a b s t r a c t

Objectives: To analyze the differences in clinical presentation and characteristics of Community-onset bacteremia between neutropenic and nonneutropenic adults visiting the emergency department.

Methods: A case-control study with a ratio of 1:2 was conducted retrospectively over a 6-year period. Demo- graphic characteristics, microorganisms, severity of illness, and clinical outcomes determined from medical records were analyzed.

Results: In total, 116 neutropenic adults (case patients) and 232 nonneutropenic adults (control patients) were examined. Significant differences in the source of bacteremia, susceptibility, and species of bacteremia-causing organisms between the case patients and control patients were observed by univariate analyses. Significantly more patients presenting with an initial syndrome of severe sepsis or septic shock at the emergency department, having high Pittsburgh bacteremia scores (>= 4 points) or having Severe comorbidities (McCabe classification), and high 28-day mortality rates were discovered in the case group, compared with the control group. Of note, Pseudomonas aeruginosa (32/137 [23.4%] vs 8/272 [2.9%], P b .001) was more often isolated from the case patients. In a further analysis using a multivariate regression to demonstrate the independent predictors of P aeruginosa infection, patients with neutropenia remained as an independent risk factors (odds ratio, 7.48; P b .001).

Conclusions: This study demonstrated obvious differences of community-onset bacteremia in severity, the distribu- tion of microorganisms, and susceptibility between neutropenic and nonneutropenic patients. Antipseudomonas therapy was empirically suggested for neutropenic patients with community-onset bacteremia and reducing the need for a glycopeptide.

(C) 2015

Introduction

Despite recent advances in care of patients with hematologic or auto- immune disease, infections account for significant morbidity and mortality in patients with neutropenia [1]. Approximately 10% to 38% of febrile neutropenic episodes are associated with bacteremia [2,3], which is a serious, life-threatening condition with high mortality, as high as 40% [4,5]. Traditionally, both gram-negative and Gram-positive bacteria are responsible for bacteremia in neutropenic patients, but there has been a shift in etiology from gram-negative rods to gram-positive cocci during the past years [6,7]. Although many studies have investigated the cause of the changing etiology of the responsible bacterial spectrum, the updated distribution of bacteremia-causing microorganism and susceptibility of bacteremic isolates are lacking, especially that the Clinical and Laboratory Standards Institute guidelines have suggested

* Corresponding author. Division of Infectious Disease, Department of Internal Medicine, National Cheng Kung University Hospital, No. 138, Sheng Li Rd, 70403 Tainan, Taiwan. Tel.: +886 6 2353535×3596; fax: +886 6 2752038.

E-mail address: [email protected] (W.-C. Ko).

a new cephalosporin-interpretive criteria for bacteremia-causing Enterobacteriaceae since 2009. In addition, based on previous inves- tigations, community-onset bacteremia comprises most of bacteremic episodes in the emergency department (ED) [8]. However, no ED- based study dealing with updated surveillance of pathogens and sus- ceptibility of community-onset bacteremia in neutropenic patients has been reported. Therefore, to delineate the clinical features and offer optimal management of community-onset bacteremia in neutro- penic adults at the ED, the clinical characteristics, cause of neutropenia, bacterial species, in vitro susceptibility data, and clinical outcome were analyzed in the present study. Furthermore, the different presentation of community-onset bacteremia between neutropenic adults and nonneutropenic adults was also discussed.

Materials and methods

Study design and population

A retrospective case-control study was conducted between January 2007 and December 2012 at the ED of a medical center in southern

