofRecurrences

Total N No. ofRecurrences

Total N

None 7 210 11 163 .15I 2 37 2 35 1.00II 11 133 10 124 .95III 31 140 40 145 .29IV 16 55 21 49 .14





Action Statement 6b

● Aggregate quality of evidence: X (ex-ceptional situation).

● Benefits: VCUG after a second UTIshould identify infants with veryhigh-grade reflux.

● Harms/risks/costs: VCUG is an un-comfortable, costly procedure thatinvolves radiation, including to theovaries of girls.


● Value judgments: The committeejudged that patients with high-grade reflux and other abnormali-ties may benefit from interventionsto prevent further scarring. Furtherstudies of treatment for grade VVUR are not underway and are un-likely in the near future, because thecondition is uncommon and ran-domization of treatment in thisgroup generally has been consid-ered unethical.

FIGURE 3A, Recurrences of febrile UTI/pyelonephritis in 373 infants 2 to 24 months of age without VUR, with andwithout antimicrobial prophylaxis (based on 3 studies; data provided by Drs Craig, Garin, and Mon-tini). B, Recurrences of febrile UTI/pyelonephritis in 72 infants 2 to 24 months of age with grade I VUR,with and without antimicrobial prophylaxis (based on 4 studies; data provided by Drs Craig, Garin,Montini, and Roussey-Kesler). C, Recurrences of febrile UTI/pyelonephritis in 257 infants 2 to 24months of age with grade II VUR, with and without antimicrobial prophylaxis (based on 5 studies; dataprovided by Drs Craig, Garin, Montini, Pennesi, and Roussey-Kesler). D, Recurrences of febrile UTI/pyelonephritis in 285 infants 2 to 24 months of age with grade III VUR, with and without antimicrobialprophylaxis (based on 6 studies; data provided by Drs Brandström, Craig, Garin, Montini, Pennesi, andRoussey-Kesler). E, Recurrences of febrile UTI/pyelonephritis in 104 infants 2 to 24 months of age withgrade IV VUR, with and without antimicrobial prophylaxis (based on 3 studies; data provided by DrsBrandström, Craig, and Pennesi). M-H indicates Mantel-Haenszel; CI, confidence interval.

TABLE 5 Rates of VUR According to Grade inHypothetical Cohort of Infants AfterFirst UTI and After Recurrence

Rate, %

After FirstUTI

(N� 100)

AfterRecurrence(N� 10)

No VUR 65 26Grades I–III VUR 29 56Grade IV VUR 5 12Grade V VUR 1 6

FIGURE 4Relationship between renal scarring and num-ber of bouts of pyelonephritis. Adapted fromJodal.59





● Role of patient preferences: As men-tioned previously, the judgment ofparents may come into play, be-cause VCUG is an uncomfortableprocedure involving radiation expo-sure. In some cases, parents mayprefer to subject their children tothe procedure even when thechance of benefit is both small anduncertain. The benefits of treatmentof VUR remain unproven, but thepoint estimates suggest a small po-tential benefit. Similarly, parentsmay want to avoid VCUG even afterthe second UTI. Because the benefitof identifying high-grade reflux isstill in some doubt, these prefer-ences should be considered. It is thejudgment of the committee thatVCUG is indicated after the secondUTI.


● Intentional vagueness: Further eval-uation will likely start with VCUG butmay entail additional studies de-pending on the findings. The detailsof further evaluation are beyond thescope of this guideline.


Action Statement 7

After confirmation of UTI, the cli-nician should instruct parents orguardians to seek prompt medicalevaluation (ideally within 48hours) for future febrile ill-nesses, to ensure that recurrentinfections can be detected andtreated promptly (evidence qual-ity: C; recommendation).

Early treatment limits renal damagebetter than late treatment,1,2 and therisk of renal scarring increases as thenumber of recurrences increase (Fig4).59 For these reasons, all infants whohave sustained a febrile UTI shouldhave a urine specimen obtained at theonset of subsequent febrile illnesses,so that a UTI can be diagnosed andtreated promptly.


● Benefits: Studies suggest that earlytreatment of UTI reduces the risk ofrenal scarring.

● Harms/risks/costs: There may beadditional costs and inconvenienceto parents withmore-frequent visitsto the clinician for evaluation offever.


● Value judgments: None.

● Role of patient preferences: Parentswill ultimatelymake the judgment toseek medical care.




CONCLUSIONS

The committee formulated 7 key actionstatements for the diagnosis andtreatment of infants and young chil-dren 2 to 24months of age with UTI andunexplained fever. Strategies for diag-nosis and treatment depend onwhether the clinician determines thatantimicrobial therapy is warranted im-mediately or can be delayed safely un-til urine culture and urinalysis resultsare available. Diagnosis is based onthe presence of pyuria and at least50 000 CFUs per mL of a single uro-pathogen in an appropriately collectedspecimen of urine; urinalysis alonedoes not provide a definitive diagnosis.After 7 to 14 days of antimicrobialtreatment, close clinical follow-upmonitoring should bemaintained, withevaluation of the urine during subse-quent febrile episodes to permitprompt diagnosis and treatment of re-current infections. Ultrasonography ofthe kidneys and bladder should be per-formed to detect anatomic abnormali-ties that require further evaluation(eg, additional imaging or urologicconsultation). Routine VCUG after the

first UTI is not recommended; VCUG isindicated if RBUS reveals hydrone-phrosis, scarring, or other findingsthat would suggest either high-gradeVUR or obstructive uropathy, as well asin other atypical or complex clinicalcircumstances. VCUG also should beperformed if there is a recurrence offebrile UTI.

AREAS FOR RESEARCH

One of the major values of a compre-hensive literature review is the identi-fication of areas in which evidence islacking. The following 8 areas are pre-sented in an order that parallels theprevious discussion.

1. The relationship between UTIs in in-fants and young children and re-duced renal function in adults hasbeen established but is notwell characterized in quantitativeterms. The ideal prospective cohortstudy from birth to 40 to 50 years ofage has not been conducted and isunlikely to be conducted. There-fore, estimates of undesirableoutcomes in adulthood, such ashypertension and end-stage renaldisease, are based on the mathe-matical product of probabilitiesat several steps, each of which issubject to bias and error. Otherattempts at decision analysis andthoughtful literature review haverecognized the same limitations.Until recently, imaging tools avail-able for assessment of the effectsof UTIs have been insensitive. Withthe imaging techniques now avail-able, it may be possible to identifythe relationship of scarring to re-nal impairment and hypertension.

2. The development of techniques thatwould permit an alternative to inva-sive sampling and culture would bevaluable for general use. Special at-tention should be given to infantgirls and uncircumcised boys, be-cause urethral catheterization may



be difficult and can produce con-taminated specimens and SPA nowis not commonly performed. Incu-bation time, which is inherent in theculture process, results in delayedtreatment or presumptive treat-ment on the basis of tests that lackthe desired sensitivity and specific-ity to replace culture.

3. The role of VUR (and therefore ofVCUG) is incompletely understood.It is recognized that pyelonephritis(defined through cortical scintigra-phy) can occur in the absence ofVUR (defined through VCUG) andthat progressive renal scarring(defined through cortical scintigra-phy) can occur in the absence ofdemonstrated VUR.52,53 The pre-sumption that antimicrobial pro-phylaxis is of benefit for individualswith VUR to prevent recurrences ofUTI or the development of renal scarsis not supported by the aggregate ofdata from recent studies and cur-rently is the subject of the Random-ized Intervention for Children WithVesicoureteral Reflux study.58

4. Although the effectiveness of anti-microbial prophylaxis for the pre-vention of UTI has not been demon-strated, the concept has biologicalplausibility. Virtually all antimicro-bial agents used to treat or to pre-vent infections of the urinary tractare excreted in the urine in highconcentrations. Barriers to the ef-fectiveness of antimicrobial pro-phylaxis are adherence to a dailyregimen, adverse effects associ-ated with the various agents, andthe potential for emergence of anti-

microbial resistance. To overcomethese issues, evidence of effective-ness with a well-tolerated, safeproduct would be required, andparents would need sufficient edu-cation to understand the value andimportance of adherence. A urinaryantiseptic, rather than an antimi-crobial agent, would be particularlydesirable, because it could be takenindefinitely without concern thatbacteria would develop resistance.Another possible strategy might bethe use of probiotics.

5. Better understanding of the ge-nome (human and bacterial) mayprovide insight into risk factors(VUR and others) that lead to in-creased scarring. Blood specimenswill be retained from children en-rolled in the Randomized Interven-tion for Children With Vesi-coureteral Reflux study, for futureexamination of genetic determinantsof VUR, recurrent UTI, and renal scar-ring.58 VUR is recognized to “run infamilies,”60,61 and multiple investiga-tors are currently engaged in re-search to identify a genetic basis forVUR. Studies may also be able to dis-tinguish the contribution of congeni-tal dysplasia from acquired scarringattributable to UTI.

6. One of the factors used to assessthe likelihood of UTI in febrile in-fants is race. Data regarding ratesamong Hispanic individuals are lim-ited and would be useful for predic-tion rules.

7. This guideline is limited to the initialmanagement of the first UTI in febrileinfants 2 to 24monthsof age. Someof

the infants will have recurrent UTIs;some will be identified as having VURor other abnormalities. Further re-search addressing the optimalcourse ofmanagement in specific sit-uations would be valuable.

8. The optimal duration of antimicro-bial treatment has not been deter-mined. RCTs of head-to-head com-parisons of various duration wouldbe valuable, enabling clinicians tolimit antimicrobial exposure towhat is needed to eradicate the of-fending uropathogen.

LEAD AUTHORKenneth B. Roberts, MD

SUBCOMMITTEE ON URINARY TRACTINFECTION, 2009–2011Kenneth B. Roberts, MD, ChairStephen M. Downs, MD, MSS. Maria E. Finnell, MD, MSStanley Hellerstein, MDLinda D. Shortliffe, MDEllen R. Wald, MDJ. Michael Zerin, MD

OVERSIGHT BY THE STEERINGCOMMITTEE ON QUALITYIMPROVEMENT AND MANAGEMENT,2009–2011

STAFFCaryn Davidson, MA

ACKNOWLEDGMENTSThe committee gratefully acknowl-edges the generosity of the research-ers who graciously shared their datato permit the data set with data for1091 infants aged 2 to 24 months ac-cording to grade of VUR to be com-piled, that is, Drs Per Brandström,Jonathan Craig, Eduardo Garin, Gio-vanni Montini, Marco Pennesi, andGwenaelle Roussey-Kesler.

REFERENCES

1. Winter AL, Hardy BE, Alton DJ, Arbus GS,Churchill BM. Acquired renal scars in chil-dren. J Urol. 1983;129(6):1190–1194

2. Smellie JM, Poulton A, Prescod NP. Retro-spective study of children with renal scar-ring associated with reflux and urinary in-fection. BMJ. 1994;308(6938):1193–1196

3. American Academy of Pediatrics, Commit-tee on Quality Improvement, Subcommitteeon Urinary Tract Infection. Practiceparameter: the diagnosis, treatment, andevaluation of the initial urinary tract infec-tion in febrile infants and young children.Pediatrics. 1999;103(4):843–852

4. Finnell SM,Carroll AE,DownsSM,etal. Technicalreport: diagnosis andmanagement of an initialurinary tract infection in febrile infants andyoung children. Pediatrics. 2011;128(3):e749

5. American Academy of Pediatrics, SteeringCommittee on Quality Improvement andManagement. Classifying recommenda-





tions for clinical practice guidelines. Pedi-atrics. 2004;114(3):874–877

6. Leong YY, Tan KW. Bladder aspiration fordiagnosis of urinary tract infection in in-fants and young children. J Singapore Pae-diatr Soc. 1976;18(1):43–47

7. Pryles CV, Atkin MD, Morse TS, Welch KJ.Comparative bacteriologic study of urineobtained from children by percutaneoussuprapubic aspiration of the bladder andby catheter. Pediatrics. 1959;24(6):983–991

8. Djojohadipringgo S, Abdul Hamid RH, ThahirS, Karim A, Darsono I. Bladder puncture innewborns: a bacteriological study. PaediatrIndones. 1976;16(11–12):527–534

9. Gochman RF, Karasic RB, Heller MB. Use ofportable ultrasound to assist urine collec-tion by suprapubic aspiration. Ann EmergMed. 1991;20(6):631–635

10. Buys H, Pead L, Hallett R, Maskell R. Supra-pubic aspiration under ultrasound guid-ance in children with fever of undiagnosedcause. BMJ. 1994;308(6930):690–692

11. Kramer MS, Tange SM, Drummond KN, MillsEL. Urine testing in young febrile children: arisk-benefit analysis. J Pediatr. 1994;125(1):6–13

12. Bonadio WA. Urine culturing technique in fe-brile infants. Pediatr Emerg Care. 1987;3(2):75–78

13. Lohr J. Pediatric Outpatient Procedures.Philadelphia, PA: Lippincott; 1991

14. Taylor CM, White RH. The feasibility ofscreening preschool children for urinarytract infection using dipslides. Int J PediatrNephrol. 1983;4(2):113–114

15. Sørensen K, Lose G, Nathan E. Urinary tractinfections and diurnal incontinence in girls.Eur J Pediatr. 1988;148(2):146–147

16. Shannon F, Sepp E, Rose G. The diagnosis ofbacteriuria by bladder puncture in infancyand childhood. Aust Pediatr J. 1969;5(2):97–100

17. Hoberman A, Chao HP, Keller DM, Hickey R,Davis HW, Ellis D. Prevalence of urinary tractinfection in febrile infants. J Pediatr. 1993;123(1):17–23

18. Haddon RA, Barnett PL, Grimwood K, HoggGG. Bacteraemia in febrile children present-ing to a paediatric emergency department.Med J Aust. 1999;170(10):475–478

19. Wiswell TE, Roscelli JD. Corroborative evi-dence for the decreased incidence of uri-nary tract infections in circumcised maleinfants. Pediatrics. 1986;78(1):96–99

20. To T, Agha M, Dick PT, Feldman W. Cohortstudy on circumcision of newborn boys andsubsequent risk of urinary-tract infection.Lancet. 1998;352(9143):1813–1816

21. Wiswell TE, Hachey WE. Urinary tract infec-tions and the uncircumcised state: an up-date. Clin Pediatr (Phila). 1993;32(3):130–134

22. Wiswell TE, Smith FR, Bass JW. Decreasedincidence of urinary tract infections in cir-cumcised male infants. Pediatrics. 1985;75(5):901–903

23. Ginsburg CM, McCracken GH Jr. Urinarytract infections in young infants. Pediatrics.1982;69(4):409–412

24. Craig JC, Knight JF, Sureshkumar P, MantzE, Roy LP. Effect of circumcision on inci-dence of urinary tract infection in pre-school boys. J Pediatr. 1996;128(1):23–27

25. Levine DA, Platt SL, Dayan PS, et al. Risk ofserious bacterial infection in young fe-brile infants with respiratory syncytial vi-rus infections. Pediatrics. 2004;113(6):1728–1734

26. Roberts KB, Charney E, Sweren RJ, et al. Uri-nary tract infection in infants with unex-plained fever: a collaborative study. J Pedi-atr. 1983;103(6):864–867

27. Gorelick MH, Hoberman A, Kearney D, WaldE, Shaw KN. Validation of a decision ruleidentifying febrile young girls at high riskfor urinary tract infection. Pediatr EmergCare. 2003;19(3):162–164

28. Gorelick MH, Shaw KN. Clinical decision ruleto identify febrile young girls at risk for uri-nary tract infection. Arch Pediatr AdolescMed. 2000;154(4):386–390

29. Shaw KN, Gorelick M, McGowan KL, YakscoeNM, Schwartz JS. Prevalence of urinarytract infection in febrile young children inthe emergency department. Pediatrics.1998;102(2). Available at: www.pediatrics.org/cgi/content/full/102/2/e16

30. Shaikh N,Morone NE, Lopez J, et al. Does thischild have a urinary tract infection? JAMA.2007;298(24):2895–2904

31. Powell HR, McCredie DA, Ritchie MA. Urinarynitrite in symptomatic and asymptomaticurinary infection. Arch Dis Child. 1987;62(2):138–140

32. Kunin CM, DeGroot JE. Sensitivity of a nitriteindicator strip method in detecting bacteri-uria in preschool girls. Pediatrics. 1977;60(2):244–245

33. Hoberman A, Wald ER, Reynolds EA, Pen-chansky L, Charron M. Is urine culture nec-essary to rule out urinary tract infection inyoung febrile children? Pediatr Infect Dis J.1996;15(4):304–309

34. Herr SM, Wald ER, Pitetti RD, Choi SS. En-hanced urinalysis improves identificationof febrile infants ages 60 days and youngerat low risk for serious bacterial illness. Pe-diatrics. 2001;108(4):866–871

35. Kunin C. A ten-year study of bacteriuria inschoolgirls: final report of bacteriologic,urologic, and epidemiologic findings. J In-fect Dis. 1970;122(5):382–393

36. Kemper K, Avner E. The case against screen-ing urinalyses for asymptomatic bacteri-uria in children. Am J Dis Child. 1992;146(3):343–346

37. Wald E. Genitourinary tract infections: cysti-tis and pyelonephritis. In: Feigin R, CherryJD, Demmler GJ, Kaplan SL, eds. Textbook ofPediatric Infectious Diseases. 5th ed. Phila-delphia, PA: Saunders; 2004:541–555

38. Mayo S, Acevedo D, Quiñones-Torrelo C,Canós I, Sancho M. Clinical laboratory auto-mated urinalysis: comparison among auto-mated microscopy, flow cytometry, two teststrips analyzers, and manual microscopicexamination of the urine sediments. J ClinLab Anal. 2008;22(4):262–270

39. Kass E. Asymptomatic infections of the uri-nary tract. Trans Assoc Am Phys. 1956;69:56–64

40. Hoberman A, Wald ER, Reynolds EA, Pen-chansky L, Charron M. Pyuria and bacteri-uria in urine specimens obtained by cathe-ter from young children with fever. JPediatr. 1994;124(4):513–519

41. Downs SM. Technical report: urinary tract infec-tions in febrile infants and young children. Pedi-atrics. 1999;103(4). Availableat:www.pediatrics.org/cgi/content/full/103/4/e54

42. Hoberman A, Wald ER, Hickey RW, et al. Oralversus initial intravenous therapy for uri-nary tract infections in young febrile chil-dren. Pediatrics. 1999;104(1):79–86

43. Hodson EM, Willis NS, Craig JC. Antibioticsfor acute pyelonephritis in children. Co-chrane Database Syst Rev . 2007;(4):CD003772

44. Bloomfield P, Hodson EM, Craig JC. Antibiot-ics for acute pyelonephritis in children. Co-chrane Database Syst Rev . 2005;(1):CD003772

45. Hoberman A, Charron M, Hickey RW, BaskinM, Kearney DH, Wald ER. Imaging studies af-ter a first febrile urinary tract infection inyoung children. N Engl J Med. 2003;348(3):195–202

46. Jahnukainen T, Honkinen O, Ruuskanen O,Mertsola J. Ultrasonography after the firstfebrile urinary tract infection in children.Eur J Pediatr. 2006;165(8):556–559

47. Economou G, Egginton J, Brookfield D. Theimportance of late pregnancy scans for re-nal tract abnormalities. Prenat Diagn. 1994;14(3):177–180

48. Roberts J. Experimental pyelonephritis inthe monkey, part III: pathophysiology of ure-


www.pediatrics.org/cgi/content/full/102/2/e16





teral malfunction induced by bacteria. In-vest Urol. 1975;13(2):117–120

49. Smith T, Evans K, Lythgoe MF, Anderson PJ,Gordon I . Rad iat ion dos imetry oftechnetium-99m-DMSA in children. J NuclMed. 1996;37(8):1336–1342

50. Ward VL. Patient dose reduction duringvoiding cystourethrography. Pediatr Radiol.2006;36(suppl 2):168–172

51. Pennesi M, Travan L, Peratoner L, et al. Isantibiotic prophylaxis in children with vesi-coureteral reflux effective in preventing py-elonephritis and renal scars? A random-ized, controlled trial. Pediatrics. 2008;121(6). Available at: www.pediatrics.org/cgi/content/full/121/6/e1489

52. Garin EH, Olavarria F, Garcia Nieto V, Valen-ciano B, Campos A, Young L. Clinical signifi-cance of primary vesicoureteral reflux andurinary antibiotic prophylaxis after acutepyelonephritis: a multicenter, randomized,controlled study. Pediatrics. 2006;117(3):626–632

53. Montini G, Rigon L, Zucchetta P, et al. Pro-phylaxis after first febrile urinary tract in-fection in children? A multicenter, random-ized, controlled, noninferiority trial.Pediatrics. 2008;122(5):1064–1071

54. Roussey-Kesler G, Gadjos V, Idres N, et al.Antibiotic prophylaxis for the prevention ofrecurrent urinary tract infection in childrenwith low grade vesicoureteral reflux: re-sults from a prospective randomized study.J Urol. 2008;179(2):674–679

55. Craig J, Simpson J, Williams G. Antibioticprophylaxis and recurrent urinary tract in-fection in children. N Engl J Med. 2009;361(18):1748–1759

56. Brandström P, Esbjorner E, Herthelius M,Swerkersson S, Jodal U, Hansson S. TheSwedish Reflux Trial in Children, part III: uri-nary tract infection pattern. J Urol. 2010;184(1):286–291

57. National Institute for Health and Clinical Ex-cellence.Urinary Tract Infection in Children:Diagnosis, Treatment, and Long-term

Management: NICE Clinical Guideline 54.London, England: National Institute forHealth and Clinical Excellence; 2007. Avail-able at: www.nice.org.uk/nicemedia/live/11819/36032/36032.pdf. Accessed March14, 2011

58. Keren R, Carpenter MA, Hoberman A, et al.Rationale and design issues of the Random-ized Intervention for Children With Vesi-coureteral Reflux (RIVUR) study. Pediatrics.2008;122(suppl 5):S240–S250

59. Jodal U. The natural history of bacteriuria inchildhood. Infect Dis Clin North Am. 1987;1(4):713–729

60. Eccles MR, Bailey RR, Abbott GD, SullivanMJ.Unravelling the genetics of vesicouretericreflux: a common familial disorder. HumMol Genet. 1996;5(Spec No.):1425–1429

61. Scott JE, Swallow V, Coulthard MG, LambertHJ, Lee RE. Screening of newborn babies forfamilial ureteric reflux. Lancet. 1997;350(9075):396–400





www.nice.org.uk/nicemedia/live/11819/36032/36032.pdf

www.nice.org.uk/nicemedia/live/11819/36032/36032.pdf



APPENDIXClinical practice guideline algorithm.



DOI: 10.1542/peds.2011-1330; originally published online August 28, 2011;Pediatrics

COMMITTEE ON QUALITY IMPROVEMENT AND MANAGEMENTSUBCOMMITTEE ON URINARY TRACT INFECTION and STEERING

Management of the Initial UTI in Febrile Infants and Children 2 to 24 MonthsUrinary Tract Infection: Clinical Practice Guideline for the Diagnosis and

ServicesUpdated Information &

/peds.2011-1330http://pediatrics.aappublications.org/content/early/2011/08/24including high resolution figures, can be found at:

Citations

/peds.2011-1330#related-urlshttp://pediatrics.aappublications.org/content/early/2011/08/24This article has been cited by 7 HighWire-hosted articles:

Permissions & Licensing

mlhttp://pediatrics.aappublications.org/site/misc/Permissions.xhttables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints http://pediatrics.aappublications.org/site/misc/reprints.xhtml

Information about ordering reprints can be found online:

rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Grove Village, Illinois, 60007. Copyright © 2011 by the American Academy of Pediatrics. All and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elkpublication, it has been published continuously since 1948. PEDIATRICS is owned, published, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by guest on January 2, 2012pediatrics.aappublications.orgDownloaded from


http://pediatrics.aappublications.org/content/early/2011/08/24/peds.2011-1330#related-urls

http://pediatrics.aappublications.org/site/misc/Permissions.xhtml

http://pediatrics.aappublications.org/site/misc/reprints.xhtml


Health & Medicine

Itu 2011 aap