Abstract
Limited clinical information is available regarding community onset infections caused by extended-spectrum β-lactamase (ESBL)-producing Escherichia coli. A case-control study was performed to evaluate the epidemiology and risk factors of these types of infections. A case patient was defined as a person whose clinical sample yielded ESBL-producing E. coli. For each case patient, one control was randomly chosen from a group of outpatients from whom non-ESBL-producing E. coli had been isolated and for whom a clinical sample had been sent to the same laboratory for culturing during the following week. Of 108 cases of ESBL-producing E. coli, 56 (51.9%) were classified as health care associated (HCA). Univariate analysis showed male gender, HCA infection, severe underlying illness, and a prior receipt of antibiotics to be associated with ESBL-producing E. coli. In the multivariate analysis, HCA infection (odds ratio [OR], 3.18; 95% confidence interval [CI], 1.67 to 6.06; P < 0.001) and previous use of antibiotics (OR, 4.88; 95% CI, 2.08 to 11.48; P < 0.001) were found to be significantly associated with the ESBL group. In a multivariate analysis that included each antibiotic, previous use of fluoroquinolone (OR, 7.32; 95% CI, 1.58 to 34.01; P = 0.011) was significantly associated with ESBL-producing E. coli. Of 101 isolates in which ESBLs and their molecular relationships were studied, all isolates produced ESBLs from the CTX-M family (CTX-M-14, 40 isolates; CTX-M-15, 39 isolates; and other members of the CTX-M family, 22 isolates). In conclusion, this study confirms that ESBL-producing E. coli strains are a notable cause of community onset infections in predisposed patients. HCA infection and previous use of fluoroquinolone were significant factors associated with ESBL-producing E. coli in community onset infections.
INTRODUCTION
During the past 2 decades, antimicrobial-resistant strains that produce extended-spectrum β-lactamases (ESBLs) have emerged among Enterobacteriaceae, predominantly among Escherichia coli and Klebsiella pneumoniae (5, 20, 21). Subsequently, ESBL-producing E. coli has gained worldwide recognition as a significant group of community-acquired pathogens (3, 22). Although most patients with community onset infections caused by these organisms have urinary tract infections (UTIs), some patients present with intra-abdominal infections and concomitant bacteremia (12, 23). Nevertheless, there is still limited information about the clinical epidemiology of community onset ESBL-producing E. coli infections. Previous studies describing the clinical and microbiological epidemiology of ESBL-producing E. coli infections have focused mainly on UTIs (2, 4, 16, 22).
Because ESBL-producing organisms are frequently resistant to multiple antimicrobial agents, therapeutic options for these infections are limited (18). At present, carbapenems are recommended for the treatment of serious infections caused by ESBL-producing organisms. Given the increasing incidence and pathogenic significance of ESBL-producing E. coli in community onset infections, it is necessary to gather additional data on clinical risk factors for ESBL-producing E. coli in community onset infections. This knowledge is essential for selection of the most appropriate empirical antimicrobial therapy for these infections, and it is important to identify the risk factors for infections caused by ESBL-producing E. coli so that effective infection management strategies can be developed. Thus, this study was performed to evaluate the clinical features and epidemiology of community onset ESBL-producing E. coli infections and to identify risk factors for infections.
MATERIALS AND METHODS
Study design and population.
A case-control study was performed to evaluate risk factors associated with ESBL-producing E. coli in community onset infections. The medical records of individuals diagnosed with E. coli infection from September 2010 to May 2011 were reviewed. Samples were collected from Samsung Medical Center, a 1,950-bed tertiary care university hospital in Seoul, Republic of Korea, and Samsung Changwon Hospital, a 700-bed community-based university-affiliated hospital in Changwon, Republic of Korea. Patients were included in the study if their cultures were drawn in the Emergency Department or an outpatient clinic and if the culture results were positive for E. coli. A case patient was defined as a person whose clinical sample yielded ESBL-producing E. coli. For each case patient, one control was randomly chosen from a group of outpatients from whom non-ESBL-producing E. coli had been isolated from a clinical sample that had been sent to the same laboratory for culturing during the following week. The controls were matched according to the culture specimen and acquisition unit. All patients older than 15 years were enrolled, and only the first infection episode of each patient was included in the analysis. The clinical significance of ESBL-producing E. coli was investigated in the case patients, and those with asymptomatic colonization of ESBL-producing E. coli were excluded.