http://dx.doi.org/10.1016/j.ajem.2015.03.058

0735-6757/(C) 2015

Taiwan. The local institutional review board approved this study. Of the patients with blood culture sampling in the ED, age, sex, laboratory data, and the results of blood cultures with bacterial growth were screened based on a computer database. The study initially included all neutrope- nic individuals with microorganism growth in blood culture and then excluded patients with contaminated blood cultures, those with fungemia or mycobacteremia, those with hospital-onset bacteremia, or those with bacteremia diagnosed before visiting the ED. To create a matched case-control study with a ratio of 1:2, eligible patients with neutropenic bacteremia were regarded as the case group, and 2 nonneutropenic patients with community-onset bacteremic episodes that occurred temporally near the onset of the cases were regarded as the control patients. Medical records were reviewed retrospectively for Pittsburgh bacteremia severity scores within 24 hours after bac- teremic onset, clinical characteristics, comorbidities and its severity, microbiological results, the source of bacteremia, and further hospitali- zation. Also, they were derived from the medical records using a predetermined case record form. As previously described [8], multiple bacteremic episodes in a single patient were considered as distinct events if separated by at least 7 days or if different causes were listed for the respective ED visits. The primary outcome assessed was overall mortality up to 28 days after visiting the ED or at discharge. Patients were excluded if they were not followed up at the study hospital within 28 days after ED arrival.

Microbiological studies

Nurses performed blood sampling to collect 2 sets of blood cultures. Each set of blood samples routinely consisted of one bottle for aerobic culture and another for anaerobic culture, with approximately 5 to 8 mL of blood in each bottle. After the collection, the culture bottles were immediately transported to the clinical laboratory department, loaded into the BACTEC 9240 system (Becton Dickinson and Company, Franklin Lakes, NJ), and incubated for 5 days or until the instrument detected bacterial growth. Biochemical tests and automatic identifica- tion systems were used for the final pathogen identification. Antimicro- bial susceptibility to each antibiotic was interpreted according to the updated document issued by Clinical Laboratory Standard Institute guidelines [9].

Definitions Neutropenia were defined as a Neutrophil count less than 500 cells/mm³,

as previously described [10]. Bacteremia was defined as 2 separate

blood cultures growing the same microorganism or as a single blood culture associated with a clinically identified source growing the same microorganism(s). Blood culture samples with potential contaminating pathogens (such as coagulase-negative Staphylococcus, Micrococcus, Propionibacterium acnes, Peptostreptococcus, or Bacillus species) were considered to be contaminated, in accordance with the previously described criteria [11]. Polymicrobial bacteremia was defined as the isolation of more than one microbial species from a single bacteremic episode. Community-onset bacteremia indicated that the place of onset of the bacteremic episode was the community, including long-term health care facility- and community-acquired bacteremia, as previously described [12]. Underlying malignancies included both hematologic malignancies and solid tumors; comorbidi- ties were defined as previously described [13]. The sources of bacteremia were classified as lower respiratory tract infections, urinary tract infec- tions, skin and soft tissue infections, intra-Abdominal infections, or primary bacteremia, according to definitions from the Centers for Disease Control and Prevention [10].

Severe sepsis was defined as the coexistence of sepsis and at least one of the following signs or symptoms of acute organ dysfunction or hypoperfu- sion: metabolic acidosis, arterial hypoxemia (PaO2 b 75 mm Hg or PaO2/FIO2 b 250), oliguria (b 0.03 L/h for 3 hours or 0.7 L/24 h), coagulopathy

(increase in prothrombin time or a drop of platelet count by 50% or to b 100 x 10⁷ L), or encephalopathy (Glasgow coma score b 14) [14]. The severity of bloodstream infection at the time of blood sampling was assessed using the Pittsburgh bacteremia score, a validated scoring system based on vital signs, mental status, mechanical ventilation, and the presence of cardiac arrest. [15] The estimated prognosis of preexisting underlying diseases was described using the classification system of McCabe [16].

Data analysis

Statistical analyses were performed using the Statistical Package for the Social Sciences for Windows (SPSS, Chicago, IL), Version 20.0. Continuous variables were expressed as the means +- SDs and com- pared using Student t test. Categorical variables, expressed as numbers and percentages, were compared using a ?² test or Fisher exact test. All significant variables (P values b .05) in the univariate analysis were incorporated into a hierarchical logistic regression model. A P value less than .05 was considered statistically significant.

