Article

Impact of inappropriate empirical antibiotic therapy on outcome of bacteremic adults visiting the ED

Unlabelled imageAmerican Journal of Emergency Medicine (2012) 30, 1447-1456

Original Contribution

Impact of inappropriate empirical antibiotic therapy on outcome of bacteremic adults visiting the ED

Ching-Chi Lee MDa, Chung-Hsun Lee MDb, Ming-Che Chuang MDb, Ming-Yuan Hong MDb, Hsiang-Chin Hsu MDb, Wen-Chien Ko MDa,c,?

aDepartment of Internal Medicine, National Cheng Kung University Hospital, Tainan 704, Taiwan bDepartment of Emergency Medicine, National Cheng Kung University Hospital, Tainan 704, Taiwan cDepartment of Medicine, National Cheng Kung University Medical College, Tainan 704, Taiwan

Received 26 September 2011; revised 15 November 2011; accepted 15 November 2011

Abstract

Objectives: To investigate the clinical impact of inappropriate Empirical antibiotics on patient outcome and determine the risk factors for mortality in bacteremic adults who visited the emergency department (ED). Methods: Bacteremic adults visiting the ED from January 2007 to June 2008 were identified retrospectively. Demographic characteristics, clinical conditions, bacteremic pathogens, antimicrobial agents, and outcomes were determined from chart records.

Results: The total of 454 eligible bacteremic adults were included in the analysis; excluded from the study were another 261 patients with contaminated blood cultures and 64 patients with ED stays of less than 24 hours. Among the included individuals, the mean age was 64.6 years, with a small predominance of males (230 patients, 50.7%). Of a total 494 bacteremic isolates, Escherichia coli (206, 41.7%) and Klebsiella species (81, 16.4%) were the most frequently encountered microorganisms. A lower 28-day mortality rate was demonstrated in bacteremic patients treated with appropriate antibiotics than that in those with inappropriate antibiotics or that in those with no antibiotic therapy, as judged by Kaplan-Meier survival curves (P = .01). Moreover, the differences among these three groups achieved higher significance (P = .002)

in critically ill patients (Pittsburgh bacteremia scores of >=4 points). In multivariate analyses, inappropriate antibiotic therapy in the ED was associated independently with mortality at 28 days (odds ratio, 2.26; 95%

confidence interval, 1.01-5.13; P = .04).

Conclusions: For bacteremic adults visiting the ED, their outcomes were favorable following appropriate antibiotics, compared to treatment with inappropriate antibiotics or no antibiotics.

(C) 2012

Introduction

Bacteremia is a serious, life-threatening condition that is associated with significant Healthcare costs and in-hospital

* Corresponding author. Department of Internal Medicine, National Cheng Kung University Hospital, Tainan 70403, Taiwan. Tel.: +886 6 2353535×3596; fax: +886 6 2752038.

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

mortality rates of up to 30% [1-4]. Bacteremia is also a common and deadly problem in patients visiting emergency departments (EDs) [5]. Early administration of appropriate empirical antibiotic therapy has been shown repeatedly to decrease mortality in patients with community-acquired and nosocomial Bloodstream infections [2,6-8]. Notably, early initiation of treatment with appropriate antibiotics has a favorable effect on outcome in critically ill patients with bacteremia [9,10]. However, previous investigations focused

0735-6757/$ – see front matter (C) 2012 doi:10.1016/j.ajem.2011.11.010

on bacteremic patients upon hospital or intensive care unit admission. Although ED clinicians serve as an earlier line in the diagnosis of bacteremia or sepsis and choice of antibiotic, an ED-based study dealing with the effect of inappropriate antibiotic therapy has not (to our knowledge) been reported in the English language literature. Thus, to demonstrate the importance of appropriate empirical therapy in the ED, we report here the impact of inappropriate empi- rical antibiotic therapy on clinical outcome, and assess the risk factors for mortality in bacteremic adults visiting the ED.

Materials and methods

Study design and population

This retrospective study compared the clinical outcomes between bacteremic patients treated in the ED with inappropriate or appropriate empirical antibiotic therapy. The local institutional review board approved this study.

This study was conducted at a teaching hospital of approximately 1,000 beds in southern Taiwan. During the study period, there were approximately 65,000 annual visits to the ED; the average length (+-standard deviation) of stay

for all ED visits was 18.2 (+-17.7) hours. Medical records were retrospectively reviewed for adult patients (age >=18 years) who visited the ED between January 2007 and June 2008 and had a positive blood culture.

Study protocol

This analysis incorporated the records of bacteremic adults who visited the ED during the study period. However, the study excluded any patients with contaminated blood cultures, or in whom bacteremia was diagnosed prior to visiting the ED. To evaluate the influence of empirical antibiotics in the ED, patients with ED stays of less than 24 hours also were excluded. For the remaining (eligible) bacteremic adults visiting the ED, medical records were reviewed, including the collection of clinical characteristics, vital signs, Pittsburgh bacteremia severity scores, comorbid- ities, initial syndrome, and clinical pathology laboratory data immediately following each patient’s visit to the ED. At that time, the therapeutic duration and type of antimicrobial agent; microbiological results and source of bacteremia; and further hospitalization and length of stay also were derived from the chart records. Moreover, recent events within four weeks before the ED arrival, including hospitalizations, invasive procedures, or surgeries, were also recorded. Multiple bacteremic episodes in a single patient were considered as distinct events, if they were separated by at least seven days, or if different causes were listed for the respective ED visits.

