Clinical Significance of Infections Caused by Extended-Spectrum β-Lactamase-Producing Enterobacteriaceae Blood Isolates with Inducible AmpC β-Lactamase

Abstract

Few studies have investigated the clinical features and outcomes for extended-spectrum β-lactamase (ESBL)-producing Enterobacter spp., Citrobacter spp., Serratia spp., and Morganella morganii (ECSM) bloodstream infections. This study was performed to investigate the clinical features and outcomes for ESBL-producing ECSM bloodstream infections. Patients with ECSM bloodstream infection were enrolled from October 2006 to March 2008. Of 124 patients with ECSM bacteremia, 30 cases (24.2%) were ESBL-producing isolates. Immunosuppressive drugs use within 30 days (p=0.028), indwelling device at the time of bacteremia (p=0.042) and antibiotics use within 3 months (p=0.022) were independently associated with ESBL production in multivariate analysis. Overall 30-day mortality rate was 19.4% (24/124). When the 30-day mortality rate was evaluated, no significant difference was found between the ESBL group (16.6%; 5/30) and non-ESBL group (20.2%; 19/94). Hospitalization was longer in the ESBL group than in the non-ESBL group (65.4±92.8 vs. 32.9±37.8 days, respectively; p=0.007). The recent use of antibiotics (especially broad-spectrum cephalosporins and other β-lactam antibiotics) was an important risk factor for ESBL among ECSM bacteremia. ESBL production of ECSM isolates was not significantly associated with mortality but ESBL-producing organisms have an important impact on the duration of hospital stay and subsequent medical cost.

Introduction

E nterobacter spp., Citrobacter spp., Serratia spp., and Morganella morganii (ECSM) are significant causes of nosocomial infections and are characterized by chromosomally encoded AmpC β-lactamase.⁴ β-lactam exposure is capable of inducing expression of AmpC β-lactamase in ECSM with consequent resistance to third-generation cephalosporins; the mutations may result in permanent hyperproduction and persistent resistance.¹⁸ Recent studies have suggested that extended-spectrum β-lactamase (ESBL) producers have become prevalent in ECSM species (10%–43%).^4,11,16,17 The importance of ESBL-producing ECSM has increased over the last decade because the therapeutic options currently available for these organisms are severely limited.^{2,4,10,11,16,17} However, little data are available regarding clinical significance and clinical outcome of patients with ECSM bacteremia during this era of increasing rates of ESBL-producing ECSMs. Furthermore, most previous studies regarding ECSM have involved all clinical isolates, including colonized pathogens. A single study for ECSM blood isolates involved only children.¹⁶ No previously published study has shown the clinical significance and outcome of patients, including adults, with ECSM bacteremia.

The present study was undertaken to address this clinical shortcoming. We conducted this study to identify the risk factors of ESBL and mortality; thus, we evaluated the effect of ESBL-producing ECSMs on outcome.

Materials and Methods

Study design and data collection

We reviewed the medical records of individuals found to have ECSM bacteremia from October 2006 to March 2008 at Samsung Medical Center, Seoul, Republic of Korea, a 1,950-bed tertiary-care university hospital. Patients were included in the study if their blood culture results were positive for ECSM. Only the first bacteremic episode for each patient was included in this study. The data collected included age, gender, underlying disease (as calculated by Charlson's weighted index of morbidity), site of infection, severity of illness (as calculated by Pitt bacteremia score),^5,14 and antimicrobial regimen. The presence of the following comorbid conditions was also documented: neutropenia, presentation with septic shock, acute renal failure, receipt of immunosuppressive agents within 30 days before presentation with bacteremia, admission location at the time of bacteremia, and indwelling device at the time of bacteremia.

