Published online September 17, 2007
PEDIATRICS
Vol. 120
No. 4
October 2007, pp.
e922-e928
(doi:10.1542/peds.2006-2417)
Does Early Treatment of Urinary Tract Infection Prevent Renal Damage?
Dimitrios Doganis, MDa,
Konstantinos Siafas, MDb,
Myrsini Mavrikou, MDa,
George Issaris, MDb,
Anna Martirosova, MDa,
Grigorios Perperidis, MDa,
Andreas Konstantopoulos, MDc and
Konstantinos Sinaniotis, MDc
a First Department of Pediatrics
b Third Department of Pediatrics, "P&A Kyriakou" Children's Hospital, Athens, Greece
c Second Department of Pediatrics, University of Athens, "P&A Kyriakou" Children's Hospital, Athens, Greece
 |
ABSTRACT
|
|---|
OBJECTIVE. Therapeutic delay has been suggested as the most important factor that is likely to have an effect on the development of scarring after acute pyelonephritis. However, this opinion has not been supported by prospective studies, so we tested it.
METHODS. In a prospective clinical study, we evaluated whether the time interval between the onset of the renal infection and the start of therapy correlates with the development of acute inflammatory changes and the subsequent development of renal scars, documented by dimercaptosuccinic acid scintigraphy. A total of 278 infants (153 male and 125 female) aged 0.5 to 12.0 months with their first urinary tract infection were enrolled in the study.
RESULTS. The median time between the onset of infection and the institution of therapy was 2 days (range: 1–8 days). Renal inflammatory changes were documented in 57% of the infants. Renal defects were recorded in 41% of the patients treated within the first 24 hours since the onset of fever versus 75% of those treated on day 4 and onward. Renal scarring was developed in 51% of the infants with an abnormal scan in the acute phase of infection. The frequency of scarring in infants treated early and in those whose treatment was delayed did not differ, suggesting that once acute pyelonephritis has occurred, ultimate renal scarring is independent of the timing of therapy. Acute inflammatory changes and subsequent scarring were more frequent in the presence of vesicoureteral reflux, especially that which is high grade. However, the difference was not significant, which suggests that renal damage may be independent of the presence of reflux.
CONCLUSIONS. Early and appropriate treatment of urinary tract infection, especially during the first 24 hours after the onset of symptoms, diminishes the likelihood of renal involvement during the acute phase of the infection but does not prevent scar formation.
Key Words: urinary tract infection acute pyelonephritis renal scarring vesicoureteral reflux
Abbreviations: UTI—urinary tract infection DMSA—dimercaptosuccinic acid Tc-99mDMSA—technetium-99m-dimercaptosuccinic acid VUR—vesicoureteral reflux CRP—C-reactive protein
Urinary tract infection (UTI) is among the more common acute illnesses and the most common bacterial infection in infants and young children.1 The infection may be limited to the lower tract or may involve the kidney; it may result in permanent renal damage and scarring, which may lead to development of hypertension and chronic renal impairment.2
Most cases of the first episode of urinary infection occur in the first year of life, and it is generally believed that infants are more susceptible to development of renal parenchymal damage after pyelonephritis than are older children.3–5 However, recent data have shown no difference in age at the time of infection between children who develop scars and those who do not,6–10 and some studies have even shown that scarring was more common in children >1 year of age.11–13
Technetium-99m-dimercaptosuccinic acid (Tc-99mDMSA) renal scintigraphy, performed during the acute phase of infection, is considered the most sensitive test for the diagnosis of renal involvement and the subsequent development of renal scarring.14 Among several risk factors likely to have an effect on the development of renal damage during the acute episode of renal infection, the time between the onset of symptoms of the infection and the beginning of the appropriate therapy is considered by many experts to be 1 of the most significant.3,15–20 This opinion is supported by experimental studies, which demonstrated a correlation between the duration of infection until the start of treatment and the extent of renal damage.21,22 On the other hand, this opinion has not been demonstrated by prospective clinical studies.9 The aim of our prospective study was to correlate the findings of the Tc-99mDMSA renal scintigraphy with the time passed since the onset of fever until the commencement of therapy in infants with a first episode of UTI.
 |
METHODS
|
|---|
Enrollment and Eligibility Criteria
We undertook a 5-year (2000–2005) prospective study in a cohort of 278 children <12 months of age who were admitted to the hospital with a first documented episode of UTI. Children were eligible to be included in the study if they had a temperature of
38°C and a positive culture of urine collected by suprapubic puncture. The growth of any number of colonies of Gram-negative bacilli was considered as positive. All of the children were treated with antibiotics immediately after the urine and blood samples were obtained. Patients were treated with intravenous antibiotics given until they had been afebrile for 24 to 36 hours. An oral antibiotic was provided to complete a 10- to 14-day course. The time between the onset of fever and the institution of therapy was recorded.