The primary outcome was the overall mortality in the 28 days after ED arrival. If patients were discharged within

28 days after ED arrival and were not followed-up at our hospital, the required information was retrieved by tele- phone. Any patients who could not be reached by telephone were excluded.

Microbiological studies

Nurses performed the blood sampling by collecting two sets of blood cultures from different peripheral veins or arteries simultaneously. Each set of blood samples routinely consisted of one bottle for aerobic culturing and another bottle for anaerobic culturing, with approximately 5-8 mL of blood per bottle. Following collection, the culture bottles were transported immediately to the hospital laboratory, loaded into the BACTEC 9240 system (Becton Dickinson and Company (BD), Franklin Lakes, NJ, USA), and incubated for 5 days or until the instrument detected bacterial growth. Culture bottles that exhibited bacterial growth were Gram-stained, and the contents of the bottles were sub- cultured onto plates with blood agar (Trypticase soy agar II 5% sheep blood; BD), Levine eosin-Methylene blue agar (BD), chocolate agar, and Centers for Disease Control and Prevention (CDC) anaerobic blood agar (BD) for further identification. Biochemical tests and automatic identification systems were used for final identification. Isolates were tested for in vitro antibiotic susceptibility using the Kirby- Bauer method on Mueller-Hinton agar. Antimicrobial susceptibility to each antibiotic was interpreted and pheno- typic confirmation for ESBL production was performed by cephalosporin/clavulanate combination disks for E. coli, Klebsiella species, and Proteus species, according to the Clinical Laboratory Standard Institute guidelines [11].

Definitions

Polymicrobial bacteremia was defined as the isolation of more than one microbial species from a single bacteremic episode. Appropriate antibiotic was defined as an antibiotic to which the isolated pathogen was susceptible in vitro; the inappropriate antibiotic was defined as an antibiotic to which the isolated pathogen was non-susceptible in vitro. The severity of bloodstream infection at the time of onset 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 [12]. Malignancy includes both Hematological malignancies and solid tumors. The definitions of comorbidities were as previously described [13]. Blood culture samples with potential contaminating pathogens (such as coagulase- negative Staphylococcus, Micrococcus, Propionebacterium acnes, Peptostreptococcus, or Bacillus species), in any blood culture bottle, were regarded as contaminated according to previously described criteria [14]. The sources of bacteremia were classified as lower respiratory tract infections, Urinary tract infections , skin and soft-tissue infections, intra-

Abdominal infections, or primary bloodstream infections according to the definitions of the CDC [15].

Data analysis

Statistical analyses were performed using the Statistical Package for the Social Sciences for Windows (SPSS, Chicago, Illinois, USA), Version 15.0. Continuous variables were expressed as means +- standard deviations (SDs) and compared with Student’s t-test. Categorical variables, expressed as numbers and percentages, were compared by chi-square test or Fisher’s exact test. All variables with P-values of less than

.1 in univariate analysis were incorporated into a stepwise, backward logistic regression model. A P-value of less than

.05 was considered statistically significant.

Results

Demographics and clinical characteristics of all bacteremic patients

During the 18-month study period, blood culture samples were drawn from 19,308 patients. The medical records of 454 eligible bacteremic adults (2.4%) who visited the ED were collected; another 261 patients with contaminated blood cultures and 64 patients with ED stays of less than

24 hours were excluded from the study. Among the 454 eligible bacteremic patients, the majority (386, 85.0%) visited the ED from the community; 36 (7.9%) of these eligible patients were transferred from long-term health- care facilities; and 32 (7.0%) of these eligible patients

Table 1 Univariate analysis of clinical characteristics, demography, source of bacteremia, severity, and laboratory data among patients treated with inappropriate antibiotics, appropriate antibiotics, or no antibiotic in the emergency department (ED) a

Clinical variables Patient number (%) P value

Inappropriate antibiotic, n = 98

Appropriate antibiotic, n = 330

No antibiotic, n = 26

old age, ?65years

53 (54.1)

184 (55.8)

12 (46.2)

.62

Gender, female

58 (59.2)

150 (45.5)

16 (61.5)

.03

Initial syndrome in the ED

Sepsis

63 (64.3)

223 (67.6)

12 (46.2)

.08

febrile neutropenia

1 (1.0)

5 (1.5)

0 (0)

.77

Characteristics of bacteremia

Community-onset

58 (59.2)

306 (92.7)

22 (84.6)

b.0001

Polymicrobial

15 (15.3)

21 (6.4)

1 (3.8)

.01

Major comorbidities, n (%)

Cardiovascular disease

53 (54.1)

151 (45.8)

14 (53.8)

.29

Diabetes mellitus

39 (39.8)

124 (37.6)

5 (19.2)

.14

Malignancy

26 (26.5)

66 (20.0)

6 (23.1)

.37

Chronic renal insufficiency

24 (24.5)

46 (13.9)

7 (26.9)

.02

Liver cirrhosis

11 (11.2)

48 (14.5)

2 (7.7)

.47

Geriatric stroke

7 (7.1)

