Risk Factors for Mortality in Patients with Pseudomonas aeruginosa Bacteremia: Clinical Impact of Antimicrobial Resistance on Outcome

Abstract

Despite the high prevalence of antimicrobial resistance among Pseudomonas aeruginosa bacteremia, the clinical consequence of resistance remains unclear. The purpose of this study was to identify predictors of mortality and evaluate the clinical impact of antimicrobial resistance on outcome in P. aeruginosa bacteremia. A retrospective cohort study including patients with P. aeruginosa bacteremia was performed. The risk factors for antimicrobial resistances were evaluated, and the impact of the respective resistances on mortality was assessed. Of 202 P. aeruginosa bacteremia cases, the resistance rates to ceftazidime, piperacillin, imipenem, fluoroquinolone, and aminoglycoside were 36.6%, 22.3%, 22.8%, 23.8%, and 17.8%, respectively. A prior use of fluoroquinolones and an indwelling urinary catheter were common risk factors for all types of antimicrobial resistance. The overall 30-day mortality rate was 25.2% (51/202), and the risk factors for mortality were corticosteroid use, nosocomial acquisition, polymicrobial infection, an increasing Charlson's weighted co-morbidity index, and intensive care unit care (p < 0.05). As compared with the susceptible group, ceftazidime-, piperacillin-, or imipenem-resistant groups had a higher mortality (p < 0.05). A multivariate analysis showed that resistance to ceftazidime or imipenem remained a significant factor associated with mortality (odds ratio, 2.96; 95% confidential interval, 1.20–7.31; and odds ratio, 2.74; 95% confidential interval, 1.02–7.31, respectively). Antimicrobial resistance, especially to ceftazidime or imipenem, adversely affected outcome in patients with P. aeruginosa bacteremia.

Introduction

The clinical impact of antimicrobial resistance on the outcome of serious infections caused by Gram-negative pathogens has been the subject of highly controversial debates over the past decade. Antimicrobial-resistant Gram-negative bacterial infections are associated with prolonged hospital stay and an increased mortality.^10,23 Among Gram-negative pathogens, Pseudomonas aeruginosa remains a significant cause of morbidity and mortality in nosocomial infections, and its antimicrobial resistance is closely related to difficulties in controlling and treating infection because of intrinsic resistance and rapid acquisition of additional resistance.⁶ One possible explanation for the increased mortality associated with antimicrobial-resistant P. aeruginosa infections is delayed administration of appropriate antimicrobial therapy.^9,10,13,17

It is presumed that infections caused by antimicrobial-resistant bacterial infections result in higher mortality, longer hospitalizations, and greater costs than infections caused by susceptible infections. Although carbapenem resistance in P. aeruginosa infections was associated with a longer duration of hospital stay and a greater hospital cost,^15,20,25 the impact of antimicrobial resistance on treatment outcome might depend on the severity of the underlying disease and the presence of comorbid conditions. Despite the high prevalence of antimicrobial resistance among P. aeruginosa that cause bacteremia, the clinical consequence of resistance remains unclear.²⁴ Thus, this study was performed to identify the predictors of mortality and evaluate the clinical impact of antimicrobial resistance on outcome in patients with P. aeruginosa bacteremia.

Materials and Methods

Patients and bacterial strains

The database at our Clinical Microbiology Laboratory was reviewed to identify patients with P. aeruginosa bacteremia. Cases that were identified to have P. aeruginosa infection were screened, and isolates recovered from blood culture specimens were reviewed. Patients older than 16 years with P. aeruginosa bacteremia were included in this study. We reviewed the electronic medical records of individuals found to have P. aeruginosa bacteremia from October 2006 to March 2009 at Samsung Medical Center (Seoul, Republic of Korea), a 1,900-bed tertiary care university hospital. Patients who stayed in hospital more than 100 days before P. aeruginosa bacteremia occurred and those who died within a day were excluded. Only the first bacteremic episode for each patient was included in the analysis.

Microbiological identification was performed using a standard identification card, and antimicrobial susceptibility testing was performed on the VITEK II automated system (bioMérieux, Marcy l'Etoile, France) using the modified broth microdilution method. Minimum inhibitory concentration breakpoints and quality-control protocols were used according to standards established by the Clinical and Laboratory Standard Institute. Strains showing intermediate antimicrobial susceptibility testing were considered to be resistant.

Study design and data collection

A retrospective cohort study was conducted to identify the predictors of mortality and evaluate the clinical impact of antimicrobial resistance on outcome in patients with P. aeruginosa bacteremia. The risk factors for antimicrobial resistance to piperacillin, ceftazidime, imipenem, fluoroquinolones, and aminoglycosides were evaluated, and the impact of the respective antimicrobial resistance on mortality was assessed. Data from patients with antimicrobial-resistant P. aeruginosa bacteremia were compared with those from patients with susceptible bacteremia. The collected data included age, sex, the date of bacteremia, underlying diseases, category of infection, presence of healthcare-associated (HCA) risk factors, empirical antibiotic regimens, appropriateness of empirical antibiotics, patterns of antimicrobial resistance, severity of illness, primary sites of infection, portal of entry, and clinical outcomes. The presence of devices such as central venous and urinary catheters was evaluated at the date when index blood samples were obtained. Severity of illness was calculated by the Pitt bacteremia score and Charlson's weighted co-morbidity index (WCI). Admission to intensive care units (ICUs) and use of a mechanical ventilator were also assessed at the time of acquisition of P. aeruginosa bloodstream infection. For comparison of outcomes, initial response to treatment was assessed at 72–96 hours after the initiation of antimicrobial therapy and was classified as follows: “improvement”, patients who had resolution of fever, leukocytosis, and all signs of infection, and “failure”, patients who experienced no improvement or experienced deterioration in any of their clinical parameters as well as those who died. The main outcome measure was the 30-day mortality rate.