Imaging Studies
Renal ultrasonography and renal scintigraphy with Tc-99mDMSA were performed within 1 to 18 days (median: 5 days) of admission to determine anatomic abnormalities and the presence or absence of acute pyelonephritis, respectively. DMSA scan is performed 4 to 6 hours after intravenous injection of an age-adjusted dose (minimum dose: 40 MBq; maximum dose: 100 MBq). Posterior, anterior, and lateral posterior oblique planar views were collected for 6 minutes each with a camera equipped with a high-resolution low-energy collimator (computer matrix: 256 x 256). Differential renal function was calculated by using the geometric mean method to compensate for differences in the position of each kidney. Pyelonephritis was defined by the presence of focal or diffuse areas of decreased uptake of DMSA. A second cortical scan was performed after 5 to 26 months (median: 6.5 months) in infants who had a positive DMSA scan result in the acute phase of infection. An abnormal second scan was defined by the presence of decreased uptake of DMSA associated with loss of the contour of the kidney or by the presence of cortical thinning.
Voiding cystourethrogram was performed in the majority of cases while the child was in the hospital before discharge.23,24 Vesicoureteral reflux (VUR) was graded according to the classification proposed by the International Reflux Study Committee.25 Children who were found to have VUR were placed on chemoprophylaxis with trimethoprim-sulfamethoxazole (2 mg/kg for the trimethoprim component). Infants <2 months of age were given prophylaxis with cefuroxime-axetil (10 mg/kg).
Laboratory Tests
The laboratory tests in these patients at the time of admission included obtaining a white blood cell count and differential and the C-reactive protein (CRP) level.
Statistics
2 test, Fisher's exact test, and the nonparametric Mann-Whitney test were used for statistical analysis. A P value of .05 was considered to indicate statistical significance.
 |
RESULTS
|
|---|
Patient Characteristics
A total of 278 children,153 boys and 125 girls, were enrolled in the study. The median age was 3.5 months (range: 0.5–12.0 months). Of the 278 children, 128 (96 boys and 32 girls) were
3 months old, 80 (34 boys and 46 girls) were 3 to 6 months old, and 70 (23 boys and 47 girls) were >6 months old.
Treatment Delay and Timing of Scintigraphy
The median time between the onset of fever and the institution of treatment was 2 days (range: 1–8 days). The median time between the onset of fever and renal scintigraphy was 7 days (range: 4–21 days) and in 266 (96%) of 278 within 14 days or less. Furthermore, the median time between the onset of treatment and renal scintigraphy was 5 days (range: 1–18 days) and in 272 (98%) of 278 within 14 days or less.
DMSA Renal Scan Results
An abnormal renal scan was observed in 158 (57%) of 278 children, and the number of affected kidneys was 179. Twenty-one children (13%) had bilateral involvement and 137 (87%) unilateral. Renal defects were observed in 43 (41%) of the 105 infants treated within the first 24 hours since the onset of fever, in 43 (59%) of the 73 infants treated during the second day of the infection, in 28 (68%) of the 41 infants treated the third day, and in 44 (75%) of the 59 infants treated on the fourth day and onward (P = .000; Fig 1). Among infants <3 months, renal defects were observed in 29 (40%) of the 73 treated within the first day since the onset of fever and in 35 (64%) of the 55 treated on the second day and onward (P = .012; Fig 2). Among infants >3 months of age, renal defects were observed in 14 (44%) of the 32 treated within the first day since the onset of fever and in 80 (68%) of the 118 treated on the second day and onward (P = .022; Fig 3).
Abnormal DMSA scan was seen more often after the age of 6 months than in younger infants. A total of 106 infants of the 208 infants (51%) aged 0.5 to 6.0 months had an abnormal DMSA scan, whereas 52 (74%) of the 70 infants aged >6.0 months showed parenchymal damage (P = .001). The frequency of abnormal scans did not differ between male (87 of 153) and female (71 of 125) infants.