22 (6.1)

2 (7.7)

.97

Chronic obstructive pulmonary disease

5 (5.1)

15 (4.5)

0 (0)

.51

Major source of bacteremia

Urinary tract infection

25 (25.5)

135 (40.7)

2 (7.7)

b.001

Primary bacteremia

17 (17.3)

42 (12.7)

14 (53.8)

b.001

Skin and soft-tissue infection

15 (15.3)

25 (7.6)

0 (0)

.02

Lower respiratory tract infection

14 (14.3)

29 (8.8)

3 (11.5)

.27

Biliary tract infection

9 (9.2)

29 (8.8)

2 (7.7)

.97

Vascular catheter

7 (7.1)

8 (2.4)

1 (3.8)

.08

Infective endocarditis

4 (4.1)

13 (3.9)

3 (11.5)

.18

Intra-abdominal infection

3 (3.1)

15 (4.5)

1 (3.8)

.81

Liver abscess

2 (2.0)

4 (7.3)

0 (0)

.06

Laboratory data in the ED, mean +- SD

Leukocyte (103/mm3)

13.5 +- 6.8

12.1 +- 6.2

9.4 +- 4.6

.008

Thrombocyte (103/mm3)

229.9 +- 135.3

200.6 +- 111.6

198.8 +- 111.6

.09

Blood urea nitrogen (mg/dL)

35.6 +- 29.0

27.6 +- 23.6

40.2 +- 32.0

.004

Serum creatinine (mg/dL)

2.4 +- 2.3

1.8 +- 2.4

3.2 +- 3.7

.004

Serum glucose (mg/dL)

185.5 +- 107.2

193.8 +- 119.0

130.0 +- 69.1

.10

serum albumin (g/dL)

3.1 +- 0.9

3.1 +- 0.7

3.7 +- 0.6

.14

C-reactive protein (mg/L)

114.9 +- 99.2

117.1 +- 117.5

52.1 +- 61.3

.03

SD indicates standard deviation.

a Data are expressed as numbers of the patients (percentages), unless specifically indicated.

were from other hospitals. The mean age was 64.6 years, and 224 patients (49.3%) were female. Comorbidities included cardiovascular disease (218 patients, 48.0%), dia- betes mellitus (168, 37.0%), malignancy (98, 21.6%),

chronic renal insufficiency (77, 17.0%), liver cirrhosis (61,

13.4%), old stroke (31, 6.8%), and chronic pulmonary

disease (20, 4.4%).

Of the 454 ED patients with true bacteremia, 323 patients (71.1%) were admitted to the general wards, and 57 (12.6%) to the ICU, whereas only eight (1.7%) were transferred to other hospitals. Ten (2.2%) patients died during the stay in the ED, and 36 (7.9) died after hospitalization; no deaths were reported in the 28 days following discharge from the ED. Therefore, the crude mortality at discharge was 10.1% (46 patients), and 9.2% (42 patients) died within 28 days after the ED visit.

Clinical variables of bacteremic patients with inappropriate, appropriate, and no antibiotic therapy

Univariate analyses were used to compare groups of patients who had been treated with inappropriate antibi- otics, with appropriate antibiotics, and with no antibiotics. Variables for comparison included clinical characteristics,

Table 2 Comparisons of clinical outcome, severity, and duration of hospital stay among patients treated with inappropriate antibiotics, appropriate antibiotics, or no antibiotic in the emergency department (ED) a

Clinical variables

Patient number (%)

Inappropriate antibiotic,

n = 98

P value

Appropriate antibiotic,

n = 330

No antibiotic, n = 26

demography, source of bacteremia, severity, and the results of clinical pathology laboratory examinations in the ED, as shown in Table 1. Patients who were female, or who had cardiovascular disease, primary bacteremia, high Blood urea nitrogen , or high serum creatinine in the ED were more likely to have received no antibiotic therapy during the ED stay. Patients with Community-onset bacteremia, bacteremia due to UTI, or high C-reactive protein in the ED were more likely to have received appropriate antibiotic therapy in the ED. Moreover, high proportions of patients with polymicrobial bacteremia, bacteremia due to skin and soft-tissue infections, or high leukocyte counts were observed among patients receiving inappropriate antibiotic therapy.

Clinical outcome of bacteremic patients with inappropriate, appropriate, and no antibiotic therapy

Univariate analyses also were used to compare the clinical outcome, severity, and duration of hospital stay among groups of patients treated with inappropriate antibiotics, with appropriate antibiotics, and with no antibiotic, as shown in Table 2. Higher 28-day mortality rates and higher propor- tions of hospitalization through the ED were observed in the

28-day crude mortality rate

16 (16.3)

23 (7.0)

3 (11.5)

.01

Severity-of-illness markers in the ED

Hospitalization through the ED

84 (85.7)

280 (84.8)

16 (61.5)

.007

Pittsburgh bacteremia score >=4 points

20 (20.4)

62 (18.8)

2 (7.7)

.32

Admitted to ICU through the ED

16 (16.3)

37 (11.2)

4 (15.4)