To identify the risk factors for antimicrobial resistance, the following data were also collected: recent operations, type of operation, history of solid organ transplantation, immunosuppressant use, corticosteroid use, prior use of antibiotics, prior hospitalization, and duration of hospital stay before bacteremia. The decision on the diagnosis and management of the primary source of infection was made solely by the attending physicians without any intervention by investigators.

Definitions

P. aeruginosa bacteremia was defined as an infection with P. aeruginosa isolates from blood cultures. Antimicrobial resistance was defined as in vitro resistance to ceftazidime, piperacillin, imipenem, fluoroquinolones, or aminoglycosides. In vitro resistance to levofloxacin or ciprofloxacin represented fluoroquinolone resistance, and resistance to amikacin, tobramycin, or gentamicin represented aminoglycoside resistance. A multidrug-resistant strain was defined as a strain resistant to three or more of the five categorized classes.^3,22 The initial empirical antimicrobial therapy was considered appropriate if initial empiric antibiotics were administered within 24 hours after acquisition of blood culture samples and included at least one antibiotic that was active in vitro against P. aeruginosa and when the dosage and route of administration conformed with current medical standards. The length of delay of appropriate antimicrobial therapy was also defined as the interval with 48, 72, and 96 hours between the time the blood culture samples were obtained and the time effective antibiotics were administered. Inappropriate initial antimicrobial therapy referred to the administration of antimicrobial agents to which P. aeruginosa was resistant in vitro. Previous antimicrobial therapy was defined when any type of antimicrobial therapy has been administered for more than 48 hours within 90 days before P. aeruginosa bacteremia. Types of antibiotics previously used were categorized into five classes: cephalosporin, penicillin, carbapenem, fluoroquinolones, and aminoglycosides.

Community-onset bacteremia was defined when it occurred within 48 hour of admission and was classified into HCA and community-associated bacteremia. Patients were categorized as HCA if they fulfilled any of the following criteria^4,19: received intravenous therapy at home or in an outpatient clinic in the previous 30 days; received renal dialysis in a hospital or clinic during the preceding 30 days; had been hospitalized for two or more days in the previous 90 days; or resided in a nursing home or long-term care facility for two or more days in the previous 90 days. Community-onset bacteremia without these HCA risk factors was classified as community-associated. Nosocomial infection was defined as an infection that occurred 48 hours after hospital admission.

Neutropenia was defined as an absolute neutrophil count below 500/mm³. Severe sepsis was defined as sepsis with organ dysfunction, hypoperfusion, or hypotension, and septic shock was defined as refractory arterial hypotension or hypoperfusion in spite of adequate fluid resuscitation, with a systolic blood pressure <90 or >30 mmHg less than the baseline or a requirement for use of a vasopressor to maintain the blood pressure.

Statistical analysis

The Student's t test and Mann–Whitney test were used to compare continuous variables, and the χ² test or Fisher's exact test was used to compare categorical variables. To identify the independent risk factors for antimicrobial resistance and mortality, a stepwise backward multivariable logistic regression analysis model was used to control for the effects of confounding variables. Variables with p < 0.05 in the univariate analyses were candidates for multivariate analysis and variables for which p < 0.05 in the multivariate analysis were retained in the final mode. 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 p-values <0.05 were considered to be statistically significant. SPSS statistics for Windows, PASW version 17.0, was used for these analyses.

Results

Study population

During the study period, a total of 244 patients with P. aeruginosa bacteremia were screened and 42 patients excluded in the study. Sixteen isolates were identified to be obtained from other sites except bloodstream. Thirteen stayed in hospital more than 100 days before the bacteremia and 10 died within a day. Four patients experienced recurrent Gram-negative bacteremia at the time of acquisition of P. aeruginosa. Finally, a total of 202 patients were included in this study. The mean age (±standard deviation) of patients was 55 ± 15 years, and 127 patients (62.9%) were men. The most common underlying diseases were solid tumors (n = 103, 51.0%), followed by hematologic malignancy and liver disease. Of 202 patients, 123 (60.9%) had nosocomial infections and the remaining 79 (39.1%) had community-onset infections, of which 61 (77.2%) were classified as community-onset HCA infections. As for the primary sites of infection, intra-abdominal infection was the most common (n = 84, 41.6%), followed by primary bacteremia and pneumonia. Demographic and clinical characteristics of the study population are described in Table 1.

Table 1.