A second cortical scan was performed 5 to 26 months (median: 6.5 months) after the infection in 86 of the 158 infants who had a positive DMSA scan result in the acute phase of infection. Ten of the 86 infants experienced a relapse before the second DMSA scan and were excluded from the analysis of the results. Among the 76 remaining infants, there were 44 boys and 32 girls. At the second scintigraphy, 39 (51%) of 76 children had an abnormal scan, which showed scar formation in the site of the initial lesion, and in 5 of them a new scar was developed in a different site, in addition to the initial lesion.
The frequency of scarring in infants treated early in the first 24 hours (11 of 24), since the onset of the infection, and those treated later (28 of 52) did not differ significantly, but this could be the result of the small number of patients. No difference in the frequency of abnormal second scans between male (23 of 44) and female (16 of 32) infants was documented. Furthermore, no significant differences concerning the frequency of renal scars between infants
6 months of age (14 of 24) and infants <6 months of age (25 of 52) were detected.
Voiding Cystourethrogram and Renal Involvement
Voiding cystourethrogram was undertaken in 269 (97%) of 278 infants. In 9 infants, voiding cystourethrogram was not performed, because the parents denied it. VUR was recorded in 66 (24.5%) of 269 children. The presence of VUR and the relationship between the grade and the renal involvement on the first DMSA scan are shown on Table 1. Furthermore, in infants treated during the first day since the onset of fever, VUR was detected in 28 (28%) of 101 infants, during the second day in 14 (20%) of 70, and on the third day or later in 24 (24%) of the 98 infants. Therefore, the presence of VUR did not differ according to the time of starting treatment and did not influence the effect of therapeutic delay on DMSA results.
The number of infants with scarring and VUR was greater (18 of 28 [64%]) than in the absence of VUR (21 of 48 [44%]) but the difference did not reach statistical significance. The relationship between the presence of VUR and the results of the follow-up scan are shown on Table 2.
Clinical and Laboratory Findings and Renal Involvement
Escherichia coli was isolated from the urine of 242 infants, Proteus mirabilis from 4, Klebsiella spp from 18, Citrobacter spp from 1, Enterobacter spp from 5, Pseudomonas spp from 1, Enterococcus spp from 1, other Gram-negative bacterial pathogens from 1, and 2 Gram-negative bacterial pathogens in 5 infants. No correlation between any type of bacteria and renal involvement was documented.
Elevated temperature, leucocytosis, increased neutrophils, and CRP level were correlated with the presence of an abnormal scan during the acute infection (Table 3). On the other hand, no significant differences were noted comparing maximum temperature, white blood cell count, neutrophils, CRP level, and development of scarring.
 |
DISCUSSION
|
|---|
Scintigraphy with Tc-99mDMSA is considered the imaging method of choice for detecting acute renal parenchymal changes, as well as the development of renal scars after infection. Data from several studies of acute pyelonephritis using DMSA scintigraphy reveal that 50% to 90% of children with febrile UTI have abnormal DMSA renal scan findings.11,14,19,26–29 The DMSA scan in the children of our study showed parenchymal changes in 57% of the patients. Most previous studies have demonstrated a higher incidence of abnormal DMSA scintigraphy in acute UTIs.19,27 Several reasons may be responsible for the low rate of abnormal DMSA scan findings in the present study. First, the young age of the infants studied may have contributed to it.30 Almost half of them (128 of 278) were
3 months of age. It has been suggested that the immature tubular function that characterizes infants 2 to 3 months of age may be responsible for the low rate of DMSA scan findings in this age group.31 Another possibility is the timing of the DMSA scintigraphy as related to the start of antibiotic therapy. Stokland et al31 have shown that the rate of positive DMSA scintigraphy results rapidly decrease by 50% during the first 14 days after initiation of antibiotic therapy.
