Antibiotics for acute pyelonephritis in children

Summary of findings for the main comparison. Oral versus IV followed by oral (11 days) therapy for acute pyelonephritis in children

Oral versus IV followed by oral (11 days) therapy for acute pyelonephritis in children
Patient or population: children with acute pyelonephritis Intervention: oral versus IV followed by oral (11 days) therapy
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Oral	IV followed by oral (11 days) therapy
Time to fever resolution (hours)		The mean time to fever resolution (hours) in the intervention groups was 2.05 higher (0.84 lower to 4.94 higher)		808 (2)	⊕⊕⊕⊝ moderate¹
Renal parenchymal damage at 6 to 12 months: all included children with acute pyelonephritis DMSA scans Follow‐up: 6 to 12 months	Study population		RR 0.82 (0.59 to 1.12)	943 (4)	⊕⊕⊕⊝ moderate²
	224 per 1000	184 per 1000 (132 to 251)
	Moderate
	313 per 1000	257 per 1000 (185 to 351)
Renal parenchymal damage at 6 to 12 months: children with renal parenchymal damage on initial DMSA Follow‐up: 6 to 12 months	Study population		RR 0.79 (0.61 to 1.03)	681 (4)	⊕⊕⊕⊝ moderate³
	320 per 1000	253 per 1000 (195 to 330)
	Moderate
	382 per 1000	302 per 1000 (233 to 393)
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.
¹ Wide confidence intervals due to large standard deviations around the mean durations of fever ² Large number of patients excluded because of lack of follow‐up DMSA scans ³ No explanation was provided

Summary of findings 2. Short duration (3 to 4 days) versus long duration (7 to 14 days) IV therapy for acute pyelonephritis in children

Short duration (3 to 4 days) versus long duration (7 to 14 days) IV therapy for acute pyelonephritis in children
Patient or population: children with acute pyelonephritis Intervention: short duration (3 to 4 days) versus long duration (7 to 14 days) IV therapy
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Short duration (3 to 4 days)	Long duration (7 to 14 days) IV therapy
Persistent bacteriuria after treatment	Study population		RR 0.78 (0.24 to 2.55)	305 (4)	⊕⊕⊝⊝ low^1,2
	38 per 1000	30 per 1000 (9 to 98)
	Moderate
	0 per 1000	0 per 1000 (0 to 0)
Recurrent UTI within 6 months	Study population		RR 0.97 (0.58 to 1.62)	993 (5)	⊕⊕⊕⊝ moderate¹
	59 per 1000	57 per 1000 (34 to 95)
	Moderate
	56 per 1000	54 per 1000 (32 to 91)
Persistent renal damage at 3 to 6 months: all included children with acute pyelonephritis	Study population		RR 1.01 (0.8 to 1.29)	726 (4)	⊕⊕⊕⊝ moderate^1,3
	246 per 1000	249 per 1000 (197 to 318)
	Moderate
	257 per 1000	260 per 1000 (206 to 332)
Persistent renal damage at 3 to 6 months: children with initial renal parenchymal damage on initial DMSA scan Follow‐up: 6‐12 months	Study population		RR 1.1 (0.84 to 1.45)	315 (3)	⊕⊕⊕⊝ moderate^1,3
	357 per 1000	393 per 1000 (300 to 518)
	Moderate
	327 per 1000	360 per 1000 (275 to 474)
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.
¹ Unclear or inadequate allocation concealment ² Small number of patients and events leading to wide confidence intervals ³ In several studies, more than 10% patients lost to follow‐up or did not have follow‐up DMSA scans

Summary of findings 3. Different dosing regimens of aminoglycosides (daily versus 8 hourly) for acute pyelonephritis in children

Different dosing regimens of aminoglycosides (daily versus 8 hourly) for acute pyelonephritis in children
Patient or population: children with acute pyelonephritis Intervention: different dosing regimens of aminoglycosides (daily versus 8 hourly)
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Daily dose	8 hourly dose
Persistent bacteriuria after 1 to 3 days of treatment	Study population		RR 1.05 (0.15 to 7.27)	435 (3)	⊕⊕⊝⊝ low^1,2
	9 per 1000	10 per 1000 (1 to 67)
	Moderate
	0 per 1000	0 per 1000 (0 to 0)
Hearing impairment following treatment	Study population		RR 2.83 (0.33 to 24.56)	271 (3)	⊕⊕⊝⊝ low^1,2
	0 per 1000	0 per 1000 (0 to 0)
	Moderate
	0 per 1000	0 per 1000 (0 to 0)
Increase in serum creatinine during treatment	Study population		RR 0.75 (0.2 to 2.82)	419 (3)	⊕⊕⊝⊝ low^1,2
	25 per 1000	19 per 1000 (5 to 70)
	Moderate
	25 per 1000	19 per 1000 (5 to 70)
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.
¹ Unclear allocation concealment in two of three studies ² Few events resulting in wide confidence intervals

Summary of findings 4. Agent: Third generation cephalosporin versus other antibiotic for acute pyelonephritis in children

Agent: Third generation cephalosporin versus other antibiotic for acute pyelonephritis in children
Patient or population: children with acute pyelonephritis Intervention: third generation cephalosporin versus other antibiotic
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Antibiotic	Third generation cephalosporin
Persistent bacteriuria	Study population		RR 2.41 (0.98 to 5.93)	439 (3)	⊕⊕⊝⊝ low^1,2
	34 per 1000	81 per 1000 (33 to 199)
	Moderate
	0 per 1000	0 per 1000 (0 to 0)
Recurrent UTI after end of therapy	Study population		RR 1.23 (0.32 to 4.74)	491 (4)	⊕⊕⊝⊝ low^1,2
	18 per 1000	22 per 1000 (6 to 87)
	Moderate
	8 per 1000	10 per 1000 (3 to 38)
Persistent symptoms after end of treatment	Study population		RR 0.28 (0.13 to 0.62)	471 (3)	⊕⊕⊝⊝ low^1,3
	104 per 1000	29 per 1000 (14 to 64)
	Moderate
	0 per 1000	0 per 1000 (0 to 0)
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.
¹ Unclear allocation in several studies ² Few events leading to imprecision ³ Meta‐analysis dominated by single trial and results inconsistent with bacteriologic results

Background

Description of the condition

The urinary tract is a common site of bacterial infection in infants and young children. Pooled prevalence data demonstrate that approximately 7% of girls and boys are diagnosed with at least one urinary tract infection (UTI) by the age of 19 years (Shaikh 2008). Girls are more susceptible to UTIs than boys after the first six months of life with a prevalence of 11% in girls and 4% in boys (Brkic 2010). UTI is defined by the presence of bacteria in urine (bacteriuria), which when cultured is measured in colony forming units/mL (CFU/mL) of uncentrifuged urine. The diagnosis of UTI in children is generally confirmed by the pure growth of a bacteria of greater than 10³ CFU/mL from a suprapubic aspirate, 10⁴ CFU/mL from a bladder catheter specimen and 10⁵ CFU/mL from non‐invasive collection methods (clean catch and urinary bag specimens) (Bhat 2011).

UTIs can be clinically grouped into asymptomatic bacteriuria, cystitis and acute pyelonephritis.

Asymptomatic bacteriuria is the presence of bacteriuria without clinical signs and symptoms.
Cystitis is a UTI limited to the urethra and bladder and is seen most commonly in girls over two years of age. It presents with localising symptoms of dysuria (pain when passing urine), frequency, urgency, cloudy urine and lower abdominal discomfort. Pyuria (white cells in the urine) and haematuria (blood in the urine) may also be found.
Acute pyelonephritis refers to infection of the kidneys, and is the most severe form of UTI in children. Clinically, this is associated with systemic features such as high fever, malaise, vomiting, abdominal and loin pain and tenderness, poor feeding and irritability in infants. Together with urine culture, diagnosis may be assisted by imaging using technetium 99m labelled dimercaptosuccinic acid (DMSA) renal scan and markers of inflammation in the blood such as erythrocyte sedimentation rate (ESR) and C‐reactive protein (CRP).
Acute pyelonephritis is associated with significant short‐term morbidity, including shock and septicaemia, especially in infants. Acute kidney parenchymal injury has been demonstrated on DMSA scan in about 60% of children shortly after UTI diagnosis (Shaikh 2010). Permanent kidney damage may occur following acute pyelonephritis and is more frequent in children who have multiple episodes or who have vesicoureteric reflux (VUR) (Shaikh 2010; Smellie 1985). Serial DMSA scans of children after a first episode of acute pyelonephritis show that 15% of children with acute changes on DMSA scans have permanent kidney scarring at follow up (Shaikh 2010). However, the long term significance of kidney damage following acute pyelonephritis is debatable. In children with previously normal kidneys, the amount of damage is small and unlikely to cause disease (Salo 2011; Toffolo 2012).

Description of the intervention

A wide variety of antibiotic agents have been used to treat acute pyelonephritis in children. Antibiotics that have been used include aminoglycosides, cephalosporins, penicillins and trimethoprim/sulphamethoxazole (TMP/SMX). Historically, children who are judged by clinicians to be in poor general condition are given parenteral antibiotics and those who appear less sick have been given oral antibiotics, without clarity whether one route of administration is superior. An antibiotic course of seven to 14 days is generally recommended although the optimal duration of therapy is not known. Shorter courses may be associated with treatment failure while longer courses may unnecessarily expose children to the adverse effects of treatment. This review evaluated antibiotic therapies used to treat acute pyelonephritis, with consideration of different antibiotic agents, different dosing regimens of the same antibiotic, different durations of treatment and different routes of administration.

How the intervention might work

Antibiotics work in the treatment of acute pyelonephritis by eliminating bacterial infection in the urinary tract. The purpose of antibiotic therapy is to eradicate and prevent progressive infection and its consequences, including shock and septicaemia, reduce acute kidney injury and resolve the acute clinical symptoms of infection. The efficacy of treatment depends on using the antibiotic(s) to which the bacteria is sensitive. While the results of antibiotic sensitivity testing are pending, initial treatment is chosen on an empiric basis to cover the most likely cause of infection.

Why it is important to do this review

Acute pyelonephritis is a common serious infection in children. Nonetheless, there remains no consensus on the most effective antibiotic regimen for the treatment of acute pyelonephritis. There is also uncertainty regarding the optimal route of administration of antibiotic therapy. Previously, most authorities recommended commencing antibiotic therapy by the parenteral route. The most recent guidelines however recommend initial treatment with oral antibiotics for children older than two months (AAP 2011) or three months of age (NICE 2007) unless children are considered to be too unwell or unable to take oral antibiotics. This is advantageous because oral therapy is more convenient and does not require hospital admission, thereby reducing costs. There is further uncertainty about the optimal duration of antibiotic therapy. Currently, the recommended duration of therapy varies from seven to 14 days.