.36

Duration in the ED (hours), mean +- SD

time to antibiotic administration

2.5 +- 4.8

2.5 +- 5.4

NA

.99

Length of stay

33.4 +- 12.8

32.2 +- 8.4

31.0 +- 8.8

.28

Duration of hospitalization (days), mean +- SD

Hospital stay

17.3 +- 18.0

15.6 +- 16.9

20.5 +- 20.2

.44

ICU stay

2.2 +- 12.2

1.5 +- 7.2

2.2 +- 6.5

.69

Types of antibiotic administered in the ED

First generation cephalosporins

40 (40.8)

119 (36.1)

NA

.29

Second generation cephalosporins

11 (11.2)

43 (13.0)

NA

.63

Third generation cephalosporins

19 (19.4)

77 (23.3)

NA

.41

Aminopenicillin/BLI

17 (17.3)

56 (17.0)

NA

.93

Ureidopenicillin/BLI

3 (3.0)

3 (0.9)

NA

.13

Fourth generation cephalosporins

2 (2.0)

12 (3.6)

NA

.74

Fluoroquinolones

2 (2.0)

9 (2.7)

NA

1.00

Ureidopenicillins

2 (2.0)

4 (1.2)

NA

.62

Penicillin G

2 (2.0)

1 (0.3)

NA

.13

Glycopeptides

0 (0)

5 (1.5)

NA

.59

Combination with an aminoglycoside

0 (0)

16 (4.8)

NA

.03

BLI indicates ?-lactamase inhibitor; ICU, intensive care unit; NA, not applicable; SD, standard deviation.

a Data are expressed as the numbers of patients (percentages), unless specifically indicated.

patient group treated with inappropriate antibiotics. More- over, the time to antibiotic administration, length of ED stay, and type of antibiotic used in the ED were similar between the patient group with appropriate antibiotic and that with inappropriate antibiotic. Notably, similar proportions of cri-

tically ill patients (Pittsburgh bacteremia score >=4 points in the ED or admission to ICU through the ED) were observed

in these three groups (P = .32 or P = .36; respectively).

In further analyses of the survival curves (Fig. 1), the dif- ference in 28-day mortality among these three groups demons- trated significance in all bacteremic patients (P = .01); moreover, this difference appeared more significant in critically ill patients

with Pittsburgh bacteremia scores >=4 points (P = .002).

Microbiological analysis

Due to 37 episodes of polymicrobial bacteremia, a total of 494 bacteremic pathogens were isolated from the 454 eligible patients. These pathogens were predominantly Gram-negative aerobes (354 isolates, 71.6%). Of the 494 bacteremic isolates, Escherichia coli (206, 41.7%) and Klebsiella species (81, 16.4%) were the most frequently observed Gram-negative aerobes; staphylococci (57, 11.5%) and streptococci (60, 12.1%) were the most frequently observed Gram-positive aerobes, and only six anaerobes (Clostridium species, 2 isolates; Bacteroides species, 4) were identified in our population. Table 3 shows the differences in proportions of bacteremic isolates among groups of patients treated with inappropriate antibiotics, with appropriate antibiotics, and with no antibiotic. Patients with bacteremia due to E. coli or Klebsiella species were more likely to have received appropriate antibiotic therapy in the ED, whereas patients infected by extended-spectrum beta-lactamase pro-

ducers accounted for 5.3% (11 patients) of those with E. coli bacteremia and 9.3% (7) of those with Klebsiella bacteremia. However, patients with bacteremia due to Pseudomonas aeruginosa, Enterobacter cloacae, Burkholderia pseudomal- lei, methicillin-resistant Staphylococcus aureus (MRSA), or enterococci were more likely to have received inappropriate antibiotic therapy in the ED. Notably, among the patients with bacteremia due to these microorganisms receiving inappro- priate empirical therapy, most came from the community, including MRSA (11 of 18 patients, 61.1%), enterococci (10 of 13, 76.9%), P. aeruginosa (10 of 11, 90.9%), E. cloacae

(5 of 5, 100%), and B. pseudomallei (3 of 3, 100%).

Factors related to inappropriate empirical antibiotics in the ED

To determine the factors related to inappropriate empirical antibiotic therapy among bacteremic patients visiting the ED, those receiving no antibiotic therapy in the ED were pooled together with those treated by inappropriate antibiotics. The relationship of inappropriate empirical antibiotic therapy and clinical variables in the ED was analyzed by univariate ana- lyses (Table 4). The following variables were significantly (positively) associated with inappropriate empirical therapy: Female gender; recent hospitalization within 4 weeks before the ED arrival; comorbidity with chronic renal insufficiency; and high serum creatinine (N1.5 mg/dL). Community-onset bacteremia was negatively associated with inappropriate empirical therapy. However, only one factor independently related to inappropriate therapy was community-onset bacteremia (odds ratio, 0.16; 95% confidence interval, 0.09- 0.28; P b .001) in the multivariate regression analysis.

Fig. 1 Kaplan-Meier survival curves comparing clinical outcomes among bacteremic patients treated in the emergency department (ED) with inappropriate antibiotics, appropriate antibiotics, or no antibiotic. (A) All bacteremic patients (P = .01). (B) Bacteremic patients with critical illness (Pittsburgh bacteremia score >=4 points) (P = .002).