The importance of early treatment of pyelonephritis has been shown in experimental pyelonephritis in different animals. In these studies, a correlation between the duration of infection until start of treatment and the extent of renal damage was detected.22,32,33
Therapeutic delay has been considered by many experts as a clinical factor in determining renal damage.3,14,16–18,34 Unfortunately, data in regard of the beneficial effect of early and aggressive treatment to prevent renal scarring in children mostly derive from retrospective analysis,3,16–18,34 and this opinion has not been supported by prospective clinical trials.9
In a retrospective analysis of patients in Goteborg, Sweden, when treatment was delayed, 4 times as many of the girls in the study developed a scar as when treatment was prompt and adequate.3 However, it is noteworthy that no details of what constituted delayed treatment were reported in this study, as well as in other reports.16,18 The children in our study demonstrating parenchymal changes on the DMSA scan were positively correlated with the day that they received their first dose of antibiotic. Forty-three patients (41%) of 105 who were treated within 24 hours since the onset of fever showed inflammatory changes, but the incidence increased to 59% (43 of 73) when they were treated the second day of the infection and to 72% (72 of 100) when they received therapy between days 3 and 8. Exactly how long a delay in therapy is harmful is not known. Hiraoka et al35 studied 22 children with first-time UTIs. They found that renal defects formed only in those children who received treatment 24 hours after the onset of the disease. In agreement with these findings are the results of a study involving 158 children by Fernandez-Menendez et al,20 who found that uptake defects in the acute infection were recorded only when the treatment delay time was >48 hours. The results of both studies are in agreement with the findings of our study. Others have not found a correlation between treatment delay time and early DMSA findings. In one study, no difference in the frequency of DMSA abnormalities was noted between those patients treated within 2 days since the onset of symptoms of infection and those treated after
2 days.31
The follow-up scan performed 5 to 26 months after the first in 76 of our patients showed that 39 (51%) of 76 developed scarring. It is of interest that no correlation between the incidence of scarring and the treatment delay time was recorded in our patients, suggesting that renal scars are caused by the infection itself and that if the kidney is involved, the risk for development of scarring is independent of the timing of therapy.
This observation is in agreement with the findings of other authors who found no difference in therapy delay between those with or without scars.8,26,29,36 In a recent study, Hoberman et al9 reported that scarring was more common among children who received treatment after
24 hours of fever compared with those treated earlier, but the difference was not significant.
Among other factors, age has been reported as a risk factor to develop pyelonephritis and subsequent scarring and that infants under the age of 1 year are at greater risk.3,37 These findings have been challenged by other reports that have recorded a higher incidence of renal involvement in older children.10,11,28 Renal involvement in the acute phase of infection was more common with advancing age in our patients (Tables 3 and 4). The effect of age in the results of DMSA scintigraphy may be related to the immature renal function of the very young infants.31 Another factor is probably the early treatment that the younger children received. The median age of those treated on the first, second, and third days was 1.9, 3.5, and 6.1 months, respectively, and of those treated during the fourth day and onward was 5.9 months (Table 5). A reason for the early treatment that the younger infants received is the fact that parents are more concerned when a young infant is febrile and they ask sooner for medical advice. In our study, the effect of early treatment of the infection in the development of renal changes was found to be independent of the age of patients (Figs 2 and 3).
Scarring was recorded more often when VUR was present. Scars were seen in 64% of the children with VUR and in 44% of those who did not show VUR, but this difference did not reach a significant level. Moreover, we did not find any correlation between scarring and grade of reflux. These findings are in accordance with those of previous reports that showed that VUR is a poor predictor of renal scarring.6,13,38,39
 |
CONCLUSIONS
|
|---|
In infants with UTI, therapeutic delay for >24 hours is associated with an increased frequency of renal involvement but not with permanent renal damage. The presence of VUR or grade of reflux does not increase the incidence of renal involvement and scarring.
 |
FOOTNOTES
|
|---|
Accepted Mar 14, 2007.
Address correspondence to Dimitrios Doganis, MD, 12 Kapetan Petroutsou St, 115 23 Athens, Greece. E-mail: doganisd{at}otenet.gr
The authors have indicated they have no financial relationships relevant to this article to disclose.
 |
REFERENCES
|
|---|
- Wald E. Urinary tract infections in infants and children: a comprehensive overview.
Curr Opin Pediatr. 2004;16
:85
–88[CrossRef][ISI][Medline]
- Hellerstein S. Long-term consequences of urinary tract infections.
Curr Opin Pediatr. 2000;12
:125
–128[CrossRef][ISI][Medline]
- Winberg J, Bollgren I, Kallenius G, Mollby R, Svenson SB. Clinical pyelonephritis and focal renal scarring: a selected review of pathogenesis, prevention, and prognosis.