Definitions

ECSM bacteremia was defined as growth of ECSM in a blood culture specimen. Nosocomial infection was defined as an infection that occurred >72 hr after admission to the hospital. Neutropenia was defined as an absolute neutrophil count below 500/mm.³ Septic shock was defined as sepsis with evidence of hypotension (arterial blood pressure <90 mmHg, or 40 mmHg less than patient's normal blood pressure) for at least 1 hr despite adequate fluid resuscitation, or need for vasopressor to maintain systolic blood pressure ≥90 mmHg or mean arterial pressure ≥70 mmHg.⁸ The initial empirical antimicrobial therapy was considered appropriate if the initial antibiotics, which were administered within 24 hr of acquisition of blood cultures, included at least one antibiotic that was active in vitro against the isolated pathogen and if the dosage and route of administration conformed to current medical standards. We considered antimicrobial therapy to be inappropriate if the drugs used did not have in vitro activity against the isolates or if the patient did not receive antimicrobial therapy.

Microbiological tests

The bacterial isolates were obtained from Asian Bacterial Bank of the Asia Pacific Foundation for Infectious Disease. The recovery of isolates from blood culture was performed with the BACTEC model 9240 system (BD Diagnostic Instrument Systems, Franklein Lakes, NJ) or the BacT/ALERT 3D system (bioMerieux, Marcy l'Etoile, France). Isolates were identified using VITEK II (bioMerieux, Hazelwood, MO), a standard identification card. In vitro antimicrobial susceptibility testing of the isolates was performed in Mueller-Hinton broth (Becton Dickinson, Sparks, MD) by the broth microdilution method in accordance with the guidelines established by the Clinical and Laboratory Standards Institute (CLSI). For all isolates, ESBL production was tested by comparing the inhibitory zone diameter of disks containing 30 μg of cefotaxime, ceftazidime, or cefepime, either alone or in combination with 10 μg of clavulanate. The phenotypical ESBL production was defined as a ≥5-mm increase in zone diameter for at least one of the combination disks relative to its corresponding single antibiotic disk.⁴ Quality control was performed using Escherichia coli ATCC 25922 and Klebsiella pneumoniae ATCC 700603. ESBL genes were detected by polymerase chain reaction (PCR) amplification, using primers based on published sequences, followed by DNA sequencing.¹¹

Statistical analyses

Student's t test was used to compare continuous variables, and the χ² test or Fisher's exact test was used to compare categorical variables. In identifying the risk factors for ESBL acquisition and 30-day mortality, stepwise logistic regression was used to control for the effects of confounding variables. Variables with p<0.25 in the univariate analyses were candidates for multivariate analysis. We used backward elimination of any variable that did not contribute to the model on the grounds of the likelihood ratio test, using a statistical significance cutoff of 0.05. All p-values were two-tailed, and p-values <0.05 were considered to be statistically significant. SPSS for Windows, version 11.5 (SPSS, Chicago, IL), was used for these analyses.

Results

Study population and antimicrobial susceptibility

During the study period, a total of 124 patients with ECSM bacteremia were identified and included in the analysis. Of these isolates, 94 were of Enterobacter spp. (66 E. cloacae, 25 E. aerogenes, 2 E. asburiae, and 1 E. cancerogenus), 18 were Citrobacter spp. (12 C. freundii, 3 C. koseri, and 3 C. brakii), 8 were of Serratia spp. (7 S. marcescens, and 1 S. fonticola), and 4 were of M. morganii. We tested double-disk synergy test in all isolates. Thirty of the 124 patients (24.2%) were positive for ESBL ECSM. The ESBL-positive rate in Enterobacter spp., Citrobacter spp., Serratia spp., and Morganella spp. was 23.4% (22/94), 16.7% (3/18), 62.5% (5/8), and 0%, respectively. The demographic and clinical characteristics of the study population are shown in Table 1. The mean (±standard deviation) age of all patients was 50.6±22.8 years (range 1–91 years) and 69 (55.6%) patients were male. There were no differences in the underlying diseases and comorbid conditions (assessed by Charlson's weighted index of comorbidity) and severity status on admission (assessed by Pitt bacteremia score) between the ESBL-producing group and the non-ESBL-producing group. In the ESBL-producing group, the history of antibiotic use within 3 months was significantly more frequent than in the non-ESBL-producing group.

Table 1.

Comparison of Risk Factors in Infection with ESBL-Producing and Nonextended-Spectrum β-Lactamase-Producing ECSM Bacteremia

	Overall (n=124)	ESBL (n=30)	Non-ESBL (n=94)	p
Age (mean±SD), years (range)	50.6±22.8 (1–91)	51.4±22.2 (1–89)	48.1±24.6 (1–91)	0.481
Sex (M/F), n	69/55	17/13	52/42	1.000
Nosocomial or healthcare-associated	105 (84.7)	27 (90.0)	78 (83.0)	0.269
Diabetes mellitus	33 (26.6)	10 (33.3)	23 (24.5)	0.351
Underlying disease
Solid tumor	47 (37.9)	10 (33.3)	37 (39.4)	0.667
Hematologic malignancy	26 (21.0)	1 (3.3)	25 (26.6)	0.004
Solid organ transplantation	12 (9.7)	7 (23.3)	5 (5.3)	0.008
LC including HCC	11 (8.9)	3 (10.0)	8 (8.5)	0.525
No underlying disease	6 (4.8)	0 (0.0)	6 (6.4)	0.182
Cerebrovascular disease	2 (1.6)	2 (6.7)	0 (0.0)	0.057
Others	17 (13.7)	7 (23.3)	10 (10.6)	0.077
Location at the time of bacteremia
General ward	61 (49.2)	13 (43.3)	48 (51.1)	0.532
Emergency department	49 (39.5)	10 (33.3)	39 (41.5)	0.522
Intensive care unit	14 (11.3)	7 (23.3)	7 (7.4)	0.024
CWIC (mean±SD)	3.1±1.9	3.1±1.9	3.1±1.9	0.843
0–2	75 (60.5)	13 (43.3)	36 (38.3)	0.671
≥3	49 (39.5)	17 (56.7)	59 (62.8)
Immunosupressant use within 30 days	46 (37.1)	13 (43.3)	33 (35.1)	0.393
Neutropenia	28 (22.6)	2 (6.67)	26 (27.7)	0.022
Pitt bacteremia score (mean±SD)	1.4±1.5	1.2±1.2	1.5±1.6	0.374
0–3	111 (89.5)	28 (93.3)	83 (88.3)	0.308
≥4	13 (10.5)	2 (6.6)	28 (29.8)
Presentation with ARF	27 (21.7)	8 (26.7)	19 (20.2)	0.456
Indwelling device at the time of bacteremia	78 (62.9)	20 (66.7)	58 (61.7)	0.128
Presentation with septic shock	47 (37.9)	8 (26.7)	39 (41.5)	0.195
Antibiotics use within 3 months	84 (67.7)	25 (83.3)	59 (62.8)	0.044
Broad-spectrum cephalosporins	58 (46.7)	20 (66.7)	38 (40.4)	0.038
Penicillin derivatives	31 (25.0)	14 (46.7)	17 (18.1)	0.004
Fluoroquinolones	14 (11.3)	7 (23.3)	7 (7.4)	0.047

Data indicate numbers (%) of patients unless otherwise indicated.

ESBL, extended-spectrum β-lactamase; ECSM, Enterobacter spp., Citrobacter spp., Serratia spp., and Morganella morganii; SD, standard deviation; LC, liver cirrhosis; HCC, hepatocellular carcinoma; CWIC, Charlson's weighted index of comorbidity; ARF, acute renal failure.

As expected, the ESBL-producing isolates displayed significantly higher resistance rates than the non-ESBL-producing isolates to the majority of antimicrobial agents, with the exception of ampicillin and imipenem (Table 2). The genotypic distribution of the ESBL-producing isolates (Table 3) has been published before.³

Table 2.

