JAC Advance Access published online on October 8, 2007
Journal of Antimicrobial Chemotherapy, doi:10.1093/jac/dkm378
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Clinical significance of healthcare-associated infections in community-onset Escherichia coli bacteraemia
1 Division of Infectious Diseases, Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 ILwon-dong, Gangnam-gu, Seoul 135-710, Republic of Korea 2 Divisions of Infectious Disease, Daegu Fatima Hospital, Daegu, Republic of Korea 3 Divisions of Infectious Disease, Gyeongsang National University Hospital, Jinju, Republic of Korea 4 Department of Molecular Cell Biology, Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 ILwon-dong, Gangnam-gu, Seoul 135-710, Republic of Korea 5 Asian-Pacific Research Foundation for Infectious Diseases (ARFID), Seoul, Republic of Korea 6 Department of Laboratory Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 ILwon-dong, Gangnam-gu, Seoul 135-710, Republic of Korea
* Corresponding author. Tel: +82-2-3410-0329; Fax: +82-2-3410-0041; E-mail: krpeck{at}skku.edu
Received 19 July 2007; returned 15 August 2007; revised 5 September 2007; accepted 6 September 2007
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
Background: The increasing antimicrobial resistance of Escherichia coli infection is of great concern, even for community-onset infections.
Methods: We conducted a retrospective cohort study of patients with E. coli bacteraemia who visited the emergency department of the Samsung Medical Center from February 2002 to December 2005 to identify the risk factors for mortality and association between healthcare-associated (HCA) infection and mortality. We classified community-onset E. coli bacteraemia into community-acquired (CA) and HCA infections.
Results: A total of 508 patients with E. coli bacteraemia were enrolled (mean age, 61.8 ± 14.3 years; male/female, 191:317). The HCA E. coli bacteraemia had significantly higher severity of illness and higher antimicrobial resistance rate than CA bacteraemia. The overall 30-day mortality rate was 13.6% (69/508) and the mortality of HCA infections was significantly higher than that of CA infections (26.4% versus 9.6%, P < 0.001). In multivariate analysis, high Charlson's co-morbidity index (OR 4.84, 95% CI 2.14–10.95, P < 0.001), high Pitt bacteraemia score (OR 32.03, 95% CI 13.08–74.43, P < 0.001), presentation with acute renal failure (OR 4.11, 95% CI 1.90–8.89, P < 0.001) and HCA bacteraemia (OR 2.34, 95% CI 1.09–5.01, P = 0.030) were found to be the significant risk factors for 30-day mortality in E. coli bacteraemia.
Conclusions: The mortality rate of HCA E. coli bacteraemia was higher than twice that of CA bacteraemia and HCA bacteraemia was one of the significant risk factors for mortality, even after adjusting for a large number of potential confounders.
Key Words: mortality , antimicrobial resistance , inappropriate therapy
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
Escherichia coli is a normal colonizer of the human gastrointestinal tract and a common cause of bacteraemia.1–3 Bacteraemia caused by Gram-negative organisms is a cause of significant morbidity and mortality. Gram-negative bacteria such as E. coli are common precipitants of sepsis and septic shock by virtue of the inflammatory response activated by endotoxin (lipopolysaccharide) present in the Gram-negative cell wall.1
Many recent studies have focused on antimicrobial resistance (especially to ciprofloxacin) of E. coli.4–7 The initiation of an antimicrobial agent is almost always empirical, requiring knowledge of the likely pathogen and usual antimicrobial susceptibility patterns. Resistance among E. coli has become increasingly common, making empirical therapy decisions more difficult, even in community-onset infections. However, there have been few reports about influences of antimicrobial resistance associated with mortality of E. coli bacteraemia.8 Furthermore, no studies had been conducted to investigate E. coli bacteraemia in community settings. The prevalence of antimicrobial resistance in E. coli bacteraemia is closely associated with exposure to the healthcare system.
A healthcare-associated (HCA) status is a recently described epidemiological category that is distinct from both community-acquired (CA) and nosocomial status.9 Friedman et al.10 have described the epidemiology, microbiology and outcomes of a cohort of patients with HCA bacteraemia. However, no previously published study has shown the differences between CA and HCA E. coli bacteraemia.
Thus, we aimed to delineate the differences in clinical manifestations of E. coli bacteraemia and the antimicrobial resistance pattern of E. coli between CA and HCA settings and we investigated further to identify the risk factors for mortality.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
Study population
The database at our Clinical Microbiology Laboratory was reviewed in order to identify patients with E. coli bacteraemia. Patients were included in the study if their blood cultures were drawn in the emergency department (ER) within 48 h of admission and the culture results were positive for E. coli. The patients with polymicrobial bloodstream infection were excluded. A review of the medical records of individuals diagnosed from February 2002 to December 2005 at Samsung Medical Center, Republic of Seoul, Korea, a 1350 bed tertiary care university hospital, identified 634 patients with community-onset E. coli bacteraemia. The annual ER census in our hospital is 
55 000 visits. Only the first bacteraemic episode for each patient was included in the analysis.
Study design and data collection
A retrospective observational cohort study was conducted. We compared data from patients with HCA bacteraemia with data from patients with CA bacteraemia.
We reviewed the medical records of the patients who were >15 years old. The data collected included: age, gender, underlying disease, site of infection, severity of illness (as calculated by Pitt bacteraemia score and Charlson's weighted index of morbidity)11,12 and antimicrobial regimen. The presence of the following co-morbid conditions was also documented: neutropenia, presentation with septic shock, the receipt of immunosuppressive agents within 30 days prior to presentation with bacteraemia and the presence of a central venous catheter. As this study was retrospective, the physicians of the patients, not researchers, had chosen the antimicrobial therapy regimens. The main outcome measure used was the 30-day mortality rate.
E. coli bacteraemia was defined as a growth of E. coli in a blood culture specimen. Clinically significant bacteraemia was defined as at least one positive blood culture, together with clinical features compatible with systemic inflammatory response syndrome.
Community-onset bacteraemia was defined when it occurred within 48 h of admission and was classified into HCA and CA bacteraemia. If the patient fulfilled any of the following criteria, they were categorized as HCA:9,10 (i) received intravenous therapy at home or in an outpatient clinic in the previous 30 days; or (ii) received renal dialysis in a hospital or clinic in the previous 30 days; or (iii) had been hospitalized for 2 or more days in the previous 90 days; or (iv) resided in a nursing home or long-term care facility for 2 or more days in the previous 90 days.
Bacteraemia was classified as CA when it did not fit criteria for HCA bacteraemia.
Neutropenia was defined as an absolute neutrophil count below 500/mm3. Septic shock was defined as sepsis associated with evidence of organ hypoperfusion and a systolic blood pressure <90 or >30 mmHg less than the baseline or a requirement for the use of a vasopressor to maintain the blood pressure. The initial empirical antimicrobial therapy was considered ‘appropriate’ if the initial antibiotics, which were administered within 24 h after acquisition of blood culture samples, included at least one antibiotic that was active in vitro and when the dosage and route of administration confirmed with current medical standards. Otherwise, initial antimicrobial therapy was considered ‘inappropriate’.13
The recovery of E. coli isolates from blood was accomplished by the processing of blood cultures in a BACTEC Model 9240 (BD Diagnostic Instrument Systems, Sparks, MD, USA) or BacT/ALERT 3D (bioMerieux, Inc., Hazelwood, MO, USA). Identification of E. coli was carried out using a standard identification card. Antibiotic susceptibility testing of E. coli isolates was performed on the VITEK II automated system (bioMerieux, Inc.) using the modified broth microdilution method. MIC breakpoints and quality control protocols were used according to standards established by the CLSI.14 A multidrug-resistant (MDR) strain was defined if the microorganism was not susceptible to three or more of the following antimicrobial agents: ampicillin, amikacin or gentamicin or tobramycin, aztreonam, cefotaxime or ceftriaxone or ceftazidime or cefepime, ciprofloxacin, imipenem, piperacillin/tazobactam and trimethoprim/sulfamethoxazole.
Student's t-test was used to compare continuous variables, and 
2 or Fisher's exact test was used to compare categorical variables. In identifying the risk factors for mortality, a stepwise logistic regression analysis was used to control for the effects of confounding variables. Variables with a P value <0.05 in the univariate analysis 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 significance cut-off of 0.05. The Kaplan–Meier method was used for survival analysis. All P values were two-tailed, and P values <0.05 were considered statistically significant. SPSS for Windows (version 11.5 software package; SPSS Inc., Chicago, IL, USA) was used for this analysis.
Ethics approval was not required for this study.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
Study population
During the study period, a total of 634 patients were identified. Among these patients, a total of 508 patients with E. coli bacteraemia were enrolled. One hundred and twenty-six (19.9%) patients were excluded from the analysis because 42 patients were aged <15 years, 38 patients were lost to follow-up after discharge, 32 patients transferred to another hospital and 14 patients had incomplete medical records or unevaluable medical records.
The mean (±SD) age of patients was 61.8 ± 14.3 years (range, 19–91) and 317 (62.4%) patients were female. The demographic and clinical features are described in Table 1. As for the primary sites of infection, the most common sites were the urinary tract (n = 238; 46.9%) and hepatobiliary tract (n = 138; 27.2%). A total of 123 (24.2%) patients presented with septic shock, and 37 (7.3%) patients were neutropenic. Of the 508 patients, 121 (23.8%) patients were classified as having HCA bacteraemia. The remaining 387 (76.2%) patients were classified as having CA bacteraemia.
|
CA versus HCA E. coli bacteraemia
The clinical characteristics of HCA E. coli bacteraemia were compared with those of CA bacteraemia (Table 1). As for the underlying disease, solid tumour and haematological malignancy were more prevalent in the HCA group. When the primary sites of infection were evaluated, urinary tract infection and hepatobiliary infection were more common in the CA group (52.5% versus 28.9%, P < 0.001; 29.7% versus 19.0%, P = 0.021, respectively). In the HCA group, Charlson's weighted index of co-morbidity and the Pitt bacteraemia score were significantly higher than those in the CA group.
Among CA bacteraemia cases, 3.2% (10/314) and 23.3% (90/387) were caused by extended-spectrum ß-lactamase (ESBL)-producing isolates and ciprofloxacin-resistant isolates, respectively. In contrast, among HCA bacteraemia cases, 6.7% (7/105) and 37.2% (45/121) were caused by ESBL-producing isolates and ciprofloxacin-resistant isolates, respectively (Table 2).
|
30-day mortality and predictors of mortality
The overall 30-day mortality rate of E. coli bacteraemia was 13.6% (69/508) and the mortality of HCA infections was significantly higher than that of CA infections [26.4% (32/121) versus 9.6% (37/387), P < 0.001]. The survival curve also shows that the HCA group had a significantly lower probability of survival than the CA group (Figure 1).
|
Factors associated with 30-day mortality are shown in Tables 3 and 4. By univariate analysis, urinary tract infection was the negative predictive factor of 30-day mortality. As for individual antimicrobial resistance, no significant risk factors were found. However, MDR E. coli was a statistically significant risk factor. ‘Inappropriate’ initial antimicrobial therapy showed a trend of high mortality although it did not reach statistical significance as a risk factor for mortality in univariate analysis.
|
|
Multivariate analysis using a logistic regression model including the variables associated with mortality by univariate analysis (P < 0.05) showed that the significant independent risk factors for mortality were a high Charlson's weighted index of co-morbidity, high Pitt bacteraemia score, presentation with acute renal failure and HCA bacteraemia. In addition, urinary tract infection was found to be a protective factor for mortality based on multivariate analysis (Table 4).
When HCA infection was excluded in the multivariate analysis, MDR E. coli (OR 2.87, 95% CI 1.32–6.24, P = 0.008) and initial inappropriate antimicrobial therapy (OR 4.23, 95% CI 1.42–12.62, P = 0.010) were the significant risk factors for 30-day mortality of E. coli bacteraemia.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
This study was undertaken to evaluate the differences in clinical manifestations of E. coli bacteraemia and the antimicrobial resistance pattern of E. coli between CA and HCA settings. We found that HCA E. coli bacteraemia had a higher severity of illness on admission as well as a higher mortality rate. To our knowledge, this is the first study that compared the clinical characteristics and outcome of patients with CA E. coli bacteraemia with those of HCA E. coli bacteraemia. Our results suggest that a history of hospitalization or HCA exposure may influence the choice of empirical antibiotics for treating patients with suspected E. coli bacteraemia. Previous studies support the above ideas.10,15 Friedman et al.10 showed that the characteristics of HCA bacteraemia closely resemble those of nosocomial infections.
As the underlying illnesses in the HCA group were more severe (as determined by Charlson's weighted index of morbidity) than in the CA group, it may be presumed that HCA bacteraemia may have a worse prognosis because of the greater severity of underlying illness. However, after adjusting for other prognostic factors associated with mortality, HCA infection was identified as an independent risk factor for mortality. Not surprisingly, the severity of underlying illness and an initial inappropriate antimicrobial therapy were prognostic factors of mortality. Inappropriate antimicrobial therapy had significantly higher mortality than appropriate therapy by multivariate analysis (healthcare association excluded) after adjusting for confounding variables although statistical significance was not reached by univariate analysis. These findings have important implications because an initial choice of antimicrobial therapy should be changed according to the condition of the patient at admission.
In our study, it was noted that resistances to individual antimicrobials (especially ciprofloxacin) were not associated with mortality. There have been few reports on the mortality rates of patients with quinolone-resistant E. coli bacteraemia compared with quinolone-susceptible E. coli bacteraemia. Cheong et al.8 reported that there was no statistical difference in the mortality rate between the patients with quinolone-resistant and quinolone-susceptible E. coli bacteraemia and suggested that quinolone-resistant strains seem not to be more virulent than quinolone-susceptible strains. Patients infected with ESBL-producing E. coli tended to be associated with higher mortality in our study (7.2% versus 2.5%, P = 0.053). The question of whether ESBL production significantly increases the risk of death remains unresolved.16 However, a number of studies observed that patients with infection caused by ESBL-producing bacilli tended to have poorer outcomes than did comparable patients with infection caused by pathogens that did not produce ESBLs.17,18 However, in this study, MDR E. coli was associated with 30-day mortality based on multivariate analysis when the healthcare association was excluded in the multivariate analysis.
Our data suggest that HCA infection may have higher mortality through multidrug resistance and consequent inappropriate therapy. Previous studies have identified an important association between inappropriate therapy and hospital mortality in different groups of patients.13,19,20
The main limitation of the present investigation is that it was observational and, thus, unknown risk factors for mortality might have been unequally distributed between the two groups. However, although the evidence of a statistical association between HCA bacteraemia and mortality does not prove causality, our data suggest that mortality in patients with HCA bacteraemia cannot be attributed to co-morbid conditions, the severity of illness or baseline organ dysfunction. In addition, as this study was of a retrospective nature, the possibility of the limitations in precluding accurate comparisons should be kept in mind. Actually, 19.9% of patients with community-onset E. coli bacteraemia were excluded during review of medical records and some selection bias may have occurred during this review process. And not all antimicrobials were tested against all isolates because all isolates were not tested using the same antimicrobial susceptibility test card. Therefore, there are potential limitations in interpretation of antimicrobial resistance and its impact. Finally, our study was conducted in a large tertiary care medical centre, and thus many of our patients had serious underlying illnesses, including neoplastic diseases and chronic liver diseases. Therefore, these data might not be generalizable in other institutes.
In conclusion, this investigation shows that HCA E. coli bacteraemia in community-onset bacteraemia is associated with a higher severity of illness and antimicrobial resistance rate than CA E. coli bacteraemia. HCA bacteraemia was found to be one of the independent risk factors for mortality.
|
Funding |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
None.
|
Transparency declarations |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
None to declare.
|
Acknowledgements |
|---|
This study was presented in part at the Forty-sixth Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, CA, USA, 2006 (Abstract 3046).
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
1 . Diekema DJ, Pfaller MA, Jones RN, et al. Survey of bloodstream infections due to gram-negative bacilli: frequency of occurrence and antimicrobial susceptibility of isolates collected in the United States, Canada, and Latin America for the SENTRY Antimicrobial Surveillance Program, 1997. Clin Infect Dis (1999) 29:595–607.[Web of Science][Medline]
2 . Weinstein MP, Towns ML, Quartey SM, et al. The clinical significance of positive blood cultures in the 1990s: a prospective comprehensive evaluation of the microbiology, epidemiology, and outcome of bacteremia and fungemia in adults. Clin Infect Dis (1997) 24:584–602.[Web of Science][Medline]
3 . Diekema DJ, Pfaller MA, Jones RN. Age-related trends in pathogen frequency and antimicrobial susceptibility of bloodstream isolates in North America: SENTRY Antimicrobial Surveillance Program, 1997–2000. Int J Antimicrob Agents (2002) 20:412–8.[CrossRef][Web of Science][Medline]
4
.
Kern WV, Andriof E, Oethinger M, et al. Emergence of fluoroquinolone-resistant Escherichia coli at a cancer center. Antimicrob Agents Chemother (1994) 38:681–7.
5 . Lautenbach E, Strom BL, Bilker WB, et al. Epidemiological investigation of fluoroquinolone resistance in infections due to extended-spectrum ß-lactamase-producing Escherichia coli and Klebsiella pneumoniae. Clin Infect Dis (2001) 33:1288–94.[CrossRef][Web of Science][Medline]
6
.
Garau J, Xercavins M, Rodriguez-Carballeira M, et al. Emergence and dissemination of quinolone-resistant Escherichia coli in the community. Antimicrob Agents Chemother (1999) 43:2736–41.
7
.
Pena C, Albareda JM, Pallares R, et al. Relationship between quinolone use and emergence of ciprofloxacin-resistant Escherichia coli in bloodstream infections. Antimicrob Agents Chemother (1995) 39:520–4.
8 . Cheong HJ, Yoo CW, Sohn JW, et al. Bacteremia due to quinolone-resistant Escherichia coli in a teaching hospital in South Korea. Clin Infect Dis (2001) 33:48–53.[CrossRef][Web of Science][Medline]
9
.
McDonald JR, Friedman ND, Stout JE, et al. Risk factors for ineffective therapy in patients with bloodstream infection. Arch Intern Med (2005) 165:308–13.
10
.
Friedman ND, Kaye KS, Stout JE, et al. Health care-associated bloodstream infections in adults: a reason to change the accepted definition of community-acquired infections. Ann Intern Med (2002) 137:791–7.
11 . Chow JW, Yu VL. Combination antibiotic therapy versus monotherapy for gram-negative bacteraemia: a commentary. Int J Antimicrob Agents (1999) 11:7–12.[CrossRef][Web of Science][Medline]
12 . Lesens O, Methlin C, Hansmann Y, et al. Role of comorbidity in mortality related to Staphylococcus aureus bacteremia: a prospective study using the Charlson weighted index of comorbidity. Infect Control Hosp Epidemiol (2003) 24:890–6.[CrossRef][Web of Science][Medline]
13
.
Kang CI, Kim SH, Park WB, et al. 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 (2005) 49:760–6.
14 . Clinical and Laboratory Standards Institute. In: Performance Standards for Antimicrobial Susceptibility Testing: Fifteenth Informational Supplement M100-S15 (2005) Wayne, PA, USA: CLSI.
15 . Sun HY, Chen SY, Chang SC, et al. Community-onset Escherichia coli and Klebsiella pneumoniae bacteremia: influence of health care exposure on antimicrobial susceptibility. Diagn Microbiol Infect Dis (2006) 55:135–41.[CrossRef][Web of Science][Medline]
16 . Ramphal R, Ambrose PG. Extended-spectrum ß-lactamases and clinical outcomes: current data. Clin Infect Dis (2006) 42(Suppl 4):S164–72.[CrossRef][Web of Science][Medline]
17
.
Kim YK, Pai H, Lee HJ, et al. Bloodstream infections by extended-spectrum ß-lactamase-producing Escherichia coli and Klebsiella pneumoniae in children: epidemiology and clinical outcome. Antimicrob Agents Chemother (2002) 46:1481–91.
18 . Ho PL, Chan WM, Tsang KW, et al. Bacteremia caused by Escherichia coli producing extended-spectrum ß-lactamase: a case-control study of risk factors and outcomes. Scand J Infect Dis (2002) 34:567–73.[CrossRef][Web of Science][Medline]
19 . Zaragoza R, Artero A, Camarena JJ, et al. The influence of inadequate empirical antimicrobial treatment on patients with bloodstream infections in an intensive care unit. Clin Microbiol Infect (2003) 9:412–8.[CrossRef][Web of Science][Medline]
20 . Valles J, Rello J, Ochagavia A, et al. Community-acquired bloodstream infection in critically ill adult patients: impact of shock and inappropriate antibiotic therapy on survival. Chest (2003) 123:1615–24.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||