Clinical Characteristics of Patients with Pseudomonas aeruginosa Bacteremia

Characteristics	No. of patients (%) (n = 202)
Age, years (mean ± SD)	55 ± 15
Male/female	127/75 (62.9/37.1)
Underlying disease
Solid tumor	103 (51.0)
Hematologic malignancy	72 (35.6)
Liver disease	71 (35.1)
Renal failure	37 (18.3)
Lung disease	30 (14.9)
Heart disease	29 (14.4)
Diabetes mellitus	25 (12.4)
Cerebrovascular disease	13 (6.4)
Comorbid conditions
Prior hospitalization within 90 days	128 (63.4)
Immunosuppressants use	109 (54.0)
Corticosteroid use	94 (46.5)
Recent operation within a year	54 (26.7)
Solid organ transplantation	28 (13.9)
Neutropenia	81 (40.1)
Polymicrobial infection	28 (13.9)
Intubated state	13 (6.4)
Indwelling urinary catheter	35 (17.3)
Central venous catheterization	90 (44.6)
Percutaneous catheterization	47 (23.3)
Invasive procedure within previous 72 hours	80 (39.6)
Category of infection
Community-associated infection	18 (8.9)
Healthcare-associated infection	61 (30.2)
Hospital-acquired infection	123 (60.9)
Prior receipt of antibiotics within 3 months before bacteremia
Any types of antibiotics	148 (73.3)
Extended-spectrum cephalosporins	131 (64.9)
Penicillins	54 (26.7)
Carbapenems	36 (17.8)
Fluoroquinolones	35 (17.3)
Aminoglycosides	16 (7.9)
Severity of illness
Pitt bacteremia score (mean ± SD)	2.03 ± 2.65
Charlson's weighted index of co-morbidity (mean ± SD)	4.51 ± 2.71
ICU admission	50 (24.8)
Septic shock	72 (35.6)
Severe sepsis or septic shock	88 (43.6)
Site of infection
Primary bacteremia	44 (21.8)
Intra-abdominal infection	84 (41.6)
Respiratory tract infection	30 (14.9)
Catheter-related infection	19 (9.4)
Urinary tract infection	18 (8.9)
Skin and soft tissue infection	7 (3.5)

SD, standard deviation; IQR, interquartile rate; MDR, multidrug resistance; ICU, intensive care units.

For the P. aeruginosa blood isolates, the resistance rates to ceftazidime, piperacillin, imipenem, fluoroquinolone, and aminoglycoside were 36.6% (74/202), 22.3% (45/202), 22.8% (46/202), 23.8% (48/202), and 17.8% (36/202), respectively. The multidrug resistance rate was 20.8% (42/202).

Risk factors associated with antimicrobial resistance

The factors associated with antimicrobial resistance to ceftazidime, piperacillin, imipenem, fluoroquinolones, or aminoglycosides were evaluated. From the univariate analysis, the common risk factors for resistance to ceftazidime or piperacillin were nosocomial acquisition, prior receipt of extended-spectrum penicillin or fluoroquinolones within the previous 90 days, indwelling urinary catheters, percutaneous tubes, and having had an invasive procedure within the previous 72 hours. The risk factors for imipenem resistance were prior use of carbapenems, fluoroquinolones, or aminoglycosides, and an indwelling urinary catheter. Risk factors associated with resistance to fluoroquinolone or aminoglycoside were prior use of fluoroquinolones, an indwelling urinary catheter or percutaneous tubes, and having had an invasive procedure within the previous 72 hours. We next performed a multivariate analysis using variables that were significantly associated with the respective antimicrobial resistances in the univariate analysis (p < 0.05). The independent risk factors associated with the respective antimicrobial resistances are shown in Table 2. A prior use of fluoroquinolones and indwelling urinary catheter were found to be common risk factors significantly associated with all types of antimicrobial resistance.

Table 2.

Independent Risk Factors Associated with the Respective Antimicrobial Resistances in Pseudomonas aeruginosa Bacteremia

Variables	Adjusted OR (95% CI)	p
Resistance to ceftazidime (n = 74)
Prior receipt of penicillins	2.25 (1.06–4.75)	0.034
Prior receipt of fluoroquinolones	3.13 (1.32–7.38)	0.009
Indwelling urinary catheter	2.52 (1.06–6.00)	0.037
Percutaneous catheterization	2.99 (1.32–6.78)	0.009
Invasive procedure within previous 72 hours	2.56 (1.21–5.42)	0.014
Resistance to piperacillin (n = 45)
Prior receipt fluoroquinolones	2.16 (0.87–5.36)	0.096
Indwelling urinary catheter	2.43 (1.02–5.77)	0.045
Resistance to imipenem (n = 46)
Prior receipt carbapenems	2.87 (1.26–6.56)	0.012
Prior receipt of fluoroquinolones	2.54 (1.08–5.96)	0.033
Prior receipt of aminoglycosides	3.60 (1.39–7.31)	0.025
Indwelling urinary catheter	3.19 (1.39–7.31)	0.006
Multidrug resistance (n = 42)
Prior receipt of fluoroquinolones	3.01 (1.29–7.01)	0.011
Indwelling urinary catheter	4.28 (1.14–5.53)	0.001
Percutaneous catheterization	2.51 (1.14–5.53)	0.023
Resistance to fluoroquinolones (n = 48)
Prior receipt of fluoroquinolones	4.45 (1.95–10.14)	<0.001
Indwelling urinary catheter	2.68 (1.19–6.06)	0.018
Invasive procedure within previous 72 hours	2.12 (1.02–4.38)	0.043
Resistance to aminoglycosides (n = 36)
Prior receipt of fluoroquinolones	3.40 (1.42–8.13)	0.006
Indwelling urinary catheter	4.09 (1.74–9.60)	0.001
Percutaneous catheterization	2.56 (1.12–5.87)	0.026