Comparison of Antimicrobial Susceptibilities Between Nonextended-Spectrum β-Lactamase-Producing and Nonextended-Spectrum β-Lactamase-Producing Isolates

	ESBL (n=30)				non-ESBL (n=94)
	MIC₅₀	MIC₉₀	Range	R (%)	MIC₅₀	MIC₉₀	Range	R (%)	p^a
Ampicillin	>64	>64	>64	100.0	>64	>64	≤0.06–>64	94.7	0.244
Gentamicin	0.25	>64	0.12–>64	33.3	0.5	2	≤0.06–>64	7.4	0.001
Cefoxitin	>64	>64	16–>64	100.0	>64	>64	1–>64	86.2	0.022
Ciprofloxacin	0.12	32	≤0.06–>64	30.0	0.12	0.5	≤0.06–>64	4.3	<0.001
Ceftazidime	>64	>64	2–>64	93.3	1	64	≤0.06–>64	29.8	<0.001
Cefotaxime	>128	>128	0.12–>128	96.7	0.5	64	≤0.12–>128	38.3	<0.001
Cefepime	8	64	0.06–>64	40.0	≤0.06	1	≤0.06–32	3.2	<0.001
Aztreonam	64	>64	0.12–>64	93.3	0.25	64	≤0.06–>64	23.4	<0.001
Imipenem	0.5	2	0.12–16	3.3	0.5	2	0.12–8	2.1	0.568
TMP-STX	1	>32	0.06–>32	43.3	0.25	4	0.03–>32	9.6	<0.001
Pip/taz	256	>256	2–>256	73.3	4	64	0.5–256	21.3	<0.001

Strains found to be intermediate by antimicrobial susceptibility testing were considered to be resistant.

Comparison of ESBL-producing isolates' and non-ESBL isolates' resistance rates by χ² test.

R, resistant; TMP-STX, trimethoprim-sulfamethoxazole; Pip/taz, piperacillin-tazobactam.

Table 3.

Distribution of Nonextended-Spectrum β-Lactamases Among Enterobacteriacae Isolates

	TEM				CTX	SHV
Species (no. of isolated tested)	TEM-1 ^a	TEM-17	TEM-52	TEM-128	CTX-M-12	SHV-12	No PCR detection	Multiple detection
Enterobacter cloacae (16)	4		1		1	3	8	1
Enterobacter aerogenes (5)	1	1	2			1	1	1
Citrobacter freundii (2)							2
Serratia marcescens (4)	1		1	1			1
Citrobacter brakii (1)							1
Enterobacter cancerogenus (1)						1	0
Serratia fonticola (1)							1
Total	6	1	4	1	1	5	14	2

TEM-1 is not ESBL enzyme.

PCR, polymerase chain reaction.

Risk factors for ESBL bacteremia

A further analysis was conducted to identify the risk factor for ESBL bacteremia (Table 3). Variables with p<0.25 in the univariate analyses (Table 1) were candidates for multivariate analysis. In the multivariate analysis, indwelling device at the onset time of bacteremia and antibiotic use within 3 months were independent factors associated with ESBL bacteremia (p<0.05) (model I in Table 4). When antibiotic use within 3 months was excluded in this analysis, immunosuppressive drugs use within 30 days, broad-spectrum cephalosporins, and use of penicillin derivatives were independent factors associated with ESBL acquisition (p<0.05) (model II in Table 4).

Table 4.

Multivariate Analysis of Risk Factors for Nonextended-Spectrum β-Lactamase-Producing ECSM Bacteremia

	Model I		Model II
Variables	Adjusted OR (95% CI)	p	Adjusted OR (95% CI)	p
Immunosupressant use within 30 days	—	—	3.68 (1.15–11.74)	0.028
Indwelling device at the time of bacteremia	2.49 (1.04–5.98)	0.042	—	—
Antibiotics use within 3 months	3.56 (1.21–10.52)	0.022	Not applicable
Broad-spectrum cephalosporins	Not applicable		3.80 (1.38–10.43)	0.010
Penicillin derivatives	Not applicable		3.73 (1.32–10.57)	0.013
Fluoroquinolones	Not applicable		—	—

Data indicate numbers (%) of patients unless otherwise indicated.

OR, odds ratio; CI, confidential interval.