Gram-negative aerobes Escherichia coli Pseudomonas aeruginosa Klebsiella species Enterobacter cloacae Morganella morganii Proteus species Burkholderia pseudomallei Acinetobacter species Salmonella enteritidis Serratia marcescens Vibrio vulnificus Aeromonas species Moraxella catarrhalis Chryseobacterium species Citrobactor species Haemophilus influenzae

Gram-positive aerobes Staphylococci

Methicillin-resistant S. aureus

Streptococci

Viridans streptococci Streptococcus pneumoniae Streptococcus bovis Streptococcus pyogenes Streptococcus agalactiae Streptococcus intermedius Enterococci

Anaerobes Bacteroides species Clostridium species

76 (65.5)

28 (24.1)

11 (9.4)

8 (6.9)

5 (4.3)

5 (4.3)

4 (3.4)

3 (2.5)

3 (2.5)

2 (1.7)

2 (1.7)

1 (0.9)

1 (0.9)

1 (0.9)

1 (0.9)

0 (0)

0 (0)

38 (32.7)

19 (16.4)

17 (14.7)

8 (6.9)

3 (2.5)

2 (1.7)

1 (0.9)

0 (0)

0 (0)

0 (0)

11 (9.4)

2 (1.7)

1 (0.9)

1 (0.9)

262 (74.4)

165 (46.8)

1 (0.3)

70 (19.9)

1 (0.3)

4 (1.1)

10 (2.8)

0 (0)

2 (0.6)

0 (0)

2 (0.6)

2 (0.6)

1 (0.3)

0 (0)

0 (0)

3 (0.9)

1 (0.3)

86 (24.4)

35 (9.9)

8 (2.3)

47 (13.3)

8 (2.3)

4 (1.1)

3 (0.9)

6 (1.7)

6 (1.7)

3 (0.9)

4 (1.1)

4 (1.1)

3 (0.9)

1 (0.3)

16 (61.5)

13 (50.0)

0 (0)

3 (11.5)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

10 (38.5)

3 (11.5)

1 (3.8)

5 (19.2)

2 (7.7)

2 (7.7)

1 (3.8)

0 (0)

0 (0)

0 (0)

2 (7.7)

0 (0)

0 (0)

0 (0)

.09 b.001 b.001

.004

.008

.07

.63

.01

.14

.05

.44

.80

.66

.23

.23

.54

.71

.08

.17

b.001

.09

.25

.04

.33

.34

.34

.59

b.001

.74

.89

.66

a Includes all microorganisms isolated in the episodes of monomicrobial and polymicrobial bacteremia.

Risk of mortality for bacteremic patients visiting the ED

Table 3 Comparison of bacteremic isolates among patients treated with inappropriate antibiotics, appropriate antibiotics, or no antibiotic in the emergency department (ED) a

Microorganisms Numbers of isolates (%)

Inappropriate antibiotic, Appropriate antibiotic, n=116 n=352

No antibiotic, n=26

P

value

The association between several clinical variables, including age, gender, major comorbidities, initial syndrome, clinical pathology laboratory parameters in the ED, severity- of-illness markers, appropriate antibiotic treatment in the ED, source of bacteremia, and 28-day mortality were examined by univariate analyses, as shown in Table 5. The following variables were significantly (positively) associated with

28-day mortality: old age (>=65 years); female gender; inappropriate antibiotic therapy in the ED; high Pittsburgh bacteremia score in the ED (>=4 points); admission to the ICUs; initial syndrome with sepsis in the ED; comorbidity

with malignancy; bacteremia due to lower respiratory tract infection; primary bacteremia; and high serum creatinine (N1.5 mg/dL), high BUN (N20 mg/dL), hypoglycemia (b45 mg/dL), or hypoalbuminemia (b3.5 g/dL) in the ED.

Moreover, bacteremia due to UTI was negatively associated with 28-day mortality.

In multivariate analyses (Table 6), five variables were independently associated with 28-day mortality: high Pittsburgh bacteremia score (>=4 points; odds ratio [OR], 11.16; 95% confidence interval [CI], 4.98-24; P b .001);

high BUN (N20 mg/dL; OR, 2.75; 95% CI, 1.13-6.72; P =

.03); inappropriate antibiotic therapy in the ED (OR, 2.26; 95% CI, 1.01-5.13; P = .04); malignancy (OR, 3.17; 95% CI,

1.38-7.31; P b .007); and bacteremia due to UTI (OR, 0.14; 95% CI, 0.04-0.49; P = .002).

Discussion

In the present study, bacteremic patients treated with appropriate empirical antibiotics in the ED had better

Old age, ?65 years Female

Events during the previous 4 weeks Prior hospitalization

Invasive procedures Surgery

Community-onset bacteremia Initial syndrome

Spesis

Febrile Neuropenia

Pittsburgh bacteremia score >=4 points

Comorbidities, n (%) Hypertension Diabetes mellitus Malignancy

Chronic renal insufficiency Liver cirrhosis

Coronary artery diseases Congestive heart failure Old stroke

Chronic obstructive pulmonary disease

Laboratory data in the ED Leukocyte N9,000/mm3 Platelet b100,000/mm3

Blood urea nitrogen N20 mg/dL Serum creatinine N1.5 mg/dL

C-reactive protein N100 mg/L Serum glucose b45 mg/dL Serum albumin b3.5g/dL

65

(52.4)

184 (55.8)