OR, odds ratio, CI, confidential interval.

Treatment outcome and predictors of mortality

When clinical outcomes of P. aeruginosa bacteremia were evaluated, the initial clinical failure rate was 51.5% (104/202), and the overall 30-day mortality rate was 25.2% (51/202). No significant difference was found between patients who received initial inappropriate antimicrobial therapy and those who received appropriate therapy (26.2% [32/122] vs. 23.8% [19/80], p = 0.692). The 30-day mortality rates, when stratified according to intervals of 24, 48, 72, and 96 hours between the time of acquisition of blood samples and the time of initiation of appropriate antimicrobial therapy, were not significantly different between the respective intervals. By univariate analysis, the factors associated with mortality were corticosteroid use, nosocomial acquisition, polymicrobial infections, an increase in Pitt bacteremia score and Charlson's WCI, ICU care, presentation with severe sepsis or septic shock, having pneumonia, and having endotracheal tubes, urinary catheters, or central venous catheters (all p < 0.05; Table 3). Multivariate analysis using a logistic regression model that included the variables associated with mortality by univariate analysis (p < 0.05) showed that the significant risk factors for mortality were corticosteroid use, nosocomial acquisition, polymicrobial infection, an increasing Charlson's WCI, and ICU care (Table 3).

Table 3.

Factors Associated with 30-day Mortality in Patients with Pseudomonas aeruginosa Bacteremia

			Univariate analysis		Multivariate analysis
	Survivor (n = 151)	Nonsurvivor (n = 51)	OR (95% CI)	p	Adjusted OR (95% CI)	p-value
Age (mean years ± SD)	56 ± 16	56 ± 16	56 ± 16	0.210
Male/female	99/52	28/23	0.64 (0.34–1.22)	0.173
Comorbid conditions
Prior hospitalization within 90 days	96 (63.6)	32 (62.7)	0.97 (0.50–1.96)	0.915
Receipt of immunosuppressants	78 (51.7)	31 (60.8)	1.45 (0.76–2.77)	0.258
Receipt of corticosteroid	58 (38.4)	36 (70.6)	3.85 (1.94–7.64)	<0.001	4.01 (1.15–13.94)	0.029
Recent operation within a year	38 (25.2)	16 (31.4)	1.36 (0.68–2.73)	0.387
Solid organ transplantation	19 (12.6)	9 (17.6)	1.49 (0.63–3.54)	0.366
Intubation	3 (2.0)	10 (19.6)	12.03 (3.16–45.76)	<0.001
Indwelling urinary catheter	19 (12.6)	16 (31.4)	3.18 (1.48–6.81)	0.002
Central venous catheterization	60 (39.7)	30 (58.8)	2.17 (1.14–4.13)	0.018
Percutaneous catheterizations	32 (21.2)	15 (29.4)	1.55 (0.76–3.18)	0.230
Invasive procedure within 72 hours	62 (41.1)	18 (35.3)	0.78 (0.41–1.51)	0.467
Neutropenia	61 (40.4)	20 (39.2)	0.95 (0.50–1.82)	0.882
Polymicrobial infection	13 (8.6)	15 (29.4)	4.42 (1.93–10.13)	<0.001	3.52 (1.13–11.01)	0.030
Nosocomial acquisition	81 (53.6)	42 (82.4)	4.03 (1.83–8.87)	<0.001	5.01 (1.39–17.99)	0.014
Appropriateness of empirical therapy	90 (59.4)	32 (62.7)	0.88 (0.46–1.69)	0.692
Severity of illness
Pitt bacteremia score (mean ± SD)	1.18 ± 1.46	4.55 ± 3.63	1.69 (1.42–2.00)	<0.001
Charlson's WIC (mean ± SD)	4.23 ± 2.58	5.33 ± 2.92	1.16 (1.03–1.30)	0.013	1.31 (1.04–1.65)	0.023
ICU admission	17 (11.3)	33 (64.7)	14.45 (6.73–31.04)	<0.001	3.84 (1.17–12.60)	0.027
Severe sepsis or septic shock	45 (29.4)	43 (84.3)	12.66 (5.51–29.08)	<0.001
Site of infection
Primary bacteremia	34 (22.5)	10 (19.6)	0.84 (0.38–1.85)	0.663
Intra-abdominal infection	67 (44.4)	17 (33.3)	0.63 (0.32–1.22)	0.167
Respiratory tract infection	14 (9.3)	16 (31.4)	4.47 (2.00–10.03)	<0.001
Catheter-related infection	16 (10.6)	3 (5.9)	0.53 (0.14–1.89)	0.413
Urinary tract infection	17 (11.3)	1 (2.0)	0.16 (0.02–1.22)	0.048
Skin and soft tissue infection	3 (2.0)	4 (7.8)	4.20 (0.91–19.44)	0.069

WIC, weighted index of co-morbidity.