Outcomes of ESBL-producing ECSM bacteremia

Clinical features and outcomes are summarized in Table 5. The most common sites of infection were hepatobiliary infection, primary bacteremia, and urinary tract infection (29.8%, 25.8%, and 13.7%, respectively). There was no significant difference in the primary site of infection between the ESBL-producing group and the non-ESBL-producing group. Of the 30 patients with ESBL-producing bacteremia, 19 patients (63.3%) received inappropriate initial antimicrobial therapy, whereas two (21.3%) of 94 patients with non-ESBL-producing bacteremia received inappropriate initial antimicrobial therapy (p<0.001). The mortality rate between ESBL-producing bacteremic patients and non-ESBL-producing bacteremic patients was not significantly different. However, the hospital stay after bacteremia was significantly longer in cases in the ESBL-producing group than in cases in the non-ESBL-producing group (46.9±90.5 vs. 20.7±27.3 days, respectively; p=0.014).

Table 5.

Clinical Features and Outcomes of 162 Cases with ECSM Bacteremia

	Overall (n=124)	ESBL (n=30)	Non-ESBL (n=94)	p
Primary site of infection
Hepatobiliary tract infection	37 (29.8)	12 (40.0)	25 (25.6)	0.176
Primary bacteremia	32 (25.8)	4 (13.3)	28 (29.8)	0.094
Urinary tract infection	17 (13.7)	3 (10.0)	14 (14.9)	0.368
Catheter-related infection	10 (8.1)	0 (0.0)	10 (10.6)	0.055
Gastrointestinal infection	9 (7.3)	3 (10.0)	6 (6.4)	0.376
Pneumonia	6 (4.8)	2 (6.7)	4 (4.3)	0.449
Wound infection	5 (4.0)	2 (6.7)	3 (3.2)	0.350
Peritonitis	4 (3.2)	1 (3.3)	3 (3.2)	0.675
Others	4 (3.2)	3 (10.0)	1 (1.1)	0.044
Inappropriate initial antimicrobial therapy	39 (31.5)	19 (63.3)	20 (21.3)	<0.001
Hospital stay after bacteremia (mean±SD), days	27.1±51.2	46.9±90.5	20.7±27.3	0.014
30-day mortality	24 (19.4)	5 (16.6)	19 (20.2)	0.794

Data indicate numbers (%) of patients unless otherwise indicated.

Predictors of mortality in patients with ECSM bacteremia

The overall 30-day mortality was 19.4% (24/124). The factors associated with 30-day mortality are shown in Table 6. By univariate analysis, the factors associated with 30-day mortality in patients with ECSM bacteremia were intensive care unit stay at the onset time of bacteremia, high Charlson's weighted index of comorbidity, high Pitt bacteremia score, and presentation with acute renal failure (all p<0.05). Multivariate analysis using a logistic regression model, which included the variables associated with mortality by univariate analysis (p≤0.25), showed that intensive care unit stay at the time of bacteremia, high Charlson's weighted index of comorbidity, high Pitt bacteremia score, presentation with acute renal failure, and pneumonia were significantly associated with 30-day mortality.

Table 6.

Risk Factors for 30-Day Infection-Related Mortality of Patients with ECSM Bacteremia