Data collected included age, gender, underlying disease, site of infection, severity of underlying diseases as classified by McCabe and Jackson criteria (19), and antimicrobial regimen. The presence of the following comorbid conditions was documented: neutropenia, a recent surgical procedure (within 3 months), corticosteroid use within 1 month (≥20-mg daily dose taken for ≥2 weeks), immunosuppressant use within 1 month, and the presence of an indwelling urinary catheter or percutaneous tube. Because the study was observational, the attending physician decided on indications for cultures, other tests, and therapy. This study was approved by the Institutional Review Board of Samsung Medical Center, Seoul, South Korea. Written informed consent was not required because of the observational character of the study.
Definition.
The sites of infection were determined by the patients' physicians on the basis of clinical evaluation and the isolation of E. coli from the presumed portal of entry (7, 12). Community onset infection was defined as an infection diagnosed within the first 48 h of hospitalization. Because many cases of infections that are present or incubating upon admission to the hospital are nonetheless health care associated (HCA), nonnosocomial bacteremia is referred to as community onset rather than community-acquired bacteremia. Episodes of community onset bacteremia were further classified as HCA if any of the following criteria were present: a history of a >48-h hospital admission in the previous 90 days, hemodialysis, intravenous medication, home wound care in the previous 30 days, or residence in a nursing home or long-term care facility (6, 12). Cases that did not meet these criteria were considered to be community associated (CA). Neutropenia was defined as an absolute neutrophil count of <500 neutrophils/mm3.
Microbiological analysis.
Of the stored bacterial isolates collected by the clinical microbiology laboratory in the study hospitals, 101 isolates of ESBL-producing E. coli were successfully recovered for inclusion in the study. Species identification was performed using Vitek II (bioMérieux, Hazelwood, MO), a standard identification card. Antimicrobial susceptibility testing and ESBL confirmatory testing were performed using the broth microdilution method or the disk diffusion method according to the recommendations of the Clinical and Laboratory Standards Institute. ESBL-related genes, such as TEM, SHV, CTX-M, and OXA, were amplified by PCR from clinical isolates as described in previous studies (13, 14). MIC values of the antibiotics tested were determined using the broth microdilution method. Quality control was performed using strains E. coli 25922 and Pseudomonas aeruginosa 27853.
Statistical analysis.
Student's t test was used to compare continuous variables, and the χ2 or Fisher's exact test was used to compare categorical variables. A stepwise conditional logistic regression analysis was used to control for the effects of confounding variables and to identify the independent risk factors of infection. All risk factors with a P value of <0.1 at the bivariate level were included in the multivariate logistic model predicting ESBL-producing E. coli. All variables for which P was <0.05 in the multivariate analysis were retained in the final model. Interactions between variables were not introduced into the models. Odds ratios (ORs) and their 95% confidential intervals (CIs) were calculated. All P values were two tailed, and a P of <0.05 was considered statistically significant.
RESULTS
Clinical features and risk factors for community onset infection caused by ESBL-producing E. coli.
During the study period, ESBL-producing E. coli was isolated from 108 outpatients, all of whom were included in the study. ESBL-producing E. coli was isolated from urine culture in 55 cases (50.9%), from blood culture in 45 cases (41.7%), from pus in 4 cases (3.7%), from ascites fluid in 2 cases (1.9%), and from other fluids in 2 cases (1.9%). The clinical characteristics of the case patients are shown in Table 1. The mean age (± standard deviation) was 62.4 ± 16.0 years, and 64.8% of the case patients were female. Of 108 patients with ESBL-producing E. coli infections, 56 (51.9%) were classified as HCA. The most common underlying diseases were diabetes mellitus (n = 29; 26.9%), followed by a solid tumor (n = 24; 22.2%), renal disease (n = 14; 13.0%), and liver disease (n = 13; 12.0%). The most common types of infection were UTIs (n = 73; 67.6%), followed by biliary tract infections (n = 12; 11.1%) and intra-abdominal infections (n = 10; 9.3%).
Table 1.
