Article Text

This article has a correction. Please see:

Download PDFPDF

Clinical and laboratory characteristics of non-E coli urinary tract infections
  1. S Friedman1,
  2. S Reif1,
  3. A Assia1,
  4. I Levy2
  1. 1Department of Pediatrics, Dana Children’s Hospital, Sourasky Medical Center, Tel Aviv, Israel
  2. 2Infectious Disease Unit, Schneider Children’s Medical Center of Israel, Petah Tiqva, and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
  1. Correspondence to:
    Dr I Levy
    Infectious Disease Unit, Schneider Children’s Medical Center of Israel, Petah Tiqva 49202, Israel; itzhakl{at}; lavyguy{at}


Comparison of the clinical and laboratory characteristics of infants and children with urinary tract infection caused by E coli (n = 107) or other pathogens (n = 32) yielded a significantly higher association of non-E coli disease with urinary tract anomalies, younger age, and previous antibiotic treatment. Underlying urinary tract anomalies were noted in 18 patients, of whom 14 (77%) were infected by non-E coli pathogens. The most frequent anomaly was grade 3–4 vesicoureteral reflux (50%), followed by hydronephrosis (22.7%), ureteropelvic junction obstruction (9%), hypospadias (4.5%), pinpoint meatus (4.5%), and dysplastic kidney (4.5%).

  • non-E-coli
  • UTI
  • infants

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Renal scarring is an important complication of urinary tract infections, which predisposes patients to hypertension and renal impairment.1 Known risk factors for renal scarring are younger age, anatomical malformations, especially vesicoureteral reflux, delayed treatment, and a causative pathogen other than Escherichia coli.2–,5 Honkinen et al found that non-E coli UTI is associated with more urinary tract abnormalities than E coli UTI.6,7 The aim of this study was to further characterise the clinical and laboratory features of non-E coli UTI compared to E coli UTI.


The medical records and databases of the Department of Pediatrics and the Microbiology Laboratory of Dana Children’s Hospital, Tel Aviv were reviewed for all infants and children (age 1 week to 16 years) hospitalised for UTI between January 2000 and 31 March 2004. The diagnosis of UTI was based on the isolation of uropathogens in suprapubic aspiration, pure growth of >103 colony forming units in a catheter sample, or isolation of 105 pathogens in a midstream urine specimen.

The initial laboratory work-up of all patients with UTI included complete blood count (CBC), erythrocyte sedimentation rate (ESR), C reactive protein (CRP), blood urea nitrogen, serum creatinine, urinalysis, and blood and urine cultures. Empiric therapy for suspected UTI consists of intravenous ampicillin and gentamicin in neonates, and cefuroxime in older infants and children pending urine culture results. Renal and urinary tract ultrasound is performed in patients with a first episode of UTI.

The patients were divided into those with E coli and non-E coli UTI and compared for demographic data, clinical data (degree and duration of fever and chills, time to defervescence, and length of hospital stay), underlying urinary tract anomalies (by history and ultrasound), recurrent infection (by history and past medical records), use of antibiotics of more than 48 hours for any reason in the two weeks prior to the present episode, and laboratory data (urinalysis, ESR, WBC, and serum creatinine).


Of the 257 children initially diagnosed with UTI during the study period, 139 fulfilled the diagnostic criteria for UTI, 107 (76.9%) with E coli UTI and 32 (23.1%) with UTI caused by other pathogens: Klebsiella pneumoniae, n = 10 (31%); Pseudomonas aeruginosa, n = 7 (21%); Enterobacter spp., n = 5 (16%); Citrobacter spp., n = 3 (9%); Serratia spp., Proteus mirabilis, Enterococcus spp., n = 2 each (6%), and Morganella morganii, n = 1 (3%). Nineteen per cent of the E coli group and 21.8% of the non-E coli group were neonates. The clinical and laboratory characteristics of the groups are shown in table 1.

Table 1

 Comparison of demographic clinical and laboratory characteristics of E coli and non-E coli UTI

Underlying urinary tract anomalies were noted in 18 patients, of whom 14 (77%) were infected by non-E coli pathogens (Pseudomonas aeruginosa, 50%; Klebsiella pneumoniae, 35%; Morganella morganii, 7%; and Enterobacter spp., 7%). The difference in the prevalence of urinary tract anomalies between the non-E coli and E coli groups (43.8% and 5.6%, respectively) was statistically significant (p < 0.0001). Four of the 18 patients had more than one anomaly. The most frequent anomaly was grade 3–4 vesicoureteral reflux (50%), followed by hydronephrosis (22.7%), ureteropelvic junction obstruction (9%), hypospadias (4.5%), pinpoint meatus (4.5%), and dysplastic kidney (4.5%). Non-E coli UTI was also significantly associated with younger age, previous antibiotic therapy, recurrent UTIs, lower maximal temperature, shorter time to defervescence, prolonged hospital stay, and lower mean WBC and ESR (p < 0.05 for all variables), but not with urinalysis. On multivariate regression analysis, significant factors were younger age, underlying urinary tract anomalies, and previous antibiotic treatment (table 2).

Table 2

 Multivariate logistic regression for variables associated with non-E coli UTI


In this study of infants and children with E coli and non-E coli UTI, we detected significant differences in the clinical and laboratory characteristics between the two groups. Patients with non-E coli UTI were younger, had milder clinical signs, longer hospitalisation, and higher rates of urinary tract anomalies. Our higher rate of urinary tract anomalies in patients with non-E coli UTI agrees with an earlier study showing that UTI caused by Klebsiella or Enterococcus was significantly associated with ureterovesical reflux compared to E coli UTI.6 The severity of the UTI is related to host susceptibility and bacterial virulence. Uropathogenic E coli belong to specific capsular and flagellar strains and have multiple virulence factors, such as alpha haemolysins, P fimbriae, aerobactins, Sat proteins, and necrotising factors, which enable them to colonise and invade the normally sterile urinary tract and bloodstream. By contrast, except for endotoxins, no constant virulence factors were found in Klebsiella or Enterobacter.8,9 Therefore, we assume that the higher rate of E coli UTI in the patients with a normal urinary tract in our study is associated with the higher virulence of E coli compared to non-E coli strains and that the abnormal urinary tract may have made it possible for the less virulent non-E coli bacteria to cause infection.9

Indeed, the more enhanced inflammatory response of fever, increased WBC, and increased ESR in the patients with E coli UTI, in addition to their significantly longer time to defervescence, may also be associated with the high virulence of this pathogen compared to non-E coli strains. Nevertheless, hospitalisation was longer in the non-E coli UTI group, owing to their greater number of urinary tract anomalies which required prolonged antibiotic therapy being defined as complicated UTI, the absence of an approved oral treatment for certain pathogens such as Pseudomonas aeruginosa, or inappropriate initial antibiotic therapy. Antibiotic treatment prior to hospitalisation was also a risk factor for non-E coli UTI. This treatment may have selected for non-E coli bacteria and colonisation by more resistant organisms. This may also have accounted for the milder nature of the UTI episode in children with E coli compared to non-E coli infections.

There are limitations in this study. First, we included in the study a selected group of hospitalised patients who did not necessarily represent the whole population of patients with UTI. Second, due to the relative small number of children with non-E coli UTI, no further subdivision according to age or causative organism was performed.

In summary, non-E coli may have milder clinical and laboratory features than E coli UTI and a higher prevalence of urinary tract anomalies. Future larger prospective studies are needed to confirm these results, determine the possible causes of these differences, and integrate these observations into clinical guidelines.



  • Published Online First 22 May 2006

  • Competing interests: none declared

Linked Articles

  • Précis
    BMJ Publishing Group Ltd and Royal College of Paediatrics and Child Health
  • Atoms
    Howard Bauchner
  • Corrections
    BMJ Publishing Group Ltd and Royal College of Paediatrics and Child Health