URINARY TRACT INFECTION IN CHILDREN: DIAGNOSIS, TREATMENT

Introduction

UTI is the most common bacterial infection in children aged <2 years [1], and it may be the first symptom of congenital anomaly of the kidneys and urinary tract (CAKUT), with VUR
being the most prevalent. The assumption that recurrent UTIs in patients with VUR lead to renal scarring and consecutive chronic kidney disease (CKD) had been the indication for ac- curate diagnosis and specific treatment of VUR. However, recently, this aggressive approach has been questioned, due to numerous studies undermining the clinical importance and effectiveness of VUR treat- ment [2]. Damage to the kidney tissue, which was previously attributed to UTIs or reflux nephropathy, has been found to be congenital in nature [3,4]. Similarly, antibiotic prophy- laxis in CAKUT has also been recently chal- lenged [5]. The current review summarized current knowledge and recommendations concerning UTI in children.
The first document included in the review was the National Institute for Health and

The guidelines substantially differ in rating evidence quality and strength of recommendations. The NICE authors revised evidence using separate criteria for interventions and diagnostic test accuracy, but they did not rate their recommendations. Similarly, neither CPS nor EAU/ESPU guidelines included any rating system for strength of rec- ommendations. The AAP guidelines are graded according to AAP policy; the Italian authors used Strength of Recommen- dation Taxonomy (SORT) criteria, whereas Polish guidelines were based on Grades of Recommendation Assessment, Development and Evaluation (GRADE) system. Thus, the strength of recommendations was unsuitable for comparison.

Definitions

A few guidelines included definitions of atypical/compli- cated or recurrent UTI. According to NICE, atypical UTI include: seriously ill patients, children with poor urine flow, abdominal or bladder mass, elevated serum creatinine, septicemia, failure to respond to treatment with suitable antibiotics within 48 h, and infections with non-Escherichia coli organisms. In both CPS and Polish guidelines, almost identical characteristics are listed as features of complicated UTI, and both latter guidelines cite NICE as a reference.

In addition, the NICE guideline separately lists risk factors for serious underlying pathology, which include: poor urine flow, history of previous UTI, recurrent fever of uncertain origin, antenatal diagnosis of renal abnormality, family history of VUR or renal disease, constipation, dysfunctional voiding, enlarged bladder, abdominal mass, evidence of spinal lesion, poor growth, and high blood pressure.

Italian guidelines similarly define risk factors for CAKUT as follows: prenatal or postnatal ultrasonographic abnor- malities, family history of VUR, septicemia, renal insuffi- ciency, male infants aged <6 months, suspected non- compliant family, micturition abnormalities or thickened

bladder wall, absence of a clinical response to antibiotics within 72 h, and pathogens other than E. coli.

A recurrent UTI is defined in the NICE guideline as two or more episodes of pyelonephritis, or one episode of pyelo- nephritis plus one or more episodes of cystitis, or three or more episodes of cystitis. The other guidelines provide no other definition of recurrent UTI.

Most of the risk factors listed above seem to be based on experts’ opinion, which explains the slight differences be- tween the guidelines. Scientific data on risk factors pre- dicting CAKUT in children with a first-time UTI are scarce and conflicting. Male gender, young age, positive family history [12], fever of 38 ○C, elevated CRP [13], high neutrophil ratio [14], pathogens other than E. coli, and positive blood culture [15,16] were all found to predict VUR in children with UTI. However, a recent study by Yılmaz

evaluated 300 children with their first UTI, and no clinical or laboratory data were found to correlate with the pres- ence of VUR [17]. On the other hand, Ristola et al., in a study evaluating 282 children with their first UTI, found that non-E. coli UTIs correlated with abnormal kidney ul- trasound, whereas risk factors predicting VUR included abnormal ultrasound, atypical infection, non-E. coli infec- tion, and recurrent infections [18].

Diagnosis

All guidelines agree that UTI in children may be difficult to diagnose, especially in children aged <2e3 years, because symptoms and signs in this age group are non-specific. For this reason, urine tests are warranted not only in children

with typical UTI symptoms, but also in cases of unexplained fever. This approach was reinforced by an AAP technical report from 2011, which stated that among children aged 2e24 months with fever of unknown origin, 5% had a UTI [19]. In addition, AAP, CPS, Polish and EAU/ESPU guidelines state that urine tests must be performed before antimi- crobial treatment is introduced, whereas CPS guidelines

emphasize that in cases of fever with a known source (e.g. diarrhea or rhinitis), urine tests should not be performed.

In toilet-trained children, the method of choice for diagnosing a UTI is a clean voided midstream urine sample [20]. The difficulty is obtaining a urine sample from a child who does not control voiding. There are four techniques for collecting a urine sample in those children. Non-invasive techniques include a bag applied to the perineum and clean catch midstream void, whereas invasive methods include bladder catheterization and suprapubic aspiration (SPA). The NICE guidelines recommend a clean catch sam- ple as the method of choice, but they also permit a urine collection bag. However, an AAP technical report states that up to 85% of positive culture results obtained by using a collection bag can be false positives [19]. For this reason, most guidelines published since 2011 agree that a urine sample collected in a bag applied to the perineum is only reliable when negative. Most guidelines recommend clean catch void as the preferred method, although, according to AAP, only invasive techniques may be used to confirm diagnosis of a UTI. The reason for this is probably that AAP guidelines relate to children aged up to 24 months, in whom mid-stream void has a high contamination rate of up to 26% [21], and is technically difficult and time-consuming. The CPS guidelines also recommend invasive techniques in children up to 24 months of age.

Some guidelines also suggest that invasive urine sam- pling methods are particularly appropriate in children who appear ill or have poor general health when there is an urgent need for antimicrobial treatment. The AAP technical report also states that SPA has higher pain scores and lower success rates than bladder catheterization, which in turn presents 95% sensitivity and 99% specificity in comparison with SPA [19]. This suggests catheterization as a preferred invasive urine sampling method.

Urinalysis

Urinalysis is a quick dipstick test for nitrite and leukocyte esterase, and microscopic examination for white blood cells (WBC) and bacteria. The nitrite dipstick test repre- sents the conversion of dietary nitrate by Gram-negative bacteria, and has high specificity (98%) for UTI [22]. Its major limitation is that it gives negative results when the bladder is emptied frequently or if the underlying pathogen is Gram-positive [23]. Therefore, most guidelines agree that a urine dipstick test is not recommended for the youngest children who void frequently. Leukocyte ester

a surrogate marker of pyuria, has sensitivity of 79% and specificity of 87% for UTIs [22].

White blood cells present in the urine on microscopic ex- amination are a useful indicator of inflammation associated with UTI, although there is no standardized definition of pyuria in the literature. In the microscopic analysis of centrifuged urine, five WBCs per high-power field is a usual threshold for pyuria. Another method is automatic counting in uncentrifuged urine, with 10 WBCs being a threshold value.

Cultures

The definition of significant bacteriuria varies slightly be- tween guidelines. According to AAP, significant bacteriuria is defined as 5 104 colony forming units (CFU) per milli- liter (CFU/ml) of urine obtained by catheterization. This definition is derived from a study published in 1956. Urine cultures from women with and without pyelonephritis symptoms revealed that the threshold range, in which the proportion of patients with symptomatic UTI exceeded those without symptoms, was between 104 and 105 CFU/ml [24]. Hoberman et al. confirmed this threshold value in children with UTI in a study in which urine culture results were verified by a renal scan with DMSA [25]. Other guidelines present broader definitions considering the urine collection method; they are presented in Table 1.

The definition included in EAU/ESPU guidelines is based on the finding that pyelonephritis may also be present with lower CFU counts on cultures. In 2016, Swerkersson et al. published an interesting study evaluating 430 infants with first-time  UTI,  diagnosed  by  SPA,  revealing  that  19% of

children had low bacterial counts of <104 CFU/ml. The authors suggested that UTI with low bacterial count  might

be a separate entity, associated with non-E. coli etiology and low inflammatory response, but with the same risk of VUR and renal scarring [26]. In another study, it was found that 19% of infants with UTI diagnosed by SPA had bacte-

riuria <105 CFU/ml in clean catch voided urine [27] and could have been missed with a higher cut-off value. These

arguments suggest that the EAU/ESPU bacteriuria defini- tions are probably the most appropriate for diagnosing UTI. Children with positive urine culture and normal urinal- ysis, without symptoms, are regarded as having asymp- tomatic bacteriuria that, in otherwise healthy subjects, is not an indication for any intervention. This applies to all

guidelines analyzed in the current review.

Management

The varying approaches to management of UTI included in guidelines are summarized in Table 1. The AAP authors stated that there is no difference in efficacy between oral and intravenous treatment of UTI, which has been proven in numerous studies both in children and adults [28,29]. Thus, most children with a UTI can be treated orally. Parenteral treatment is only required in children who are severely ill or unable to retain oral intake; however, sequence treat- ment is recommended even in those children.

The AAP guidelines suggest that pyelonephritis treat- ment should last 7e14 days. This broad range is due to the lack of sufficient data identifying optimal treatment dura- tion [30]. There is evidence that a 1e3-day regimen for

febrile children with a UTI is less effective than a 7-day regimen, which is the reason for the recommended mini- mum treatment duration of 7 days.

The choice of antibiotic should be based on locally developed current resistance patterns of urinary patho- gens. The AAP guidelines state that antibiotics excreted in the urine, which do not reach therapeutic concentrations in the blood (e.g. nitrofurantoin), should not be used in py- elonephritis treatment. Another issue to be considered in the treatment of UTI is the increasing frequency of in- fections with extended-spectrum beta-lactamases (ESBL)- producing pathogens, which is reported to be 20% and is more common in younger children [19].

Further diagnostics

Further diagnostics in children with febrile UTI is undoubt- edly the most controversial issue. The general tendency is to restrict indications to VCUG and DMSA scintigraphy. Sig- nificant radiation exposure, the risk of catheter-induced UTI, stress for a young patient and their parents, and the cost of the imaging techniques must be considered. The main objective of performing imaging tests following a  UTI is to identify children with CAKUT, mainly VUR, who may be more susceptible to recurrent UTI and further renal scarring. Some of those patients may benefit from surgical interventions. They may also benefit from antimicrobial prophylaxis, which  used  to  be  routinely  administered  in children with CAKUT, and had been proven effective in reducing the risk of recurrent UTI in the Prevention of Recurrent Urinary Tract Infection in Children with Ves- icoureteric Reflux and Normal Renal Tracts (PRIVENT) study [31]. This approach, however, has recently been chall- enged by several studies demonstrating that antimicrobial prophylaxis in those children neither avoids subsequent infections nor influences further renal scarring [5,32]. Roussey-Kesler et al. found no benefit from antimicrobial

prophylaxis in children with grade IeIII VUR, excluding boys

with grade III VUR in whom it may avoid further UTI [33], whereas a Swedish reflux study revealed that in a group of infant girls with grade III or IV VUR, antimicrobial prophy- laxis is effective in preventing renal scarring [34]. On the other hand, the recent Randomized Intervention for Chil- dren with Vesicoureteral Reflux (RIVUR) trial has proven that antimicrobial prophylaxis reduces risk of UTI recur- rence, but not of renal scarring in children with VUR [35].

The benefit of diagnosing VUR has also been questioned for other reasons. Several studies have stated that the risk of VUR in children with UTI is similar to the rest of the population at around 30% [19,36]. Moreover, a large pro- portion of children with UTI and VUR, particularly low- grade, reach spontaneous resolution without medical intervention [37] and mild/moderate VUR does not increase the risk of recurrent UTI or renal scarring [38]. The question is: how to identify those children with VUR who would benefit from surgical treatment?