			Univariate	Multivariate
	Survivors (n=100)	Nonsurvivors (n=24)	p	OR (95% CI)	p
Age (mean±SD), years (range)	49.2±23.2	56.3±20.1	0.348
Sex (M/F), n	58/42	11/13	0.281
Nosocomial or healthcare-associated infection	82 (82.0)	23 (95.8)	0.119
Invasive procedure within 30 days	45 (45.0)	7 (29.2)	0.158
Location at the time of bacteremia
General ward	49 (49.0)	11 (45.8)	0.780
Emergency department	43 (43.0)	7 (29.2)	0.248
Intensive care unit	8 (8.0)	6 (25.0)	0.029	9.81 (1.80–53.53)	0.008
CWIC (mean±SD)	2.8±1.8	4.3±1.9	<0.001
0–2	68 (68.0)	7 (29.2)	<0.001
≥3	32 (32.0)	17 (70.8)		7.81 (2.03–30.10)	0.003
Immunosupressant use within 30 days	39 (39.0)	7 (29.2)	0.370
Neutropenia	79 (79.0)	17 (70.8)	0.390
Pitt bacteremia score (mean±SD)	1.0±1.2	2.8±1.7	<0.001
0–3	96 (96.0)	15 (62.5)	<0.001
≥4	4 (4.0)	9 (37.5)		21.08 (4.15–107.04)	<0.001
Presentation with acute renal failure	16 (16.0)	12 (50.0)	<0.001	4.23 (1.07–16.65)	0.039
Indwelling device at the time of bacteremia	61 (61.0)	17 (70.8)	0.370
Primary site of infection
Hepatobiliary tract infection	33 (33.0)	4 (16.7)	0.116
Primary bacteremia	24 (24.0)	8 (33.3)	0.348
Urinary tract infection	13 (13.0)	4 (16.7)	0.741
Catheter-related infection	9 (9.0)	1 (4.2)	0.685
Gastrointestinal infection	8 (8.0)	1 (4.2)	1.000
Pneumonia	3 (3.0)	3 (12.5)	0.086	13.88 (1.66–116.32)	0.015
Wound infection	5 (5.0)	0 (0.0)	0.582
Peritonitis	3 (3.0)	1 (4.2)	0.582
Others	2 (2.0)	2 (8.3)	0.168
ESBL-producing organisms	25 (25.0)	5 (20.8)	0.918
Antibiotics use within 3 months	67 (67.0)	17 (70.8)	0.811
Inappropriate initial antimicrobial therapy	32 (32.0)	7 (29.2)	1.000

Data indicate numbers (%) of patients unless otherwise indicated.

Discussion

The present study evaluated the clinical significance related to ESBL-producing ECSM infections. In contrast to the previous reports that included all specimens of ECSM,^2,4,11,17 we only selected cases with ECSM bacteremia. This was done to avoid the difficulties in differentiating infection and colonization.⁶ To our knowledge, this is the first study to evaluate the clinical features and treatment outcomes of adult patients with ESBL-producing ECSM bacteremia.

The recent use of antibiotics was a strong risk factor of ESBL bacteremia. Broad-spectrum cephalosporins, penicillin derivatives, and fluoroquinolones were used more in the ESBL-positive group than in the ESBL-negative group. The results confirm that the ESBLs may emerge as result of excessive antibiotic use, and also indicate the need for interventions to restrict antibiotic use to reduce the level of antibiotic resistance.

Resistant bacteria, such as ESBL-producing pathogens, may result in much higher mortality among patients with infection.^1,9,12 However, there was no significant association between ESBL production and mortality in our study, despite the finding of more inappropriate therapy in the ESBL-producing group. The impact of ESBL-producing organisms on outcomes of infections caused by Gram-negative organisms (mainly, E. coli and K. pneumoniae) has been the subject of debates over the past decade. A number of studies found no significant association between ESBL production and treatment failure or crude mortality.^6,7,21 In contrast, other studies observed that patients with infections due to ESBL-producing organisms tend to have poorer outcomes than comparable patients with infection caused by non-ESBL-producing organisms.^1,9,12,20 Conflicting results might be statistically influenced by different patient characteristics, underlying comorbidities, patient managements (excluding antimicrobials), and as yet unknown factors. Even if infections due to ESBL-producing ECSMs had little impact on mortality, such infections were associated with significantly longer durations of hospital stay (46.9±90.5 vs. 20.7±27.3 days, p=0.014). This finding might have a clinical significance because longer hospital stay can increase hospital charges.¹³

Also, since AmpC β-lactamase-producing organisms, such as ECSM, can act as hidden reservoirs for ESBLs, it is important for clinical microbiology laboratories.¹⁹ However, there is no CLSI recommendation from governing the phenotypic detection for ESBL among AmpC β-lactamase-producing Enterobacteriaceae because clavulanic acid may induce a high-level expression of chromosomal AmpC enzyme and may then antagonize, rather than protect, the antibacterial activity of the partner β-lactam, masking the synergy required to detect ESBL production.^15,22 Since presently a significant proportion of clinical isolates of AmpC β-lactamase-producing Enterobacteriaceae were positive for ESBL, the selection of antibiotics for the treatment of infections caused by ESBL-producing organisms is an important issue for clinicians. Carbapenems are the optimal choice for severe infections such as bloodstream infections caused by AmpC β-lactamase-producing Enterobacteriacae. To use antibiotics safely for treatment, an ESBL confirmatory test is necessary for clinical isolates of AmpC β-lactamase-producing Enterobacteriaceae.