.52

74

(59.7)

150 (45.5)

.007

96

(77.4)

285 (86.4)

.02

7

(5.6)

7 (2.1)

.06

4

(3.2)

2 (0.6)

.05

80

(64.5)

306 (92.7)

b.001

75

(60.5)

223 (67.6)

.15

1

(0.5)

5 (1.5)

1.00

22

(17.7)

62 (18.8)

.79

61

(49.2)

138 (41.8)

.15

44

(35.5)

124 (37.6)

.68

32

(25.8)

66 (20.0)

.18

31

(25.0)

46 (13.9)

.005

13

(10.5)

48 (14.5)

.25

13

(10.5)

21 (6.4)

.13

11

(8.9)

22 (6.1)

.29

9

(7.3)

22 (6.7)

.82

5

(4.0)

15 (4.5)

.81

78

(62.9)

217 (65.8)

.57

15

(12.1)

61 (18.5)

.10

73

(58.9)

167 (50.6)

.11

50

(40.3)

80 (24.2)

.001

39

(31.5)

119 (36.1)

.35

3

(2.4)

2 (0.6)

.12

24

(19.4)

56 (17.0)

.55

a 26 patients receiving no antibiotic therapy in the ED were categorized as receiving inappropriate antibiotics.

outcomes than those treated with inappropriate antibiotics and those receiving no antibiotic therapy. Especially in critically ill patients, an appropriate antibiotic therapy yielded a favorable clinical outcome. In the further analyses by multivariate regression, inappropriate antibiotic therapy was also an independent risk factor for mortality. In addition, the current study also demonstrated that the presence of any of several pathogens (including P. aeruginosa, E. cloacae, B. pseudomallei, MRSA, and enterococci) correlated with treatment with inappropriate antibiotics.

Table 4 Univariate analysis of clinical factors available in the emergency department (ED) for inappropriate empirical antibiotics among all bacteremic patients visited the ED

Variables in the ED Patient with inappropriate empirical antibiotics, (%)

Yes, n = 124 a

No, n = 330

P

values

Previous studies have demonstrated consistently that the administration of appropriate empirical antibiotics strongly decreases the mortality rate from bacteremia in patients with community-acquired and nosocomial bloodstream infections [2,6-8]. As in previous reports, the present study revealed lower mortality rates in bacteremic patients treated with appropriate empirical antibiotics. However, those previous investigations employed hospital-based populations (e.g., upon hospital admission or in ICU settings), while the

present study focused solely on bacteremic patients who visited the ED.

Several investigations have demonstrated the clinical impact of delays in administering the first dose of antibiotics in ED patients with various infections, including severe sepsis [16], pneumonia [17] and meningitis [18]. However, to our knowledge, no ED-based study focusing on bacteremia has been published in the English literature. Moreover, no investigation dealing with the relation between the length of delay in administering appropriate antibiotics and clinical outcome in infected patients is available. In order to avoid difficulty in evaluating the effect of antibiotics due to short therapeutic duration in the ED, the bacteremic patients with ED stays of less than 24 hours were excluded from our population. In addition, the present study revealed similar periods of time to antibiotic administration and duration of ED stay among patients with the appropriate antibiotic therapy compared to those receiving inappropriate antibiotic therapy. Therefore, the potential influence of

Old age, ?65years

30

(71.4)

219 (53.2)

2.2

(1.1 – 4.4)

.02

Gender, female

28

(66.7)

196 (47.6)

2.2

(1.1 – 4.3)

.02

Inappropriate antibiotic therapy in the ED a

19

(45.2)

105 (25.2)

1.36

(1.0 – 1.8)

.006

Severity-of-illness markers in the ED

Pittsburgh bacteremia score >=4 points

28

(66.7)

56 (13.6)

12.7

(6.3 – 25.6)

b.001

Admitted to ICU through the ED

11

(26.2)

46 (11.2)

2.8

(1.3 – 5.9)

.005

Initial syndrome in the ED

Sepsis

34

(81.0)

264 (64.1)

2.4

(1.1 – 5.3)

.03

Febrile neutropenia

2

(4.8)

4 (1.0)

5.1

(0.9 – 28.7)

1.00

Major comorbidities

Cardiovascular disease

24

(57.1)

194 (47.1)

1.5

(0.8 – 2.8)

.61

Diabetes mellitus

18

(42.9)

150 (36.4)

1.3

(0.7 – 2.5)

.41

Malignancy

17

(40.5)

81 (19.7)

2.8

(1.4 – 5.4)

.002

Chronic renal insufficiency

10

(23.8)

67 (16.3)

1.6

(0.8 – 3.4)

.21

Liver cirrhosis

7

(16.7)

54 (13.1)

1.3

(0.6 – 3.1)

.52

Major source of bacteremia

Lower respiratory tract infection

10

(23.8)

36 (8.7)

3.3

(1.5 – 7.2)

.005

Primary bacteremia

12

(28.6)

61 (14.8)

2.3

(1.1 – 4.7)

.02

Urinary tract infection

4

(9.5)

158 (38.3)

0.2

(0.1 – 0.5)

b.001

Skin and soft-tissue infection

4

(9.5)

36 (8.7)

1.1

(0.4 – 3.3)

.77

Vascular catheter-related infection

4

(9.5)

12 (2.9)

3.5

(1.0 – 11.4)

.05

Liver abscess

2

(4.8)

24 (5.8)

0.8

(0.2 – 3.5)

1.00

Biliary tract infection

2

(4.8)

38 (9.2)

0.5

(0.1 – 2.1)

.52

Laboratory data in the ED

Leukocyte N9,000/mm3

24

(57.1)

271 (61.8)

0.7

(0.4 – 1.3)

.26

Platelet b100,000/mm3

11

(26.2)

65 (15.8)

1.9

(0.9 – 3.9)

.09

Blood urea nitrogen N20 mg/dL

33

(78.6)

207 (50.2)

3.6

(1.6 – 7.7)

b.001

Serum creatinine N1.5 mg/dL

22

(52.4)

108 (26.2)

3.1

(1.6 – 5.9)

b.001

C-reactive protein N100 mg/L

20

(47.6)

138 (33.5)

1.8

(0.9 – 3.4)

.07

Serum glucose b45 mg/dL

5

(11.9)

0 (0)

57.1

(6.5 – 501.9)

b.001

Serum albumin b3.5g/dL

13

(31.0)

67 (16.3)

2.3

(1.1 – 4.7)

.02

ICU indicates intensive care unit.

a For analysis of this variable, 26 patients receiving no antibiotic therapy in the ED were regarded as receiving inappropriate antibiotics.

distinct antibiotic therapeutic periods in these two patient groups was diminished, although the type of antimicrobial agent used in the ED may have contributed to the difference of clinical outcome between these two groups.

Table 5 Univariate analysis of risk factors of 28-day mortality in bacteremic adults visiting the emergency department (ED)

Variables Numbers of patients (%)

Non-survivor, n = 42

Survivor, n = 412

Odds ratio (95% confidence interval)

P

value

The current study also demonstrated the clinical factors related to inappropriate antibiotic therapy among the

Table 6 Multivariate analyses of risk factors of 28-day mortality in all bacteremic adults visiting the emergency department (ED)

bacteremic patients visiting the ED. However, only an independent factor, community-onset bacteremia, was neg- atively associated with inappropriate therapy in the multi- variate analyses. It is not surprised that healthcare-associated infections were usually caused by antimicrobial-resistant microorganisms, such as P. aeruginosa and MRSA, and

Clinical variables

Odds ratio

95% confidence interval

P value

ED variables

Pittsburgh bacteremia score >=4 points

11.16

4.98-24.99

b.001

Blood urea nitrogen N20 mg/dL

2.75

1.13-6.72

.03

Inappropriate antibiotic therapy a

2.26

1.01-5.13

.04

Comorbidities with malignancy

3.17

1.38-7.31

.007

Bacteremia due to urinary tract infections

0.14

0.04-0.49

.002

a For analysis of this variable, 26 patients receiving no antibiotic therapy in the ED were regarded as receiving inappropriate antibiotics.

often associated with inappropriate empirical therapy, as seen in previous hospital-based investigations [6,19,20]. Therefore, using hospital protocols designed for their specific hospital, to determine which initial antibiotic to be uses in specific circumstances, likely would improve the quality of care and decrease the chance of initial inappro- priate antibiotic use in the ED. However, to create the hospital-based protocol, the extent of antibiotic resistance in the community and long-term healthcare facilities and presumed sources of the infections should be taken into consideration. So local surveillance of antibiotic resistance in the community and healthcare institutions should be considered for such protocols first.

In addition to inappropriate antibiotics in the ED, several independent risk factors for mortality were identified by multivariate analysis of our population, including high Pittsburgh bacteremia scores, high BUN, and comorbidity with malignancy. Of these risk factors, the most powerful was the Pittsburgh bacteremia score. The value of this score for determining the severity of illness at the onset of bacteremia by microorganisms, such as E. coli [21], K. pneumonia [12,21], streptococci [22], and staphylococci [23], has been discussed in previous investigations, and these microorganisms were the most frequently seen pathogens in our population. Therefore, it is not surprising that the Pittsburgh bacteremia score was the best predictor of mortality among all bacteremic patients visiting the ED. As in a previous ED-based investigation of bacteremia [24], ma- lignancy was also an independent risk factor for mortality. Moreover, consistent with previous hospital- based studies focusing on Gram-negative bacteremia [25,26], high BUN levels were found to be a good predictor of mortality; the predictive value of elevated BUN may reflect the predominance of Gram-negative bacteremia (354/494, 71.6%) in our population.

Limitations

There are several limitations inherent to the design of this study. First, to assess adequate the clinical effect of empirical antibiotic therapy in the ED, patients with ED stays of less than 24 hours were excluded and it may led to some selective bias. However, only 12% (64) of 518 bacteremic patients were excluded from our study, so little influence on the results of current study should be considered. Second, to avoid any potential ethical conflict, it was necessary to review medical records and patient outcomes retrospectively; such an approach has inherent limits due to possible confounding data and bias. Finally, although we examined all bacteremic patients who visited the ED for a long period, the study was conducted in a single ED of a tertiary hospital. Therefore, our findings may not be generalizable to other populations.

Conclusions

By focusing on bacteremic patients visiting the ED, this study demonstrated the possible disadvantage of inappropri- ate antibiotic treatment in our ED population, especially in critically ill patients. Notably, patients in our population with bacteremia due to P. aeruginosa, MRSA, or enterococci were more likely to receive inappropriate empirical antibiotic therapy. To avoid inappropriate antibiotic therapy, further studies should focus on establishing the hospital protocols designed for the hospital where initial antibiotic therapy was recommended in an individual circumstance.

References

  1. Ispahani P, Pearson NJ, Greenwood D. An analysis of community and hospital-acquired bacteraemia in a large teaching hospital in the United Kingdom. Q J Med 1987;63:427-40.
  2. Leibovici L, Greenshtain S, Cohen O, Mor F, Wysenbeek AJ. Bacteremia in febrile patients. A clinical model for diagnosis. Arch Intern Med 1991;151:1801-6.
  3. Bates DW, Pruess KE, Lee TH. How bad are bacteremia and sepsis? Outcomes in a cohort with suspected bacteremia. Arch Intern Med 1995;155:593-8.
  4. Haug JB, Harthug S, Kalager T, Digranes A, Solberg CO. Bloodstream infections at a Norwegian university hospital, 1974-1979 and 1988- 1989: changing etiology, clinical features, and outcome. Clin Infect Dis 1994;19:246-56.
  5. Strehlow MC, Emond SD, Shapiro NI, Pelletier AJ, Camargo Jr CA. National study of emergency department visits for sepsis, 1992 to 2001. Ann Emerg Med 2006;48:326-31.
  6. Leibovici L, Shraga I, Drucker M, Kongigsberger H, Samra Z, Pitlik SD. The benefit of appropriate empirical antibiotic treatment in patients with bloodstream infection. J Intern Med 1998;244:379-86.
  7. Harbarth S, Garbino J, Pugin J, Romand JA, Lew D, Pittet D. Inappropriate initial antimicrobial therapy and its effect on survival in a clinical trial of immunomodulating therapy for severe sepsis. Am J Med 2003;115:529-35.
  8. 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.
  9. Ibrahim EH, Sherman G, Ward S, Fraser VJ, Kollef MH. The influence of inadequate Antimicrobial treatment of bloodstream infections on patient outcomes in the ICU setting. Chest 2000;118:146-55.
  10. Leone M, Bourgoin A, Cambon S, Dubuc M, Albanese J, Martin C. empirical antimicrobial therapy of septic shock patients: adequacy and impact on the outcome. Crit Care Med 2003;31:462-7.
  11. Clinical and Labortaory Standards Institute. Performance standards for antimicrobial disk susceptibility tests, 17 th informational sup- plement. document M100-S16. Wayne (Pa): CLSI; 2007.
  12. 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.
  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. 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.
  15. Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988;16:128-40.
  16. Nguyen HB, Rivers EP, Abrahamian FM, Moran GJ, Abraham E, Trzeciak S. Severe sepsis and septic shock: review of the literature and emergency department management guidelines. Ann Emerg Med 2006;48:28-54 .
  17. Pines JM, Morton MJ, Datner EM, Hollander JE. Systematic delays in antibiotic administration in the emergency department for adult patients admitted with pneumonia. Acad Emerg Med 2006;13:939-45.
  18. Proulx N, Frechette D, Toye B, Karvcik S. Delays in the administration of antibiotics are associated with mortality from adult acute Bacterial meningitis. QJM 2005;98:291-8.
  19. McDonald JR, Friedman ND, Stout JE, Sexton DJ, Kaye KS. Risk factors for ineffective therapy in patients with bloodstream infection. Arch Intern Med 2005;165:308-13.
  20. Kang CI, Kim SH, Park WB, Lee KD, Kim HB, Kim EC, et al. Blood- stream infections caused by antibiotic-resistant gram-negative bacilli: risk factors for mortality and impact of inappropriate initial antimicro- bial therapy on outcome. Antimicrob Agents Chemother 2005;49:760-6.
  21. Lee CI, Lee NY, Yan JJ, Lee HC, Ko NY, Chang CM. Extended- spectrum beta-lactamase-producing phenotype signifies a poor prognosis for patients with cefpodoxime-resistant Escherichia coli or Klebsiella pneumoniae bacteremia. J Microbiol Immunol Infect 2009;42:303-9.
  22. Yu VL, Chiou CC, Feldman C, Ake O, Jordi R, Arthur JM, et al. An international prospective study of pneumococcal bacteremia: correla- tion with in vitro resistance, antibiotics administered, and clinical outcome. Clin Infect Dis 2003;37:230-7.
  23. Chang FY, Peacock Jr JE, Musher DM, Triplett P, MacDonald BB, Mylotte JM, et al. Staphylococcus aureus bacteremia: recurrence and the impact of antibiotic treatment in a prospective multicenter study. Medicine (Baltimore) 2003;82:333-9.
  24. 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.
  25. Kreger BE, Craven DE, McCabe WR. Gram-negative bacteremia. IV. Re-evaluation of clinical features and treatment in 612 patients. Am J Med 1980;68:344-55.
  26. Hodgin UG, Sanford JP. Gram-negative rod bacteremia. An analysis of 100 patients. Am J Med 1965;39:952-60.

Leave a Reply

Your email address will not be published. Required fields are marked *