When the impact of the respective antimicrobial resistances on mortality was assessed, ceftazidime, piperacillin, and imipenem resistances were associated with mortality in the univariate analysis (Table 4). A multivariate analysis showed that ceftazidime and imipenem resistances remained significant factors associated with mortality after adjustment for other variables that were found to be risk factors for mortality (OR, 2.96; 95% CI, 1.20–7.31; and OR, 2.74; 95% CI, 1.02–7.31, respectively) (Table 4).

Table 4.

Impact of the Respective Antimicrobial Resistances on Mortality in Pseudomonas aeruginosa Bacteremia

	R to Ceftazidime (n = 74)		R to Piperacillin (n = 45)		R to Imipenem (n = 46)		MDR (n = 42)		R to FQ (n = 48)
	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p
Overall	2.78 (1.45–5.33)	0.002	2.20 (1.08–4.48)	0.028	2.40 (1.18–4.86)	0.014	2.20 (1.06–4.55)	0.031	1.94 (0.96–3.91)	0.063
Adjusted^a	3.35 (1.38–8.10)	0.007	1.77 (0.66–4.74)	0.257	2.74 (1.02–7.37)	0.046	2.24 (0.80–6.29)	0.126	1.76 (0.65–4.74)	0.263

Adjusted for corticosteroid use, nosocomial acquisition, polymicrobial infection, Charlson's weighted index of co-morbidity, and admission to ICUs.

No., number; R, resistance; FQ, fluoroquinolones.

Clinical outcome of antimicrobial-resistant P. aeruginosa bacteremia

When the initial clinical response to treatment was assessed, patients with antimicrobial-resistant bacteremia had higher clinical failure rates than those with antimicrobial-susceptible bacteremia, regardless of the type of antimicrobial resistance (all p < 0.05; Table 5). With the exception of fluoroquinolone resistance and aminoglycoside resistance, the 30-day mortality rate was higher, and the duration of hospital stay was longer in the resistant group than in the susceptible group (Table 5).

Table 5.

Clinical Outcome of Antimicrobial-Resistant Pseudomonas aeruginosa Bacteremia

	Resistant group	Susceptible group	p
Ceftazidime
Clinical failure	52/74 (70.3)	52/128 (40.6)	0.000
30-day mortality	28/74 (37.8)	23/128 (18.0)	0.002
Hospital stay, median day (IQR)^a	23 (14–41)	15 (9–30)	0.023
Piperacillin
Clinical failure	33/45 (73.3)	71/157 (45.2)	0.001
30-day mortality	17/45 (37.8)	34/157 (21.7)	0.028
Hospital stay, median day (IQR)^a	25 (14–48)	16 (10–30)	0.018
Imipenem
Clinical failure	37/46 (80.4)	67/156 (42.9)	0.000
30-day mortality	18/46 (39.1)	33/156 (21.2)	0.014
Hospital stay, median day (IQR)^a	26 (14–42)	16 (9–29)	0.016
MDR
Clinical failure	32/42 (76.2)	72/160 (45.0)	0.000
30-day mortality	16/42 (38.1)	35/160 (21.9)	0.031
Hospital stay, median day (IQR)^a	25 (13–38)	16 (10–31)	0.102
Fluoroquinolone
Clinical failure	33/48 (68.8)	71/154 (46.1)	0.006
30-day mortality	17/48 (35.4)	34/154 (22.1)	0.063
Hospital stay, median day (IQR)^a	21 (14–35)	16 (9–22)	0.106
Aminoglycosides
Clinical failure	25/36 (69.4)	79/166 (47.6)	0.017
30-day mortality	9/36 (25.0)	42/166 (25.3)	0.970
Hospital stay, median day (IQR)^a	21 (11–36)	17 (10–31)	0.457

Data are presented as no. of events/no. of patients (%), unless otherwise indicated.

The patients who died within 30 days after bacteremia were excluded in the calculation of hospital stay, and Mann–Whitney test was used for the comparison.