This review update was necessary to provide additional information about the optimum antibiotic regimens, route of administration and duration of treatment for acute pyelonephritis in children and about adverse effects of treatment.

Objectives

To evaluate the benefits and harms of antibiotics used to treat children with acute pyelonephritis. The aspects of therapy considered were:

different antibiotics
different dosing regimens of the same antibiotic
different duration of treatment
different routes of administration.

Methods

Criteria for considering studies for this review

Types of studies

All randomised controlled trials (RCTs) and quasi‐RCTs (RCTs in which allocation to treatment was obtained by alternation, use of alternate medical records, date of birth or other predictable methods) in which antibiotics were used in the treatment of children (birth to 18 years) with acute pyelonephritis were included. Where studies included both children with acute pyelonephritis and those with cystitis, these were included if data for participants with acute pyelonephritis could be extracted separately; otherwise, these studies were excluded.

Types of participants

Inclusion criteria

Children from birth to 18 years with acute pyelonephritis treated either in hospital or as outpatients with antibiotics were included. For this review, the diagnosis of acute pyelonephritis required UTI (as specified in the included studies but generally requiring a bacterial growth on urine culture of more than 10⁵ CFU/mL or 10⁸ CFU/L) with at least one symptom or sign of systemic illness such as fever, loin pain or toxicity and additional diagnostic criteria as defined by the authors of the included studies. Children with previously diagnosed urinary tract abnormalities including VUR or previous UTI could be included.

Exclusion criteria

Patients considered to have asymptomatic bacteriuria or cystitis (UTI as defined in Inclusions with no symptom or sign of systemic illness) were excluded.

Types of interventions

Different antibiotic agents
IV antibiotic versus oral antibiotic
Different doses or duration or both of the same antibiotic
Antibiotic versus placebo, no therapy or alternative non‐antibiotic therapy.

Types of outcome measures

Primary outcomes

Short‐term outcome measures.

Duration of fever
Persistent symptoms (e.g. UTI at 72 hours; inflammatory markers at 72 hours (ESR, WCC, CRP)
Acute kidney parenchymal damage on DMSA scan
Length of hospital stay for inpatients
Persistent bacteriuria after completion of antibiotics
Recurrent UTI
Adverse effects of treatment including minor (e.g. vomiting, discomfort from IV cannula) and major (e.g. anaphylaxis, hearing impairment)
Economic costs of treatment (if data available).

Secondary outcomes

Long‐term outcome measures.

Persistent kidney damage (as defined by authors of included studies)
Hypertension
Chronic kidney disease.

Search methods for identification of studies

Electronic searches

For the 2014 update, we searched the Cochrane Renal Group's Specialised Register through contact with the Trials' Search Co‐ordinator using search terms relevant to this review. The Cochrane Renal Group’s Specialised Register contains studies identified from the following sources.

Monthly searches of the Cochrane Central Register of Controlled Trials (CENTRAL)
Weekly searches of MEDLINE OVID SP
Handsearching of renal‐related journals and the proceedings of major renal conferences
Searching of the current year of EMBASE OVID SP
Weekly current awareness alerts for selected renal journals
Searches of the International Clinical Trials Register (ICTRP) Search Portal and ClinicalTrials.gov.

Studies contained in the Specialised Register are identified through search strategies for CENTRAL, MEDLINE, and EMBASE based on the scope of the Cochrane Renal Group. Details of these strategies, as well as a list of handsearched journals, conference proceedings and current awareness alerts, are available in the Specialised Register section of information about the Cochrane Renal Group.

See Appendix 1 for search terms used in strategies for this review update.

For previous search strategies please refer to our earlier reviews (Bloomfield 2003; Bloomfield 2005; Hodson 2007).

Searching other resources

Reference lists of review articles, relevant studies and clinical practice guidelines.
Letters seeking information about unpublished or incomplete studies to investigators known to be involved in previous studies.

Data collection and analysis

Selection of studies

The search strategy described was used to obtain titles and abstracts of studies relevant to the review. Titles and abstracts were screened independently by four authors, who discarded studies that were not applicable. However, studies and reviews that included relevant data or information on studies were retained initially. Four authors independently assessed retrieved abstracts, and if necessary the full text, to determine which studies satisfied the inclusion criteria.

Data extraction and management

Data extraction was carried out independently by four authors using standard data extraction forms. Studies reported in non‐English language journals were translated before assessment. Where more than one publication of one study existed, reports were grouped together and the publication with the most complete data was used in the analyses. Where relevant outcomes were only published in earlier versions those data were used. Any discrepancy between published versions was highlighted.

Assessment of risk of bias in included studies

The following items were independently assessed by four authors using the risk of bias assessment tool (Higgins 2011) (see Appendix 2).

Was there adequate sequence generation (selection bias)?
Was allocation adequately concealed (selection bias)?
Was knowledge of the allocated interventions adequately prevented during the study (detection bias)?
- Participants and personnel
- Outcome assessors
Were incomplete outcome data adequately addressed (attrition bias)?
Are reports of the study free of suggestion of selective outcome reporting (reporting bias)?
Was the study apparently free of other problems that could put it at a risk of bias?

Measures of treatment effect

For dichotomous outcomes (persistent bacteriuria, recurrent UTI, persistent clinical symptoms, presence of kidney parenchymal damage, adverse effects) results were expressed as risk ratio (RR) with 95% confidence intervals (CI). Where continuous scales of measurement were used to assess the effects of treatment (duration of fever, inflammatory markers, extent of kidney parenchymal damage), the mean difference (MD) was used.

Unit of analysis issues

The unit of analysis was the study participant and not events, that is, the number of children with acute pyelonephritis rather than the number of episodes of acute pyelonephritis per child.

Dealing with missing data

Where data were missing or unclear, we contacted the original authors of studies to request additional data. An attempt to obtain the preliminary results of the terminated study (NCT00724256) was made by contacting the lead investigator. We did not impute missing data.

Assessment of heterogeneity

Heterogeneity was analysed using a Chi squared test on N‐1 degrees of freedom with an alpha of 0.1 used for statistical significance and the I² statistic (Higgins 2003). I² values of 25%, 50% and 75% correspond to low, medium and high levels of heterogeneity.

Assessment of reporting biases

We planned to assess publication bias by constructing funnel plots; however, there were insufficient data in each meta‐analysis to enable this analysis to be conducted.

Data synthesis

Data were pooled using the random‐effects model but the fixed‐effect model was also used to ensure robustness of the model chosen and susceptibility to outliers.

Subgroup analysis and investigation of heterogeneity

Subgroup analysis was planned to explore possible sources of heterogeneity (participants, treatments and study quality) but could not be undertaken because of the small number of studies for each comparison. Heterogeneity among participants could be related to age (infants versus adolescents) and pre‐existing renal tract pathology. Heterogeneity in treatment could be related to inpatient versus outpatient management, prior antibiotics used and the antibiotic, dose, duration and route of administration of therapy.

Sensitivity analysis

Sensitivity analysis was planned to identify individual studies that were contributing to significant heterogeneity (I² value greater than 75%) but the I² values of all meta‐analyses were less than 50% so sensitivity analysis was not undertaken.

Results

Description of studies

Results of the search

First version published 2003

The initial search in September 2002 identified 1520 titles and abstracts of which 51 were screened. We found that 16 parallel RCTs (16 reports) involving 1872 children fulfilled the eligibility criteria and were included in the review. We excluded 11 studies (11 reports) (Bloomfield 2003).

Review updates published 2005 and 2007

A search in June 2004 identified two additional studies (three reports) (Chong 2003; Montini 2007). A total of 18 studies (19 reports) involving 2612 children were included in our 2005 review update (Bloomfield 2005).

A search from 2004 to July 2007 identified 26 reports of which 16 were excluded (not randomised, mixed populations or wrong interventions). We included five new studies (nine reports) (Banfi 1993; Cheng 2006; Fujii 1987; Neuhaus 2008; Noorbakhsh 2004). The final results of the multicentre study by Montini 2007 were published and included in this update. We included 23 studies (29 reports) that involved 3407 children in the 2007 update (Hodson 2007).

Review update 2014

A search in 10 April 2014 identified 32 reports. In addition, a previously excluded study was re‐evaluated and included because it had been excluded incorrectly based on how outcomes were reported. Khan 1981 had been previously excluded because results were reported as episodes of acute pyelonephritis rather than number of patients with an episode of acute pyelonephritis. We excluded 17 studies (17 reports) and identified one eligible study that was terminated (see Characteristics of ongoing studies); contact with the triallists confirmed that no results were available (NCT00724256). We found that nine new records were further reports of four previously included studies (Benador 2001; Cheng 2006; Montini 2007; Neuhaus 2008). Reports relating to Benador 2001, Cheng 2006 and Montini 2007 did not provide any new data; however, the final results of Neuhaus 2008 (two new reports) were published in September 2008 and results were included in this review update. Four reports were of three newly identified studies (Bocquet 2012; Bouissou 2008; Marild 2009). This update included 27 studies (42 reports) that involved 4452 children.

See Figure 1.

Figure 1

Study flow diagram.

Included studies

The characteristics of the 27 included studies are summarised in Characteristics of included studies.

Participants

Studies recruited participants from the ages of two weeks to 16 years. Three studies did not specify age range (Bakkaloglu 1996; Levtchenko 2001; Pylkkänen 1981).

Definition of acute pyelonephritis

All studies required positive urine culture. Additional criteria required for diagnosis of acute pyelonephritis in children with UTI varied among studies:

Four studies required fever > 38°C (Baker 2001; Bocquet 2012; Hoberman 1999; Khan 1981).
Eight required fever and at least one additional clinical feature (Bakkaloglu 1996; Carapetis 2001; Chong 2003; Grimwood 1988; Noorbakhsh 2004; Repetto 1984; Schaad 1998; Toporovski 1992).
Nine required fever, clinical features and/or laboratory abnormalities (CRP, ESR, white blood count) (Bouissou 2008; Fischbach 1989; Francois 1997; Kafetzis 2000; Levtchenko 2001; Marild 2009; Montini 2007; Pylkkänen 1981; Vigano 1992).
Three required fever, clinical features and acute kidney parenchymal injury on DMSA scan (Benador 2001; Neuhaus 2008; Vilaichone 2001).
Five other studies (Bocquet 2012; Chong 2003; Hoberman 1999; Levtchenko 2001; Montini 2007) provided information on the number of children with acute pyelonephritis based on clinical characteristics, who had DMSA abnormalities at study entry.
One study required fever with computer tomography scan evidence of acute lobular nephronia (Cheng 2006).