Clinical features and predisposing factors for community onset infections caused by ESBL-producing E. coli
| Parametera | No. (%) of patients with: |
P | |
|---|---|---|---|
| ESBL-producing E. coli infection (n = 108) | Non-ESBL-producing E. coli infection (n = 108) | ||
| Female | 70 (64.8) | 86 (79.6) | 0.015 |
| Old age (≥65 yr) | 56 (51.9) | 49 (45.4) | 0.376 |
| Healthcare-associated infection | 56 (51.9) | 21 (19.4) | <0.001 |
| Underlying disease | |||
| Diabetes mellitus | 28 (25.9) | 24 (22.2) | 0.524 |
| Liver disease | 13 (12.0) | 10 (9.3) | 0.508 |
| Renal disease | 14 (13.0) | 15 (13.9) | 0.842 |
| Solid tumor | 24 (22.2) | 20 (18.5) | 0.499 |
| Severity of underlying diseases by McCabe and Jackson criteria | |||
| Nonfatal | 76 (70.4) | 93 (86.1) | 0.014 |
| Ultimately fatal | 27 (25.0) | 14 (13.0) | |
| Rapidly fatal | 5 (4.6) | 1 (0.9) | |
| Comorbid condition | |||
| Recent operation | 6 (5.6) | 3 (2.8) | 0.307 |
| Corticosteroid use | 2 (1.9) | 0 (0) | 0.155 |
| Immunosuppressant use | 3 (2.8) | 3 (2.8) | 1.000 |
| Central venous catheterization | 6 (5.6) | 3 (2.8) | 0.498 |
| Indwelling urinary catheter | 8 (7.4) | 5 (4.6) | 0.391 |
| Percutaneous tube | 11 (10.2) | 5 (4.6) | 0.161 |
| Previous use of antibiotics | 39 (36.1) | 8 (7.4) | <0.001 |
| Cephalosporin | 18 (16.7) | 4 (3.7) | 0.002 |
| Penicillin derivative | 8 (7.4) | 1 (0.9) | 0.017 |
| Fluoroquinolone | 20 (18.5) | 2 (1.9) | <0.001 |
| Severity of infection | |||
| Sepsis | 23 (21.3) | 22 (22.4) | 0.953 |
| Severe sepsis | 5 (4.6) | 5 (4.6) | |
| Septic shock | 9 (8.3) | 7 (6.5) | |
Male, HCA infection, severity of underlying diseases by McCabe and Jackson criteria, and prior use of antibiotics (all P were <0.1) were included in the multivariate logistic model predicting ESBL-producing E. coli.
One hundred eight case patients were compared with 108 control patients to analyze the risks factors associated with ESBL-producing E. coli. Crude analysis showed male gender, HCA infection, severe underlying illness, and a prior receipt of antibiotics to be associated with ESBL-producing E. coli (Table 1). Two multivariate models were performed. In the first model, a variable called “previous use of antibiotics” was included, but the individual components (e.g., specific antimicrobials) of the antibiotics were not considered. The second model included the same variables as the first group, except that specific types of antimicrobials (cephalosporins, a penicillin derivative, and fluoroquinolones) replaced the “previous use of antibiotic” variable. In the multivariate analysis, HCA infection (OR, 3.18; 95% CI, 1.67 to 6.06; P < 0.001) and previous use of antibiotics (OR, 4.88; 95% CI, 2.08 to 11.48; P < 0.001) were found to be significantly associated with the ESBL group. In the second model, previous use of fluoroquinolone (OR, 7.32; 95% CI, 1.58 to 34.01; P = 0.011) and the presence of HCA infections (OR, 2.86; 95% CI, 1.48 to 5.55; P = 0.002) were significantly associated with the ESBL group. The overall 30-day mortality rate was 4.8% (10/208); that in the ESBL group was 7.4% (8/108), and that in the non-ESBL group was 2.0% (2/100) (P = 0.103).
A subgroup analysis that included 52 patients with CA infection and excluded those with HCA infections was performed. ESBL-producing E. coli was isolated from urine culture in 32 cases (61.5%), from blood culture in 16 cases (30.8%), from pus in two cases (3.8%), and from other fluids in two cases (3.8%). The most common types of infection were UTIs (n = 38; 73.1%), followed by biliary tract infections (n = 5; 9.6%) and intra-abdominal infections (n = 3; 5.8%). Fifty-two cases with ESBL-producing E. coli were compared to 87 control cases with non-ESBL-producing E. coli, and the comparison of clinical characteristics is shown in Table 2. Previous use of cephalosporins (OR, 6.61; 95% CI, 1.32 to 33.16; P = 0.022) was the only independent factor associated with ESBL-producing E. coli in CA infections.
Table 2.