Discussion

Our study has provided a comprehensive evaluation of the clinical impact of antimicrobial resistance on the outcome of patients with P. aeruginosa bacteremia. The mortality associated with P. aeruginosa bacteremia with reduced susceptibility to ceftazidime, piperacillin, or imipenem was significantly higher than that associated with susceptible bacteremia in this study. As the underlying illness in the resistant group was more severe than in the susceptible group, it may be presumed that bacteremia by antimicrobial-resistant bacteremia has a worse prognosis because of the more severe underlying illness. However, after adjusting for other prognostic factors associated with mortality, ceftazidime resistance and imipenem resistance were identified as independent risk factors for mortality in our study of patients with P. aeruginosa bacteremia. Our findings were also in accordance with a previous study which reported that fluoroquinolone resistance was not an independent risk factor for mortality even though fluoroquinolone-resistant P. aeruginosa was associated with increased hospital charges; rather, other resistance patterns, such as imipenem resistance, had a more significant impact on mortality.⁵

It is presumed that infection with antimicrobial-resistant P. aeruginosa results in higher mortality, longer hospitalizations, and greater costs than infection with susceptible strains. Although several studies have described higher mortality rates in infections caused by antimicrobial-resistant Gram-negative bacilli, the causal link between antimicrobial resistance and fatal bacteremia remains unclear. In previous studies that focused on clinical impact of antimicrobial resistance of P. aeruginosa, it was found that resistance to imipenem, multidrug or piperacillin/tazobactam, adversely affected the outcome in patient with P. aeruginosa bacteremia alike with our findings.^1,15,23,27 Even though well-designed, large-population studies have already addressed the risk factors associated with P. aeruginosa bloodstream infection, the main purpos of these studies was identification of factors for development of P. aeruginosa bacteremia or the evaluation of the adverse impact of inappropriate therapy on outcome in P. aeruginosa infection, rather than focusing on impact of antimicrobial resistance.^9,21 In our study, we aimed to evaluate the impact of respective antimicrobial resistance on mortality in P. aeruginosa infection, and identified that ceftazidime and imipenem resistance adversely affects the outcome in patients with P. aeruginosa bacteremia. One challenge in our study was how to adjust adequately for the severity of underlying illness in quantifying the difference in outcomes between patients with infections caused by susceptible versus resistant organisms. To control for the inherent biases due to clinical heterogeneity between patients in the resistant and susceptible groups, we used multivariate logistic regression analysis to determine the factors associated with 30-day mortality. As expected, we found significant differences between patients with resistant and susceptible infections when the resistance was stratified according to the respective type of antimicrobials, and showed that antimicrobial-resistant P. aeruginosa infection itself was independently associated with poorer outcome than was susceptible one. For more comprehensive evaluation of exact impact of antimicrobial resistance on the outcome in patients with P. aeruginosa infection, further large-scale case–control study is needed by matching with similar severity of illness.

Inappropriateness of empirical therapy has been found to be the strong prognostic factor for mortality in P. aeruginosa bacteremia, even after adjustment for severity of illness and underlying conditions, emphasizing the importance of appropriate antimicrobial therapy.^9,10,26 Contrary to findings of previous studies, we could not find the adverse impact of inappropriate therapy on mortality in P. aeruginosa bacteremia. It might result from the fact that the outcomes of infections may be dependent on the primary site of infection, and the virulence of the pathogens.^7,12 Similarly, the impact of antimicrobial resistance on outcome might depend on the severity of underlying disease and the primary site of infection.

Our data showed that prior receipt of fluoroquinolones and an indwelling urinary catheter were significant risk factors for antimicrobial resistance to various anti-pseudomonal antibiotics. Several studies have already provided the risk factors for antimicrobial-resistant P. aeruginosa infection, but mainly focused on imipenem resistance or multidrug resistance.^2,8,15,20,28 In our study, we evaluated the risk factors for respective antimicrobial resistance and found the association of fluoroquinolone and indwelling urinary catheters with all types of antimicrobial resistance. Alike with our findings, a previous study also showed that P. aeruginosa bacteremic isolates from patients who have been exposed to ciprofloxacin during the 30 days before the development of bacteremia had an increased risk of being resistant to ceftazidime, imipenem, meropenem, piperacillin-tazobactam, or ciprofloxacin.¹⁸ This result, regarding risk factors associated with antimicrobial resistance, has significant clinical implications for the utility of fluoroquinolone, which might result in high levels of cross-resistance to other antibiotics. To reduce the frequency of P. aeruginosa bacteremia due to antimicrobial-resistant organisms, it would be important to administer fluoroquinolone prudently and minimize invasive procedures, including the insertion of urinary catheters or percutaneous catheterization if possible.¹¹

It is known that ceftazidime and piperacillin share the same resistance mechanisms of P. aeruginosa carrying an inducible AmpC cephalosporinase. The resistance rate of ceftazidime (36.6%) in our study, however, was much higher than that of piperacillin (22.3%), suggesting the existence of different resistance mechanisms. Whereas resistance to the β-lactams emerges as a result of AmpC overproduction, a definitive relationship between P. aeruginosa AmpC and certain β-lactams remains unknown. It is notable that our study population has received cephalosporin more frequently within 90 days before P. aeruginosa bacteremia than have penicillin (65% vs. 26.7%). Given the ability of resistant P. aeruginosa to emerge during the course of therapy, the previous prolonged exposure to cephalosporin could play a role of higher resistance of ceftazidime than that of piperacillin, combined with the variability in the hydrolytic activity of AmpC from P. aeruginosa to both antibiotics.¹⁶ Interestingly, the resistance of fluoroquinolone (23.8%) was quite lower than that of ceftazidime in our data. It might result from more frequent use of cephalosporin than that of fluoroquinolone.