Results

Study population and demographics

During the 6-year study period, there were a total of 405 565 ED visits with a mean annual visit of 67 594.2. Of 80 058 (19.7%) patients with blood culture sampling, bacterial growth of blood cultures were noted in 3934 (4.9%) patients. The majority (2445 patients, or 3.1%) had true bacteremia, 1471 (1.8%) had contaminated blood cultures, and 18 (0.02%) had a mixture of true bacteremic pathogens and con- taminated isolates. Of 2445 bacteremic patients, 120 patients (4.9%) have neutropenia. After exclusion of 1 case with an unknown 28-day mortality and 3 cases with hospital-onset bacteremia, 116 neutropenic patients with community-onset bacteremia were included in the case group. Correspondingly, 232 matched nonneutropenic patients with an onset of community-onset bacteremia that was temporally near the onset of the neutropenic cases served as the control group.

Of the case patients, most patients (112, or 96.6%) directly visited the ED from the community, but 4 were transferred from the EDs of other hospitals. Comorbidities of these patients included malignancy (99 patients, or 85.3%), hypertension (33, or 28.4%), diabetes mellitus (27, or 23.3%), Chronic renal insufficiency (16, or 13.8%), liver cirrhosis (11, or 9.5%), neurologic disorders (6, or 5.2%), coronary artery disease

(5, or 4.3%), and chronic obstructive pulmonary disease (3, or 2.6%). The causes for neutropenia included chemotherapy (72, or 62.1%), co- morbidity (22, or 19.0%), sepsis (19, or 16.4%), or immunosuppressive

agents (3, or 2.5%).

Clinical characteristics, severity, and outcomes in the case and control patients

For the case and control patients, clinical variables, including demo- graphics, sources of bacteremia, comorbidities, and laboratory data at the ED, were compared in Table 1. Having bacteremia due to respiratory tract or vascular catheter, having primary bacteremia, and having ma- lignancy were factors more prevalent among the case patients. Having bacteremia due to urinary tract infection or intra-abdominal infection; or having hypertension, diabetes mellitus, or neurologic disorders; or having low leukocyte count, platelet count, serum creatinine, or Serum glucose were more often noted in control patients.

Initial presentation at the ED, severity of illness, and clinical outcome were compared between case and control patients by univariate analy- sis, as shown in Table 2. Significantly more patients presenting with an initial syndrome of severe sepsis or septic shock, having high Pittsburgh bacteremia scores (>=4 points) or having severe comorbidities (based on McCabe classification), and high 28-day mortality rates or discharge

Table 1 Comparisons of clinical characteristics, major comorbidity, source of bacteremia, and clinical laboratory in adults with community-onset neutropenic bacteremia (NB group) and those with nonneutropenic bacteremia (non-NB group) in the ED

Variable	Patients		P
	NB group	Non-NB group
	(n = 116)	(n = 232)
old age, >= 65 y, no. (%)	60 (51.7)	131 (56.5)	.40
Sex, male, no. (%)	72 (62.1)	121 (52.2)	.08
Polymicrobial bacteremia, no. (%)	17 (14.7)	33 (14.2)	.91
Major source of bacteremia, no. (%)
Respiratory tract	38 (32.8)	29 (12.5)	b.001
Primary bacteremia	38 (32.8)	16 (6.9)	b.001
Vascular catheter	18 (15.5)	10 (4.3)	b.001
Urinary tract	7 (6.0)	74 (31.9)	b.001
Skin and soft tissue	6 (5.2)	20 (8.6)	.25
Intra-abdomen	3 (2.6)	66 (28.4)	b.001
Infective endocarditis	2 (1.7)	8 (3.4)	.51
Bone and joint	0 (0)	7 (3.0)	1.00
Major comorbidities
Malignancy	99 (85.3)	71 (30.6)	b.001
Hypertension	33 (28.4)	109 (47.0)	.001
Diabetes mellitus	27 (23.3)	79 (34.1)	.04
Chronic renal insufficiency	16 (13.8)	41 (17.7)	.36
Liver cirrhosis	11 (9.5)	34 (14.7)	.18
Neurologic disorder	6 (5.2)	43 (18.5)	.001
Coronary artery disease	5 (4.3)	24 (10.3)	.06
Chronic obstructive pulmonary disease	3 (2.6)	12 (5.2)	.26
Laboratory data in the EDa, mean +- SD Neutrophil (/mm3)	188.4 +- 181.9	9902.9 +- 6167.5	b.001
Platelets (x103/mm3)	87.8 +- 119.9	190.7 +- 170.2	b.001
Serum creatinine (mg/dL), n = 311	1.6 +- 1.5	4.9 +- 44.8	.02
C-reactive protein (mg/L), n = 342	138.5 +- 125.4	106.3 +- 107.7	.43
Serum glucose (mg/dL), n = 325	158.7 +- 86.2	184.8 +- 123.1	.03
serum albumin (g/dL), n = 113	3.7 +- 4.7	3.2 +- 0.7	.34

^a Not all patients had the indicated laboratory data.

mortality rates were discovered in the case group, compared with the control group.

Microbiological and susceptibility data from the case and control patients

Seventeen episodes of polymicrobial bacteremia were detected among the case patients and 33 polymicrobial episodes were found in the control group, leaving a total of 137 isolates obtained from the

Table 2

Comparisons of severity of bacteremia and comorbidity, length of hospital stay, and clinical outcome in adults with community-onset neutropenic bacteremia (NB) and those with nonneutropenic bacteremia (non-NB group) in the ED

Variable Patients P

	NB group (n = 116)	Non-NB group (n = 232)
Severity-of-illness marker in the ED, no. (%) Pittsburgh bacteremic score >= 4 points	38 (32.8)	39 (16.8)	.001
Initial syndrome at bacteremic onset Severe sepsis	67 (57.8)	87 (37.5)	b.001
Septic shock	43 (37.1)	37 (15.9)	b.001
Severity of underlying disease (McCabe classification), no. (%) Rapidly fatal	21 (18.1)	5 (2.2)	b.001
Ultimately fatal	67 (57.8)	58 (25.0)
Nonfatal	28 (24.1)	169 (72.8)
Clinical outcome Length of stay (d), mean +- SD Total hospitalization	16.7 +- 18.7	13.7 +- 15.1	.11
Intensive care unit, n = 42 Crude mortality, no. (%)	1.6 +- 6.6	1.1 +- 5.2	.40
28-d	38 (32.8)	34 (14.7)	b.001
At discharge	39 (33.6)	38 (16.4)	b.001

Fig. 1. The distribution of major bacteremia-causing microorganisms in adults with community-onset bacteremia, categorized as neutropenic (137 isolates) and nonneutropenic (272 isolates) patients. The percentage data were the proportion of all indicated patients, and asterisk indicates P b .05, calculated by the ?² test.

case patient group and 272 obtained from the controls. In both groups, Escherichia coli was the most common gram-positive microorganism and Staphylococcus aureus was the most common gram-positive aer- obes. The distribution of major bacterial species was compared between the case and the control groups, as shown in Fig. 1. E coli, S aureus, and Enterococcus species were more often isolated from the control patients, whereas Pseudomonas aeruginosa were more often isolated from the case patients.

In vitro susceptibility for the commonly used antibiotics in the ED was compared between the case and the control groups (Fig. 2), including the following: ampicillin/sulbactam, cefazolin, cefuroxime, ceftazidime, ceftriaxone, cefepime, ertapenem, imipenem, and levofloxacin. The susceptibility for all antibiotics but cefazolin and cefuroxime was dissimilar between the case and the control groups. The susceptibility of ampicillin/sulbactam, ceftriaxone, and ertapenem was higher in bacteremic isolates detected from the control patients, whereas suscep- tibility for antipseudomonas agents (ceftazidime, levofloxacin, cefe- pime, and imipenem) was higher in bacteremic isolates from the case patients. Furthermore, methicillin-resistance S aureus (MRSA) was not discovered in the case patients, but a higher proportion of MRSA (10 [3.7%], P = 0.04) isolates were noted in the control group. The isolate with extended-spectrum ?-lactamase production was similarly evident in 2.9% (4 isolates) of all bacteremic pathogens isolated in the case patients and 3.3% of all isolates in the control patients.

Risk factors of P aeruginosa infection among bacteremic patients visiting the ED

By univariate analysis, the variables that were significantly associated with P aeruginosa infection included sex with male, the presence of neu- tropenia at the ED, recent chemotherapy receipt, a Pittsburgh bacteremia score of 4 or higher points at the ED, an initial syndrome with septic shock, fatal comorbidity based on McCabe classification, bacteremia due to respiratory tract, vascular catheter, intra-abdomen and urinary tract, comorbidity with malignancy, or liver cirrhosis (Table 3). Using multivariate analysis, only 2 variables, including the presence of neutro- penia at the ED and comorbidity with liver cirrhosis, were independently associated with P aeruginosa infection, as shown in Table 4. Notably, the presence of neutropenia at the ED was the only positive predictor.

Discussion

The present study demonstrates the obvious differences of community-onset bacteremia in severity at onset, the distribution of

Fig. 2. In vitro susceptibility of a total 409 bacteremia-causing microorganisms in adults with community-onset bacteremia, categorized as neutropenic (137 isolates) and nonneutropenic (272 isolates) patients. The percentage data were the proportion of all indicated patients, and asterisk indicates P b .05, calculated by the ?² test.

resource and microorganisms, susceptibility, patient’s comorbidity, severity of comorbidity, and clinical outcome between neutropenic and nonneutropenic patients. Notably, a high proportion of P aeruginosa infection was observed in patients with neutropenic bacteremia. In a further analysis using a multivariate regression, patients with neutropenia at bacteremic onset remained as an independent risk factor

Table 3 Univariate analysis of risk factors for P aeruginosa (PA) infection in adults with community- onset bacteremia in the ED

Variable Patients, no. (%) P

associated with P aeruginosa infection among patients with community- onset bacteremia.

Bacteremia is a subject that has been extensively discussed in the medical literature. The study populations examined in articles have varied greatly, from the general population to specific patient groups, including the elderly [17], the critically ill [18], and individuals with various comor- bidities [19,20]; from all bacteremia-causing microorganisms to specific pathogens, including Streptococcus pneumoniae [21], P aeruginosa [12], and Klebsiella pneumoniae [15]; and from nosocomial to community origins of acquisition [22]. Given the increased incidence of hematologic patients worldwide and recent advances in care of patients with hemato- logic and oncologic disease [23-25], we suspect that these reasons may

PA group (n = 40)

Non-PA group

(n = 308)

have led to the increased frequency of ED visits with hemato-oncologic

disorder and to epidemiologic changes in the bacteremic population seen in the ED. The current study was the first ED-based investigation

Old age, >= 65 y	27 (67.5)	164 (53.2)	.09	focusing on the clinical impact of neutropenia on bacteremic patients in
Sex, male	30 (75.0)	163 (52.9)	.008	the ED, to assist ED clinicians in facing the growing neutropenic popula-

tion with suspected bacteremic onset. Our study emphasized the vast influence of neutropenia on severity of acute illness, the distribution and susceptibility of bacteremia-causing microorganism, and patient’s outcome in patients with community-onset bacteremia.

Presence of neutropenia at bacteremic onset Events during the previous 4 wk	32 (80.0)	84 (27.3)	b.001
Chemotherapy	20 (50.0)	95 (19.2)	b.001
Hemodialysis	2 (5.0)	19 (6.2)	1.00
Invasive procedure	2 (5.0)	15 (4.9)	1.00
Surgery	1 (2.5)	16 (5.2)	.71
Severity-of-illness marker in the ED Pittsburgh bacteremic score >= 4 points	15 (37.5)	62 (20.1)	.01
Initial syndrome at bacteremic onset
Severe sepsis	23 (57.5)	131 (42.5)	.07
Septic shock	16 (40.0)	61 (19.8)	.007
Severity of underlying disease
(McCabe classification)
Rapidly and Ultimately fatal	25 (62.5)	126 (40.9)	.01
Major comorbidities
Malignancy	30 (75.0)	140 (45.5)	b.001
Hypertension	13 (32.5)	129 (41.9)	.26
Diabetes mellitus	12 (30.0)	94 (30.5)	.95
Chronic renal insufficiency	6 (15.0)	51 (16.6)	.80
Neurologic disorder	5 (12.5)	44 (14.3)	.76
Liver cirrhosis	1 (2.5)	44 (14.3)	.04
Major source of bacteremia
Respiratory tract	17 (42.5)	50 (16.2)	b.001
Vascular catheter	9 (22.5)	19 (6.2)	.002
Primary bacteremia	6 (15.0)	48 (15.6)	.92

Empirical broad-spectrum antibiotic therapy has been the cornerstone of initial management of fever or bacteremia in patients with neutrope- nia [26]. In the past years, the etiology of bacteremia in febrile neutropenic patients has shifted from gram-negative to Gram-positive organisms. [6,7] These changes affect empirical therapy as it was practiced past years ago. Moreover, as most of community-onset bacteremia, the susceptibility of Enterobacteriaceae based on the Clinical and Laboratory Standards Institute criteria has suggested new cephalosporin-interpretive criteria since 2009. Thus, to assist the ED physician, it is necessary to offer timely the distribution of bacteremia-causing organisms and susceptibility

Table 4

Multivariate analysis of risk factors for P aeruginosa infection in adults with community- onset bacteremia in the ED

Clinical variables Adjusted odds ratio P

(95% confidence interval)

Urinary tract	3 (7.5)	78 (25.3)	.01
Intra-abdomen	2 (5.0)	67 (21.8)	.01	Presence of neutropenia at bacteremic onset	7.48 (3.20-17.47)	b.001
Skin and soft tissue	1 (2.5)	25 (8.1)	.34	Sex, male	2.22 (0.98-5.03)	.06
Infective endocarditis	0 (0)	10 (3.2)	.61	Comorbidity with liver cirrhosis	0.32 (0.14-0.74)	.007

among neutropenic patients. Based on the current study, we empha- sized that the neutropenia is one of the critical factors when choosing empiric antimicrobial therapy and that the strategy of empirical therapy for community-onset bacteremia was different between patients with neutropenia and general population. Because of a high proportion of P aeruginosa in neutropenic patients, antipseudomonas agents (ceftazidime, levofloxacin, cefepime, and imipenem) were empirically warranted for such patients. As previously described [6], among the most commonly used agents, cefepime and carbapenems continue to show the highest rates of in vitro susceptibility, providing coverage against most gram-positive and Gram-negative organisms in our popu- lation. Of note, no MRSA was observed in neutropenic patients, and thus, reducing the need for a glycopeptide was reasonable, in contrast with previous reports [6,7].

Consistent with previous study discussed with etiology of bacteremia in febrile neutropenic patients [6,7], P aeruginosa were one of the preva- lent agents in our population. However, in contrast with these reports, our study demonstrated E coli as the leading microorganism. From our suspect, this diversity was based on the differently selected study population, because E coli were the leading pathogens of community- onset bacteremia in the ED [8], and our study was focusing on the community-onset bacteremia. For P aeruginosa, it is traditionally a lead- ing nosocomial pathogen often causing severe infections, whereas community-onset infections are rare and mostly affect immunocompro- mised hosts, neutropenia, and malignancy patients [27]. Furthermore, the strong relationship of P aeruginosa infection and comorbidity with malignancy was also well established in bacteremic adults visiting the ED [28]. Our finding was consistent with these previous studies discussed with the risk factors of P aeruginosa infection in the communi- ty. We hypothesize that there was a relatively high percentage of P aeruginosa infection in neutropenic population because of its high colonization rate and the use of long-term indwelling catheters.

The control patients in the current study were chosen to represent community-onset bacteremia in the ED and were randomly selected based on the timing of their infection rather than matching patients by age or sex, as has been done in previous studies. In the present study, the overall 28-day mortality in the control group was 14.7%, and the proportion of high Pittsburgh bacteremia scores (>= 4 points) was 16.8%. Of the gram-negative microorganisms in our control patients, the most common was E coli, followed by K pneumoniae, P aeruginosa, and Proteus mirabilis. These features were all similar to previous ED-based studies with a predominance of community-onset bacteremia [8,28].

Limitations

We interpreted our data in light of several limitations inherent in the study design. First, to demonstrate the different presentations and features of bacteremia, the baseline characteristics of the neutropenic and nonneutropenic patients were not truly comparable. The neutrope- nic patients were severe bacteremic onset and comorbidity, thus making this group vulnerable to various infections and poor outcomes. Therefore, lacking a multivariate analysis for risk factor of mortality, the impact of neutropenia on patients’ outcome remained unclear. Nevertheless, our study aimed to emphasize the different presentations and features of bacteremia between the neutropenic and nonneutropenic patients. Second, we selected only bacterial pathogens. Although fungal and mycobacterial species are also important pathogens in nonneutropenic patients, cultures for fungi and mycobacteria were not routinely performed in the ED of the study hospitals. To determine the clinical impact and epidemiology of these pathogens, a prospective study will need to be conducted. Third, in our case patients, neutropenia caused majorly by chemotherapy (62.1%); only few proportions (16.4%) were complicated with sepsis. Also, hemato-oncologic patients account for most (85.3%) of the population. Therefore, our finding may not be generalized to other population with sepsis-related neutropenia. Finally,

although we examined all neutropenic patients who visited the EDs over an extended period, the study was conducted only in metropolitan areas. Therefore, our findings, especially susceptibility data, may not be generalizable to other locations.

Conclusions

The present study demonstrates obvious differences of community- onset bacteremia in severity, the distribution of resource and microor- ganisms, susceptibility, comorbidity, and clinical outcome between neutropenic and nonneutropenic patients. Notably, a strong relation- ship of P aeruginosa infection and neutropenia was established in patients with community-onset bacteremia. Based on our finding, antipseudomonas agents (ceftazidime, levofloxacin, cefepime, and imipenem) were empirically warranted for neutropenic patients with community-onset bacteremia, providing coverage against most gram- positive and gram-negative organisms, and reducing the need for a glycopeptide.

References

1. Yadegarynia D, Tarrand J, Raad I, Rolston K. Current spectrum of bacterial infections in patients with cancer. Clin Infect Dis 2003;37:1144-5.
2. Klastersky J, Ameye L, Maertens J, Georgala A, Muanza F, Aoun M, et al. Bacteraemia in febrile neutropenic cancer patients. Int J Antimicrob Agents 2007;30(Suppl 1): S51-9.
3. Madani TA. Clinical infections and bloodstream isolates associated with fever in pa- tients undergoing chemotherapy for Acute myeloid leukemia. Infection 2000;28: 367-73.
4. Norgaard M, Larsson H, Pedersen G, Schonheyder HC, Sorensen HT. Risk of bacteraemia and mortality in patients with haematological malignancies. Clin Microbiol Infect 2006;12:217-23.
5. Collin BA, Leather HL, Wingard JR, Ramphal R. Evolution, incidence, and susceptibility of bacterial bloodstream isolates from 519 bone marrow transplant patients. Clin Infect Dis 2001;33:947-53.
6. Ramphal R. Changes in the etiology of bacteremia in febrile neutropenic patients and the susceptibilities of the currently isolated pathogens. Clin Infect Dis 2004; 39(Suppl. 1):S25-31.
7. Kanamaru A, Tatsumi Y. Microbiological data for patients with febrile neutropenia. Clin Infect Dis 2004;39(Suppl. 1):S7-S10.
8. Lin JN, Tsai YS, Lai CH, Chen YH, Tsai SS, Lin HL, et al. risk factors for mortality of bacteremic patients in the emergency department. Acad Emerg Med 2009;16: 749-55.
9. Clinical and Laboratory Standards Institute. Performance standards for antimicrobial disk susceptibility tests, 24th informational supplement. Document M100-S24. Wayne, PA: Clinical and Laboratory Standards Institute; 2014.
10. Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988;16:128-40.
11. Lee CC, Lin WJ, Shih HI, Wu CJ, Chen PL, Lee HC, et al. Clinical significance of potential contaminants in blood cultures among patients in a medical center. J Microbiol Immunol Infect 2007;40:438-44.
12. Cheong HS, Kang CI, Wi YM, Kim ES, Lee JS, Ko KS, et al. Clinical significance and pre- dictors of community-onset Pseudomonas aeruginosa bacteremia. Am J Med 2008; 121:709-14.
13. Schellevis FG, van der Velden J, van de Lisdonk E, van Eijk JT, van Weel C. Comorbidity of chronic diseases in General practice. J Clin Epidemiol 1993;46:469-73.
14. Brun-Buisson C, Doyon F, Carlet J, Dellamonica P, Gouin F, Lepoutre A, et al. Inci- dence, risk factors, and outcome of severe sepsis and septic shock in adults. A multicenter prospective study in intensive care units. French ICU Group for Severe Sepsis. JAMA 1995;274:968-74.
15. Paterson DL, Ko WC, Von Gottberg A, Mohapatra S, Casellas JM, Goossens H, et al. International prospective study of Klebsiella pneumoniae bacteremia: implications of extended-spectrum beta-lactamase production in Nosocomial infections. Ann Intern Med 2004;140:26-32.
16. McCabe WR. Gram-negative bacteremia. Adv Intern Med 1974;19:135-58.
17. Leibovici L, Pitlik SD, Konisberger H, Drucker M. Bloodstream infections in patients older than eighty years. Age Ageing 1993;22:431-42.
18. Valles J, Leon C, Alvarez-Lerma F. Nosocomial bacteremia in critically ill patients: a multicenter study evaluating epidemiology and prognosis. Spanish Collaborative Group for Infections in Intensive Care Units of Sociedad Espanola de Medicina Intensiva y Unidades Coronarias (SEMIUC). Clin Infect Dis 1997;24:387-95.
19. Kuo CH, Changchien CS, Yang CY, Sheen IS, Liaw YF. Bacteremia in patients with cirrhosis of the liver. Liver 1991;11:334-9.
20. Leibovici L, Samra Z, Konisberger H, Kalter-Leibovici O, Pitlik SD, Drucker M. Bacteremia in adult Diabetic patients. Diabetes Care 1991;14:89-94.
21. Yu VL, Chiou CC, Feldman C, Ortqvist A, Rello J, Morris AJ, et al. An international pro- spective study of pneumococcal bacteremia: correlation with in vitro resistance, antibiotics administered, and clinical outcome. Clin Infect Dis 2003;37:230-7.
22. Salomao R, Castelo Filho A, Pignatari AC, Wey SB. Nosocomial and community acquired bacteremia: variables associated with outcomes. Rev Paul Med 1993;111:456-61.
23. Rodriguez-Abreu D, Bordoni A, Zucca E. Epidemiology of hematological malignan- cies. Ann Oncol 2007;18(Suppl 1):i3-8.
24. Yancik R, Ries LA. Aging and cancer in America. Demographic and epidemiologic perspectives. Hematol Oncol Clin North Am 2000;14:17-23.
25. Vrdoljak E, Wojtukiewicz MZ, Pienkowski T, Bodoky G, Berzinec P, Finek J, et al. Cancer epidemiology in Central, South and Eastern European countries. Croat Med J 2011;52:478-87.
26. Pizzo PA. Management of fever in patients with cancer and treatment-induced neutropenia. N Engl J Med 1993;328:1323-32.
27. Bodey GP, Jadeja L, Elting L. Pseudomonas bacteremia. Retrospective analysis of 410 episodes. Arch Intern Med 1985;145:1621-9.
28. Lee CC, Lee CH, Hong MY. Risk factors and outcome of Pseudomonas aeruginosa bacteremia among adults visiting the ED. Am J Emerg Med 2012;30: 852-60.