Clinical factors associated with community-associated infections caused by ESBL-producing E. coli
| Parametera | No. (%) of patients with: |
P | |
|---|---|---|---|
| ESBL-producing E. coli (n = 52) | Non-ESBL-producing E. coli (n = 87) | ||
| Female | 41 (78.8) | 70 (80.5) | 0.818 |
| Old age (≥65 yr) | 21 (41.2) | 35 (40.2) | 0.913 |
| Underlying disease | |||
| Diabetes mellitus | 14 (26.9) | 17 (19.5) | 0.312 |
| Liver disease | 6 (11.5) | 7 (8.0) | 0.494 |
| Renal disease | 3 (5.8) | 10 (11.5) | 0.262 |
| Solid tumor | 6 (11.5) | 16 (18.4) | 0.284 |
| Severity of underlying diseases by McCabe and Jackson criteria | |||
| Nonfatal | 44 (84.6) | 75 (86.2) | 0.796 |
| Ultimately fatal | 8 (15.4) | 12 (13.8) | |
| Comorbid condition | |||
| Recent operation | 1 (1.9) | 2 (2.3) | 1.000 |
| Immunosuppressant use | 2 (3.8) | 3 (3.4) | 1.000 |
| Central venous catheterization | 4 (7.7) | 1 (1.1) | 0.045 |
| Indwelling urinary catheter | 1 (1.9) | 4 (4.6) | 0.413 |
| Percutaneous tube | 4 (7.7) | 2 (2.3) | 0.196 |
| Previous use of antibiotics | 9 (17.3) | 6 (6.9) | 0.056 |
| Cephalosporin | 7 (13.5) | 2 (2.3) | 0.014 |
| Penicillin derivative | 3 (5.8) | 1 (1.1) | 0.147 |
| Fluoroquinolone | 1 (1.9) | 2 (2.3) | 0.883 |
| Severity of infection | |||
| Sepsis | 7 (13.5) | 18 (20.7) | 0.636 |
| Severe sepsis | 2 (3.8) | 2 (2.3) | |
| Septic shock | 2 (3.8) | 5 (5.7) | |
Central venous catheterization and prior use of cephalosporins (P < 0.1) were included in the multivariate logistic model predicting ESBL-producing E. coli.
Clinical features and risk factors for community onset UTIs by ESBL-producing E. coli.
Seventy-three cases with ESBL-producing E. coli were compared to 67 controls with non-ESBL-producing E. coli to identify risk factors of ESBL-producing E. coli in community onset UTIs. ESBL-producing E. coli was isolated from urine culture in 55 cases (75.3%) and from blood culture in 18 cases (24.7%). Univariate analysis showed that health care-associated infection and previous use of antibiotics (e.g., cephalosporin and fluoroquinolone) were significantly associated with the ESBL group (Table 3). HCA infection (OR, 4.80; 95% CI, 1.91 to 12.05; P = 0.001) and previous use of fluoroquinolone (OR, 4.27; 95% CI, 0.85 to 21.41; P = 0.077) were found to be independent factors associated with ESBL-producing E. coli. However, in a subgroup analysis, including 38 patients with CA UTIs and excluding patients with HCA UTIs, no identified risk factor for ESBL-producing E. coli was found, although the number of patients who had previously received cephalosporins was slightly higher in the ESBL group than in the non-ESBL group (10.5% [4/38] versus 1.7% [1/59]; P = 0.075).
Table 3.
Clinical features and predisposing factors for community onset urinary tract infections caused by ESBL-producing E. coli
| Parametera | No. (%) of patients with: |
P | |
|---|---|---|---|
| ESBL-producing E. coli infection (n = 73) | Non-ESBL-producing E. coli infection (n = 67) | ||
| Female | 56 (76.7) | 55 (82.1) | 0.433 |
| Old age (≥65 yr) | 39 (54.2) | 27 (40.3) | 0.102 |
| Healthcare-associated infection | 35 (47.9) | 8 (11.9) | <0.001 |
| Underlying disease | |||
| Diabetes mellitus | 16 (21.9) | 16 (23.9) | 0.782 |
| Liver disease | 6 (8.2) | 4 (6.0) | 0.747 |
| Renal disease | 9 (12.3) | 10 (14.9) | 0.654 |
| Solid tumor | 12 (16.4) | 11 (16.4) | 0.997 |
| Comorbid condition | |||
| Recent operation | 3 (4.1) | 2 (3.0) | 1.000 |
| Corticosteroid use | 1 (1.4) | 0 (0) | 1.000 |
| Immunosuppressant use | 2 (2.7) | 3 (4.5) | 0.670 |
| Central venous catheterization | 2 (2.7) | 1 (1.5) | 1.000 |
| Indwelling urinary catheter | 4 (5.5) | 3 (4.5) | 1.000 |
| Percutaneous tube | 4 (5.5) | 1 (1.5) | 0.368 |
| Previous use of antibiotics | 26 (35.6) | 6 (9.0) | <0.001 |
| Cephalosporin | 10 (13.7) | 2 (3.0) | 0.024 |
| Penicillin derivative | 3 (4.1) | 1 (1.5) | 0.621 |
| Fluoroquinolone | 17 (23.3) | 2 (3.0) | <0.001 |
HCA infection and prior use of antibiotics (P < 0.1) were included in the multivariate logistic model predicting ESBL-producing E. coli.