Our study had several limitations. First, the retrospective nature of this study may preclude accurate comparisons. The results of clear-up of P. aeruginosa from the blood were not available in many patients, which would be a very helpful clinical factor on mortality. Further, any hidden bias that was not adjusted for in our cohort may lead to underestimation or overestimation of the true relationship between antimicrobial resistance and mortality, even though we performed multivariate logistic analysis to control for other confounding factors. Second, although information concerning in-hospital antibiotic use was available from the medical records, the assessment of the use of antibiotics outside the hospital may not be accurate. Third, we reported susceptibility as a categorical variable, based on medical records provided by VITEK II automated system. Since P. aeruginosa blood isolates were not available at the time of initiation of our study, it was impossible to perform confirmative tests for identification and susceptibility of P. aeruginosa to antimicrobial agents. Similarly, the presence of resistance mechanisms was not able to be investigated either, especially in imipenem-resistant pathogens, which were usually mediated with metallo-β-Lactamase production.¹⁴ Finally, our study was conducted in a large referral center. Thus, many of our patients had serious underlying illnesses, and these data might not be generalizable to other institutes, particularly community hospitals.

In conclusion, antimicrobial resistance, especially ceftazidime resistance and imipenem resistance, adversely affected outcome in patients with P. aeruginosa bacteremia, although fluoroquinolone resistance was not associated with adverse outcome. To our knowledge, this is the largest study to assess the impact of antimicrobial resistance on outcome in patients with P. aeruginosa bacteremia after adjustment for host variables.

Footnotes

Acknowledgments

This study was supported by a grant from the Korean Health 21 R&D Project Ministry of Health, Welfare, & Family Affairs, Republic of Korea (Grant No. A084063).

Disclosure Statement

No competing financial interests exit.

References

Aloush

, Navon-Venezia

, Seigman-Igra

, Cabili

, Carmeli

2006. Multidrug-resistant Pseudomonas aeruginosa: risk factors and clinical impact. Antimicrob. Agents Chemother., 50:43–48.

Cao

, Wang

, Sun

, Zhu

, Chen

2004. Risk factors and clinical outcomes of nosocomial multi-drug resistant Pseudomonas aeruginosa infections. J. Hosp. Infect., 57:112–118.

Falagas

M.E.

, Karageorgopoulos

D.E.

2008. Pandrug resistance (PDR), extensive drug resistance (XDR), and multidrug resistance (MDR) among Gram-negative bacilli: need for international harmonization in terminology. Clin. Infect. Dis., 46:1121–1122; author reply 1122.

Friedman

N.D.

, Kaye

K.S.

, Stout

J.E.

, McGarry

S.A.

, Trivette

S.L.

, Briggs

J.P.

, Lamm

, Clark

, MacFarquhar

, Walton

A.L.

, Reller

L.B.

, Sexton

D.J.

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.

Gasink

L.B.

, Fishman

N.O.

, Weiner

M.G.

, Nachamkin

, Bilker

W.B.

, Lautenbach

2006. Fluoroquinolone-resistant Pseudomonas aeruginosa: assessment of risk factors and clinical impact. Am. J. Med., 119:526.e519–525.

Gaynes

, Edwards

J.R.

2005. Overview of nosocomial infections caused by gram-negative bacilli. Clin. Infect. Dis., 41:848–854.

Harbarth

, Nobre

, Pittet

2007. Does antibiotic selection impact patient outcome? Clin. Infect. Dis., 44:87–93.

Harris

A.D.

, Smith

, Johnson

J.A.

, Bradham

D.D.

, Roghmann

M.C.

2002. Risk factors for imipenem-resistant Pseudomonas aeruginosa among hospitalized patients. Clin. Infect. Dis., 34:340–345.

Kang

C.I.

, Kim

S.H.

, Kim

H.B.

, Park

S.W.

, Choe

Y.J.

, Oh

M.D.

, Kim

E.C.

, Choe

K.W.

2003. Pseudomonas aeruginosa bacteremia: risk factors for mortality and influence of delayed receipt of effective antimicrobial therapy on clinical outcome. Clin. Infect. Dis., 37:745–751.

10.

Kang

C.I.

, Kim

S.H.

, Park

W.B.

, Lee

K.D.

, Kim

H.B.

, Kim

E.C.

, Oh

M.D.

, Choe

K.W.

2005. Bloodstream infections caused by antibiotic-resistant gram-negative bacilli: risk factors for mortality and impact of inappropriate initial antimicrobial therapy on outcome. Antimicrob. Agents Chemother., 49:760–766.

11.

Kang

C.I.

, Kim

S.H.

, Park

W.B.

, Lee

K.D.

, Kim

H.B.

, Kim

E.C.

, Oh

M.D.

, Choe

K.W.

2005. Risk factors for antimicrobial resistance and influence of resistance on mortality in patients with bloodstream infection caused by Pseudomonas aeruginosa. Microb. Drug Resist., 11:68–74.

12.

Kim

S.H.

, Park

W.B.

, Lee

K.D.

, Kang

C.I.

, Kim

H.B.

, Oh

M.D.

, Kim

E.C.

, Choe

K.W.