There are some potential limitations to our study. First, this study was observational, and thus unknown risk factors for ESBL acquisition and mortality might have affected the outcome. Second, our study was performed at a single large institution and the results may or may not be applicable to other hospital settings. Third, other multidrug-resistance mechanisms, such as derepressed AmpC mutants, were not considered, because most ESBL-producing organisms were already derepressed AmpC mutants (21 of 30 strains; 70%). Finally, because other PCR primers, such as other CTX-M types, were not tested in this study, there were many no-PCR-detecting isolates.

In conclusion, the data implicate recent use of antibiotics (especially broad-spectrum cephalosporins and penicillin derivatives) as an independent risk factor for ESBL among ECSM bacteremia. Although there was no significant association between ESBL production and mortality, ESBL-producing ECSMs were associated with a significantly longer duration of hospital stay, thereby indicating that the impact of these infections on medical costs may be significant.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

Ariffin

, Navaratnam

, Mohamed

, Arasu

, Abdullah

W.A.

, Lee

C.L.

, Peng

L.H.

2000. Ceftazidime-resistant Klebsiella pneumoniae bloodstream infection in children with febrile neutropenia. Int. J. Infect. Dis., 4:21–25.

Bell

J.M.

, Turnidge

J.D.

, Jones

R.N.

2003. Prevalence of extended-spectrum beta-lactamase-producing Enterobacter cloacae in the Asia-Pacific region: results from the SENTRY Antimicrobial Surveillance Program, 1998 to 2001. Antimicrob. Agents Chemother., 47:3989–3993.

Cheong

, Ko

, Kang

, Chung

, Peck

, Song

2010. Prevalence of extended-spectrum beta-lactamase among Enterobacteriacae blood isolates with inducible AmpC beta-lactamase. Infect. Chemother., 42:280–284.

Choi

S.H.

, Lee

J.E.

, Park

S.J.

, Kim

M.N.

, Choo

E.J.

, Kwak

Y.G.

, Jeong

J.Y.

, Woo

J.H.

, Kim

N.J.

, Kim

Y.S.

2007. Prevalence, microbiology, and clinical characteristics of extended-spectrum beta-lactamase-producing Enterobacter spp., Serratia marcescens, Citrobacter freundii, and Morganella morganii in Korea. Eur. J. Clin. Microbiol. Infect. Dis., 26; 557–561.

Chow

J.W.

, Yu

V.L.

1999. Combination antibiotic therapy versus monotherapy for gram-negative bacteraemia: a commentary. Int. J. Antimicrob. Agents, 11:7–12.

, Long

, Liu

, Chen

, Liu

, Xu

, Xie

2002. Extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella pneumoniae bloodstream infection: risk factors and clinical outcome. Intensive Care Med., 28:1718–1723.

Emery

C.L.

, Weymouth

L.A.

1997. Detection and clinical significance of extended-spectrum beta-lactamases in a tertiary-care medical center. J. Clin. Microbiol., 35:2061–2067.

Fauci

A.S.

, Braunwald

, Kasper

D.L.

, Hauser

S.L.

, Longo

D.L.

, Jameson

J.L.

, Jones

R.N.

2008. Harrison's Principles of Internal Medicine. 17th. McGraw-Hill: New York.

P.L.

, Chan

W.M.

, Tsang

K.W.

, Wong

S.S.

, Young

2002. Bacteremia caused by Escherichia coli producing extended-spectrum beta-lactamase: a case-control study of risk factors and outcomes. Scand. J. Infect. Dis., 34:567–573.

10.

Kim

B.N.

, Woo

J.H.

, Kim

M.N.

, Ryu

, Kim

Y.S.

2002. Clinical implications of extended-spectrum beta-lactamase-producing Klebsiella pneumoniae bacteraemia. J. Hosp. Infect., 52:99–106.

11.

Kim

, Lim

Y.M.

2005. Prevalence of derepressed ampC mutants and extended-spectrum beta-lactamase producers among clinical isolates of Citrobacter freundii, Enterobacter spp., and Serratia marcescens in Korea: dissemination of CTX-M-3, TEM-52, and SHV-12. J. Clin. Microbiol., 43:2452–2455.

12.

Kim

Y.K.

, Pai

, Lee

H.J.

, Park

S.E.

, Choi

E.H.

, Kim

J.H.

, Kim

E.C.

2002. Bloodstream infections by extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella pneumoniae in children: epidemiology and clinical outcome. Antimicrob. Agents Chemother., 46:1481–1491.

13.

Lautenbach

, Patel

J.B.

, Bilker

W.B.

, Edelstein

P.H.

, Fishman

N.O.

2001. Extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella pneumoniae: risk factors for infection and impact of resistance on outcomes. Clin. Infect. Dis., 32:1162–1171.

14.

Lesens

, Methlin

, Hansmann

, Remy

, Martinot

, Bergin

, Meyer

, Christmann

2003. Role of comorbidity in mortality related to Staphylococcus aureus bacteremia: a prospective study using the Charlson weighted index of comorbidity. Infect. Control Hosp. Epidemiol., 24:890–896.

15.

Livermore

D.M.

, Brown

D.F.

2001. Detection of beta-lactamase-mediated resistance. J. Antimicrob. Chemother., 48,Suppl 1:59–64.

16.

Pai

, Hong

J.Y.

, Byeon

J.H.

, Kim

Y.K.

, Lee

H.J.

2004. High prevalence of extended-spectrum beta-lactamase-producing strains among blood isolates of Enterobacter spp. collected in a tertiary hospital during an 8-year period and their antimicrobial susceptibility patterns. Antimicrob. Agents Chemother., 48:3159–3161.

17.

Park

Y.J.

, Park

S.Y.

, Oh

E.J.

, Park

J.J.

, Lee

K.Y.

, Woo

G.J.

, Lee

2005. Occurrence of extended-spectrum beta-lactamases among chromosomal AmpC-producing Enterobacter cloacae, Citrobacter freundii, and Serratia marcescens in Korea and investigation of screening criteria. Diagn. Microbiol. Infect. Dis., 51:265–269.

18.

Paterson

D.L.

2006. Resistance in gram-negative bacteria: enterobacteriaceae. Am. J. Med., 119:S20–S28discussion S62–S70.

19.

Pitout

J.D.

, Reisbig

M.D.

, Venter

E.C.

, Church

D.L.

, Hanson

N.D.

2003. Modification of the double-disk test for detection of enterobacteriaceae producing extended-spectrum and AmpC beta-lactamases. J. Clin. Microbiol., 41:3933–3935.

20.

Ramphal

, Ambrose

P.G.

2006. Extended-spectrum beta-lactamases and clinical outcomes: current data. Clin. Infect. Dis., 42,Suppl 4:S164–S172.

21.

Schiappa

D.A.

, Hayden

M.K.

, Matushek

M.G.

, Hashemi

F.N.

, Sullivan

, Smith

K.Y.

, Miyashiro

, Quinn

J.P.

, Weinstein

R.A.

, Trenholme

G.M.

1996. Ceftazidime-resistant Klebsiella pneumoniae and Escherichia coli bloodstream infection: a case-control and molecular epidemiologic investigation. J. Infect. Dis., 174:529–536.

22.

Yang

J.L.

, Wang

J.T.

, Lauderdale

T.L.

, Chang

S.C.

2009. Prevalence of extended-spectrum beta-lactamases in Enterobacter cloacae in Taiwan and comparison of 3 phenotypic confirmatory methods for detecting extended-spectrum beta-lactamase production. J. Microbiol. Immunol. Infect., 42:310–316.