Two studies did not report the definition used for acute pyelonephritis (Banfi 1993; Fujii 1987).

Commonly reported exclusion criteria

Impaired kidney function (12 studies: Carapetis 2001; Chong 2003; Fischbach 1989; Francois 1997; Kafetzis 2000; Khan 1981; Noorbakhsh 2004; Repetto 1984; Schaad 1998; Toporovski 1992; Vigano 1992; Vilaichone 2001).
Known severe urinary tract abnormality (14 studies: Baker 2001; Benador 2001; Bocquet 2012; Bouissou 2008; Francois 1997; Hoberman 1999; Khan 1981; Levtchenko 2001; Marild 2009; Neuhaus 2008; Noorbakhsh 2004; Repetto 1984; Vigano 1992; Vilaichone 2001).
Known sensitivity to study medications (17 studies: Banfi 1993; Baker 2001; Benador 2001;Bocquet 2012; Carapetis 2001; Chong 2003; Fischbach 1989; Francois 1997; Hoberman 1999; Kafetzis 2000; Marild 2009; Noorbakhsh 2004; Repetto 1984; Schaad 1998; Toporovski 1992; Vigano 1992; Vilaichone 2001).

Other exclusion criteria

Recent antibiotic use (10 studies: Banfi 1993; Baker 2001; Bocquet 2012; Chong 2003; Fischbach 1989; Kafetzis 2000; Marild 2009; Montini 2007; Noorbakhsh 2004; Vilaichone 2001).
Previous UTI (seven studies: Bouissou 2008; Fischbach 1989; Francois 1997; Hoberman 1999; Marild 2009; Montini 2007; Vilaichone 2001).
Clinical signs of shock at presentation (six studies: Baker 2001; Bocquet 2012; Francois 1997; Hoberman 1999; Montini 2007; Neuhaus 2008).
Immune compromise (six studies: Banfi 1993; Bouissou 2008; Carapetis 2001; Francois 1997; Noorbakhsh 2004; Schaad 1998).
Known hearing impairment (four studies: Carapetis 2001; Chong 2003; Kafetzis 2000; Vigano 1992).
Uncomplicated acute pyelonephritis (APN) (one study: Cheng 2006).

Four studies did not specify any exclusion criteria (Bakkaloglu 1996; Fujii 1987; Grimwood 1988; Pylkkänen 1981).

Study comparisons

The 27 included studies evaluated eight different comparisons.

Four studies compared oral therapy with short duration IV therapy followed by oral therapy (Bocquet 2012; Hoberman 1999; Montini 2007; Neuhaus 2008).
In six studies, short duration IV therapy (three to four days) followed by oral therapy was compared with long duration IV therapy (seven to 14 days) (Benador 2001; Bouissou 2008; Francois 1997; Levtchenko 2001; Noorbakhsh 2004; Vilaichone 2001).
A single dose of parenteral antibiotic added to oral therapy was compared to oral therapy alone in one study (Baker 2001).
Three studies compared different dosing frequencies of the same antibiotic agents (Carapetis 2001; Chong 2003; Vigano 1992).
Seven studies compared different antibiotics (Banfi 1993; Bakkaloglu 1996; Fischbach 1989; Kafetzis 2000; Marild 2009; Schaad 1998; Toporovski 1992). Toporovski 1992 included two experimental groups who received different doses of antibiotic. Because treatment response did not differ, experimental group data were combined.
Three studies compared different durations of antibiotics (Cheng 2006; Khan 1981; Pylkkänen 1981).
Two studies assessed single dose parenteral therapy against seven to 10 days of oral antibiotic therapy (Grimwood 1988; Repetto 1984).
One study compared ampicillin suppositories with oral ampicillin (Fujii 1987).

Reported outcomes

Not all studies reported the same outcomes. Table 1shows the outcomes reported for each study comparison.

See: Summary of findings for the main comparison Oral versus IV followed by oral (11 days) therapy for acute pyelonephritis in children; Summary of findings 2 Short duration (3 to 4 days) versus long duration (7 to 14 days) IV therapy for acute pyelonephritis in children; Summary of findings 3 Different dosing regimens of aminoglycosides (daily versus 8 hourly) for acute pyelonephritis in children; Summary of findings 4 Agent: Third generation cephalosporin versus other antibiotic for acute pyelonephritis in children

Table 1. Reported outcomes of included studies

Study/ comparisons	Reported outcomes
Study/ comparisons	Persistent bacteriuria at 48 to 72 hours	Bacteriuria at the end or ≥ 5 days of treatment	UTI at follow‐up	Resolution of clinical symptoms	Symptomatic recurrence of UTI	Parenchymal renal damage on DSMA scan	Adverse effects
Oral therapy versus sequential short duration IV therapy and oral therapy
Bocquet 2012				•		•	•
Hoberman 1999			•	•		•
Montini 2007	•			•		•	•
Neuhaus 2008						•	•
Sequential short duration (3 to 4 days) IV therapy and oral therapy versus long duration (7to 14 days) IV therapy
Benador 2001			•			•
Bouissou 2008						•
Francois 1997		•	•				•
Levtchenko 2001		•	•			•
Noorbakhsh 2004		•
Vilaichone 2001		•	•			•	•
Single dose parenteral therapy and oral therapy versus oral therapy alone
Baker 2001	•		•	•			•
Different dosing regimens of aminoglycoside therapy
Carapetis 2001	•			•			•
Chong 2003	•					•	•
Vigano 1992		•	•				•
Third generation cephalosporins versus other antibiotics
Banfi 1993		•	•	•	•		•
Fischbach 1989	•		•	•			•
Marild 2009		•		•			•
Toporovski 1992		•	•				•
Third generation cephalosporins versus fourth generation cephalosporins
Schaad 1998	•	•	•	•			•
Ceftriaxone versus cefotaxime
Bakkaloglu 1996	•	•	•				•
Aminoglycosides versus aminoglycosides
Kafetzis 2000		•	•				•
Different durations of the same oral antibiotic
Pylkkänen 1981			•
Single dose parenteral therapy versus oral therapy alone
Grimwood 1988		•	•
Repetto 1984		•	•
Different durations of different antibiotics
Cheng 2006		•			•
Different routes of antibiotic administration
Fujii 1987		•		•
Three days versus 10 days of oral therapy
Khan 1981			•

DMSA ‐ Tc99m‐dimercaptosuccinic acid nuclear scan; UTI ‐ urinary tract infection

Excluded studies

We excluded a total of 43 studies because: data from children with acute pyelonephritis could not be separated from those with lower UTI (18), children had lower UTI only (8), antibiotics were studied as prophylactic agents (5), the studies involved ineligible interventions or populations (5) or the study was not randomised (7).

Risk of bias in included studies

The assessment of risk of bias is shown in Figure 2 and Figure 3. Figure 2 shows the proportion of studies assessed as low, high or unclear risk of bias for each risk of bias indicator. Figure 3 shows the risk of bias indicators for individual studies.

Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Sequence generation was considered to be at low risk of bias in 12 studies (Baker 2001; Benador 2001; Bocquet 2012; Bouissou 2008; Carapetis 2001; Fischbach 1989; Francois 1997; Grimwood 1988; Marild 2009; Montini 2007; Neuhaus 2008; Toporovski 1992) and assessed at high risk in three studies (Cheng 2006; Khan 1981; Noorbakhsh 2004). Randomisation methods were not reported in 12 studies.

Allocation concealment was considered to be at low risk of bias in six studies (Baker 2001; Benador 2001; Bouissou 2008; Marild 2009; Montini 2007; Neuhaus 2008) and high risk in three studies (Cheng 2006; Khan 1981;Noorbakhsh 2004). Allocation concealment was assessed as unclear in 18 studies.

Blinding

Participants and investigators were not blinded in any of the included studies. The absence of blinding was considered to be a high risk of bias because symptom reporting and clinical management could be influenced by knowledge of the treatment group (performance bias). Bakkaloglu 1996 was reported to be double‐blinded but antibiotics were administered at different frequencies and no placebos were given.

Outcome assessors (detection bias) were blinded in 17 studies (Baker 2001; Benador 2001; Bocquet 2012; Bouissou 2008; Chong 2003; Francois 1997; Grimwood 1988; Hoberman 1999; Kafetzis 2000; Khan 1981; Levtchenko 2001; Montini 2007; Neuhaus 2008; Pylkkänen 1981; Schaad 1998; Vigano 1992; Vilaichone 2001). Blinding of outcome assessors was not carried out in nine studies (Bakkaloglu 1996; Banfi 1993; Carapetis 2001; Cheng 2006; Fischbach 1989; Marild 2009; Noorbakhsh 2004; Repetto 1984; Toporovski 1992). There was no reporting of outcome assessment blinding in Fujii 1987 (abstract only available).

Incomplete outcome data

Not all studies reported all outcomes. The reported outcomes from each of the included studies are summarised in Table 1. Incomplete outcome data was considered to be at low risk of bias in 19 studies because they reported outcomes in more than 90% of participants (Baker 2001; Bakkaloglu 1996; Benador 2001; Carapetis 2001; Cheng 2006; Chong 2003; Fischbach 1989; Francois 1997; Fujii 1987; Grimwood 1988; Kafetzis 2000; Levtchenko 2001; Marild 2009; Noorbakhsh 2004; Repetto 1984; Schaad 1998; Toporovski 1992; Vigano 1992; Vilaichone 2001). Attrition and exclusion of participants after the randomisation process was considered to be at a high risk of bias in seven studies (Banfi 1993; Bocquet 2012; Bouissou 2008; Hoberman 1999; Montini 2007; Neuhaus 2008; Pylkkänen 1981) and unclear in Khan 1981.

Selective reporting

There were 13 studies that reported bacteriological, clinical and adverse outcomes and were considered at low risk of bias (Baker 2001; Bakkaloglu 1996; Banfi 1993; Benador 2001; Carapetis 2001; Chong 2003; Fischbach 1989; Francois 1997; Kafetzis 2000; Marild 2009; Montini 2007; Schaad 1998; Toporovski 1992). Fujii 1987 was an abstract and did not clearly indicate outcomes investigated. The remaining 13 studies did not report all three types of outcomes and were considered at high risk of bias.

Other potential sources of bias

We assessed that 11 studies were at high risk of bias because they reported receiving funding from pharmaceutical companies (Baker 2001; Banfi 1993; Bouissou 2008; Hoberman 1999; Kafetzis 2000; Marild 2009; Neuhaus 2008; Noorbakhsh 2004; Pylkkänen 1981; Schaad 1998; Toporovski 1992). Four received funding from hospital grants (Bocquet 2012; Chong 2003; Grimwood 1988) or a government grant (Montini 2007) and were considered at low risk of bias. The source of funding in the remaining 12 studies was unclear.

