Clinical characteristics associated with mortality of patients with anaerobic bacteremia

Anaerobe 39 (2016) 45e50

Contents lists available at ScienceDirect

Anaerobe journal homepage: www.elsevier.com/locate/anaerobe

Clinical microbiology

Clinical characteristics associated with mortality of patients with anaerobic bacteremia Takumi Umemura a, b, Yukihiro Hamada c, Yuka Yamagishi a, c, Hiroyuki Suematsu c, Hiroshige Mikamo a, c, * a b c

Department of Clinical Infectious Diseases, Aichi Medical University Graduate School of Medicine, Japan Department of Pharmacy, Tosei General Hospital, Japan Department of Infection Control and Prevention, Aichi Medical University Hospital, Japan

a r t i c l e i n f o

a b s t r a c t

Article history: Received 17 March 2015 Received in revised form 15 February 2016 Accepted 18 February 2016 Available online 20 February 2016

The presence of anaerobes in the blood stream is known to be associated with a higher rate of mortality. However, few prognostic risk factor analyses examining whether a patient's background characteristics are associated with the prognosis have been reported. We performed a retrospective case-controlled study to assess the prognostic factors associated with death from anaerobic bacteremia. Seventy-four patients with anaerobic bacteremia were treated between January 2005 and December 2014 at Aichi Medical University Hospital. The clinical information included drug susceptibility was used for analysis of prognostic factors for 30-day mortality. Multivariate logistic analyses revealed an association between the 30-day mortality rate and malignancy (OR: 3.64, 95% CI: 1.08e12.31) and clindamycin resistance (OR: 7.93, 95% CI: 2.33e27.94). The result of KaplaneMeier analysis of mortality showed that the 30-day survival rate was 83% in clindamycin susceptible and 38.1% in clindamycin resistant anaerobes causing bacteremia. The result of log-rank test also showed that susceptibility to clindamycin affected mortality (P < 0.001). Our results indicated that malignancy and clindamycin susceptibility could be used to identify subgroups of patients with anaerobic bacteremia with a higher risk of 30-day mortality. The results of this study are important for the early and appropriate management of patients with anaerobic bacteremia. © 2016 Elsevier Ltd. All rights reserved.

Handling Editor: Elisabeth Nagy Keywords: Anaerobe Anaerobic bacteremia Blood stream infection Prognostic factor Clindamycin

1. Introduction An overall increase in the number of immunocompromised patients with various types of disease has reflected some of the changes in patient demographics over recent years. In addition to this phenomenon, an increase in the frequency of the isolation of anaerobic bacteria has been noted, with special emphasis on Bacteroides species from blood cultures [1,2]. It is important to recognize the risk factors of anaerobic bacteremia in order to enable quick and appropriate actions. Anaerobic bacteremia has been associated with associated with higher mortality, thus requiring appropriate therapy [3,4]. Risk factors for anaerobic bacteremia include an underlying malignancy, diseases of the gastrointestinal and genitourinary tracts, diabetes, and a history of gastrointestinal

* Corresponding author. Yazakokarimata, Nagakute, Aichi, 480-1195, Japan. Tel.: þ81 561623311; fax: þ81 561611842. E-mail address: [email protected] (H. Mikamo). http://dx.doi.org/10.1016/j.anaerobe.2016.02.007 1075-9964/© 2016 Elsevier Ltd. All rights reserved.

surgery [2,5e10]. The presence of anaerobes in the blood stream is known to be associated with mortality [11,12]. However, few risk factor analyses have been performed with the goal of investigating whether a patient's background characteristics are correlated with the prognosis. Therefore, we performed a retrospective case-controlled study to assess the risk factors associated with death from anaerobic bacteremia.

2. Materials and methods 2.1. Setting This study was conducted from January 2005 to December 2014 at Aichi Medical University Hospital (1014 beds). For blood culture, BD rezun bottles (Becton Dickinson, Tokyo, Japan) were used, and the BD Bactec™ FX blood culture system (Becton Dickinson, Tokyo, Japan) was utilized for the growth and detection of