Pediatr Clin North Am. 1982;29
:801
–814[ISI][Medline]
- Pylkkanen J, Vilska J, Koskimies O. The value of level diagnosis of childhood urinary tract infection in predicting renal injury.
Acta Paediatr Scand. 1981;70
:879
–883[ISI][Medline]
- Berg UB, Johansson SB. Age as a main determinant of renal functional damage in urinary tract infection.
Arch Dis Child. 1983;58
:963
–969[Abstract]
- Rushton HG, Majd M, Jantausch B, Wiedermann BL, Belman AB. Renal scarring following reflux and nonreflux pyelonephritis in children: evaluation with 99m-technetium-dimercaptosuccinic acid scintigraphy.
J Urol. 1992;147
:1327
–1332[ISI][Medline]
- Stokland E, Hellstrom M, Jacobsson B, Jodal U, Sixt R. Renal damage one year after first urinary tract infection: role of dimercaptosuccinic acid scintigraphy.
J Pediatr. 1996;129
:815
–820[CrossRef][ISI][Medline]
- Jakobsson B, Jacobson SH, Hjalmas K. Vesico-ureteric reflux and other risk factors for renal damage: identification of high- and low-risk children.
Acta Paediatr. 1999;(suppl)88
:31
–39[CrossRef]
- Hoberman A, Wald ER, Hickey RW, et al. Oral versus initial intravenous therapy for urinary tract infections in young febrile children.
Pediatrics. 1999;104
:79
–86[Abstract/Free Full Text]
- Lin KY, Chiu NT, Chen MJ, et al. Acute pyelonephritis and sequelae of renal scar in pediatric first febrile urinary tract infection.
Pediatr Nephrol. 2003;18
:362
–365[ISI][Medline]
- Benador D, Benador N, Slosman D, Mermillod B, Girardin E. Are younger children at highest risk of renal sequelae after pyelonephritis?
Lancet. 1997;349
:17
–19[CrossRef][ISI][Medline]
- Jakobsson B, Svensson L. Transient pyelonephritic changes on 99m-Technetium-dimercaptosuccinic acid scan for at least five months after infection.
Acta Paediatr. 1997;86
:803
–807[ISI][Medline]
- Taskinen S, Ronnholm K. Post-pyelonephritic renal scars are not associated with vesicoureteral reflux in children.
J Urol. 2005;173
:1345
–1348[CrossRef][ISI][Medline]
- Rushton HG. The evaluation of acute pyelonephritis and renal scarring with technetium 99m-dimercaptosuccinic acid renal scintigraphy: evolving concepts and future directions.
Pediatr Nephrol. 1997;11
:108
–120[CrossRef][ISI][Medline]
- Winter AL, Hardy BE, Alton DJ, Arbus GS, Churchill BM. Acquired renal scars in children.
J Urol. 1983;129
:1190
–1194[ISI][Medline]
- Smellie JM, Ransley PG, Normand IC, Prescod N, Edwards D. Development of new renal scars: a collaborative study.
BMJ. 1985;290
:1957
–1960[ISI][Medline]
- Jodal U, Winberg J. Management of children with unobstructed urinary tract infection.
Pediatr Nephrol. 1987;11
:647
–656
- South Bedfordshire Practitioners Group. Development of renal scars in children: missed opportunities in management.
BMJ. 1990;301
:1082
–1084[ISI][Medline]
- Jakobsson B, Nolstedt L, Svensson L, Soderlundh S, Berg U. 99m-Technetium-dimercaptosuccinic acid scan in the diagnosis of acute pyelonephritis in children: relation to clinical and radiological findings.
Pediatr Nephrol. 1992;6
:328
–334[CrossRef][ISI][Medline]
- Fernandez-Menendez JM, Malaga S, Matesanz JL, et al. Risk factors in the development of early technetium-99m- dimercaptosuccinic acid renal scintigraphy lesions during first urinary tract infection in children.
Acta Paediatr. 2003;92
:21
–26[ISI][Medline]
- Glauser MP, Lyons JM, Braude AI. Prevention of chronic experimental pyelonephritis by suppression of acute suppuration.
J Clin Invest. 1978;61
:403
–407[ISI][Medline]
- Miller T, Phillips S. Pyelonephritis: the relationship between infection, renal scarring, and antimicrobial therapy.