Effects of interventions

Because results from random and fixed‐effect models did not differ, only results from the random‐effects model were reported. Few studies were available for pooling in meta‐analyses. No pre‐planned subgroup analyses for outcomes according to patient age (infant, child, adolescent) were possible from the available data. Post hoc subgroup analyses were reported for age (less than or greater than one year of age, Benador 2001) and VUR (Benador 2001; Hoberman 1999; Vilaichone 2001) and delay in treatment (less than or greater than seven days, Levtchenko 2001).

Oral therapy versus sequential IV therapy and oral therapy

We found four studies (Bocquet 2012; Hoberman 1999; Montini 2007; Neuhaus 2008) involving 1131 children compared oral antibiotics (cefixime, ceftibuten or amoxicillin/clavulanic acid) for 10 to 14 days with IV cefotaxime (Hoberman 1999) or ceftriaxone (Bocquet 2012; Montini 2007; Neuhaus 2008) for three to four days or until resolution of fever followed by oral antibiotics to complete the course of therapy.

Time to resolution of fever did not differ significantly between groups (Analysis 1.1 (2 studies, 808 children): MD 2.05 hours, 95% CI ‐0.84 to 4.94; I² = 0%). Neuhaus 2008 reported the number of children with fever on day three did not differ significantly between groups (Analysis 1.2 (1 study, 152 children): RR 0.79, 95% CI 0.30 to 2.06).
The number of children with persistent UTI at 72 hours after commencing therapy did not differ significantly between groups (Analysis 1.3 (2 studies, 542 children): RR 1.10, 95% CI 0.07 to 17.41).
Montini 2007 reported no significant difference between groups in the mean levels of inflammatory markers: WCC (Analysis 1.4.1 (1 study, 473 children): MD 0.30 x 10⁹/L, 95% CI ‐0.30 to 0.90), ESR (Analysis 1.4.2 (1 study, 338 children): MD ‐1.80 mm/60 min, 95% CI‐8.20 to 4.60]) or CRP (Analysis 1.4 (1 study, 486 children): MD 1.10 mg/L, 95% CI ‐2.18 to 4.38).
Hoberman 1999 reported no significant difference between groups in the rate of recurrences of bacteriuria (Analysis 1.5.1 (1 study, 287 children): RR 0.65, 95% CI 0.28 to 1.51) or symptomatic UTI within six months (Analysis 1.5.2 (1 study, 287 children): RR 0.67, 95% CI 0.27 to 1.67).
There were no significant differences among treatment groups in the rate of persistent kidney parenchymal defects on DMSA scan whether considered in relation to the total number of children with acute pyelonephritis (Analysis 1.6.1 (4 studies, 943 children): RR 0.82, 95% CI 0.59 to 1.12; I² = 41%) or only those with defects on the initial DMSA scan (Analysis 1.6.2 (4 studies, 681 children): RR 0.79, 95% CI 0.31 to 1.03; I² = 19%). Hoberman 1999 reported no significant difference between groups in the size of persistent kidney parenchymal defects on DMSA scan (Analysis 1.7 (1 study, 272 children): MD ‐0.70, 95% CI ‐1.74 to 0.34).
Post hoc subgroup analysis (Analysis 1.8) by Hoberman 1999 found no difference in the number of kidney parenchymal defects on DMSA scan at six months between children with VUR (Analysis 1.8.1 (1 study, 107 children): RR 1.88, 95% CI 0.83 to 4.24) and those without VUR (Analysis 1.8.2 (1 study, 107 children): RR 0.80, 95% CI 0.23 to 2.73). However, post hoc analysis (Analysis 1.8.4) raised the possibility that among children with VUR (grades III‐V), persistent kidney parenchymal defects on DMSA scan at six months occurred more frequently after oral than IV therapy (RR 7.33, 95% CI 1.00 to 54.01).
The average cost of treatment for each patient was USD 3630 and USD 7382 for oral and IV groups respectively (Hoberman 1999).
Adverse effects were reported in three studies (Bocquet 2012; Montini 2007; Neuhaus 2008). No children experienced therapy‐related adverse effects in Neuhaus 2008. In Bocquet 2012 two children experienced vomiting with oral cefixime and required change to parenteral therapy. In Montini 2007 15 children experienced diarrhoea or vomiting (13), erythema (1) and leucopenia (1) with oral amoxicillin and clavulanic acid; 10 required a change of antibiotics. In the same study three children experienced diarrhoea (1), erythema (1) and candida (1) with ceftriaxone; none required change of treatment. Hoberman 1999 did not report on adverse effects.

Sequential short duration (three to four days) IV therapy and oral therapy versus long duration (seven to 14 days) IV therapy

There were six studies (Benador 2001; Bouissou 2008; Francois 1997; Levtchenko 2001; Noorbakhsh 2004; Vilaichone 2001) involving 917 children that compared oral therapy after an initial three to four days of IV therapy with a long duration of IV therapy alone. Two studies compared IV ceftriaxone (three to four days) followed by oral cefixime (Benador 2001) or ceftibuten (Vilaichone 2001) with IV ceftriaxone (10 days). Levtchenko 2001 compared IV temocillin (three days) followed by oral amoxicillin or amoxicillin/clavulanic acid with IV temocillin (seven days). Noorbakhsh 2004 compared IV ceftriaxone (two to three days) followed by oral ceftibuten with IV amikacin or gentamicin with IV ampicillin (14 days). Francois 1997 compared IV cefotaxime (four days) followed by oral amoxicillin/clavulanic acid with IV cefotaxime (14 days). Bouissou 2008 compared IV netilmicin (two days) and ceftriaxone (three days) followed by oral antibiotics (cefixime, amoxicillin/clavulanic acid, TMP/SMX) chosen according to sensitivity with IV netilmicin (two days) and ceftriaxone (eight days). Benador 2001 and Levtchenko 2001 also converted the IV group to oral therapy after seven to 10 days to complete 15 to 21 days of treatment.

There was no significant difference between the risk of persistent bacteriuria at the end of treatment (Analysis 2.1 (4 studies, 305 children): RR 0.78, 95% CI 0.24 to 2.55; I² = 0%).
There was no significant difference between groups for recurrent UTI within six months (Analysis 2.2 (5 studies, 993 children): RR 0.97, 95% CI 0.58 to 1.62; I² = 0%).
The number of persisting kidney parenchymal defects seen on DMSA scan at three to six months did not differ significantly between treatment groups when considered in relation to the total number of children with acute pyelonephritis (Analysis 2.3.1 (4 studies, 726 children): RR 1.01, 95% CI 0.80 to 1.29; I² = 0%) or only those with defects on the initial DMSA scan (Analysis 2.3.2 (3 studies, 315 children): RR 1.10, 95% CI 0.84 to 1.45; I² = 0%).
Post hoc subgroup analysis showed that the number of children with persisting kidney parenchymal defects on DMSA scan did not differ between those with VUR (Analysis 2.4.1 (2 studies, 81 children): RR 0.99, 95% CI 0.69 to 1.43; I² = 0%) and without VUR (Analysis 2.4.2 (2 studies, 173 children): RR 1.19, 95% CI 0.81 to 1.76; I² = 0%), those aged under one year (Analysis 2.4.3 (1 study, 22 children): RR 1.46, 95% CI 0.71 to 3.01) and aged one year and over (Analysis 2.4.4 (1 study, 54 children): RR 0.89, 95% CI 0.59 to 1.34), and those who had a delay of treatment of less than seven days (Analysis 2.4.5 (1 study, 13 children): RR 1.52, 95% CI 0.59 to 3.92) or more than seven days (Analysis 2.4.6 (1 study, 8 children): RR 2.10, 95% CI 0.92 to 4.77).
Adverse effects were reported in Francois 1997 and Vilaichone 2001; both related to gastrointestinal upsets, and frequency did not differ between therapy routes (Analysis 2.5.1 (2 studies, 175 children): RR 1.29, 95% CI 0.55 to 3.05; I² = 0%). Four studies did not report on adverse effects (Benador 2001; Bouissou 2008; Levtchenko 2001; Noorbakhsh 2004).
Duration of hospitalisation was 4.9 days for the IV and oral group compared with 9.8 days for the IV group (Vilaichone 2001).
Costs of treatment for four days of IV therapy followed by six days of oral therapy were 513 French Francs (range 176 to 896) compared with 3545 French Francs (range 2478 to 4673) for 10 days of IV therapy (Francois 1997).

Single dose parenteral therapy and oral treatment versus oral therapy alone

Baker 2001 (69 children) compared the addition of a single intramuscular dose of the third generation cephalosporin, ceftriaxone, to an oral course of TMP/SMX. There was no significant difference in:

persistence of bacteriuria after 48 hours (Analysis 3.1: RR 0.77, 95% CI 0.19 to 3.20)
persistence of clinical symptoms (Analysis 3.2: RR 0.82, 95% CI 0.24 to 2.81), or
total adverse events (Analysis 3.4.1: RR 1.37, 95% CI 0.33 to 5.68) between groups.

No child developed symptomatic UTI during one month after treatment.

Different dosing regimens of aminoglycoside therapy

Three studies that involved 495 children compared daily parenteral administration of gentamicin (Carapetis 2001; Chong 2003) or netilmicin (Vigano 1992) to eight‐hourly administration of aminoglycosides.

There was no significant difference in the risk for persisting bacteriuria one to three days after commencing treatment with either dose frequency (Analysis 4.1 (3 studies, 435 children): RR 1.05, 95% CI 0.15 to 7.27).
Carapetis 2001 reported no difference in numbers of children with persisting clinical symptoms after three days of gentamicin (Analysis 4.2 (1 study, 179 children): RR 1.98, 95% CI 0.37 to 10.53).
Vigano 1992 reported persisting bacteriuria one week after (Analysis 4.3 (1 study, 144 children): RR 2.84, 95% CI 0.12 to 68.57) and recurrent UTI within one month (Analysis 4.4 (1 study, 144 children): RR 1.18, 95% CI 0.33 to 4.23) after completing netilmicin treatment did not differ between treatment groups.
There was no significant difference in numbers of children with hearing impairment (Analysis 4.5 (3 studies, 271 children): RR 2.83, 95% CI 0.33 to 24.56; I² = 0%) or kidney dysfunction (Analysis 4.6 (3 studies, 419 children): RR 0.75, 95% CI 0.20 to 2.82; I² = 0%).
Chong 2003 reported mean time to resolution of fever with gentamicin did not differ between groups (Analysis 4.7 (1 study, 172 children): MD 2.40 hours, 95% CI ‐7.90 to 12.70). Median time to resolution of fever was 27 hours (interquartile range 15 to 48 hours) with daily dosing and 33 hours (interquartile range 12 to 48 hours) with eight‐hourly dosing in a second study (Carapetis 2001).

