Community-acquired urinary tract pathogens and their resistance patterns in hospitalized children in southeastern Ontario between 2002 and 2006

Abstract

BACKGROUND:

The purpose of the present study was to assess the epidemiology and resistance patterns of bacteria causing urinary tract infections in children who were admitted to Kingston General Hospital (Kingston, Ontario) – the regional tertiary care hospital of southeastern Ontario.

METHODS:

A retrospective chart review of patients one to 18 years of age who were admitted to Kingston General Hospital with a discharge diagnosis of urinary tract infection between 2002 and early 2006 was undertaken.

RESULTS:

One hundred forty-two patient charts were reviewed, of which 56.3% of patients were female. The mean age of the patients was 12.3 months. The most common bacteria identified on urine culture over a five-year period were Escherichia coli (71.6%), Enterococcus species (5.7%) and Klebsiella species (5.0%). Bacteria were frequently resistant to ampicillin (54.4%) and trimethoprim-sulfamethoxazole (TMP-SMX) (40.4%). During the three months before admission, bacteria resistant to ampicillin were cultured from the urine of 75.6% of patients who were receiving some antibiotic, compared with 44% of children with no documented use of antibiotics (P<0.0001). Resistance to TMP-SMX in those with pre-existing genitourinary disease was 72.2% versus 31.8% in those without (P<0.0001). Patients who had previous admissions for urinary tract infections also showed greater resistance to TMP-SMX (70.6% versus 32.7%; P<0.005), cefazolin (64.7% versus 20.0%; P<0.0001) and nitrofurantoin (58.8% versus 18.2%; P<0.0001).

CONCLUSIONS:

There was a high resistance to ampicillin. Risk factors for resistant bacteria included the use of antibiotics three months before admission, previous genitourinary disease and previous admissions for urinary tract infections. In the presence of these risk factors, a third-generation cephalosporin as first-line antimicrobial therapy is recommended. However, the combination of ampicillin plus gentamicin can be considered for empirical therapy in low-risk patients.

Urinary tract infections (UTIs) are a common cause of fever in the paediatric population. UTIs are among the most common causes of serious bacterial infections in children in North America (1–3). The classic signs of serious UTIs, such as pyelonephritis present in older children, are difficult to discern in young children. Nonspecific signs such as fever, vomiting and abdominal pain can make the diagnosis of UTI challenging. In young children, fever without a source is frequently the admitting diagnosis. This is further confounded by the possibility of other sources of infection (eg, acute otitis media, upper respiratory tract infections and gastroenteritis) that do not exclude the presence of a UTI. Ineffectively treating these patients suspected of having a UTI puts them at risk for renal injury. Of particular concern is renal scarring leading to hypertension, renal insufficiency and end-stage renal disease (4–6). Thus, rapid and effective antibiotic treatment is necessary to avoid both immediate and long-term complications.

Antibiotic resistance has also been a concern in Canada and internationally (7–13). Escherichia coli, the most prevalent urinary tract pathogen, has shown high resistance to commonly used antibiotics, such as amoxicillin and trimethoprim-sulfamethoxazole (14–17). To effectively treat children with UTIs, knowledge of local bacterial epidemiology and their resistance patterns is important. Effective treatment includes well-chosen empirical antimicrobials. The purpose of the present study was to assess the epidemiology and resistance patterns of bacteria causing UTIs in children who were admitted to Kingston General Hospital (KGH) – the regional tertiary care hospital of southeastern Ontario.

METHODS

The present study was a retrospective chart review of patients one month to 18 years of age who were admitted to KGH between January 2002 and January 2006, and who had UTI as the primary discharge diagnosis. This included the terms – urinary tract infection, urosepsis, pyelonephritis and cystitis. Patients with other primary discharge diagnoses, and UTI as a secondary diagnosis were excluded. The study was approved by Queen’s University’s (Kingston, Ontario) ethics board.

Statistical analysis

SPSS (SPSS Inc, USA) was the primary statistics program used for analysis. Comparisons of categorical variables were performed with χ² tests and Fisher’s exact test. P≤0.05 was considered statistically significant.

RESULTS

Descriptive data

A total of 142 patient charts were reviewed, of which 56.3% were female and 43.7% were male. The median age of the patients was 12.3 months. There was a high proportion (62.7%) of patients younger than two years of age (Figure 1).

Figure 1).

Age distribution of patients. Age ranges of the categories are not equal

Most (52.8%) of the patients lived in Kingston. With the exception of two patients, the remainder lived in communities in southeastern Ontario. One patient lived in Hamilton, Ontario, and another lived in Vancouver, British Columbia. Twelve per cent of patients were referred from hospitals that did not admit children. The most common presenting symptom among children diagnosed with UTI was fever (90.8%), followed by nausea or vomiting (44.4%), and abdominal pain (23.2%). Urine cultures were reported for 89.4% of patients. The urine culture collection methods are shown in Figure 2.

Figure 2).

Urine culture collection method

Urinary pathogens and their resistance patterns

E coli was the predominant urinary pathogen, making up 71.6% of all urine cultures. Enterococcus species and Klebsiella species occurred with similar frequencies (5.7% and 5%). There were nine patients (6.4%) in whom urine culture yielded no organism. The basis for the diagnosis of UTI was clinical, in addition to pyuria, nitrates or blood on dipstick.

On admission, patients were most frequently treated with ampicillin (59.2%), gentamicin (53.5%) or ceftriaxone (25.4%) empirically. Ampicillin and gentamicin were usually used together and always administered intravenously. Five patients were treated with intramuscular ceftriaxone. Antibiotics were administered intravenously in 91.4% of patients, and oral antibiotics were administered in 5% of patients. The remaining 3.5% of patients received antibiotics intramuscularly. The resistance patterns of all the organisms are shown in Figure 3. Although several antibiotics were tested by the laboratory, including second- and third-generation cephalosporins, piperacillin, tobramycin, vancomycin, cloxacillin, fluoroquinolones and amoxicillin-clavulanate, not all sensitivities to each of these antibiotics were reported. Testing of sensitivities and the availability of such data depended on the organism cultured and resistance to the common antibiotics. The five most commonly reported antibiotics for paediatric UTIs are presented in Figures 3 to 6.

Figure 3).

Antibiotic resistance pattern of all cultured urinary organisms. TMP-SMX Trimethoprim-sulfamethoxazole

Figure 6).

Difference in resistance pattern with and without antibiotic use less than three months before admission. *P<0.05. TMP-SMX Trimethoprim-sulfamethoxazole

Of the 85 patients initially treated with ampicillin, including those treated in combination with gentamicin, 28 (32.9%) were discovered to be infected with an organism resistant to ampicillin. Of the 77 patients initially treated with gentamicin, four (5.2%) were resistant to gentamicin. The total number of patients empirically treated with both ampicillin and gentamicin was 73; 26 (35.6%) were resistant to ampicillin, three (4.1%) were resistant to gentamicin and only two (2.7%) were resistant to both ampicillin and gentamicin. During the study period, E coli was sensitive to greater than 95% of third-generation cephalosporins (ceftriaxone and ceftazidime) according to the KGH’s annual antibiogram. Antibiotics were changed to the appropriate antibiotics following culture results.

Possible risk factors for antibiotic resistance were also identified. These included being on prophylactic antibiotics at the time of admission, pre-existing genitourinary (GU) disease, previous admission for UTI and antibiotic use in the three months before admission. Pre-existing GU disease included any congenital GU malformation, renal dysplasia, renal hypoplasia, renal cystic diseases, hydronephrosis, ureteropelvic junction obstruction, urinary calculi, vesicoureteric reflux and neurological disease affecting the GU system (neurogenic bladder and non-neurogenic neurogenic bladder). Patients who used antibiotics less than three months before admission included those on prophylactic antibiotics at the time of admission and those started on antibiotics in the community. Figure 7 demonstrates the presence of these risk factors in the patients reviewed.

Figure 7).

Possible risk factors for antibiotic resistance. *P<0.05; ^†Use of antibiotics less than three months before admission. GU Genitourinary; UTI Urinary tract infection

Figure 4 shows a significant increase in pathogen resistance to trimethoprim-sulfamethoxazole, cefazolin, gentamicin and nitrofurantoin when a pre-existing GU disease was present. Resistance to ampicillin also increased, although not significantly, due to the high baseline resistance already present (51.4%). There was also a significant (P<0.05) increase in pathogen resistance when a patient had been admitted for a UTI in the past (Figure 5), and when antibiotics were used within three months of admission (Figure 6). The exception to the latter was gentamicin, which has no oral counterpart, and was used only if the patient had been admitted to hospital in the three months before admission for UTI.

Figure 4).

Difference in resistance pattern with and without pre-existing genitourinary (GU) disease. *P<0.05. TMP-SMX Trimethoprim-sulfamethoxazole

Figure 5).

Difference in resistance pattern with and without previous admissions for urinary tract infection (UTI). * P<0.05. TMP-SMX Trimethoprim-sulfamethoxazole

DISCUSSION

There are many choices of empirical antibiotics for the treatment of UTIs. Depending on the local flora, there have been several recommendations published in different parts of the world. Haller et al (14) performed a retrospective study of paediatric inpatients who received empirical intravenous therapy at the University Hospital of Freiburg (Germany). They concluded that ampicillin and an aminoglycoside were appropriate empirical treatments for UTIs. Prais et al (16) examined both inpatient and outpatients for urinary pathogens, and resistance patterns in tertiary paediatric centres in Israel. They concluded that effective empirical oral antibiotics for the treatment of community-acquired UTIs include amoxicillin-clavulanate or cefuroxime. Hoberman et al (18) recommended cefixime as a first-line empirical antibiotic for outpatient therapy of community-acquired UTIs.

The present analysis of the distribution of bacteria causing UTIs and their resistance patterns in the Kingston area are similar to many areas in North America and internationally (14–17,19,20). Our findings that E coli was the predominant organism was consistent with other reviews (14–17,21). Urinary pathogens affecting children admitted to our regional referral centre were highly resistant to ampicillin (54.4%). Centres in Canada and internationally have demonstrated resistance to ampicillin ranging from 45% to 73.3% (14,15,17,19,20). However, due to the low resistance of pathogens to gentamicin (22.1%), most patients were treated until the resistance profile of their infective organism was known. In the subanalysis of patients who were treated initially with ampicillin and gentamicin, there was again a high resistance to ampicillin (35.6%) and a low resistance to gentamicin (4.1%). Two patients were resistant to both empirical antibiotics.

No changes to the resistance profile of the antibiotics tested were seen over the four-year period included in the present study. However, increasing resistance has been demonstrated in other studies performed over five years (15) and nine years (16).

Risk factors for resistant bacteria in Kingston and the surrounding areas included the use of antibiotics within three months before admission, GU abnormalities and previous admissions for UTIs. This is echoed in the current literature (15,17), although there is variation in the amount of risk each of these factors pose. Additional risk factors have been identified, including children admitted to hospital one or more times in the previous year (17), children who received antibiotics for more than four weeks in the preceding six months and children two to six years of age (19). Our study did not demonstrate a significant increase in bacterial resistance in the presence of prophylactic antibiotic use. This may be due to the smaller sample size of the present study, because another Canadian study (17) demonstrated antibiotic prophylaxis to be a risk factor for resistant UTIs.

CONCLUSIONS

Based on the results of the present study, in the presence of antibiotic use in the previous three months, GU disease and previous admissions for UTIs, a third-generation cephalosporin is recommended as a first-line antimicrobial for the treatment of paediatric UTIs requiring admission to hospital in the Kingston area. Ampicillin plus gentamicin can be considered for empirical parenteral therapy for low-risk patients, as long as a urine culture is performed and appropriate adjustments are made when organisms and antibiotic sensitivities are identified. It is important for physicians to be familiar with the antibiotic resistance profiles within their community.