<scp>KHA‐CARI</scp> guideline: Diagnosis and treatment of urinary tract infection in children

Abstract

Specialist assessment and management is required for children who are considered at high risk of serious illness (underlying structural urinary tract abnormalities or neurogenic bladder or kidney transplant recipients). These children are beyond the scope of these guidelines, and it is important that they are excluded from the recommendations detailed below. UTI in children is common, about 6% of girls and 2% of boys will experience an episode before their 7th birthday.1 Having had one infection, the child is at a 13−19% risk of having another UTI.2-4 UTI causes pain, discomfort and irritability to the child, and anxiety, stress and inconvenience to the family. Prompt diagnosis and early treatment are central to good clinical care. Initial treatment of UTI is guided by clinical presentation; however, reliance on clinical symptoms may result in under treatment. The practical definition of UTI for clinical practice is the combination of symptomatic evidence of infection of the urinary tract (including one or more of: abdominal pain, dysuria, frequency, fever, loin pain and irritability in infants) in association with a urine sample containing a positive bacterial culture. A pure growth of a single bacterial species with a significant colony number is generally regarded as evidence of a true UTI. Mixed growths, growth of skin commensals or low colony counts are regarded as evidence of contamination. However, there may be occasions when these may represent true UTIs, especially in individuals with abnormal urinary tracts. There is no universal definition of a contaminated positive culture compared with a ‘true’ UTI, and none of the criteria for contamination are sufficient to exclude the possibility that the culture result does indeed reflect a true UTI. Culture results should therefore be interpreted in the context of the individual child. Pro-calcitonin may aid in the identification of children with UTI, warranting more intense evaluation and management. Although a systematic review found that clinical and laboratory features generally performed poorly in localizing the site of UTI,1 the majority of studies use either one or both of these characteristics to define patient populations. As such, evidence for the treatment of UTIs is generally based upon classifying the nature of urinary infection according to clinical characteristics. Acute pyelonephritis refers to infection within the kidney parenchyma and is characterized clinically by systemic symptoms such as fever (>38°C), malaise, vomiting, abdominal pain and loin tenderness. Cystitis refers to infection limited to the bladder that is not associated with systemic features but may present with localizing symptoms such as frequency, urgency, dysuria and suprapubic discomfort. Initial treatment of UTI is guided by the clinical presentation. Children with significant systemic symptoms (fever and loin pain) have a clinical diagnosis of pyelonephritis but can be treated with oral antibiotics providing they are older than 1 month of age, do not appear septic and able to tolerate oral medications. The optimal duration of therapy is unknown but 7–10 days is currently recommended. Children without systemic features can be managed as cystitis and treated with oral antibiotic therapy for 2–4 days. Renal tract imaging after UTI aims to identify children with renal tract abnormalities that increase their risk of repeated infections. This is based firstly on the assumption that recurrent infections damage the kidneys that in the long term leads to high blood pressure and end-stage kidney failure, and secondly that preventative treatments are effective. Preventative treatment options include low-dose antibiotics for long periods of time, surgical interventions to correct VUR and complementary therapies such as cranberry products. The first assumption on which renal tract imaging has been justified is the detrimental long-term effect of repeated UTI. Evidence on which this assumption is made is not convincing because retaining a large and representative sample of children with a relatively minor illness for long-term follow-up is difficult. In general, studies that have managed to follow children over long periods tend to retain those with more serious illness that leads to overestimation of rates of hypertension due to selection bias. A systematic review of the prevalence of hypertension following reflux nephropathy showed that 6–28% had hypertension.1 The risk of hypertension after childhood UTI is likely to be small. Registry data demonstrate that the frequency of end-stage renal failure following pyelonephritic scarring or reflux nephropathy is around 0.5% in the USA, 4% in Australia and Sweden, 7.3% in England and Wales and around 15% in some European countries.2 It is difficult to be precise about a risk of end-stage renal failure after UTI, but it would appear quite low given the available data. The second assumption that preventative treatment is effective is also questionable (refer to Section 4 of these guidelines). Available research findings for renal tract imaging assessment can determine which modality is optimal for identifying various aspects of the urinary tract. Renal ultrasound is useful for identifying structural abnormalities of the urinary tract. Structural abnormalities detected by ultrasound occur in 10–75% (median around 30%) of children scanned after a UTI.2 Abnormalities tend to be seen more often in younger children. Renal ultrasound may suggest an obstructive uropathy that can be diagnosed with a MAG3 renal scan (pelvi-ureteric or uretro-vesical obstruction) or MCUG (posterior urethral valves or other urethral pathology). MCUG is the preferred test to detect VUR, and between 30% and 40% of children with a UTI will have VUR detected.2 A DMSA scan is considered the most appropriate test for detection of renal parenchymal abnormalities. The National Institute for Health and Clinical Excellence evidence compilation concludes that 5% of children who have had a UTI have renal parenchymal abnormalities.2 A later systematic review of 33 studies3 showed that 57% of children had changes when measurements were made in the acute phase (<15 days after UTI) and 15% had renal parenchymal changes when measured during the follow-up phase (>5 months post UTI). There is insufficient evidence to conclude whether performing a DMSA scan to detect these abnormalities is beneficial to the patient in the long term.2, 4 There is insufficient evidence to demonstrate a clinical benefit for renal tract imaging after first UTI in children. Renal ultrasound can detect structural integrity of the renal tract system, whereas MCUG is the test of choice for detecting VUR and DMSA is the best test for identifying renal parenchymal abnormalities. Abnormalities of the renal tract are identified in 10–75% of children following a UTI. Few children with renal tract obstruction or grades 4 and 5 VUR will have normal ultrasound findings. Grades 4 and 5 VUR increase the risk of repeat UTI. Very few children (<5%) with UTI will develop hypertension and end stage renal failure. Having had one infection, the child is at a 13–19% risk of having another UTI.3-5 UTI causes pain, discomfort and irritability to the child, and anxiety, stress and inconvenience to the family. Preventing further infections would be considered beneficial from a family's perspective and may also protect the child's kidneys from damage. For many years, long-term, low-dose antibiotics were given to children at risk of recurrent infection under the assumption this treatment would prevent further UTI. Little evidence existed to support the practice and systematic reviews published in 2000 and 20016, 7 highlighted the poor quality and insufficient evidence to justify this practice. A number of larger, better designed trials were commenced and the most recent systematic reviews8, 9 analysed these and demonstrated a change in the evidence to show a small benefit of low dose antibiotics in the larger rigorously conducted trials. The small scale of the benefit (6% absolute risk reduction) and considerably increased rate of antibiotic resistance to the prophylactic drug suggests this treatment might best be reserved for those children at most risk of recurrent UTI. Studies of children with UTI3, 5 have shown that those with VUR are at increased risk of repeat infection. Interventions that focus on this abnormality include open and laparoscopic surgery to reimplant the ureter in an attempt to prevent the backward flow of urine towards the kidney. Later techniques include injecting a bulking agent that increases the stiffness of the ureter to prevent backward flow of urine. To date, evidence that stopping reflux by correction of the anatomical abnormality prevents morbidity from UTI and prevents kidney damage or hypertension is unconvincing. Such invasive treatment should be considered only for those with recurring symptomatic infections unimproved by other preventative treatment. Complementary therapies such as cranberry products, probiotics, methenamine hippurate, nasturtium and horseradish along with immuno active bacterial fractions have been trialled for the purpose of preventing recurrent UTI to some extent although primarily in adults. These trials generally demonstrate a benefit however rigour and power in study design are usually lacking. Practitioners working with children with UTI, often recommend treating constipation, increasing fluid intake, avoiding bubble baths, hygiene issues and addressing dysfunctional voiding patterns to prevent further UTI. Trials to explore the efficacy of these interventions are absent, but given the harmless nature and possible benefits of these options, parents may appreciate awareness of them. There is a small benefit in low-dose antibiotics for preventing further UTI in children, but the benefit should be weighed against harms such as increased bacterial resistance to the prophylactic drug. Cranberry product may be helpful if tolerated. Circumcision may be warranted in boys with high grade VUR or recurrent UTI. The evidence and recommendations in this KHA-CARI guideline have been evaluated and graded following the approach detailed by the GRADE working group (http://www.gradeworkinggroup.org). A description of the grades and levels assigned to recommendations is provided in Tables 1 and 2. High quality of evidence. We are confident that the true effect lies close to that of the estimate of the effect. Moderate quality of evidence. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low quality of evidence. The true effect may be substantially different from the estimate of the effect. Very low quality of evidence. The estimate of effect is very uncertain, and often will be far from the truth. For a full text version of the guideline, readers need to go to the KHA-CARI website (go to the Guidelines section (http://www.cari.org.au))