Different antibiotic agents

Six studies compared different antibiotics (Bakkaloglu 1996; Carapetis 2001; Chong 2003; Kafetzis 2000; Schaad 1998; Vigano 1992).

Third generation cephalosporins versus other antibiotics

In four studies involving 491 children treatment with third generation cephalosporins (IV cefotaxime (Fischbach 1989), oral cefetamet (Toporovski 1992) or oral ceftibuten (Banfi 1993; Marild 2009) were compared with amoxicillin/clavulanic acid (Fischbach 1989; Toporovski 1992) or TMP/SMX (Banfi 1993; Marild 2009).

There was no significant difference in the number of children with persistent bacteriuria after 48 hours of therapy (Analysis 5.1 (3 studies, 433 children): RR 2.41, 95% CI 0.98 to 5.93; I² = 0%).
There was no significant difference in numbers of children who had recurrent UTI at 4 to 10 days after treatment (Analysis 5.2 (4 studies, 419 children): RR 1.23, 95% CI 0.32 to 4.74; I² = 0%).
A significantly greater number of children treated with TMP/SMX had persistent clinical symptoms at four to 10 days after treatment compared with those treated with a third generation cephalosporin (Analysis 5.3 (3 studies, 471 children): RR 0.28, 95% CI 0.13 to 0.62; I² = 0%). The study by Marild 2009 contributed to 94% of the weight of this result.
Fischbach 1989 reported no significant difference in numbers of children with persistent fever for more than 48 hours (Analysis 5.4 (1 study, 20 children): RR 5.00, 95% CI 0.27 to 92.62).
Banfi 1993 reported no significant difference between groups in the rate of recurrences of bacteriuria (Analysis 5.5 (1 study, 28 children): RR 2.14, 95% CI 0.11 to 40.30) or symptomatic UTI at four to six weeks (Analysis 5.6; no symptomatic UTIs in either group).
All four studies reported adverse effects. There was no significant difference in numbers of children who experienced gastrointestinal adverse effects (Analysis 5.7 (4 studies, 591 children): RR 0.93, 95% CI 0.34 to 2.58; I² = 0%). Marild 2009 reported that four children in each group discontinued treatment because of adverse reactions (Analysis 5.8 (1 study, 461 children): RR 0.49, 95% CI 0.12 to 1.94).

Third generation cephalosporins versus fourth generation cephalosporins (Analysis 6)

In Schaad 1998, which included 299 children, IV cefepime (a fourth generation cephalosporin) was compared to IV ceftazidime (a third generation cephalosporin).

No significant differences between groups were detected in numbers of children with persistent or recurrent bacteriuria with the same pathogen at different time points after therapy (Analysis 6.1).
Recurrent UTI with a different pathogenic organism at four to six weeks did not differ between groups (Analysis 6.2: RR 1.19, 95% CI 0.45 to 3.18).
There were no significant differences in the occurrence of an unsatisfactory clinical response at different time points after therapy (Analysis 6.3).
The frequency of adverse effects did not differ between treatment groups (Analysis 6.4).

Ceftriaxone versus cefotaxime

Bakkaloglu 1996 compared ceftriaxone and cefotaxime in 100 children aged over 24 months.

No child had persistent bacteriuria at 48 hours (Analysis 7.1).
There were no significant differences between groups for bacteriuria at the end of treatment (Analysis 7.2.1: RR 0.87, 95% CI 0.37 to 2.03), for recurrent infection at one month after therapy (Analysis 7.3.1: RR 0.68, 95% CI 0.30 to 1.50), or for total adverse events (Analysis 7.4.1: RR 0.67, 95% CI 0.12 to 3.82).
Post hoc subgroup analysis (Analysis 7.2.2 and Analysis 7.2.3) revealed no differences in outcomes for bacteriuria at the end of treatment or recurrent UTI at one month after therapy between children with and without abnormalities on imaging studies of the urinary tract.

Aminoglycosides

Kafetzis 2000 compared the aminoglycosides isepamicin and amikacin in 16 children.

No child in either group had persistent bacteriuria after 48 hours of treatment, or seven days or 30 days after treatment (Analysis 8.1).
Mean time to resolution of fever in each group was identical (24 hours).
No child in either treatment group developed hearing impairment on testing.

Duration of antibiotic administration

Four studies compared different durations of antibiotic administration (Cheng 2006; Grimwood 1988; Pylkkänen 1981; Repetto 1984).

Ten days versus 42 days of oral sulphafurazole

The study by Pylkkänen 1981 involved 149 children and compared 10 days with 42 days of oral sulphafurazole.

Recurrence of UTI within one month of ceasing therapy was significantly higher in children treated for 10 days compared with children treated for 42 days (Analysis 9.1: RR 17.70, 95% CI 2.42 to 129.61).
The number of children with recurrent UTI from one to 12 months after ceasing therapy did not differ between groups (Analysis 9.2: RR 0.87, 95% CI 0.40 to 1.88).

Single dose parenteral antibiotic therapy versus seven to 10 days of oral therapy

Grimwood 1988 and Repetto 1984 (involving a total of 61 children) compared single dose parenteral antibiotic therapy with seven to 10 days of oral therapy.

There were no significant differences in the number of children with persistent bacteriuria after treatment (Analysis 10.1 (2 studies, 35 children): RR 1.73, 95% CI 0.18 to 16.30; I² = 15%) or with recurrent UTI within six weeks (Analysis 10.2 (2 studies, 35 children): RR 0.24, 95% CI 0.03 to 1.97).

Three weeks with two weeks of antibiotics

Cheng 2006, which involved 80 children, compared three weeks with two weeks of antibiotics for children with acute lobar nephronia. Antibiotics were chosen according to sensitivities.

Seven children treated for two weeks had persistent or recurrent bacteriuria; this was not significantly different (Analysis 11.1: RR 0.07, 95% CI 0.00 to 1.19).
Two children had recurrence of clinical symptoms with bacteriuria; this was not significantly different (Analysis 11.2: RR 0.21, 95% CI 0.01 to 4.24).

Three days with 10 days of antibiotics

Khan 1981 (54 children) compared three and 10 days of oral antibiotics. Data were reported as episodes of UTI (asymptomatic, lower tract, APN) and could not be included in a meta‐analysis. Of 31 episodes of UTI, 23 were cured in 27 children in the three day treatment group and 25 of 31 episodes were cured in 27 children in the 10 day treatment group. Of episodes of acute pyelonephritis, four were cured in five episodes in the three day treatment group and five were cured in six episodes in the 10 day treatment group.

Different routes of antibiotic administration

Fujii 1987, which reported on 105 children, compared ampicillin administered by suppository with oral administration.

There was no significant difference between treatments in the risk of persistent clinical symptoms (Analysis 12.1: RR 0.89, 95% CI 0.51 to 1.56) or bacteriuria (Analysis 12.2: RR 0.89, 95% CI 0.53 to 1.50).

Discussion

Summary of main results

This review was designed to include all RCTs addressing all aspects of antibiotic treatment for children with acute pyelonephritis. Identified studies formed a heterogeneous group with few studies addressing the same or similar comparisons to enable assessment in meta‐analyses. The 27 included studies addressed a variety of different questions related to the therapy of children with acute pyelonephritis.

Oral therapy versus IV therapy

Four studies compared an oral antibiotic (ceftibuten, cefixime or amoxicillin/clavulanic acid) alone with IV therapy (cefotaxime or ceftriaxone) followed by oral therapy. These studies found:

No significant difference in bacteriological outcomes between groups.
The number of children with kidney parenchymal damage on DMSA scan at follow‐up whether expressed as a proportion of the total number considered to have acute pyelonephritis or as a proportion of those with DMSA changes at entry did not differ significantly between groups.

Thus, there were no significant differences in efficacy between treatment with oral ceftibuten, cefixime and amoxicillin/clavulanic acid and IV therapy followed by oral therapy. Studies that support these findings enrolled children older than one month of age and hence the finding cannot be extrapolated to children aged less than one month.

Short duration versus long duration IV therapy

A meta‐analysis of six studies showed:

No significant differences in clinical or bacteriological outcomes between IV antibiotic therapy given for three to four days followed by oral therapy and IV therapy for seven to 14 days.
That the prevalence of kidney parenchymal injury on DMSA scan at three to six months after UTI therapy did not differ significantly between treatment groups.

These data show that short duration IV therapy (three to four days) can be used instead of longer courses of IV therapy to treat childhood acute pyelonephritis. The findings cannot be extrapolated to children less than one month of age as such children were excluded from the studies.

Single daily dosing with aminoglycosides

If IV therapy is required, three studies provide data to support the safety and efficacy of daily dosing with aminoglycosides (gentamicin and netilmicin) compared with eight‐hourly dosing in children with acute pyelonephritis. Once daily dosing has been studied extensively in adults and is preferred due to improved efficacy, similar or reduced toxicity, convenience and lower costs. These findings have also been supported in children justifying the use of single daily dosing of aminoglycosides (Contopoulos‐Ioannidis 2004; Jenh 2011), although aminoglycoside pharmacokinetics and toxicity differ in children from adults.

Efficacy of different antibiotics

The seven studies that compared different antibiotics did not demonstrate any advantage of one agent over another. Four studies compared a cephalosporin with amoxicillin/clavulanic acid or TMP/SMX. A meta‐analysis of three studies demonstrated that children treated with oral ceftibuten had a higher clinical cure rate than TMP/SMX. However one large study (Marild 2009) contributed most of the weight of the analysis. The study defined clinical cure as the resolution of all symptoms related to the infection within 10 days. It is possible that there was some cross‐over of symptoms related to the infection and those due to adverse effects of medication such as vomiting. Furthermore, the study did not demonstrate any difference in bacteriological elimination rates despite 15% of the pathogens responsible for the infection being resistant to TMP/SMX compared to 2% that were resistant to ceftibuten.

One study demonstrated that a single dose of parenteral medication added to oral therapy did not improve efficacy compared with oral therapy alone.

Adverse events

Adverse events resulting from antibiotics were reported in 16 studies. Events were uncommon and rarely resulted in treatment discontinuation or significant alteration.