Kidney Int. 1981;19
:654
–662[ISI][Medline]
- McDonald A, Scranton M, Gillespie R, Mahajan V, Edwards GA. Voiding cystourethrograms and urinary tract infections: how long to wait?
Pediatrics. 2000;105(4)
. Available at: www.pediatrics.org/cgi/content/full/105/4/e50
- Mahant S, To T, Friedman J. Timing of voiding cystourethrogram in the investigation of urinary tract infections in children.
J Pediatr. 2001;139
:568
–571[CrossRef][ISI][Medline]
- Lebowitz RL, Olbing H, Parkkulainen KV, Smellie JM, Tamminen-Mobius TE. International system of radiographic grading of vesicoureteric reflux. International Reflux Study in Children.
Pediatr Radiol. 1985;15
:105
–109[CrossRef][ISI][Medline]
- Rushton HG, Majd M. Dimercaptosuccinic acid renal scintigraphy for the evaluation of pyelonephritis and scarring: a review of experimental and clinical studies.
J Urol. 1992;148
:1726
–1732[ISI][Medline]
- Jakobsson B, Soderlundh S, Berg U. Diagnostic significance of 99m-Tc-dimercaptosuccinic acid (DMSA) scintigraphy in urinary tract infection.
Arch Dis Child. 1992;67
:1338
–1342[Abstract]
- Benador D, Benador N, Slosman DO, Nussle D, Mermillod B, Girardin E. Cortical scintigraphy in the evaluation of renal parenchymal changes in children with pyelonephritis.
J Pediatr. 1994;124
:17
–20[CrossRef][ISI][Medline]
- Jakobsson B, Berg U, Svensson L. Renal scarring after acute pyelonephritis.
Arch Dis Child. 1994;70
:111
–115[Abstract]
- Linne T, Fituri O, Escobar-Billing R, et al. Functional parameters and 99m-technetium-dimercaptosuccinic acid scan in acute pyelonephritis.
Pediatr Nephrol. 1994;8
:694
–699[CrossRef][ISI][Medline]
- Stokland E, Hellstrom M, Jacobsson B, Jodal U, Lundgren P, Sixt R. Early 99m-Tc-dimercaptosuccinic acid (DMSA) scintigraphy in symptomatic first-time urinary tract infection.
Acta Paediatr. 1996;85
:430
–436[ISI][Medline]
- Ransley PG, Risdon RA. Reflux nephropathy: effects of antimicrobial therapy on the evolution of the early pyelonephritic scar.
Kidney Int. 1981;20
:733
–742[ISI][Medline]
- Slotki IN, Asscher AW. Prevention of scarring in experimental pyelonephritis in the rat by early antibiotic therapy.
Nephron. 1982;30
:262
–268[ISI][Medline]
- Smellie JM, Poulton A, Prescod NP. Retrospective study of children with renal scarring associated with reflux and urinary infection.
BMJ. 1994;308
:1193
–1196[Abstract/Free Full Text]
- Hiraoka M, Hashimoto G, Tsuchida S, Tsukahara H, Ohshima Y, Mayumi M. Early treatment of urinary infection prevents renal damage on cortical scintigraphy.
Pediatr Nephrol. 2003;18
:115
–118[ISI][Medline]
- Benador D, Neuhaus TJ, Papazyan JP, et al. Randomised controlled trial of three day versus 10 day intravenous antibiotics in acute pyelonephritis: effect on renal scarring.
Arch Dis Child. 2001;84
:241
–246[Abstract/Free Full Text]
- Gleeson FV, Gordon I. Imaging in urinary tract infection.
Arch Dis Child. 1991;66
:1282
–1283[Abstract]
- Gordon I, Barkovics M, Pindoria S, Cole TJ, Woolf AS. Primary vesicoureteric reflux as a predictor of renal damage in children hospitalized with urinary tract infection: a systematic review and meta-analysis.
J Am Soc Nephrol. 2003;14
:739
–744[Abstract/Free Full Text]
- Moorthy I, Easty M, McHugh K, Ridout D, Biassoni L, Gordon I. The presence of vesicoureteric reflux does not identify a population at risk for renal scarring following a first urinary tract infection.
Arch Dis Child. 2005;90
:733
–736[Abstract/Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics