Clin Microbiol Infect 2011; 17: 539–545
Enterobacter species are the fourth most common cause of Gram-negative bloodstream infection (BSI). We examined temporal changes and seasonal variation in the incidence rate of Enterobacter spp. BSI, estimated 28-day and 1-year mortality, and determined in vitro antimicrobial resistance rates of Enterobacter spp. bloodstream isolates in Olmsted County, Minnesota, from 1 January 1998 to 31 December 2007. Multivariable Poisson regression was used to examine temporal changes and seasonal variation in incidence rate and Kaplan–Meier method was used to estimate 28-day and 1-year mortality. The median age of patients with Enterobacter spp. BSI was 58 years and 53% were female. The overall age- and gender-adjusted incidence rate of Enterobacter spp. BSI was 3.3 per 100 000 person-years (95% CI 2.3–4.4). There was a linear trend of increasing incidence rate from 0.8 (95% CI 0–1.9) to 6.2 (95% CI 3.0–9.3) per 100 000 person-years between 1998 and 2007 (p 0.002). There was no significant difference in the incidence rate of Enterobacter spp. BSI during the warmest 4 months compared to the remainder of the year (incidence rate ratio 1.06; 95% CI 0.47–2.01). The overall 28-day and 1-year mortality rates of Enterobacter spp. BSI were 21% (95% CI 8–34%) and 38% (95% CI 22–53%), respectively. Up to 13% of Enterobacter spp. bloodstream isolates were resistant to third-generation cephalosporins. To our knowledge, this is the first population-based study to describe the epidemiology and outcome of Enterobacter spp. BSI. The increase in incidence rate of Enterobacter spp. BSI over the past decade, coupled with its associated antimicrobial resistance, dictate the need for further investigation of this syndrome.
Enterobacter species are the fourth most common cause of Gram-negative bloodstream infection (BSI) [1–3]. Population-based studies that specifically address the epidemiology, outcome and in vitro antimicrobial resistance rates of Enterobacter spp. BSI are lacking and most surveys that have been published have been derived from referral tertiary care centres [4–10]. Therefore, we performed a population-based study to determine the incidence rate and examine temporal changes in the incidence rate of Enterobacter spp. BSI. In addition, we examined seasonal variation in incidence rate of Enterobacter spp. BSI because recent reports have demonstrated seasonal variation in both Escherichia coli  and Klebsiella pneumoniae BSI  and other syndromes of infections caused by Gram-negative bacilli . We also estimated the 28-day and 1-year mortality rates in patients with Enterobacter spp. BSI. Lastly, we examined the in vitro antimicrobial resistance rates of Enterobacter spp. bloodstream isolates in Olmsted County, Minnesota, from 1998 to 2007.
Materials and Methods
Olmsted County is located in southeastern Minnesota and has a population of 124 277 according to the 2000 census (US Census Bureau, Olmsted County QuickFacts; http://quickfacts.census.gov, accessed 21 April 2008). With the exception of a lower prevalence of injection drug use, a higher prevalence of middle-class individuals and a higher proportion being employed in the healthcare industry, the population characteristics of Olmsted County residents are similar to those of USA non-Hispanic whites [14,15]. The Rochester Epidemiology Project (REP) is a unique medical records-linkage system that encompasses care delivered to residents of Olmsted County, Minnesota. The microbiology laboratories at Mayo Medical Center and Olmsted Medical Center are the only two laboratories in Olmsted County. These two medical centres are geographically isolated from other urban centres, as previously described [14,16,17]; therefore, local residents are able to obtain healthcare within the community, rather than seeking healthcare at a distant geographic location.
We used complete enumeration of Olmsted County, Minnesota, from 1 January 1998 to 31 December 2007. Using the microbiology databases at the Mayo Medical Center Rochester and Olmsted Medical Center, we identified 38 unique patients with first episodes of monomicrobial Enterobacter spp. BSI during the study period. Medical records were reviewed by the primary investigator (M.N.A.) to confirm the diagnosis, determine patient residency status, and obtain baseline clinical features and outcome.
Blood cultures were processed using standard microbiology techniques according to the CLSI. Both laboratories are certified by the College of American Pathologists. CLSI methods were employed to evaluate in vitro antimicrobial susceptibility results of Enterobacter spp. bloodstream isolates. The study was approved by the institutional review boards of both institutions. The detailed case ascertainment and blood culture methods used have been described previously [17–19].
Monomicrobial Enterobacter spp. BSI was defined as the growth of only Enterobacter spp. in a blood culture, excluding coagulase-negative staphylococci, Corynebacterium spp. and Propionibacterium spp. Cases were classified according to the site of acquisition into nosocomial, healthcare-associated and community-acquired . The primary source of BSI was defined using the Centers for Disease Control and Prevention criteria .
Descriptive statistics were used to summarize the data: medians and interquartile range (IQR) for continuous variables and counts and percentages for categorical variables. Fisher’s exact test was used to evaluate associations between categorical variables and Wicoxon rank-sum test was used to test for differences in medians across continuous variables.
The incidence rate, expressed as the number of new cases of Enterobacter spp. BSI per 100 000 person-years, was calculated assuming that the entire population of Olmsted County was at risk of BSI. The 2000 Olmsted County census figures were used to compute the age-, gender- and calendar year-specific person-years denominator with a projected population growth rate after 2000 of 1.9% per year. The 10-year study period was divided into five 2-year intervals (1998–1999, 2000–2001, 2002–2003, 2004–2005 and 2006–2007) and age was divided into five groups (0–18, 19–39, 40–59, 60–79 and ≥80 years). The incidence rate was directly adjusted to the USA 2000 white population. A 95% CI for each incidence rate was estimated using a Poisson distribution.
To evaluate the association between seasonal variation and incidence rate of Enterobacter spp. BSI, the age- and gender-adjusted incidence rate was calculated for both the four warmest months (June to September) and the remaining 8 months; the person-years denominator was multiplied by 1/3 and 2/3, respectively. The incidence rate ratio is the ratio of the incidence rate for the four warmest months relative to the incidence rate for the remaining 8 months. A 95% CI for the incidence rate ratio was constructed using bootstrap resampling.
To create an additional measure of seasonal variation, the average monthly temperatures for Rochester, Minnesota, were obtained from historic city records (Weatherbase Historical Weather for Rochester, MN, USA; http://www.weatherbase.com, accessed 24 July 2008). Incidence rates were calculated for each of the 12 months assuming a fixed population within a given year. To test for an association between average monthly temperature and the incidence rate of Enterobacter spp. BSI at the same time as adjusting for gender, age and calendar year, a multivariable Poisson regression model was used.
The Kaplan–Meier method was used to estimate the 28-day and 1-year all-cause mortality rates. Patients were followed from the date of first episode of Enterobacter spp. BSI until death or last healthcare encounter; long-term follow-up was available through the REP. Patients lost to follow-up were censored on the date of their last healthcare encounter. All analyses were carried out using the sas software, version 8.2 (SAS Institute, Cary, NC, USA). p <0.05 (two-sided) was considered statistically significant.
We identified 38 unique patients with Enterobacter spp. BSI during the study period; 26 had Enterobacter cloacae, ten had Enterobacter aerogenes and two had Enterobacter sakazakii BSI. The median age of patients with Enterobacter spp. BSI was 58 years (IQR 45–75) and 20 (53%) were female. Most cases were healthcare-associated (58%) or nosocomial (21%); the remaining 21% of cases were community-acquired. The urinary tract was the most common identified primary source of infection (24%), followed by the gastrointestinal tract (18%), central venous catheter-related cases (11%), the respiratory tract (5%), skin and soft tissue (5%), bone and joint (3%), and the central nervous system (3%). Twelve patients (32%) had primary BSI of unknown source.
The overall age- and gender-adjusted incidence rate of Enterobacter spp. BSI was 3.3 (95% CI 2.3–4.4) per 100 000 person-years. The age-adjusted incidence rate of Enterobacter spp. BSI per 100 000 person-years was 3.9 (95% CI 2.0–5.8) in males and 3.2 (95% CI 1.8–4.6) in females. The incidence rate of Enterobacter spp. BSI increased linearly with age (p 0.007) (Fig. 1). After adjusting for gender and age, there was a linear increase in the incidence rate of Enterobacter spp. BSI from 1998 to 2007 (p 0.002) (Fig. 2).