Overall completeness and applicability of evidence

In this review a comprehensive and extensive literature review was performed to identify studies that assessed the benefits and harms of antibiotics to treat children with acute pyelonephritis. We found that oral antibiotic therapy alone is as effective as IV therapy followed by oral therapy, and similarly short IV therapy is as effective as longer courses of IV therapy to treat childhood acute pyelonephritis. It is unknown whether these findings apply to children less than one month of age since children aged below one month of age were excluded from studies. The exclusion criteria for participation in the included studies mean that our findings may not be generalizable to all children with acute pyelonephritis. There were 10 studies that excluded children who were severely ill or clinically unstable. Four studies did not specify any exclusion criteria. The applicability of the findings in children with uropathy may also be limited since 10 studies excluded children with known uropathy. Hoberman 1999 performed a post hoc subgroup analysis to analyse differences in efficacy between children with or without VUR but the study was not designed and had no power to detect differences between small subgroups. Further data are required to determine whether treatment efficacy differs in children with non‐dilating VUR (grades I‐II) and dilating VUR (grades III‐V).

Most of the studies examined the bacteriological efficacy of antibiotic therapy. Few studies compared the efficacy of antibiotic therapies on the resolution of clinical symptoms other than fever.

None of the 27 included studies analysed the optimal duration of antibiotic therapy for childhood acute pyelonephritis. Our review evaluated oral antibiotic regimens in which oral antibiotics were used either alone or following IV antibiotics for a total duration of eight to 42 days of therapy. Inadequate data are available on the benefits and harms of shorter duration therapies (e.g. seven days or less). Three studies compared single dose parenteral antibiotic therapy or short course oral therapy with seven to 10 days of oral therapy and showed no significant differences but the studies were small. This is unlike the evidence supporting the use of short‐course therapies for the treatment of lower urinary tract infections in children (Michael 2002; Michael 2003).

From the low reported incidence of adverse events, we were only able to detect common adverse effects e.g. gastrointestinal upsets. Generally, RCTs are not powered to detect rare but serious side effects e.g. Stevens–Johnson syndrome, so our findings of adverse effects may not be generalizable to larger groups of children.

Most of the included studies reported on short‐term outcomes. Nine studies analysed kidney parenchymal damage on DSMA scan at three to 12 months following an episode of acute pyelonephritis. Five of the nine studies had a high loss to follow up because of refusal to do a DMSA scan among well children and one study (not included) was terminated without any results for this reason (NCT00724256). The loss to follow up of well children may lead to an underestimate of the effect of treatment. As the longest follow up of children was 12 months, this review cannot provide data on the likelihood of long term kidney scarring following antibiotic therapy.

Although several studies potentially included adolescents aged to 16 years, none of the studies reported results for different age groups. Thus we could not determine whether there was any difference in results according to the patients' ages.

Quality of the evidence

The quality of the included studies was quite variable. The main limitations in the quality of the studies were concealment of allocation, blinding of participants and personnel and sponsorship from pharmaceutical companies. Of the 27 included studies, 12 reported adequate sequence generation and six demonstrated adequate allocation concealment. The lack of adequate sequence generation and allocation concealment can lead to biased estimates of treatment effects in the original study and therefore the results of a systematic review (Hollis 1999; Juni 1999; Moher 1998; Schulz 1995). All of the studies were unblinded to participants and personnel primarily because antibiotics were delivered by the parenteral route compared with oral or used different dosing regimens. This was considered a high risk of bias because clinicians’ management could be influenced by knowledge of the treatment group. For blinding of outcome assessors, blinding was adequate in 17 studies where the primary outcome was bacteriological or radiological and considered unlikely to be influenced by lack of blinding. We found that 19 studies provided complete data reporting and 13 reported on all reasonably expected outcomes (bacteriological eradication, clinical cure and adverse effects). The authors of 11 studies reported receiving sponsorship from pharmaceutical companies.

The quality of the evidence was assessed according to the GRADE approach and is displayed in summary of findings tables.

The evidence that oral therapy alone is as effective as IV therapy followed by oral therapy in children with acute pyelonephritis is considered to be of moderate quality. The quality of evidence was downgraded because of imprecision regarding the time to resolution of fever and the loss to follow‐up for DMSA scans (summary of findings Table for the main comparison).

The evidence that short duration IV therapy followed by oral therapy is as effective as long duration IV therapy is considered to be of moderate quality. The quality of evidence was downgraded because of unclear or inadequate allocation concealment in the included studies which may increase the risk of selection bias, the studies being too small and the loss to follow‐up for DMSA scans (summary of findings Table 2).

The evidence that daily dosing of aminoglycosides is as safe and effective as thrice daily dosing of aminoglycosides is considered to be of low quality. The quality of evidence was downgraded because of unclear allocation concealment in two of three studies and the small number of participants combined with the low frequency of events that made the analysis significantly underpowered to detect a difference (summary of findings Table 3).

The evidence that third generation cephalosporins are no more effective than other antibiotics (amoxicillin/clavulanic acid and TMP/SMX) is low. The quality of evidence was downgraded because of unclear allocation, imprecision from sparse data, and inconsistent bacteriologic results from one study that provided most of the weight of the meta‐analysis (summary of findings Table 4).

Potential biases in the review process

We attempted to reduce publication bias by searching multiple databases and the grey literature without language restriction. Although the Cochrane Renal Group's Specialised Register contains the handsearched reports of studies, it is possible that we missed unpublished data presented at smaller conferences or studies published in foreign language journals and low impact journals. Studies may have been added since our last search of the register. No data were available from the terminated study (NCT00724256) after personal communication with the lead author. Not all included studies reported all outcomes. Some outcomes that would be expected to be known (e.g. resolution of clinical symptoms) were not reported which may have affected the results of meta‐analyses.

Agreements and disagreements with other studies or reviews

A systematic review published in 2008 evaluated an early switch to oral antibiotics after at least one day of initial IV antibiotics with IV therapy alone for hospitalised patients with acute pyelonephritis (Vouloumanou 2008). The meta‐analysis included both children and adults with acute pyelonephritis but the data from the two populations could be separated. It identified all six RCTs in children that we included in our review for this comparison. The authors found there was no difference in the incidence of kidney scars, microbiological eradication, clinical cure, reinfection, persistence of acute pyelonephritis, or adverse events between the two treatment regimens. This is consistent with our finding that short duration IV therapy followed by oral therapy is as effective as longer courses of IV therapy for the treatment of acute pyelonephritis in children.

No other systematic reviews were found for the 11 other comparisons in our review.

This review agrees with recently published guidelines (AAP 2011; Ammenti 2012; NICE 2007) for the treatment of childhood UTI, which recommend oral antibiotics for the initial treatment of children with acute pyelonephritis unless the child is seriously ill and/or unable to tolerate oral antibiotics. Our findings can be applied to children aged over one month. The NICE 2007 guidelines apply to children aged three months or older, the AAP 2011 guidelines apply to children aged 2 to 24 months and the Ammenti 2012 recommendations apply to children aged two months to three years. The guidelines also suggest seven or more days of antibiotic treatment but recognise that this is not based on best evidence because there are no data on the optimal duration of antibiotic therapy, particularly shorter courses.

Figure 1

Study flow diagram.

Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Analysis 1.1

Comparison 1 Oral versus IV followed by oral (11 days) therapy, Outcome 1 Time to fever resolution.

Analysis 1.2

Comparison 1 Oral versus IV followed by oral (11 days) therapy, Outcome 2 Fever on Day 3.

Analysis 1.3

Comparison 1 Oral versus IV followed by oral (11 days) therapy, Outcome 3 Number with persistent UTI at 72 hours.

Analysis 1.4

Comparison 1 Oral versus IV followed by oral (11 days) therapy, Outcome 4 Inflammatory markers at 72 hours.

Analysis 1.5

Comparison 1 Oral versus IV followed by oral (11 days) therapy, Outcome 5 Recurrent UTI within 6 months.

Analysis 1.6

Comparison 1 Oral versus IV followed by oral (11 days) therapy, Outcome 6 Persistent kidney damage at 6‐12 months.

Analysis 1.7

Comparison 1 Oral versus IV followed by oral (11 days) therapy, Outcome 7 Proportion of kidney parenchyma with damage at 6 months.

Analysis 1.8

Comparison 1 Oral versus IV followed by oral (11 days) therapy, Outcome 8 Kidney damage at 6 months (post hoc subgroup analysis).

Analysis 2.1

Comparison 2 Short duration (3‐4 days) versus long duration (7‐14 days) IV therapy, Outcome 1 Persistent bacteriuria after treatment.

Analysis 2.2

Comparison 2 Short duration (3‐4 days) versus long duration (7‐14 days) IV therapy, Outcome 2 Recurrent UTI within 6 months.

Analysis 2.3

Comparison 2 Short duration (3‐4 days) versus long duration (7‐14 days) IV therapy, Outcome 3 Persistent kidney damage at 3‐6 months.

Analysis 2.4

Comparison 2 Short duration (3‐4 days) versus long duration (7‐14 days) IV therapy, Outcome 4 Persistent kidney damage at 3‐6 months (post hoc subgroup analysis).

Analysis 2.5

Comparison 2 Short duration (3‐4 days) versus long duration (7‐14 days) IV therapy, Outcome 5 Adverse effects.

Analysis 3.1

Comparison 3 Single dose parenteral therapy and oral therapy versus oral therapy alone, Outcome 1 Persistent bacteriuria at 48 hours.

Analysis 3.2

Comparison 3 Single dose parenteral therapy and oral therapy versus oral therapy alone, Outcome 2 Treatment failure after 48 hours of therapy.

Analysis 3.3

Comparison 3 Single dose parenteral therapy and oral therapy versus oral therapy alone, Outcome 3 Recurrent UTI within 1 month.

Analysis 3.4

Comparison 3 Single dose parenteral therapy and oral therapy versus oral therapy alone, Outcome 4 Adverse events.

Analysis 4.1

Comparison 4 Different dosing regimens of aminoglycosides (daily versus 8 hourly), Outcome 1 Persistent bacteriuria after 1‐3 days of treatment.

Analysis 4.2

Comparison 4 Different dosing regimens of aminoglycosides (daily versus 8 hourly), Outcome 2 Persistent symptoms at end of 3 days of IV therapy.

Analysis 4.3

Comparison 4 Different dosing regimens of aminoglycosides (daily versus 8 hourly), Outcome 3 Persistent bacteriuria at 1 week after treatment.

Analysis 4.4

Comparison 4 Different dosing regimens of aminoglycosides (daily versus 8 hourly), Outcome 4 Reinfection at 1 month after completing treatment.

Analysis 4.5

Comparison 4 Different dosing regimens of aminoglycosides (daily versus 8 hourly), Outcome 5 Hearing impairment following treatment.

Analysis 4.6

Comparison 4 Different dosing regimens of aminoglycosides (daily versus 8 hourly), Outcome 6 Increase in serum creatinine during treatment.

Analysis 4.7

Comparison 4 Different dosing regimens of aminoglycosides (daily versus 8 hourly), Outcome 7 Time to resolution of fever.

