Urinary Tract Infections in Children: Clinical and Antimicrobial Resistance Data from Bucharest Area, Romania

Abstract

Introduction: Urinary tract infections (UTIs) are among the most common bacterial diseases of childhood with an increased frequency in infants and young children. Methods: We conducted a retrospective study of children diagnosed with UTI in the Alessandrescu-Rusescu National Institute for Mother and Child Health, Bucharest, Romania between January 2017 and December 2019. We evaluated the patients’ clinical characteristics, laboratory parameters and the resistance profile of the identified uropathogens. Results: A total of 264 children were included in the analysis. Females (71.6%, n = 186) and infants (52.7%, n = 139) were more commonly affected. The recurrence rate was 27.7% and was positively associated with the presence of renal malformations. Age under 1-year, increased leukocyte and neutrophil counts, and elevated C-reactive protein were associated with hospitalization. E. coli (80.3%, n = 212) was the main etiological agent isolated, followed by Proteus mirabilis (9.8%, n = 26) and Klebsiella spp. (6.4%, n = 17). We identified increased resistance for all germs to common antibiotics used in pediatrics: ampicillin, amoxicillin/clavulanate, cefuroxime, ceftriaxone, and trimethoprim/sulfamethoxazole. Conclusions: We identified an increased resistance of uropathogens to antimicrobials commonly used in children. Reporting antimicrobial resistance from real-world clinical practice is necessary for accurate mapping and continuous updating of initial treatment recommendations until antibiogram results are received. In Romania and other countries, extensive studies are needed to follow up uropathogen resistance in both children and adults.

Introduction

Urinary tract infections (UTIs) are common among children, with an increased incidence especially in infants and young children [1]. By adolescent age, about 11% of girls and 7% of boys have had at least one episode of UTI, with recurrent infections reported in many cases [2]. The morbidity of UTIs is significantly higher among infants, who are at increased risk of kidney damage or sepsis [3,4].

Prompt diagnosis and appropriate treatment are necessary to reduce the burden that this disease has among children, but also indirectly on parents, doctors, and health systems. Identification of the etiological agent is important for the right choice of targeted antimicrobial therapy. Unfortunately, increasing antibiotic resistance makes it difficult to prescribe an empiric antibiotic treatment for UTIs until the results of urine culture and antibiogram are obtained. Antimicrobial resistance patterns differ from country to country, and even within the same country, across regions. Therefore, knowing the local antimicrobial susceptibility is crucial for the correct adaptation of therapeutic management in patients with UTI.

The aim of this study is to present the clinical and epidemiological characteristics and to report data on antimicrobial resistance of bacteria isolated in urine among children admitted to the Alessandrescu-Rusescu National Institute for Mother and Child Health (NIMCH) in Bucharest, Romania.

Methods

Alessandrescu-Rusescu NIMCH is a tertiary children’s hospital in the second district of the Romanian capital, Bucharest. The second district of Bucharest is a metropolitan area of about 70 square km with a population of nearly half a million inhabitants. NIMCH has more than 45.000 pediatric outpatients per year, and more than 7000 pediatric inpatients.

We conducted a retrospective, analytical, descriptive and comparative study corresponding to the period January 2017–December 2019. We included in the analysis all children (0–18 years) evaluated in NIMCH during the study period with positive urine culture and symptomatology (fever or dysuria or frequent urination or lower back pain) and urinalysis suggestive of UTI. Children with positive urine culture but no change in urinalysis or incomplete data in the hospital information system were excluded.

Isolation of the etiologic agent in urine culture and antimicrobial testing was done as part of routine clinical practice according to the internal protocol of the NIMCH microbiology laboratory. Data were interpreted according to CLSI guidelines for 2017 [5], 2018 [6] and 2019 [7] respectively.

Epidemiological, clinical and laboratory data were collected in an Excel database from the hospital information system.

Data analysis was performed with IBM SPSS Statistics for Windows, version 25 (IBM Corp., USA). Categorical data are presented as frequency/percentage and compared using the Chi-squared test. For continuous variables, data distribution was checked using the Shapiro-Wilk test; for non-normally distributed variables, median and interquartile range (IQR) are reported, the Mann-Whitney U test was used for statistical comparisons between the study groups. The level of statistical significance was set at p < 0.05.

This study was approved by the Ethics Committee of NIMCH with a waiver for informed consent.

Results

A total of 302 pediatric patients with positive urine cultures were evaluated in NIMCH during the period under review. Of these, only 264 were included in the analysis: 122 patients (46.2%) from 2017, 70 (26.5%) from 2018, and 72 (27.3%) from 2019. A number of 8 patients were excluded because they had normal urinalysis, and 30 others because they had incomplete data in the medical or laboratory records.

Patient characteristics

The majority of patients included in the study were female (71.6%, n = 189). The median age was 1.8 years (IQR: 0.6, 4.6 years), but infants (under 1-year) predominated (52.7%, n = 139). Among male patients, infants were significantly more commonly affected, their risk of UTI being 1.8 times higher than that of male children over 1-year [36.0% (50/139) vs. 20.0% (25/125), p = 0.004, χ² = 8.3, OR = 1.8, 95%CI: 1.2–2.7].

A percentage of 56.8% (n = 150) required hospitalization, the remaining 114 patients (43.2%) were evaluated in the hospital and received outpatient treatment. Infants more frequently required hospitalization compared to children over 1-year [77.0% (107/139) vs. 34.4% (43/125), p < 0.001, χ² = 49.8, OR = 6.6, 95%CI: 3.8–11.3]. By sex, male patients required hospitalization more frequently, but not significantly, compared to female patients [65.3% (49/75) vs. 53.4% (101/189), p = 0.086]. Median length of hospitalization was 3 days (IQR:1, 7 days). The number of days of hospitalization was influenced by age, and infants had a median hospital admission duration 4 days longer than children over 1-year [6 days (IQR: 4, 9 days) vs. 2 days (IQR: 1, 7 days), p < 0.001, U = 4972, z = −6.26].

A total of 73 children (73/264, 27.7%) had at least one prior episode of UTI in their history, and 34 patients (34/264, 12.9%) were known or were diagnosed during admission with a renal malformation. Thus, the risk of recurrence was significantly higher among those with renal malformations [27.4% (20/73) vs. 7.9% (14/177), p < 0.001, χ² = 19.0, OR = 4.8, 95%CI: 2.3–10.1]. Recurrences were not associated with patient gender (21.3% for males vs. 43.2% for females, p = 0.148), age group (25.9% for infants vs. 29.6% for children over 1-year, p = 0.502) or hospitalization (23.3% for inpatients vs. 33.3% for outpatients, p = 0.071).

Clinical and laboratory characteristics

The clinical picture at presentation was dominated by fever (79.9%, 211/264) at a median temperature of 39.4 °C (IQR: 38.9, 39.8 °C), and was found in almost all infants (

Table 1. Clinical and laboratory characteristics of pediatric patients with urinary tract infection.

Characteristics	Total Patients, n = 264	Infants, n = 139	Children Under 1-Year, n = 125	Statistical Analysis
Clinical Characteristics
Fever, n (%)	211 (79.9)	133 (95.7)	78 (62.4)	p < 0.001, χ2 = 45.4, OR = 13.4, 95%CI: 5.5–32.7
Temperature, median (IQR) °C	39.4 (38.9, 39.8)	39.4 (38.9, 39.8)	39.4 (38.8, 39.9)	p = 0.695
Malaise, n (%)	185 (70.1%)	117 (84.2%)	68 (54.4%)	p < 0.001, χ2 = 27.8, OR = 4.5, 95%CI: 2.5–7.9
Nausea, n (%)	52 (19.7)	30 (21.6)	18 (17.6)	p = 0.334
Diarrhea, n (%)	64 (24.0)	43 (30.9)	21 (16.8)	p = 0.002, χ2 = 10.1, OR = 4.6,95%CI: 1.4–14
Abdominal pain, n (%)	86 (32.6)	NA	86 (68.8)	NA
Lower back pain, n (%)	5 (1.9)	NA	5 (4.0)	NA
Dysuria, n (%)	63 (23.9)	NA	63 (50.4)	NA
Frequent urination, n (%)	36 (13.6)	NA	36 (28.8)	NA
Macroscopically cloudy urine, n (%)	57 (21.6)	18 (13.0)	39 (31.2)	p < 0.001, χ2 = 11.9, OR = 3.0,95%CI: 1.6–5.7
Macroscopic hematuria, n (%)	15 (5.7)	4 (2.9)	11 (8.8)	p = 0.060
Laboratory characteristics
Increased leukocytes [N = 203], n (%)	82 (40.4)	61/120 (50.8)	21/83 (25.3)	p < 0.001, χ2 = 12.2, OR = 3.1, 95%CI: 1.7–5.6
Increased neutrophils [N = 203], n (%)	68 (33.5)	46/120 (38.3)	22/83 (26.5)	p = 0.096
Anemia [N = 203], n (%)	63 (31.0)	46/120 (38.3)	17/83 (20.5)	p = 0.008, χ2 = 6.5, OR = 2.4, 95%CI: 1.3–4.6
Increased thrombocytes [N = 203], n (%)	55 (27.2)	46/120 (38.3)	9/83 (10.8)	p < 0.001, χ2 = 19.3, OR = 5.8, 95%CI: 2.6–13.1
Inflammatory syndrome [N = 193], n (%)	122 (63/2)	85/117 (72.6)	37/76 (48.7)	p = 0.001, χ2 = 11.4, OR = 2.8, 95%CI: 1.5–5.1
Increased urea [N = 170], n (%)	9 (5.3)	4/105 (3.8)	5/65 (7.7)	p = 0.236
Increased creatinine [N = 163], n (%)	2 (1.2)	1/101 (1.0)	1/62 (1.6)	p = 0.413
Increased AST [N = 156], n (%)	7 (4.5)	4/99 (4.0)	3/57 (5.3)	p = 0.722
Increased ALT [N = 156], n (%)	3 (1.9)	2/99 (2.0)	1/57 (1.8)	p = 0.907

ALT—alanine transaminase; AST—aspartate transaminase; NA—not applicable; IQR—interquartile range.

). Malaise was present in 70.1% of children, more common in infants (Table 1). Manifestations such as abdominal pain, lower back pain, dysuria, polyuria were quantified only in children older than 1-year (Table 1).

The majority of patients underwent assessment of their complete blood count, inflammatory profile and biochemical parameters. Characteristics of laboratory parameters are highlighted in Table 1 for all patients analyzed and by age groups. Infants more commonly had leukocytosis (p < 0.001), anemia (p = 0.008) and thrombocytosis (p < 0.001). Also, the presence of inflammatory syndrome, characterized by C-reactive protein value above 0.5 mg/dL, was significantly more common in the group of children under 1-year (p = 0.001).

We observed that increased numbers of leukocytes (47.0% vs. 22.2%, p = 0.002, χ² = 10.1, OR = 3.1, 95%CI: 1.5–6.4), and neutrophils (38.9% vs. 18.5%, p = 0.007, χ² = 7.4, OR = 2.8, 95%CI: 1.3–3. 6), as well as CRP values above normal (68.8% vs. 46.9%, p = 0.010, χ² = 7.5, OR = 2.5, 95%CI: 1.3–4.8) were significantly more common among patients requiring hospitalization (Figure 1). Furthermore, the presence of inflammatory syndrome at the time of admission was significantly associated with longer length of hospitalization [5.5 days (IQR: 3, 8 days) vs. 4 days (IQR: 1, 8 days), p = 0.040, U = 3569, z = −2.06]

Uropathogens and antimicrobial resistance profiles

The bacterial agents isolated in urine culture were Gram-negative bacilli in 98.5% of cases. Only 4 patients had Gram-positive cocci isolated: 2 strains of Enterococcus faecalis isolated from two infants, an 11-month-old female and a 10-month-old male, 1 strain of Streptococcus agalactiae isolated from a 10-year-old girl and 1 strain of Staphylococcus saprophyticus isolated from a 12-year-old girl (Figure 2). Overall, the most common bacterium isolated was Escherichia coli in 80.3% (n = 212) of cases, with similar proportions in infants and children over 1-year (Figure 2). Proteus mirabilis and Klebsiella spp. followed in frequency, being isolated in 9.8% (n = 26) and 6.4% (n = 17) of cases, respectively. Pseudomonas aeruginosa grew in 3 cases in urine culture, and Serratia spp. and Citrobacter spp. were identified in 1 case each.

E. coli showed a high rate of resistance to ampicillin (63.8%), amoxicillin/clavulanate (44.9%) and trimethoprim/sulfamethoxazole (27.0%). Similarly, for Proteus mirabilis we identified increased rates of resistance to ampicillin (38.5%) and trimethoprim/sulfamethoxazole (34.6%). In addition, Klebsiella spp. showed high resistance to cephalosporins and even carbapenems. The antibiotic susceptibility profiles of E. coli, Proteus mirabilis and Klebsiella spp. are detailed in

Table 2. Antimicrobial susceptibility profile of E. coli isolated from pediatric urinary tract infections.

Antibiotic	Susceptible n (%)	Intermediate n (%)	Resistant n (%)
Ampicillin [n = 207]	53 (25.6)	22 (10.6)	132 (63.8)
Amoxicillin/clavulanate [n = 207]	77 (37.2)	37 (17.9)	93 (44.9)
Ceftriaxone [n = 207]	180 (94.2)	0	27 (13.0)
Cefuroxime [n = 207]	164 (79.2)	5 (2.4)	38 (18.4)
Ceftazidime [n = 110]	86 (78.2)	3 (2.7)	21 (19.1)
Ciprofloxacin [n = 198]	171 (86.4)	7 (3.5)	20 (10.1)
Colistin [n = 72]	72 (100)	0	0
Gentamicin [n = 207]	187 (90.3)	5 (2.5)	15 (7.2)
Trimethoprim/sulfamethoxazole [n = 207]	149 (72.0)	2 (1.0)	56 (27.0)
Meropenem [n = 16]	16 (100)	0	0
Imipenem [n = 16]	16 (100)	0	0

,

Table 3. Antimicrobial susceptibility profile of Proteus mirabilis isolated from pediatric urinary tract infections.

Antibiotic	Susceptible n (%)	Intermediate n (%)	Resistant n (%)
Ampicillin [n = 26]	16 (61.5)	0	10 (38.5)
Amoxicillin/clavulanate [n = 26]	21 (80.8)	4 (15.4)	1 (3.8)
Ceftriaxone [n = 26]	25 (96.2)	0	1 (3.8)
Cefuroxime [n = 26]	24 (92.3)	0	2 (7.7)
Ceftazidime [n = 7]	7 (100)	0	0
Ciprofloxacin [n = 23]	23 (88.5)	2 (7.7)	1 (3.8)
Colistin [n = 7]	7 (100)	0	0
Gentamicin [n = 26]	24 (92.3)	0	2 (7.7)
Trimethoprim/sulfamethoxazole [n = 26]	17 (65.4)	0	9 (34.6)
Meropenem [n = 7]	7 (100)	0	0
Imipenem [n = 7]	7 (100)	0	0

and

Table 4. Antimicrobial susceptibility profile of Klebsiella spp. isolated from pediatric urinary tract infections.

Antibiotic	Susceptible n (%)	Intermediate n (%)	Resistant n (%)
Ampicillin [n = 17]	1 (5.9)	0	16 (94.1)
Amoxicillin/clavulanate [n = 17]	3 (17.6)	3 (17.6)	11 (64.7)
Ceftriaxone [n = 17]	10 (58.8)	0	7 (41.2)
Cefuroxime [n = 17]	10 (58.8)	0	7 (41.2)
Ceftazidime [n = 7]	3 (42.9)	1 (14.2)	3 (42.9)
Ciprofloxacin [n = 17]	13 (76.5)	1 (5.9)	3 (17.6)
Colistin [n = 7]	7 (100)	0	0
Gentamicin [n = 17]	11 (64.7)	0	6 (35.3)
Trimethoprim/sulfamethoxazole [n = 17]	14 (82.4)	0	3 (17.6)
Meropenem [n = 7]	6 (85.7)	0	1 (14.3)
Imipenem [n = 7]	5 (71.4)	0	2 (28.6)

. For the 3 isolates of P. aeruginosa antimicrobial testing showed 100% resistance to ampicillin, amoxicillin/clavulanate, cefuroxime, ceftriaxone, trimethoprim/sulfamethoxazole, and piperacillin (data for 2 isolates) and 100% susceptibility to ceftazidime, cefepime, ciprofloxacin, colistin, gentamicin, amikacin, meropenem, imipenem (data for 2 isolates) and piperacillin/tazobactam (data for 2 isolates). The isolate of Citrobacter spp. was resistant to amoxicillin/clavulanate only, being susceptible to the rest of the antibiotics tested (cefotaxime, ceftazidime, cefepime, ciprofloxacin, amikacin, trimethoprim/sulfamethoxazole, meropenem and imipenem), while the strain of Serratia spp. was resistant to ampicillin, amoxicillin/clavulanate and cefuroxime and showed susceptibility to ceftriaxone, ceftazidime, ciprofloxacin, and trimethoprim/sulfamethoxazole. One of the two Enterococcus faecalis isolates identified in this study was resistant to ampicillin, ciprofloxacin, and gentamicin and both were susceptible to linezolid, vancomycin and teicoplanin. The S. saprophyticus isolate showed resistance only to ampicillin and was susceptible to the remaining antibiotics tested, while S. agalactiae was susceptible to all antibiotics tested.

Discussion

In this paper we presented data on UTIs in children with a focus on the clinical characteristics and antimicrobial resistance profile as derived from actual clinical practice in a tertiary pediatric hospital in Bucharest, Romania. To the best of our knowledge, following literature review, there are no recent data from this region of Romania for children with UTI.

As expected, we identified females and infants as being the most likely to develop UTIs. When it comes to boys, those under 1-year of age were more commonly affected. These data are consistent with those published previously [1,4]. Hospitalization was required more frequently among infants due to the need for monitoring and treatment administration. However, recent data show that outpatient management of infants may be a good and safe choice if their progress is adequately followed [8].

Compared to other acute illnesses, patients with UTIs may experience recurrences with a frequency of 30 to 50% [1,4]. For the patients in our study we identified a recurrence rate of 27.7%. These were not associated with gender or age as reported in other studies [1], but were associated with the presence of renal malformations.

As a strength of our study, we showed that blood count assessment and inflammatory profile can be useful in guiding the decision whether to hospitalize a child with UTI. Increased leukocyte counts, neutrophils and an above normal CRP value were positively associated with the need for hospitalization.

We identified E. coli as the main etiologic agent of UTIs in children regardless of age and sex, followed by Proteus mirabilis and Klebsiella spp. E. coli remains the main pathogen implicated in UTIs globally [1]. Therefore empirical antibiotic choice should aim at adequate coverage of this germ. But we should not forget the possible involvement of other Gram-negative bacilli or, occasionally, Gram-positive cocci.

Resistance data is concerning. In our study, E. coli has shown high resistance to antibiotics routinely recommended for the pediatric population (ampicillin, amoxicillin/clavulanate, cefuroxime or trimethoprim/sulfamethoxazole). Recent data reported from Cluj (a study that included children with UTI in the central region of Romania) found slightly lower resistance rates of E. coli to ampicillin (56.5%), cefuroxime (12.5%) or trimethoprim/sulfamethoxazole (24.7%) [9]. In comparison, if we look at recent reports on the adult UTI population in Romania, the situation is even more alarming: in a multicenter study, E. coli resistance rates were 96.9% to ampicillin, 70.9% to trimethoprim/sulfamethoxazole and 58.5% to amoxicillin/clavulanate [10]. At the same time, the pattern of resistance of E. coli to 3rd generation cephalosporins (19.1%—ceftazidime, 13.0%—ceftriaxone in our study) raises concern about the potential production of extended-spectrum β-lactamases (ESBL). The study from Cluj reported lower resistance rates (10.5% for ceftazidime, 3.4% for ceftriaxone),⁹ but there are national data that identified higher values [11]. The prevalence of ESBL-positive E. coli in Europe has increased greatly in recent years compared to the 2000s [12] and is a real threat to antibiotic therapy. Therefore, regular monitoring of E. coli resistance and determination of the ESBL phenotype are important to take appropriate action.

Proteus mirabilis showed good susceptibility to the antibiotics tested except for ampicillin and trimethoprim/sulfamethoxazole, where the resistance rate was higher compared to other national reports [10]. However, our data for Klebsiella spp. are alarming. With the exception of colistin, resistance rates were high for all antibiotics tested, including 3rd generation cephalosporins, fluoroquinolones and carbapenems. The data are not surprising for Romania, as similar resistance profiles have recently been reported elsewhere in the country in the adult population [11]. Klebsiella spp. has been shown to be an opportunistic germ, capable through numerous molecular mechanisms of developing increased resistance to many classes of antibiotics, including colistin, according to recent data from Europe [13,14]. These data, taken together with the fact that this germ is frequently implicated in pediatric pathology, indicate that urgent surveillance measures for Klebsiella infections are needed to address the increased rate of antimicrobial resistance.

Gram-positive bacteria are less commonly implicated in childhood urinary pathology [15], but should not be overlooked in the empirical choice of antibiotic, especially in patients with clear UTI symptoms but no nitrites on urinalysis. Ampicillin, according to our data, does not seem to be a good empirical choice for the child with UTI, as for all identified etiological agents resistance was high for this antimicrobial.

The retrospective nature of the data and the non-systematized testing of samples are the main limitations of the study. Nevertheless, our data provide insight into actual clinical practice in a tertiary pediatric hospital in the Romanian capital and add data on antimicrobial resistance among children with UTI in this area.

Conclusions

Urinary tract infections are a common pathology, whose impact is often underestimated, but can be very difficult to treat. Reporting antimicrobial resistance from real world clinical practice is necessary for a correct mapping and continuous updating of initial treatment recommendations until antibiogram results are received. We identified increased resistance of uropathogens to antibiotics routinely used in pediatric practice. In Romania and elsewhere, extensive follow-up studies of uropathogen resistance in both children and adults are needed. At the same time, ongoing medical education and information programmes are needed among physicians and the general population alike to reduce unnecessary antibiotic consumption, which is the main driver of increased resistance rates and treatment difficulties when antibiotics are really needed.