The age- and gender-adjusted incidence rate of Enterobacter spp. BSI per 100 000 person-years was 3.5 (95% CI 1.6–5.4) during the warmest 4 months of the year (June to September) compared to 3.3 (95% CI 2.0–4.5) during the remainder of the year (incidence rate ratio 1.07; 95% CI 0.47–2.01). Additionally, there was no association between the incidence rate of Enterobacter spp. BSI and average temperature (p 0.83) (Fig. 3).
Patients with E. aerogenes BSI were older than those with E. cloacae BSI (median age 74 vs. 52 years; p 0.008) (Table 1). Additionally, patients with E. aerogenes BSI were more likely than those with E. cloacae BSI to have a community-acquired site of infection acquisition (60% vs. 8%; p 0.002) and a urinary tract primary source of infection (50% vs. 15%; p 0.08).
(n = 26)
|Enterobacter aerogenes |
(n = 10)
|Age, median (IQR)||52 (39–71)||74 (61–82)|
|Female sex, n (%)||15 (58)||4 (40)|
|Site of acquisition, n (%)|
|Community-acquired||2 (8)||6 (60)|
|Healthcare-associated||17 (65)||3 (30)|
|Nosocomial||7 (27)||1 (10)|
|Primary source, n (%)|
|Urinary tract||4 (15)||5 (50)|
|Gastrointestinal tract||5 (19)||1 (10)|
|Central venous catheter-related||4 (15)||0 (0)|
|Other||3 (12)||2 (20)|
|Unknown||10 (38)||2 (20)|
The age- and gender-adjusted incidence rates of E. cloacae and E. aerogenes BSI per 100 000 person-years were 2.2 (95% CI 1.4–3.1) and 0.9 (95% CI 0.4–1.5), respectively. The age-adjusted incidence rates of E. cloacae BSI were comparable in males and females [2.2 (95% CI 0.8–3.5) and 2.4 (95% CI 1.2–3.7) per 100 000 person-years, respectively]. The incidence rates of E. aerogenes BSI for males and females were 1.6 (95% CI 0.3–2.8) and 0.6 (95% CI 0–1.3) per 100 000 person-years, respectively.
Complete patient follow-up was obtained for most of the cohort; no patient was lost to follow-up within 28 days and only 3 (8%) were lost to follow-up within 1 year of Enterobacter spp. BSI. The overall 28-day and 1-year all-cause mortality rates of Enterobacter spp. BSI were 21% (95% CI 8–34%) and 38% (95% CI 22–53%), respectively. Although the overall 28-day all-cause mortality rate was relatively higher in patients with E. aerogenes as compared to those with E. cloacae BSI [30% (95% CI 2–58%) vs. 15% (95% CI 2–29%)] (Fig. 4a), there was no apparent difference in the long-term survival (Fig. 4b).
The in vitro antimicrobial susceptibility rates of Enterobacter spp. bloodstream isolates to all tested antimicrobial agents are shown in Table 2. Among all tested β-lactams, ceftazidime and piperacillin-tazobactam had the lowest in vitro susceptibility rates (87% and 89%, respectively). No cefepime- or carbapenem-resistant Enterobacter spp. bloodstream isolates were detected in our population over the past decade.
|Antimicrobial||Number of susceptible isolates/number of isolates tested||Susceptibility (%)|
To our knowledge, this is the first population-based study to describe the epidemiology, outcome and in vitro antimicrobial resistance rates of Enterobacter spp. BSI. We demonstrated a linear trend of an increasing incidence rate of Enterobacter spp. BSI during the past decade. The age- and gender-adjusted incidence rate of Enterobacter spp. BSI increased from 0.8 (95% CI 0–1.9) to 6.2 (95% CI 3.0–9.3) per 100 000 person-years between 1998 and 2007. There was no seasonal variation in the incidence rate of Enterobacter spp. BSI.
Based on an age- and gender-adjusted incidence rate of 2.2 (95% CI 1.4–3.1) per 100 000 person-years, E. cloacae was the fourth most common Gram-negative bacillus to cause BSI in our population; E. coli, K. pneumoniae and Pseudomonas aeruginosa [incidence rates of 41.4 (95% CI 37.6–45.3), 9.7 (95% CI 7.8–11.6) and 4.7 (95% CI 3.4–6.1) per 100 000 person-years, respectively] were the three most common causes of BSI due to Gram-negative bacilli, as previously described [11,17,22]. The incidence rate of BSI due to all four organisms increased with age. The median age of 52 years in patients with E. cloacae BSI was notably lower than that in patients with E. coli, K. pneumoniae and P. aeruginosa BSI (69, 73 and 69 years, respectively). By contrast to E. coli BSI that was more common in females [11,23], and K. pneumoniae BSI and P. aeruginosa BSI that were more common in males [17,22,24,25], the incidence rate of E. cloacae BSI was not influenced by gender. Most cases of E. cloacae BSI were healthcare-associated or nosocomial, and almost one-third had no known primary source of infection, which is consistent with previous studies [5,9,10,26].
E. aerogenes is an uncommon cause of Gram-negative BSI with an age- and gender-adjusted incidence rate of 0.9 (95% CI 0.4–1.5) per 100 000 person-years. Although the incidence rate of E. aerogenes BSI in males was nearly three-fold greater than that in females, the difference was not statistically significant as a result of the small number of patients with E. aerogenes BSI in our population. Patients with E. aerogenes were older and more likely to have a urinary tract primary source of infection than those with E. cloacae BSI, which is consistent with the results of a recent study .
Our finding of an increasing incidence rate of Enterobacter spp. BSI over the past decade is consistent with the results of a recent investigation that demonstrated increasing frequency of Enterobacter spp. as a cause of BSI in a tertiary care centre in Spain from 1991 to 2006 . It is also consistent with observations of earlier paediatric studies that suggested the emergence of Enterobacter spp. as an important cause of BSI in children [28,29]. The increase in incidence rate of Enterobacter spp. BSI in our population is unlikely to be a result of changes in physicians’ practices in obtaining blood cultures in febrile patients because the incidence rates of BSI due to E. coli, Klebsiella spp., and other Gram-negative bacilli have remained stable over the same time period [11,19,22]. Because most cases of Enterobacter spp. BSI were healthcare-associated or nosocomial, it is likely that changes within the local hospitals’ environment have resulted in this increase in incidence rate. It is conceivable that an increasing number of patients are colonized with Enterobacter spp. after hospital admission or contact with the healthcare setting for an outpatient procedure such as urinary tract instrumentation, haemodialysis or outpatient chemotherapy . In addition, some studies have identified previous exposure to broad-spectrum antimicrobial agents, particularly third-generation cephalosporins, as a risk factor for Enterobacter spp. BSI [4,9,31]. It is possible that the increasing use of these antimicrobial agents in our local population is temporally associated with the increase in the incidence rate of Enterobacter spp. BSI, although these data were not collected to permit that determination.
Our observation of an increasing incidence rate of Enterobacter spp. BSI deserves additional study. Subsequent work should determine whether the increase in incidence rate of Enterobacter spp. BSI that was seen locally is also present in other locales. The development of a multi-national population-based study to examine temporal trends in Enterobacter spp. BSI with molecular testing of bloodstream isolates would provide vital clinical and microbiological information to advance this field.
Enterobacter spp. is more likely than other commonly isolated Gram-negative bacilli, such as E. coli and Klebsiella spp., to be resistant to antimicrobial agents. For example, 13% of Enterobacter spp. bloodstream isolates in our local area were not susceptible to third-generation cephalosporins compared to only 1% of both E. coli  and Klebsiella spp. bloodstream isolates . The increased resistance is likely due to the inherent capability of Enterobacter spp. isolates to produce inducible chromosomal β-lactamases, including Amp C. Even when the isolates demonstrate in vitro antimicrobial susceptibility, exposure to certain types of β-lactam antibiotics, including cephalosporins and extended-spectrum penicillins, can cause a transient increase or induction of Amp C production, resulting in antimicrobial resistance and possibly a predisposition to treatment failures [31,33,34]. It has been demonstrated that 15–19% of Enterobacter spp. isolates develop resistance to third-generation cephalosporins during treatment [4,35]. Therefore, this increase in the incidence rate of Enterobacter spp. BSI makes the choice of an empiric antimicrobial regimen in patients with Gram-negative BSI, when awaiting identification of the Gram-negative bacillus, much more difficult, especially in patients with recent contact with the healthcare system and other identifiable risk factors for Enterobacter spp. BSI.
The lack of seasonal variation in the incidence rate of Enterobacter spp. BSI in the present study is consistent with the results of a large study from four continents . However, this is in contrast to our previous work that demonstrated a higher incidence rate of E. coli BSI during the warmest 4 months than in the reminder of the year . The factors involved in the presence or absence of seasonal variation in BSI have yet to be identified.
Patients with E. cloacae BSI were younger than those with E. aerogenes BSI in the present study and it is conceivable that the younger age accounted for the lower mortality rate in the former group. The relatively small sample size did not permit an examination of predictors of mortality in a multivariable model to confirm this notion. The 28-day all-cause mortality rate of 15% in patients with E. cloacae BSI was notably lower than the short-term mortality rates (21–69%) reported in studies extending back to the 1980s [6,8–10,26,36,37]. In more recent studies, mortality rates of 13–15% have been described and are consistent with our results [7,27]. We speculate that this decline in mortality could be due to advancements in critical care and antimicrobial management.
The unique availability of long-term patient follow-up through the REP resources in our population permitted an estimation of 1-year all-cause mortality rate subsequent to Enterobacter spp. BSI that has not been previously described. Over one-third of patients did not survive beyond 1 year after Enterobacter spp. BSI, and was most likely a result of the multiple comorbid conditions that characterized these patients. This observation was similar to that previously observed in patients with Klebsiella spp. BSI .
The in vitro antimicrobial resistance rates of Enterobacter spp. bloodstream isolates in our population-based study were notably lower than those reported previously from tertiary care centres. Resistance rate to third-generation cephalosporins was 13% in our study compared to 17–51% in hospital-based investigations [2,27,38–40]. Similarly, the 5% resistance rate to fluoroquinolones was also lower than previously reported rates of 8–14% from tertiary care centres [2,27,38–41]. Carbapenem-resistance rates among Enterobacter spp. isolates were under 1% in both population- and hospital-based studies [27,38,40]. The lower antimicrobial resistance rates reported from population-based studies compared to investigations from tertiary care centres were likely the result of a referral bias that can adversely affect isolate susceptibility data in tertiary care centres . It is conceivable that referral patients underwent more procedures, antimicrobial exposure and complications that prompted transfer to tertiary care centres and predisposed to colonization or infection with bacteria that harboured antimicrobial resistance.
The strength of the present study is the population-based design and, therefore, the lack of referral bias. By contrast to previous hospital-based studies that have estimated the incidence rate of Enterobacter spp. BSI per the number of admissions to a particular hospital, we determined the incidence rate by 100 000 person-years in a well-defined population.
The present study has limitations. First, the data are derived from one geographic area. Studies from multiple geographic locations may provide a more comprehensive view. Second, we did not perform pulse-field gel electrophoresis to investigate whether the increasing incidence rate of Enterobacter spp. BSI was a result of clonal spread of a single strain in the medical facilities in our region or patient-unique strains. For example, the peak in incidence rate of Enterobacter spp. BSI observed in the 2006–2007 interval might have been related to a nosocomial outbreak rather than reflect a true increase in incidence rate of Enterobacter spp. BSI among the general population. Third, the lack of seasonal variation in Enterobacter spp. BSI might have been the result of a lack of power because of the relatively small sample size. Finally, the population of Olmsted County consists mainly of middle class whites; therefore, the results obtained in the present study may be generalized only to communities with similar population characteristics.
In summary, Enterobacter spp. has emerged as an important cause of Gram-negative BSI. The increase in incidence rate of Enterobacter spp. BSI over the past 10 years should prompt physicians to develop a higher suspicion for Enterobacter spp. in patients with suspect or proven BSI. Empiric antimicrobial coverage should include coverage for these microorganisms, particularly in patients with frequent contact with the healthcare setting. For as yet unexplained reasons, the more recent mortality rate in patients with Enterobacter spp. BSI has declined.
The authors thank Emily Vetter and Mary Ann Butler for providing us with vital data from the microbiology laboratory databases at the Mayo Clinic, Rochester and Olmsted Medical Center. The authors thank Susan Schrage, Susan Stotz, RN, and all the staff at the Rochester Epidemiology Project for their administrative help and support.
M.N.A. and B.D.L. have full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. The study received funding from the Small Grants Program and the Baddour Family Fund at the Mayo Clinic, Rochester, MN. The funding source had no role in study design. This work was made possible by research grant R01-AR30582 from the National Institute of Arthritis and Musculoskeletal and Skin Diseases (National Institutes of Health, US Public Health Service). All the authors declare that there are no conflicts of interest.