46

T. Umemura et al. / Anaerobe 39 (2016) 45e50

anaerobes. Each pair of aerobic and anaerobic bottles was incubated for a week. Anaerobic bacteria were identified using the RapID-ANA II system (Remel, Kansas, USA) and additional biochemical tests. All the patient identifiers were removed before analysis. This study was approved by the ethical committee of Aichi Medical University. 2.2. Definition of patients with anaerobic bacteremia Patients with blood cultures positive for anaerobic bacteria were retrospectively identified using bacterial data included in electronic health records. Bacteremia was deemed clinically significant when the patient had one or more positive blood cultures and met one of the following criteria: white blood cell count <4000 and > 12,000/ mL; temperature >38  C; or physical, pathological, or surgical evidence consistent with infection (e.g., isolation of anaerobic bacteria from a source other than blood) [12,13]. Patients with Propionibacterium species isolated from blood cultures were defined as clinically significant using previously cited criteria [12,14]. 2.3. Variables The following data were obtained for the analysis of prognostic factors for 30-day mortality: age, sex, white blood cell count, Creactive protein, diabetes mellitus and malignancy as underlying diseases, intensive care unit (ICU) admission, primary disease, current administration of steroids or immunosuppressants for >4 weeks, clear source of bacteremia, initial inappropriate therapy, Bacteroides bacteremia, and multibacterial (or monobacterial) bacteremia, with drug resistance defined as intermediate or resistant based on drug susceptibility testing. Inappropriate therapy was defined as failure to administer anti-anaerobic drug therapy when a blood culture was submitted. Anti-anaerobic drugs were defined as beta-lactam/beta-lactamase inhibitor combinations, oxacephem, carbapenem, clindamycin, and other drugs identified by susceptibility testing. 2.4. Antibiotic susceptibility testing Antibiotic susceptibility testing for clindamycin, amoxicillinclavulanic acid, cefmetazole, imipenem, and moxifloxacin complied with the standards of Clinical and Laboratory Standards Institute (CLSI). The final inoculum for anaerobic bacteremia was 106 colony forming units (CFU) per mL. Brucella broth was supplemented with hemin (5 mg/mL), vitamin K1 (1 mg/mL) and lysed horse blood. All the analyses were performed in anaerobic chambers. Cultures were incubated in an anaerobic atmosphere at 37  C for 48 h. The strains were considered resistant according to break points defined by the CLSI. 2.5. Statistical analysis Qualitative and stratified continuous variables were compared using the Fisher exact or Pearson c2 test. Continuous variables were compared using the Student t-test or the ManneWhitney U test, as appropriate. Multivariate logistic analyses were used for the logistic regression models. Variables achieving a probability (P) value of <0.2 in the univariate logistic analyses were included in the multivariate analysis [15]. Survival was analyzed using the KaplaneMeier method. Plots were compared using the log-rank test. Predictive values are presented as the odds ratios (ORs) with respective 95% confidence intervals (CI). Two-tailed P values < 0.05 were considered statistically significant. Analyses were performed using IBM SPSS Statistics 19 (IBM®).

3. Results Seventy-four patients with anaerobic bacteremia were eligible for this study. Besides patients with monobacterial anaerobic bacteremia, also those were included who had multibacterial bacteremia caused by two different anaerobic species or an anaerobic and an aerobic bacteria were isolated from the blood culture at the same time. Table 1 shows the anaerobic bacteria detected in blood cultures and the results from drug susceptibility testing. The most frequently isolates anaerobic bacteria were Bacteroides, followed by Clostridium, Prevotella, and Peptoniphilus. Some anaerobes remained unidentified because the identification of anaerobic bacteria was not routinely performed. Only one deceased patient had multibacterial bacteremia involving Bacteroides distasonis and Bacteroides thetaiotaomicron. Unfortunately, not all the anaerobic blood culture isolates were tested for all antianaerobic drugs during the routine procedures in the laboratory (Table 1). However, we noticed that more resistant isolates among the Bacteroides strains to clindamycin, amoxicillin/clavulanic acid, cefmetazole, and moxifloxacin than among other anaerobic blood culture isolates. Out of 13 Bacteroides fragilis isolates, 5 isolates (38.4%) were susceptible to clindamycin, whereas 8 isolates (61.6%) were resistant. Among the non-fragilis Bacteroides isolates, 4 (33.3%) strains were susceptible to clindamycin and 8 (66.7%) showed resistance. Among the Bacteroides strains, which had resistance data for amoxicillin/clavulanic acid were available, 10 (55.6%) were susceptible and 8 (44.4%) were resistant. The same tendency was seen for cefmetazole and moxifloxacin. All the tested Bacteroides blood culture isolates were fully susceptible to imipenem. We found only one anaerobic Gram-negative bacillus (unidentified) which showed resistance to imipenem. Table 2 shows the possible sources of the anaerobic bacteremia of the patients according to whether they survived or died within 30 days, according to their other clinical data. In general, a gastrointestinal origin of infection led to anaerobic bacteremia in both patients group was dominating. Regarding the origin of the bacteremia, the incidence of the intra-abdominal foci were significantly higher in the deceased patient group than in the survivors (P ¼ 0.030), while no other source of anaerobic bacteremia was significantly linked to unfavorable patients outcome (Table 2). Table 3 shows the results of univariate analyses of the included patient characteristics. Variables with P values < 0.2 included malignancy (P ¼ 0.017), a history of ICU admission (P ¼ 0.060), inappropriate therapy (P ¼ 0.174), and clindamycin resistance (P < 0.001). Univariate analysis was carried out only for clindamycin and imipenem resistance. Susceptibility testing was not consistently conducted for other drugs. Table 4 shows the results of a multivariate logistic regression analysis of prognostic factors associated with the mortality of anaerobic bacteremia. No statistically significant differences in ICU admission (P ¼ 0.266) or inappropriate therapy (P ¼ 0.287) were observed between the groups. