ORIGINAL ARTICLE

Comparison of procalcitonin and different guidelines for firstfebrile urinary tract infection in children by imaging

Pei-Fen Liao & Min-Sho Ku & Jeng-Dau Tsai &Yu-Hua Choa & Tung-Wei Hung & Ko-Huang Lue &Ji-Nan Sheu

Received: 17 December 2013 /Revised: 25 February 2014 /Accepted: 26 February 2014 /Published online: 20 March 2014# IPNA 2014

AbstractBackground We examined the ability of a procalcitonin (PCT)protocol to detect vesicoureteral reflux (VUR) and renal scar-ring (RS), evaluated procedural costs and radiation burden,and compared four representative guidelines for children withtheir first febrile urinary tract infection (UTI).Methods Children aged 2 years with their first febrile UTIwho underwent renal ultrasonography (US), acute and latetechnetium-99m (99mTc)-dimercaptosuccinic acid scan, andvoiding cystourethrography were prospectively studied. Therepresentative guidelines applied in a retrospective simulationincluded the American Academy of Pediatrics (AAP),National Institute of Clinical Excellence, top-down approach(TDA), and Italian Society of Pediatric Nephrology (ISPN).These were compared in terms of ability to detect abnormal-ities, procedural costs and radiation.Results Of 278 children analyzed, 172 (61.9 %) had acutepyelonephritis. There was VUR in 101 (36.3 %) children,including 73 (26.3 %) with grades IIIV VUR. RS was iden-tified in 75 (27.0 %) children. To detect VUR, TDA and PCT

had the highest sensitivity for grades IV VUR (80.2 %) andIIIV VUR (94.5 %), respectively, whereas AAP had thehighest specificity for IV VUR (77.4 %) and IIIV VUR(78.0 %), respectively. TDA and PCT had the highest sensi-tivity (100 %) for detecting RS. The highest cost and radiationdose was associated with TDA, whereas AAP had the leastexpenditure and radiation exposure. By multivariate analysis,PCT and VUR, especially grades IIIV, were independentpredictors of RS.Conclusions There is no perfect guideline for first febrile UTIchildren. The PCT protocol has good ability for detectinghigh-grade VUR and RS. If based on available imaging mo-dalities and reducing cost and radiation burden, clinical sug-gestions in the AAP guidelines represent a considerableprotocol.

Keywords Acute pyelonephritis . Renal ultrasonography .

Voiding cystourethrography . Renal scarring .99mTc-dimercaptosuccinic acid scan . Vesicoureteral reflux

Introduction

Urinary tract infection (UTI) is one of the most commonbacterial infections in febrile children 2 years [1]. The asso-ciation between UTI and congenital abnormalities, such asvesicoureteral reflux (VUR), may put children at a high riskfor acute pyelonephritis (APN) and subsequent renal scarring(RS) [1, 2]. Although long-term complications of VUR andinfection-related renal damage are being questioned [3, 4], it isbelieved that postpyelonephritic RS with recurrences, espe-cially in the presence of high-grade VUR, may cause futuremedical problems, such as hypertension and/or impaired kid-ney function. This has been the major driving force for thecomprehensive investigation and treatment of the first UTI[58]. Therefore, choosing a diagnostic approach is based on

Pei-Fen Liao and Min-Sho Ku contributed equally to this work.

P.

identifying children who have VUR that may lead to addi-tional infections and ongoing renal damage if not diagnosedand managed appropriately after the first febrile UTI. Existingguidelines propose different diagnostic algorithms that mini-mize invasive procedures while maintaining an acceptablediagnostic ability to detect abnormalities [1, 911]. Renalultrasonography (US), voiding cystourethrography (VCUG),and technetium-99m (99mTc)-dimercaptosuccinic acid(DMSA) scan are the core imaging methods. However, theirvarying diagnostic ability, availability, invasiveness, andprocedure-related complications prevent any consensus onthe best approach. La Scola et al. [12] evaluated the diagnosticability of the five different guidelines for VUR and RS as wellas incurred cost and radiation exposure. The authors conclud-ed there was no ideal diagnostic protocol after their first UTI.We demonstrated that procalcitonin (PCT) was a good predic-tor of APN and high-grade VUR and that PCT 1.0 ng/mlmight be used as the first step of a diagnostic algorithm forchildren after their first febrile UTI [13, 14]. We hypothesizedthat the detection of the high-grade VUR and RS was impor-tant for children after their first UTI [58]. This study aimed toexamine the PCT protocol and compared it with four repre-sentative guidelines for its ability to detect VUR and RS [1,911]. The cost and radiation burden of the different protocolsfor children with their first febrile UTI were also evaluated.

Methods

This prospective cohort study evaluated children aged2 yearswho were admitted to an urban tertiary referral center withtheir first febrile UTI. The hospitals Institutional ReviewBoard approved the study protocol, and the parents of allparticipants provided informed consent. Diagnosis of a firstfebrile UTI was based on the following:

1. Fever with body temperature 38 C2. Presence of pyuria, defined as five or more white blood

cells per high-power field (WBC/hpf) and/or abnormaldipstick urinalysis (positive nitrite or leukocyte esterasetests)

3. Positive urine culture defined as any growth of a singlebacterium in urine from a suprapubic bladder aspiration,or growth of a singlemicroorganism 105 colony-formingunits (CFU)/ml collected from the midstream clean-voidurine specimen, or 5104 CFU/ml collected from atransurethral catheterized specimen [1]

4. No previous history of UTI, kidney, bladder, or urogenitaldisorder

Serum and urine indices for laboratory investigations, in-cluding PCT, C-reactive protein (CRP), WBC count, serumcreatinine (sCr), urinalysis, and urine and blood cultures were

taken on admission and before initiation of antibiotic treat-ment. Rapid and quantitative measurements of PCT valueswere done using enzyme-linked fluorescent assay in anautomated instrument (VIDAS PCT, B R A H M SDiagnostica, Berlin, Germany) according to the manufac-turers instructions. The detection limit was 0.05 ng/ml, andPCT values 0.5 ng/ml were considered abnormal.

All children underwent complete imaging investigation,including US, acute and late DMSA renal scan, and VCUG;USwas performed within the first 3 days of admission, and allabnormal US findings considered to be related to VUR wererecorded. This included anteroposterior diameter of the renalpelvis 5 mm and/or any grade of dilatation of calyces orureters; pelvic or ureteral wall thickening; absence ofcorticomedullary differentiation; irregular renal outline andsigns of renal hypodysplasia (i.e., small kidney, thinned orhyperechoic cortex); and duplicated renal collecting system[13, 15]. DMSA scan was performed within the first 5 days ofadmission to verify the presence of renal lesions. An abnormalacute DMSA scan suggesting APN was defined as the pres-ence of focal or diffuse areas of decreased uptake with pres-ervation of the renal contour [16, 17]. If there was an abnormalacute DMSA scan, a late DMSA scanwas performed at least 6months after the acute infection to evaluate the presence of lateRS. Previous studies showed that all children with normalinitial scan results had normal scans on follow-up [18]. Thus,it was unnecessary to perform late DMSA studies on patientswith previously normal scans. RS was defined as the occur-rence of focal or generalized areas of persistent diminisheduptake at the same locations as in the acute DMSA scan and/orassociated with loss of kidney contour or cortical thinningwith reduced volume [16, 17]. VCUG was performed 13 weeks after control of acute infection and graded IVaccording to the International Reflux Study in Children [19].Interpretations of the DMSA and VCUG were made by asingle experienced nuclear medicine physician and a singleexperienced pediatric radiologist, both of whom were un-aware of the patients clinical and laboratory findings andwho were blinded to the study.

Diagnostic guidelines

In the PCT protocol (PCT 2013) [13], a value 1.0 ng/ml wasselected as the first step of the diagnostic algorithm instead ofacute DMSA scan and VCUG for revealing APN and VUR[13]. The four more recent (published after 2007), well-known, representative guidelines in different areas, includingthe American Academy of Pediatrics (AAP 2011) [1],National Institute of Clinical Excellence (NICE 2007) [9],top-down approach (TDA 2007) [10], and Italian Society ofPediatric Nephrology (ISPN 2012) [11], were selected. Theirapplication on eligible children from this PCT cohort studywas retrospectively simulated (Table 1). Diagnostic guidelines

1568 Pediatr Nephrol (2014) 29:15671574

were applied for their ability to detect VUR and RS and toevaluate total and individual costs and radiation burden. Thecompleted studies of children who underwent US, acute andlate DMSA scans, and VCUG during the protocol were usedas a reference to compare with the five different guidelines.

Costs, radiation exposure, and measurement outcomes

Costs incurred by the different protocols were calculatedbased on price lists of the Taiwan National Health InsuranceAdministration, as follows, in New Taiwan dollars (NTD):DMSA renal scan $2,750; US $870; VCUG $1,090; PCT$1,000). Reported radiation dose associated with the VCUGprocedure varied from 0.5 to 3.2 mSv [2]. DMSA renal scanradiation dose was estimated to be 1 mSv regardless of thechilds age [2, 20]. For convenience, both imaging methodswere used as radiation burden of 1 mSv in this study. Thediagnostic ability for detecting VUR (grades IV and gradesIIIV) and RS after the first febrile UTI was examined ac-cording to each guideline. Costing and radiation dose of eachdiagnostic protocol was then evaluated.

Statistical analyses

Nonparametric data were assessed using the MannWhitneyU test; chi-square test was used to compare group proportionswith qualitative data. Diagnostic values of sensitivity, speci-ficity, and positive (PPV) and negative (NPV) predictivevalues, and positive (LR+) and negative (LR) likelihoodratios were calculated. Multivariate analysis with stepwiseprocedure was used to identify potential predictors of RS.Results are expressed as odds ratio (OR), 95 % confi-dence interval (95 % CI), and p value. Statistical sig-nificance was set at p

in four, solitary kidney in three, and hypoplastic kidney in 21.Of the 278 children, VUR (maximum degree of reflux given ifbilateral) was diagnosed in 101 (36.3 %), including 73(26.3 %) with grades IIIV VUR. There was no differencein VUR incidence between boys and girls (p=0.549).Children with APN had significantly higher VUR rates thanthose with lower UTI (47.1 % vs. 18.9 %, p

16, 21, 23]. APN rate also significantly increased with thepresence and severity of VUR, similar to previous studies [8,10, 23, 24]. When VUR was present, 8090 % of childrenwith febrile UTI had acute abnormal DMSA scans [3, 8, 25],similar to findings presented here (80.2 %). Our results showthat 27.0 % of the entire population and 72.6 % of childrenwith IIIV VUR had RS, revealing a significant associationbetween RS and VUR grade, consistent with previous reports[8, 23, 24, 26]. Our findings also show that VUR is a riskfactor for RS, particularly in grades IIIV VUR [68, 24, 27],suggesting that VUR is a clinically significant condition[24, 28].

A diagnostic test with the highest sensitivity is not neces-sarily the best in view of the uncertainty of the benefits ofdetecting abnormalities (VUR and RS) in children with UTI.

On the other hand, if identifying abnormal test results does notmatter clinically, then the highest sensitivity is not necessarilythe best. Although the long-term clinical significance of VURand UTI-related RS in previously healthy kidneys remainsincompletely understood [29], reports show a significant cor-relation between reflux grade andRS frequency, suggesting thatVUR increases the risk of RS formation after UTI [68, 24, 27],which may lead to late hypertension, proteinuria, and impairedrenal function [5, 30]. Early VUR andUTI-related RS detectionis therefore of key importance in the diagnostic assessment ofany infant and young child with a first febrile UTI. Althoughthe five different diagnostic protocols are designed in differentalgorithms to evaluate children at the time of first febrile UTI,they have the same goal of detecting any child who is at a highrisk for renal damage and also to reduce unnecessary imagingfollowing UTI. A more comprehensive protocol has a highersensitivity for detecting VUR and RS but is also costly andincurs risks of exposing children to attendant radiation andsedation. Thus, it is reasonable to design a less invasive algo-rithm by which to identify abnormalities and RS and to reducecost and radiation exposure in a selected population.

Previous studies show that PCT is a reliable marker fordifferentiating APN from lower UTI in children [14, 3133].In cases in which the PCT value in children is

probability of detecting abnormalities but also increase furtherexaminations. The study we present here also shows that theNICE guideline has the least sensitivity for detecting VURbecause further imaging is only performed in highly selectedchildren, consistent with findings of a study by Ristola et al.[34]. That study evaluated the applicability of NICE guidelinesfor imaging studies in children

Nonetheless, the study also has several limitations: First,for convenience, the same radiation dose was used for bothDMSA and VCUG, and radiation exposure may be somewhatdifferent in different procedures and individuals. Second, asthe study was limited to hospitalized children at the time of thefirst febrile UTI, findings may not be immediately generalizedor extrapolated to outpatients with UTI and to children withafebrile or recurrent UTIs. Finally, the cost effectiveness de-rived from our study may be different than that in other heath-care markets. Study results must therefore be used with cau-tion in other parts of the world.

In conclusion, there exists markedly varying diagnosticabilities among different guidelines for detecting abnormali-ties and considering costs and radiation exposure in youngchildren and infants with their first febrile UTI. The PCT-based strategy is a good diagnostic protocol with which todetect high-grade VUR and RS in that population. However,higher costs and greater radiation burden remain shortcomingsof the PCT protocol. Pediatricians should make every effort toreduce unnecessary imaging in such children. If based onreducing costs and radiation burden, and considering theavailability of imaging modalities, clinical suggestions in theAAP guidelines represent a considerable protocol.

Acknowledgments This study was supported by grants from the Na-tional Science Council, Taiwan (NSC93-2314-B-040-012) and the ChungShan Medical University Hospital, Taiwan (CSH-2012-C-010).

References

1. Subcommittee on Urinary tract Infection, Steering Committee onQuality Improvement and Management (2011) Urinary tract infec-tion: clinical practice guideline for the diagnosis and management ofthe initial UTI in febrile infants and children 2 to 24 months.Pediatrics 128:595610

2. Stefanidis CJ, Siomou E (2007) Imaging strategies for vesicoureteralreflux diagnosis. Pediatr Nephrol 22:937947

3. Rushton HG (1997) The evaluation of acute pyelonephritis and renalscarring with technetium 99 m-dimercaptosuccinic acid renal scin-tigraphy: evolving concepts and future directions. Pediatr Nephrol11:108120

4. Moorthy I, Easty M, McHugh K, Ridout D, Biassoni L, Gordon I(2005) The presence of vesicoureteric reflux does not identifying apopulation at risk for renal scarring following a first urinary tractinfection. Arch Dis Child 90:733736

5. Lahdes-Vasama T, Niskanen K, Rnnholm K (2006) Outcome ofkidneys in patients treated for vesicoureteral reflux (VUR) duringchildhood. Nephrol Dial Transplant 21:24912497

6. Faust W, Diaz M, Pohl H (2009) Incidence of post-pyelonephriticrenal scarring: a meta-analysis of the dimercapto-succinic acid liter-ature. J Urol 181:290298

7. Shaikh N, EwingAL, Bhatnagar S, HobermanA (2010) Risk of renalscarring in children with a first urinary tract infection: a systematicreview. Pediatrics 126:10841091

8. Orellana P, Baquedano P, Rangarajan V, Zhao JH, Eng ND, Fettich J,Chaiwatanarat T, Sonmezoglu K, Kumar D, Park YH, Samuel AM,Sixt R, Bhatnagar V, Padhy AK (2004) Relationship between acute

pyelonephritis, renal scarring, and vesicoureteral reflux. PediatrNephrol 19:11221126

9. National Institute for Health and Clinical Excellence (2007)Urinary tract infection in children: diagnosis, treatment and long termmanagement. Available at: www.nice.org.uk/nicemedia/pdf/CG54fullguideline.pdf. (Date accessed 15 Jun 2013)

10. Preda I, Jodal U, Sixt R, Stokland E, Hansson S (2007)Normal dimercaptosuccinic acid scintigraphy makes voidingcystourethrography unnecessary after urinary tract infection. JPediatr 151:581584

11. Ammenti A, Cataldi L, Chimenz R, Fanos V, Manna AL, Marra G,Materssi M, Pcecile P, Pennesi M, Pisanello L, Sica F, Toffolo A,Montini G (2012) Febrile urinary tract infections in young children:recommendations for the diagnosis, treatment and follow-up. ActaPaediatr 101:451457

12. La Scola C, DeMutiis C, Hewitt IK, Puccio G, Toffolo A, Zucchtta P,Mencarelli F, Marsciani M, DallAmico R, Montini G (2013)Different guidelines for imaging after first UTI in febrile infants:yield, cost, and radiation. Pediatrics 131:e665e671

13. Sun HL, Wu KH, Chen SM, Chao YH, Ku MS, Hung TH, Liao PF,Lue KH, Sheu JN (2013) Role of procalcitonin in predicting dilatingvesicoureteral reflux in young children hospitalizedwith a first febrileurinary tract infection. Pediatr Infect Dis J 32:e348e354

14. Sheu JN, Chang HM, Chen SM, Hung TW, Lue KH (2011)The role of procalcitonin on acute pyelonephritis and subse-quent renal scarring in infants and young children. J Urol186:20022009

15. Avni EF, Ayadi K, Rypens F, Hall M, Schulman CC (1997) Cancareful ultrasound examination of the urinary tract excludevesicoureteric reflux in the neonate? Br J Radiol 70:977982

16. Mandell GA, Eggli DF, Gilday DL, Heyman S, Leonard JC, MillerJH, Nadel HR, Treves ST (1997) Procedure guideline for renalcortical scintigraphy in children. J Nucl Med 38:16441646

17. Sheu JN, Wu KH, Chen SM, Tsai JD, Chao YH, Lue KH (2013)Acute 99mTc DMSA scan predicts dilating vesicoureteral reflux inyoung children with a first febrile urinary tract infection: apopulation-based cohort study. Clin Nucl Med 38:163168

18. HobermanA, CharronM, Hickey RW, BaskinM, Kearney DH,WaldER (2003) Imaging studies after a first febrile urinary tract infectionin young children. N Engl J Med 348:195202

19. Lebowitz RL, Olbing H, Parkkulainen KV, Smellie JM, Tamminen-Mbius TE (1985) International system of radiographic grading ofvesicoureteric reflux. International Reflux Study in Children. PediatrRadiol 15:105109

20. Smith T, Evans K, Lythgoe MF, Anderson PJ, Gordon I (1996)Radiation dosimetry of texhnetium-99 m-DMSA in children. J NuclMed 37:13361342

21. Tseng MH, Lin WJ, Lo WT, Wang SR, Chu ML, Wang CC (2007)Does a normal DMSA obviate the performance of voidingcystourethrography in evaluation of young children after their firsturinary tract infection. J Pediatr 150:9699

22. Lee MD, Lin CC, Huang FY, Tsai TC, Huang CT, Tsai JD (2009)Screening young childrenwith a first febrile urinary tract infection forhigh-grade vesicoureteral reflux with renal ultrasound scanning andtechnetium-99-labelled dimercaptosuccinic acid scanning. J Pediatr154:797802

23. Hansson S, Dhamey M, Sigstrm O, Sixt R, Stokland E,Wennerstrm M, Jodal U (2004) Dimercapto-succinic acid scintig-raphy instead of voiding cystourethrography for infants with urinarytract infection. J Urol 172:10711073

24. Swerkersson S, Jodal U, Sixt R, Stokland E, Hansson S (2007)Relationship among vesicoureteral reflux, urinary tract infectionand renal damage in children. J Urol 178:647651

25. Jakobsson B, Soderlundh S, Berg U (1992) Diagnostic significanceof 99mTc-dimercaptosuccinic acid (DMSA) scintigraphy in urinarytract infection. Arch Dis Child 67:13381342

Pediatr Nephrol (2014) 29:15671574 1573

26. Ditchfield MR, Grimwood K, Cook DJ, Powell HR, Sloane R, GulatiS, De Campo JF (2004) Persistent renal cortical scintigram defects inchildren 2 years after urinary tract infection. Pediatr Radiol 34:465471

27. Lee YJ, Lee JH, Park YS (2012) Risk factors for renal scar formationin infants with first episode of acute pyelonephritis: a prospectiveclinical study. J Urol 187:10321036

28. CoulthardMG (2009) Vesicoureteral reflux is not a benign condition.Pediatr Nephrol 24:227232

29. Montini G, Tullus K, Hewitt I (2011) Febrile urinary tract infection inchildren. N Engl J Med 365:239250

30. Ardissino G, Avolio L, Dacco V, Testa S, Marra G, Vigan S, Loi S,Caione P, De Castro R, De Pascale S, Marras E, Riccipetitoni G,Selvaggio G, Pedotti P, Claris-Appiani A, Giofani A, Dello StrologoL, Lama G, Montini G, Verrina E (2004) Long-term outcome ofvesicourerteral reflux associated chronic renal failure in children.Data from the Italkid project. J Urol 172:305310

31. Leroy S, Fernandez-Lopez A, Nikfar R, Romanello C, Bouissou F,Gervaix A, Gurgoze MK, Bressan S, Smolkin V, Tuerlinckx D,Stefanidis CJ, Vaos G, Leblond P, Gungor F, Gendrel D,Chalumeau M (2013) Association of procalcitonin with acutepyelonephritis and renal scars in pediatric UTI. Pediatrics 131:870879

32. Chen SM, Cheng HM, Hung TW, ChaoY-H TJD, Lue KH, Sheu JN(2013) Diagnostic performance of procalcitonin for hospitalised chil-dren with acute pyelonephritis presenting to the paediatric emergencydepartment. Emeg Med J 30:406410

33. Bressan S, Andreola B, Zucchetta P, Montini G, Burei M, PerilongoG, Dalt LD (2009) Procalcitonin as a predictor of renal scarring ininfants and young children. Pediatr Nephrol 24:11991204

34. Ristola MT, Hurme T (2013) NICE guidelines for imaging studies inchildren with UTI adequate only in boys under the age of 6 months.Pediatr Surg Int 29:215222

1574 Pediatr Nephrol (2014) 29:15671574

Comparison of procalcitonin and different guidelines for first febrile urinary tract infection in children by imagingAbstractAbstractAbstractAbstractAbstractIntroductionMethodsDiagnostic guidelinesCosts, radiation exposure, and measurement outcomesStatistical analyses

ResultsPatients and clinical characteristicsImaging findingsPrimary outcomesVUR grades IV and IIIV and late RSAssociation between RS and VUR

Secondary outcomesCosts and radiation exposure

DiscussionReferences