Microbiological characteristics of ESBL-producing E. coli causing community onset infections.
The results of antimicrobial susceptibility testing, reported as rates of resistance to ESBL-producing E. coli isolates causing community onset infection, are as follows: 63% (68/108) for amoxicillin-clavulanic acid, 31.5% (34/108) for piperacillin-tazobactam, 0.9% (1/108) for amikacin, 52.8% (57/108) for gentamicin, 52.8% (57/108) for trimethoprim-sulfamethoxazole, 79.6% (86/108) for levofloxacin, and 0% (0/108) for imipenem (Table 4).
Table 4.
Rates of resistance to antimicrobials of ESBL-producing E. coli isolates causing community-onset infectionsa
| Antibiotic | % of patients (no. taking drug/total) |
P | ||
|---|---|---|---|---|
| Overall | With HCA infections | With CA infections | ||
| Amoxicillin-clavulanic acid | 63 (68/108) | 69.6 (39/56) | 55.8 (29/52) | 0.136 |
| Piperacillin-tazobactam | 31.5 (34/108) | 35.7 (20/56) | 26.9 (14/52) | 0.326 |
| Amikacin | 0.9 (1/108) | 1.8 (1/56) | 0 (0/52) | 1.000 |
| Gentamicin | 52.8 (57/108) | 55.4 (31/56) | 50.0 (26/52) | 0.577 |
| Tobramycin | 58.3 (63/108) | 66.1 (37/56) | 50.0 (26/52) | 0.091 |
| Trimethoprim-sulfamethoxazole | 52.8 (57/108) | 51.8 (29/56) | 53.8 (28/52) | 0.830 |
| Levofloxacin | 79.6 (86/108) | 87.5 (49/56) | 71.2 (37/52) | 0.035 |
According to automatic susceptibility testing.
A total of 101 isolates of ESBL-producing E. coli were included for further microbiological study of ESBL characterization and for an antimicrobial susceptibility test (Table 5). Of 101 isolates in which ESBLs and their molecular relationships were studied, all isolates produced ESBLs from the CTX-M family (CTX-M-14, 40 isolates; CTX-M-15, 39 isolates; CTX-M-24, 2 isolates; CTX-M-27, 10 isolates; CTX-M-57, 6 isolates, and CTX-M-3, 1 isolate). Three isolates had ESBLs of both CTX-M-14 and CTX-M-15. TEM-1 was present in 50 isolates, but no TEM-type ESBLs were found. No SHV-type ESBLs were found in ESBL-producing E. coli isolates causing community onset bacteremia.
Table 5.
ESBL types and antimicrobial susceptibility results of ESBL-producing E. coli isolates causing community onset infectionsa
| Antibiotic | ESBL producers (n = 101) |
CTX-M-15 ESBL producers (n = 39) |
CTX-M-14 ESBL producers (n = 40) |
|||
|---|---|---|---|---|---|---|
| MIC50 (mg/liter) | MIC90 (mg/liter) | MIC50 (mg/liter) | MIC90 (mg/liter) | MIC50 (mg/liter) | MIC90 (mg/liter) | |
| Cefotaxime | >128 | >128 | >128 | >128 | >128 | >128 |
| Ceftazidime | 32 | >128 | 64 | >128 | 2 | 32 |
| Cefepime | 128 | >128 | >128 | >128 | 16 | >128 |
| Piperacillin-tazobactam | 4/4 | 64/4 | 16/4 | 128/4 | 2/4 | 32/4 |
| Ciprofloxacin | 64 | >128 | 128 | >128 | 16 | >128 |
| Imipenem | 0.12 | 0.12 | 0.12 | 0.12 | 0.12 | 0.25 |
| Meropenem | ≤0.03 | 0.06 | ≤0.03 | 0.06 | ≤0.03 | 0.06 |
| Ertapenem | ≤0.03 | 0.25 | 0.06 | 0.25 | ≤0.03 | 0.12 |
MIC50 and MIC90, MICs required to kill 50% and 90% of the organisms, respectively; ESBL, extended-spectrum β-lactamase.
DISCUSSION
The present study showed that ESBL-producing E. coli is a significant cause of community onset infection, particularly in UTIs and intra-abdominal infections. CTX-M enzymes were the predominant enzymes found in the community isolates. ESBL-producing E. coli has emerged in recent years as a cause of community-acquired infections worldwide (3, 12, 22, 23). Several studies attempted to identify risk factors for ESBL-producing E. coli among community onset infections (2, 4, 12, 16, 22). An association with health care and previous use of antimicrobials were found to be independently associated with ESBL-producing E. coli infection. This study provides a comprehensive analysis of community onset infection caused by ESBL-producing E. coli. The distribution by type of infection was as expected for E. coli, in that UTIs accounted for the majority of cases (67.6%). However, it is noteworthy that 41.7% of the cases were bacteremic and 20.4% had intra-abdominal infections, including biliary tract infections. The frequency of intra-abdominal infections in which ESBL-producing E. coli strains are involved is probably underestimated, since samples from patients with community-acquired secondary peritonitis are not routinely sent for culturing (22).
Even though HCA infections were associated with an increased risk of ESBL-producing E. coli, a high percentage of cases were exposed to no risk factors or to only one. Because rectal colonization with virulent strains of ESBL-producing E. coli is increasing among healthy individuals, it can be hypothesized that these infections may increasingly occur among patients with no specific risk factors (22–24).
In our two previous studies, the proportions of ESBL-producing isolates among cases of community onset E. coli bacteremia were 4.1% and 9.5%, respectively (11, 12). A recent study performed in a secondary hospital in the Republic of Korea showed that community onset bacteremia caused by ESBL-producing E. coli accounted for 6.7% of all community onset E. coli bacteremia (17). Determining the exact prevalence of these organisms in Korea would require a population-based study, but these findings suggest that ESBL producers have already begun to disseminate throughout the community. The responsible mechanisms are poorly understood, and further investigation to characterize these mechanisms is warranted. The investigation of risk factors and outcomes using a cohort design would have enabled prevalence and predictive rules to be developed, but this did not prove feasible because of the high number of patients with community onset infection due to E. coli (22, 23).
Numerous studies have demonstrated that most ESBL-producing Enterobacteriaceae are resistant to multiple antibiotic classes. Community onset ESBL-producing isolates from most regions displayed high rates of resistance to ciprofloxacin (>70%), which suggests that resistance to fluoroquinolones is prevalent among community onset ESBL-producing organisms. The rates of resistance to levofloxacin and trimethoprim-sulfamethoxazole among ESBL-producing E. coli isolates in this study were 79.6% and 52.8%, respectively, which is comparable to rates found in other studies (1–4, 22, 23). Such a high prevalence of resistance to these antimicrobial agents, particularly fluoroquinolones, is a well-known feature of ESBL producers and is of great concern. Although carbapenem is considered the drug of choice for serious infections, alternatives such as ciprofloxacin or aminoglycosides may be additional options. However, the susceptibility profile of the isolates in this study confirms that alternatives for such infections are considerably limited.
The worldwide dissemination of Gram-negative bacteria that produce ESBLs (specifically, CTX-M enzymes) has been described as an unfolding pandemic (21). This study confirmed that CTX-M-15 and CTX-M-14 were the most common (81.2%) ESBL types harbored by E. coli strains causing community onset infections in our region. Of all the risk factors found, only previous exposure to antimicrobial agents is potentially susceptible to intervention. Because fosfomycin and nitrofurantoin are frequently active against ESBL-producing isolates (15), it should be determined whether therapy (or prophylaxis) of recurrent UTIs with these agents is helpful in avoiding the selection pressure posed by fluoroquinolones, aminopenicillins, and cephalosporins (22, 23). For community onset UTIs, the most important consequence of such high resistance rates is the difficulty in the management of UTI patients. The number of drugs available for use in the outpatient setting is limited, and nitrofurantoin and fosfomycin seem to be the only choices in some cases (15). Amoxicillin-clavulanate was used with success for patients with cystitis due to ESBL-producing E. coli in a recent study (22); however, as 63% of ESBL-positive isolates in the present study were found to be resistant, amoxicillin-clavulanate may not always be a good choice.
This study has several limitations. First, clinical data were retrospectively collected though electronic medical records and chart review, although cases and bacterial isolates were enrolled prospectively. The present study was observational, and thus unknown risk factors might have been unequally distributed between the two groups. The possibility of limitations that preclude accurate comparisons should be kept in mind given the observational nature of this study. Second, although this study used a control group consisting of patients with acquisition of non-ESBL-producing E. coli, such a design may overestimate the importance of previous antimicrobial use because patients who had received antimicrobials may be underrepresented in the control population (9). The selection bias introduced by the use of control patients with antimicrobial-susceptible organisms is likely to have the largest impact on antibiotics that are active against susceptible forms of the organism but are not active against the resistant form of the organism (9). However, our research question addressed in this study is “what are the risk factors for infections due to ESBL producers among patients with infection or colonization caused by E. coli other than the risk factors for ESBL positivity among hospitalized patients in general?” (8, 10). Third, although this study was performed at two large tertiary care hospitals, a relatively small number of patients were included, resulting in limited statistical power, which may have created instability in the multivariable logistic regression model. Finally, this study was conducted mainly at large referral centers. Thus, many of the study patients had serious underlying illnesses, and our findings may not be generalizable to other institutions, particularly community hospitals.
In conclusion, this study confirms that ESBL-producing E. coli strains are a notable cause of community onset infections in predisposed patients. HCA infection and previous use of cephalosporins or fluoroquinolone were significant factors associated with ESBL-producing E. coli in community onset infections. The widespread and rapid dissemination of ESBL-producing E. coli seems to be an emerging issue worldwide, and ESBL-producing E. coli is a pathogen that is increasingly found in the community and that may drive significant changes in the empirical use of antibiotics for certain infections. Further clinical studies are needed to guide clinicians in the management of community onset infections caused by E. coli.
ACKNOWLEDGMENTS
This study was supported by the Basic Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Education, Science and Technology (grant 2010-0021572).
Bacterial isolates were obtained from the Asian Bacterial Bank (ABB) of the Asia Pacific Foundation for Infectious Diseases (APFID).
No competing financial interest exists.
Footnotes
Published ahead of print 7 December 2011
REFERENCES
- 1. Arpin C, et al. 2003. Extended-spectrum beta-lactamase-producing Enterobacteriaceae in community and private health care centers. Antimicrob. Agents Chemother. 47: 3506–3514 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Azap OK, et al. 2010. Risk factors for extended-spectrum beta-lactamase positivity in uropathogenic Escherichia coli isolated from community-acquired urinary tract infections. Clin. Microbiol. Infect. 16: 147–151 [DOI] [PubMed] [Google Scholar]
- 3. Ben-Ami R, et al. 2009. A multinational survey of risk factors for infection with extended-spectrum beta-lactamase-producing Enterobacteriaceae in nonhospitalized patients. Clin. Infect. Dis. 49: 682–690 [DOI] [PubMed] [Google Scholar]
- 4. Calbo E, et al. 2006. Risk factors for community-onset urinary tract infections due to Escherichia coli harbouring extended-spectrum beta-lactamases. J. Antimicrob. Chemother. 57: 780–783 [DOI] [PubMed] [Google Scholar]
- 5. Cheong HS, et al. 2010. Prevalence of extended-spectrum beta-lactamase among Enterobacteriaceae blood isolates with inducible AmpC beta-lactamase. Infect. Chemother. 42: 280–284 [Google Scholar]
- 6. Friedman ND, et al. 2002. Health care-associated bloodstream infections in adults: a reason to change the accepted definition of community-acquired infections. Ann. Intern. Med. 137: 791–797 [DOI] [PubMed] [Google Scholar]
- 7. Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. 1988. CDC definitions for nosocomial infections, 1988. Am. J. Infect. Control 16: 128–140 [DOI] [PubMed] [Google Scholar]
- 8. Harris AD, Karchmer TB, Carmeli Y, Samore MH. 2001. Methodological principles of case-control studies that analyzed risk factors for antibiotic resistance: a systematic review. Clin. Infect. Dis. 32: 1055–1061 [DOI] [PubMed] [Google Scholar]
- 9. Harris AD, et al. 2002. Control-group selection importance in studies of antimicrobial resistance: examples applied to Pseudomonas aeruginosa, enterococci, and Escherichia coli. Clin. Infect. Dis. 34: 1558–1563 [DOI] [PubMed] [Google Scholar]
- 10. Hyle EP, et al. 2005. Risk factors for increasing multidrug resistance among extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella species. Clin. Infect. Dis. 40: 1317–1324 [DOI] [PubMed] [Google Scholar]
- 11. Kang CI, et al. 2008. Clinical features and outcome of community-onset bloodstream infections caused by extended-spectrum beta-lactamase-producing Escherichia coli. Eur. J. Clin. Microbiol. Infect. Dis. 27: 85–88 [DOI] [PubMed] [Google Scholar]
- 12. Kang CI, et al. 2010. Risk factors and treatment outcomes of community-onset bacteraemia caused by extended-spectrum beta-lactamase-producing Escherichia coli. Int. J. Antimicrob. Agents 36: 284–287 [DOI] [PubMed] [Google Scholar]
- 13. Kim J, Lim YM, Jeong YS, Seol SY. 2005. Occurrence of CTX-M-3, CTX-M-15, CTX-M-14, and CTX-M-9 extended-spectrum beta-lactamases in Enterobacteriaceae clinical isolates in Korea. Antimicrob. Agents Chemother. 49: 1572–1575 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Kim J, et al. 2005. CTX-M and SHV-12 beta-lactamases are the most common extended-spectrum enzymes in clinical isolates of Escherichia coli and Klebsiella pneumoniae collected from 3 university hospitals within Korea. FEMS Microbiol. Lett. 245: 93–98 [DOI] [PubMed] [Google Scholar]
- 15. Ko KS, et al. 2007. In vitro activity of fosfomycin against ciprofloxacin-resistant or extended-spectrum beta-lactamase-producing Escherichia coli isolated from urine and blood. Diagn. Microbiol. Infect. Dis. 58: 111–115 [DOI] [PubMed] [Google Scholar]
- 16. Meier S, Weber R, Zbinden R, Ruef C, Hasse B. 2011. Extended-spectrum beta-lactamase-producing Gram-negative pathogens in community-acquired urinary tract infections: an increasing challenge for antimicrobial therapy. Infection 39: 333–340 [DOI] [PubMed] [Google Scholar]
- 17. Park SH, et al. 2011. Emergence of extended-spectrum beta-lactamase-producing Escherichia coli as a cause of community-onset bacteremia in South Korea: risk factors and clinical outcomes. Microb. Drug Resist. 17: 537–544 [DOI] [PubMed] [Google Scholar]
- 18. Paterson DL, Bonomo RA. 2005. Extended-spectrum beta-lactamases: a clinical update. Clin. Microbiol. Rev. 18: 657–686 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Perl TM, Dvorak L, Hwang T, Wenzel RP. 1995. Long-term survival and function after suspected gram-negative sepsis. JAMA 274: 338–345 [PubMed] [Google Scholar]
- 20. Pitout JD. 2010. The latest threat in the war on antimicrobial resistance. Lancet Infect. Dis. 10: 578–579 [DOI] [PubMed] [Google Scholar]
- 21. Pitout JD, Laupland KB. 2008. Extended-spectrum beta-lactamase-producing Enterobacteriaceae: an emerging public-health concern. Lancet Infect. Dis. 8: 159–166 [DOI] [PubMed] [Google Scholar]
- 22. Rodriguez-Bano J, et al. 2008. Community infections caused by extended-spectrum beta-lactamase-producing Escherichia coli. Arch. Intern. Med. 168: 1897–1902 [DOI] [PubMed] [Google Scholar]
- 23. Rodriguez-Bano J, et al. 2010. Community-onset bacteremia due to extended-spectrum beta-lactamase-producing Escherichia coli: risk factors and prognosis. Clin. Infect. Dis. 50: 40–48 [DOI] [PubMed] [Google Scholar]
- 24. Valverde A, et al. 2004. Dramatic increase in prevalence of fecal carriage of extended-spectrum beta-lactamase-producing Enterobacteriaceae during nonoutbreak situations in Spain. J. Clin. Microbiol. 42: 4769–4775 [DOI] [PMC free article] [PubMed] [Google Scholar]