2003. Outcome of Staphylococcus aureus bacteremia in patients with eradicable foci versus noneradicable foci. Clin. Infect. Dis., 37:794–799.

13.

Kollef

M.H.

2000. Inadequate antimicrobial treatment: an important determinant of outcome for hospitalized patients. Clin. Infect. Dis., 31,Suppl 4:S131–S138.

14.

Laupland

K.B.

, Parkins

M.D.

, Church

D.L.

, Gregson

D.B.

, Louie

T.J.

, Conly

J.M.

, Elsayed

, Pitout

J.D.

2005. Population-based epidemiological study of infections caused by carbapenem-resistant Pseudomonas aeruginosa in the Calgary Health Region: importance of metallo-beta-lactamase (MBL)-producing strains. J. Infect. Dis., 192:1606–1612.

15.

Lautenbach

, Synnestvedt

, Weiner

M.G.

, Bilker

W.B.

, Vo

, Schein

, Kim

2010. Imipenem resistance in Pseudomonas aeruginosa: emergence, epidemiology, and impact on clinical and economic outcomes. Infect. Control Hosp. Epidemiol., 31:47–53.

16.

Lister

P.D.

, Wolter

D.J.

, Hanson

N.D.

2009. Antibacterial-resistant Pseudomonas aeruginosa: clinical impact and complex regulation of chromosomally encoded resistance mechanisms. Clin. Microbiol. Rev., 22:582–610.

17.

Lodise

T.P.

Jr. , Patel

, Kwa

, Graves

, Furuno

J.P.

, Graffunder

, Lomaestro

, McGregor

J.C.

2007. Predictors of 30-day mortality among patients with Pseudomonas aeruginosa bloodstream infections: impact of delayed appropriate antibiotic selection. Antimicrob. Agents Chemother., 51:3510–3515.

18.

Lopez-Dupla

, Martinez

J.A.

, Vidal

, Almela

, Soriano

, Marco

, Lopez

, Olona

, Mensa

2009. Previous ciprofloxacin exposure is associated with resistance to beta-lactam antibiotics in subsequent Pseudomonas aeruginosa bacteremic isolates. Am. J. Infect. Control, 37:753–758.

19.

McDonald

J.R.

, Friedman

N.D.

, Stout

J.E.

, Sexton

D.J.

, Kaye

K.S.

2005. Risk factors for ineffective therapy in patients with bloodstream infection. Arch. Intern. Med., 165:308–313.

20.

Onguru

, Erbay

, Bodur

, Baran

, Akinci

, Balaban

, Cevik

M.A.

2008. Imipenem-resistant Pseudomonas aeruginosa: risk factors for nosocomial infections. J. Korean Med. Sci., 23:982–987.

21.

Parkins

M.D.

, Gregson

D.B.

, Pitout

J.D.

, Ross

, Laupland

K.B.

2010. Population-based study of the epidemiology and the risk factors for Pseudomonas aeruginosa bloodstream infection. Infection, 38:25–32.

22.

Paterson

D.L.

, Doi

2007. A step closer to extreme drug resistance (XDR) in gram-negative bacilli. Clin. Infect. Dis., 45:1179–1181.

23.

Raymond

D.P.

, Pelletier

S.J.

, Crabtree

T.D.

, Evans

H.L.

, Pruett

T.L.

, Sawyer

R.G.

2003. Impact of antibiotic-resistant Gram-negative bacilli infections on outcome in hospitalized patients. Crit. Care Med., 31:1035–1041.

24.

Suarez

, Pena

, Gavalda

, Tubau

, Manzur

, Dominguez

M.A.

, Pujol

, Gudiol

, Ariza

2010. Influence of carbapenem resistance on mortality and the dynamics of mortality in Pseudomonas aeruginosa bloodstream infection. Int. J. Infect. Dis., 14,Suppl 3:e73–e78.

25.

Suarez

, Pena

, Tubau

, Gavalda

, Manzur

, Dominguez

M.A.

, Pujol

, Gudiol

, Ariza

2009. Clinical impact of imipenem-resistant Pseudomonas aeruginosa bloodstream infections. J. Infect., 58:285–290.

26.

Tam

V.H.

, Gamez

E.A.

, Weston

J.S.

, Gerard

L.N.

, Larocco

M.T.

, Caeiro

J.P.

, Gentry

L.O.

, Garey

K.W.

2008. Outcomes of bacteremia due to Pseudomonas aeruginosa with reduced susceptibility to piperacillin-tazobactam: implications on the appropriateness of the resistance breakpoint. Clin. Infect. Dis., 46:862–867.

27.

Tam

V.H.

, Rogers

C.A.

, Chang

K.T.

, Weston

J.S.

, Caeiro

J.P.

, Garey

K.W.

2010. Impact of multidrug-resistant Pseudomonas aeruginosa bacteremia on patient outcomes. Antimicrob. Agents Chemother., 54:3717–3722.

28.

Troillet

, Samore

M.H.

, Carmeli

1997. Imipenem-resistant Pseudomonas aeruginosa: risk factors and antibiotic susceptibility patterns. Clin. Infect. Dis., 25:1094–1098.