Analysis 4.8

Comparison 4 Different dosing regimens of aminoglycosides (daily versus 8 hourly), Outcome 8 Kidney parenchymal damage at 3 months.

Analysis 5.1

Comparison 5 Third generation cephalosporin versus other antibiotic, Outcome 1 Persistent bacteriuria.

Analysis 5.2

Comparison 5 Third generation cephalosporin versus other antibiotic, Outcome 2 Recurrent UTI after end of therapy.

Analysis 5.3

Comparison 5 Third generation cephalosporin versus other antibiotic, Outcome 3 Persistent symptoms after end of treatment.

Analysis 5.4

Comparison 5 Third generation cephalosporin versus other antibiotic, Outcome 4 Number with fever for more than 48 hours.

Analysis 5.5

Comparison 5 Third generation cephalosporin versus other antibiotic, Outcome 5 Recurrent bacteriuria at 4‐6 weeks.

Analysis 5.6

Comparison 5 Third generation cephalosporin versus other antibiotic, Outcome 6 Recurrent symptomatic UTI at 4‐6 weeks.

Analysis 5.7

Comparison 5 Third generation cephalosporin versus other antibiotic, Outcome 7 Gastrointestinal adverse events.

Analysis 5.8

Comparison 5 Third generation cephalosporin versus other antibiotic, Outcome 8 Number discontinuing treatment for adverse effect.

Analysis 6.1

Comparison 6 Cefepime versus ceftazidime, Outcome 1 Persistence or recurrence of initial pathogen.

Analysis 6.2

Comparison 6 Cefepime versus ceftazidime, Outcome 2 Infection with new pathogen at 4‐6 weeks.

Analysis 6.3

Comparison 6 Cefepime versus ceftazidime, Outcome 3 Unsatisfactory clinical response.

Analysis 6.4

Comparison 6 Cefepime versus ceftazidime, Outcome 4 Adverse effects.

Analysis 7.1

Comparison 7 Ceftriaxone versus cefotaxime, Outcome 1 Persistent bacteriuria at 48 hours.

Analysis 7.2

Comparison 7 Ceftriaxone versus cefotaxime, Outcome 2 Bacteriuria 10 days after end of treatment.

Analysis 7.3

Comparison 7 Ceftriaxone versus cefotaxime, Outcome 3 UTI at 1 month after therapy.

Analysis 7.4

Comparison 7 Ceftriaxone versus cefotaxime, Outcome 4 Adverse events.

Analysis 8.1

Comparison 8 Isepamicin versus amikacin, Outcome 1 Persistent bacteriuria.

Analysis 9.1

Comparison 9 10 days versus 42 days of oral sulfafurazole, Outcome 1 Recurrent UTI within 1 month after ceasing therapy.

Analysis 9.2

Comparison 9 10 days versus 42 days of oral sulfafurazole, Outcome 2 Recurrent UTI at 1‐12 months after completing therapy.

Analysis 10.1

Comparison 10 Single dose of parenteral antibiotic versus 7‐10 days oral therapy, Outcome 1 Persistent bacteriuria 1‐2 days after treatment.

Analysis 10.2

Comparison 10 Single dose of parenteral antibiotic versus 7‐10 days oral therapy, Outcome 2 UTI relapse or reinfection within 6 weeks.

Analysis 11.1

Comparison 11 3 weeks versus 2 weeks, Outcome 1 Persistence/recurrence of bacteriuria.

Analysis 11.2

Comparison 11 3 weeks versus 2 weeks, Outcome 2 Recurrence of clinical UTI.

Analysis 12.1

Comparison 12 Suppositories versus oral ampicillin, Outcome 1 Persistence of clinical symptoms.

Analysis 12.2

Comparison 12 Suppositories versus oral ampicillin, Outcome 2 Persistence of bacteriuria.

Summary of findings for the main comparison. Oral versus IV followed by oral (11 days) therapy for acute pyelonephritis in children

Oral versus IV followed by oral (11 days) therapy for acute pyelonephritis in children
Patient or population: children with acute pyelonephritis Intervention: oral versus IV followed by oral (11 days) therapy
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Oral	IV followed by oral (11 days) therapy
Time to fever resolution (hours)		The mean time to fever resolution (hours) in the intervention groups was 2.05 higher (0.84 lower to 4.94 higher)		808 (2)	⊕⊕⊕⊝ moderate¹
Renal parenchymal damage at 6 to 12 months: all included children with acute pyelonephritis DMSA scans Follow‐up: 6 to 12 months	Study population		RR 0.82 (0.59 to 1.12)	943 (4)	⊕⊕⊕⊝ moderate²
	224 per 1000	184 per 1000 (132 to 251)
	Moderate
	313 per 1000	257 per 1000 (185 to 351)
Renal parenchymal damage at 6 to 12 months: children with renal parenchymal damage on initial DMSA Follow‐up: 6 to 12 months	Study population		RR 0.79 (0.61 to 1.03)	681 (4)	⊕⊕⊕⊝ moderate³
	320 per 1000	253 per 1000 (195 to 330)
	Moderate
	382 per 1000	302 per 1000 (233 to 393)
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.
¹ Wide confidence intervals due to large standard deviations around the mean durations of fever ² Large number of patients excluded because of lack of follow‐up DMSA scans ³ No explanation was provided

Summary of findings for the main comparison. Oral versus IV followed by oral (11 days) therapy for acute pyelonephritis in children

Summary of findings 2. Short duration (3 to 4 days) versus long duration (7 to 14 days) IV therapy for acute pyelonephritis in children

Short duration (3 to 4 days) versus long duration (7 to 14 days) IV therapy for acute pyelonephritis in children
Patient or population: children with acute pyelonephritis Intervention: short duration (3 to 4 days) versus long duration (7 to 14 days) IV therapy
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Short duration (3 to 4 days)	Long duration (7 to 14 days) IV therapy
Persistent bacteriuria after treatment	Study population		RR 0.78 (0.24 to 2.55)	305 (4)	⊕⊕⊝⊝ low^1,2
	38 per 1000	30 per 1000 (9 to 98)
	Moderate
	0 per 1000	0 per 1000 (0 to 0)
Recurrent UTI within 6 months	Study population		RR 0.97 (0.58 to 1.62)	993 (5)	⊕⊕⊕⊝ moderate¹
	59 per 1000	57 per 1000 (34 to 95)
	Moderate
	56 per 1000	54 per 1000 (32 to 91)
Persistent renal damage at 3 to 6 months: all included children with acute pyelonephritis	Study population		RR 1.01 (0.8 to 1.29)	726 (4)	⊕⊕⊕⊝ moderate^1,3
	246 per 1000	249 per 1000 (197 to 318)
	Moderate
	257 per 1000	260 per 1000 (206 to 332)
Persistent renal damage at 3 to 6 months: children with initial renal parenchymal damage on initial DMSA scan Follow‐up: 6‐12 months	Study population		RR 1.1 (0.84 to 1.45)	315 (3)	⊕⊕⊕⊝ moderate^1,3
	357 per 1000	393 per 1000 (300 to 518)
	Moderate
	327 per 1000	360 per 1000 (275 to 474)
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.
¹ Unclear or inadequate allocation concealment ² Small number of patients and events leading to wide confidence intervals ³ In several studies, more than 10% patients lost to follow‐up or did not have follow‐up DMSA scans

Summary of findings 2. Short duration (3 to 4 days) versus long duration (7 to 14 days) IV therapy for acute pyelonephritis in children

Summary of findings 3. Different dosing regimens of aminoglycosides (daily versus 8 hourly) for acute pyelonephritis in children

Different dosing regimens of aminoglycosides (daily versus 8 hourly) for acute pyelonephritis in children
Patient or population: children with acute pyelonephritis Intervention: different dosing regimens of aminoglycosides (daily versus 8 hourly)
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Daily dose	8 hourly dose
Persistent bacteriuria after 1 to 3 days of treatment	Study population		RR 1.05 (0.15 to 7.27)	435 (3)	⊕⊕⊝⊝ low^1,2
	9 per 1000	10 per 1000 (1 to 67)
	Moderate
	0 per 1000	0 per 1000 (0 to 0)
Hearing impairment following treatment	Study population		RR 2.83 (0.33 to 24.56)	271 (3)	⊕⊕⊝⊝ low^1,2
	0 per 1000	0 per 1000 (0 to 0)
	Moderate
	0 per 1000	0 per 1000 (0 to 0)
Increase in serum creatinine during treatment	Study population		RR 0.75 (0.2 to 2.82)	419 (3)	⊕⊕⊝⊝ low^1,2
	25 per 1000	19 per 1000 (5 to 70)
	Moderate
	25 per 1000	19 per 1000 (5 to 70)
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.
¹ Unclear allocation concealment in two of three studies ² Few events resulting in wide confidence intervals

Summary of findings 3. Different dosing regimens of aminoglycosides (daily versus 8 hourly) for acute pyelonephritis in children

Summary of findings 4. Agent: Third generation cephalosporin versus other antibiotic for acute pyelonephritis in children

Agent: Third generation cephalosporin versus other antibiotic for acute pyelonephritis in children
Patient or population: children with acute pyelonephritis Intervention: third generation cephalosporin versus other antibiotic
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Antibiotic	Third generation cephalosporin
Persistent bacteriuria	Study population		RR 2.41 (0.98 to 5.93)	439 (3)	⊕⊕⊝⊝ low^1,2
	34 per 1000	81 per 1000 (33 to 199)
	Moderate
	0 per 1000	0 per 1000 (0 to 0)
Recurrent UTI after end of therapy	Study population		RR 1.23 (0.32 to 4.74)	491 (4)	⊕⊕⊝⊝ low^1,2
	18 per 1000	22 per 1000 (6 to 87)
	Moderate
	8 per 1000	10 per 1000 (3 to 38)
Persistent symptoms after end of treatment	Study population		RR 0.28 (0.13 to 0.62)	471 (3)	⊕⊕⊝⊝ low^1,3
	104 per 1000	29 per 1000 (14 to 64)
	Moderate
	0 per 1000	0 per 1000 (0 to 0)
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio;
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.
¹ Unclear allocation in several studies ² Few events leading to imprecision ³ Meta‐analysis dominated by single trial and results inconsistent with bacteriologic results

Summary of findings 4. Agent: Third generation cephalosporin versus other antibiotic for acute pyelonephritis in children

Table 1. Reported outcomes of included studies

Study/ comparisons	Reported outcomes
Study/ comparisons	Persistent bacteriuria at 48 to 72 hours	Bacteriuria at the end or ≥ 5 days of treatment	UTI at follow‐up	Resolution of clinical symptoms	Symptomatic recurrence of UTI	Parenchymal renal damage on DSMA scan	Adverse effects
Oral therapy versus sequential short duration IV therapy and oral therapy
Bocquet 2012				•		•	•
Hoberman 1999			•	•		•
Montini 2007	•			•		•	•
Neuhaus 2008						•	•
Sequential short duration (3 to 4 days) IV therapy and oral therapy versus long duration (7to 14 days) IV therapy
Benador 2001			•			•
Bouissou 2008						•
Francois 1997		•	•				•
Levtchenko 2001		•	•			•
Noorbakhsh 2004		•
Vilaichone 2001		•	•			•	•
Single dose parenteral therapy and oral therapy versus oral therapy alone
Baker 2001	•		•	•			•
Different dosing regimens of aminoglycoside therapy
Carapetis 2001	•			•			•
Chong 2003	•					•	•
Vigano 1992		•	•				•
Third generation cephalosporins versus other antibiotics
Banfi 1993		•	•	•	•		•
Fischbach 1989	•		•	•			•
Marild 2009		•		•			•
Toporovski 1992		•	•				•
Third generation cephalosporins versus fourth generation cephalosporins
Schaad 1998	•	•	•	•			•
Ceftriaxone versus cefotaxime
Bakkaloglu 1996	•	•	•				•
Aminoglycosides versus aminoglycosides
Kafetzis 2000		•	•				•
Different durations of the same oral antibiotic
Pylkkänen 1981			•
Single dose parenteral therapy versus oral therapy alone
Grimwood 1988		•	•
Repetto 1984		•	•
Different durations of different antibiotics
Cheng 2006		•			•
Different routes of antibiotic administration
Fujii 1987		•		•
Three days versus 10 days of oral therapy
Khan 1981			•
DMSA ‐ Tc99m‐dimercaptosuccinic acid nuclear scan; UTI ‐ urinary tract infection

Table 1. Reported outcomes of included studies

Comparison 1. Oral versus IV followed by oral (11 days) therapy

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Time to fever resolution Show forest plot	2	808	Mean Difference (IV, Random, 95% CI)	2.05 [‐0.84, 4.94]

2 Fever on Day 3 Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

3 Number with persistent UTI at 72 hours Show forest plot	2	542	Risk Ratio (M‐H, Random, 95% CI)	1.10 [0.07, 17.41]

4 Inflammatory markers at 72 hours Show forest plot	1		Mean Difference (IV, Random, 95% CI)	Totals not selected

4.1 WCC [×10⁹/L]	1		Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
4.2 ESR [mm/60 min]	1		Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
4.3 CRP [mg/L]	1		Mean Difference (IV, Random, 95% CI)	0.0 [0.0, 0.0]
5 Recurrent UTI within 6 months Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

5.1 Total UTIs	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
5.2 Symptomatic UTIs	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
6 Persistent kidney damage at 6‐12 months Show forest plot	4		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

6.1 All included patients with acute pyelonephritis	4	943	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.59, 1.12]
6.2 Patients with kidney parenchymal damage on initial DMSA	4	681	Risk Ratio (M‐H, Random, 95% CI)	0.79 [0.61, 1.03]
7 Proportion of kidney parenchyma with damage at 6 months Show forest plot	1		Mean Difference (IV, Random, 95% CI)	Subtotals only

8 Kidney damage at 6 months (post hoc subgroup analysis) Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

8.1 Persistent damage in children with VUR	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
8.2 Persistent damage in children without VUR	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
8.3 Persistent kidney damage with VUR grades 1‐2	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
8.4 Persistent damage with VUR grades 3‐5	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

Comparison 1. Oral versus IV followed by oral (11 days) therapy

Comparison 2. Short duration (3‐4 days) versus long duration (7‐14 days) IV therapy

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistent bacteriuria after treatment Show forest plot	4	305	Risk Ratio (M‐H, Random, 95% CI)	0.78 [0.24, 2.55]

2 Recurrent UTI within 6 months Show forest plot	5	993	Risk Ratio (M‐H, Random, 95% CI)	0.97 [0.58, 1.62]

3 Persistent kidney damage at 3‐6 months Show forest plot	4		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

3.1 All included patients with acute pyelonephritis	4	726	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.80, 1.29]
3.2 Patients with renal parenchymal damage on initial DMSA scan	3	315	Risk Ratio (M‐H, Random, 95% CI)	1.10 [0.84, 1.45]
4 Persistent kidney damage at 3‐6 months (post hoc subgroup analysis) Show forest plot	3		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

4.1 VUR present	2	81	Risk Ratio (M‐H, Random, 95% CI)	0.99 [0.69, 1.43]
4.2 No VUR	2	173	Risk Ratio (M‐H, Random, 95% CI)	1.19 [0.81, 1.76]
4.3 Age less than 1 year	1	91	Risk Ratio (M‐H, Random, 95% CI)	1.46 [0.71, 3.01]
4.4 Age 1 year or over	1	129	Risk Ratio (M‐H, Random, 95% CI)	0.89 [0.59, 1.34]
4.5 Delay in treatment less than 7 days in individual kidneys	1	53	Risk Ratio (M‐H, Random, 95% CI)	1.52 [0.59, 3.92]
4.6 Delay in treatment of 7 days or more in individual kidneys	1	12	Risk Ratio (M‐H, Random, 95% CI)	2.10 [0.92, 4.77]
5 Adverse effects Show forest plot	2		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

5.1 Gastrointestinal effects	2	175	Risk Ratio (M‐H, Random, 95% CI)	1.29 [0.55, 3.05]

Comparison 2. Short duration (3‐4 days) versus long duration (7‐14 days) IV therapy

Comparison 3. Single dose parenteral therapy and oral therapy versus oral therapy alone

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistent bacteriuria at 48 hours Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

2 Treatment failure after 48 hours of therapy Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

3 Recurrent UTI within 1 month Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

4 Adverse events Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

4.1 Total adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4.2 Gastrointestinal adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

Comparison 3. Single dose parenteral therapy and oral therapy versus oral therapy alone

Comparison 4. Different dosing regimens of aminoglycosides (daily versus 8 hourly)

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistent bacteriuria after 1‐3 days of treatment Show forest plot	3	435	Risk Ratio (M‐H, Random, 95% CI)	1.05 [0.15, 7.27]

2 Persistent symptoms at end of 3 days of IV therapy Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

3 Persistent bacteriuria at 1 week after treatment Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

4 Reinfection at 1 month after completing treatment Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

5 Hearing impairment following treatment Show forest plot	3	271	Risk Ratio (M‐H, Random, 95% CI)	2.83 [0.33, 24.56]

6 Increase in serum creatinine during treatment Show forest plot	3	419	Risk Ratio (M‐H, Random, 95% CI)	0.75 [0.20, 2.82]

7 Time to resolution of fever Show forest plot	1		Mean Difference (IV, Random, 95% CI)	Totals not selected

8 Kidney parenchymal damage at 3 months Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Comparison 4. Different dosing regimens of aminoglycosides (daily versus 8 hourly)

Comparison 5. Third generation cephalosporin versus other antibiotic

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistent bacteriuria Show forest plot	3	439	Risk Ratio (M‐H, Random, 95% CI)	2.41 [0.98, 5.93]

2 Recurrent UTI after end of therapy Show forest plot	4	491	Risk Ratio (M‐H, Random, 95% CI)	1.23 [0.32, 4.74]

3 Persistent symptoms after end of treatment Show forest plot	3	471	Risk Ratio (M‐H, Random, 95% CI)	0.28 [0.13, 0.62]

4 Number with fever for more than 48 hours Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

5 Recurrent bacteriuria at 4‐6 weeks Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

6 Recurrent symptomatic UTI at 4‐6 weeks Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

7 Gastrointestinal adverse events Show forest plot	4	591	Risk Ratio (M‐H, Random, 95% CI)	0.93 [0.34, 2.58]

8 Number discontinuing treatment for adverse effect Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Comparison 5. Third generation cephalosporin versus other antibiotic

Comparison 6. Cefepime versus ceftazidime

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistence or recurrence of initial pathogen Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

1.1 At end of IV therapy	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.2 At the end of IV and oral therapy	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.3 At 5‐9 days after treatment	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.4 At 4‐6 weeks after treatment	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2 Infection with new pathogen at 4‐6 weeks Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

3 Unsatisfactory clinical response Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

3.1 At end of IV therapy	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3.2 At end of IV and oral therapy	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3.3 At 5‐9 days after treatment	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3.4 At 4‐6 weeks after treatment	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4 Adverse effects Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

4.1 Total adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4.2 Drug‐related adverse effects	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4.3 Gastrointestinal adverse effects	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4.4 Cutaneous adverse effects	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4.5 Discontinuation due to drug related adverse effects	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

Comparison 6. Cefepime versus ceftazidime

Comparison 7. Ceftriaxone versus cefotaxime

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistent bacteriuria at 48 hours Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

2 Bacteriuria 10 days after end of treatment Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

2.1 All patients	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.2 Normal renal tract imaging (post hoc analysis)	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.3 Abnormal renal tract imaging (post hoc analysis)	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3 UTI at 1 month after therapy Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

3.1 All patients	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3.2 Normal renal tract imaging	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3.3 Abnormal renal tract imaging	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4 Adverse events Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

4.1 All adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4.2 Skin eruptions	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4.3 Gastrointestinal adverse events	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

Comparison 7. Ceftriaxone versus cefotaxime

Comparison 8. Isepamicin versus amikacin

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistent bacteriuria Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

1.1 After 2‐3 days of therapy	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.2 At 7 days after completing therapy	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.3 At 30 days after completing therapy	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

Comparison 8. Isepamicin versus amikacin

Comparison 9. 10 days versus 42 days of oral sulfafurazole

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Recurrent UTI within 1 month after ceasing therapy Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

2 Recurrent UTI at 1‐12 months after completing therapy Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Comparison 9. 10 days versus 42 days of oral sulfafurazole

Comparison 10. Single dose of parenteral antibiotic versus 7‐10 days oral therapy

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistent bacteriuria 1‐2 days after treatment Show forest plot	2	35	Risk Ratio (M‐H, Random, 95% CI)	1.73 [0.18, 16.30]

2 UTI relapse or reinfection within 6 weeks Show forest plot	2	35	Risk Ratio (M‐H, Random, 95% CI)	0.24 [0.03, 1.97]

Comparison 10. Single dose of parenteral antibiotic versus 7‐10 days oral therapy

Comparison 11. 3 weeks versus 2 weeks

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistence/recurrence of bacteriuria Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

2 Recurrence of clinical UTI Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Comparison 11. 3 weeks versus 2 weeks

Comparison 12. Suppositories versus oral ampicillin

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistence of clinical symptoms Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

2 Persistence of bacteriuria Show forest plot	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected

Comparison 12. Suppositories versus oral ampicillin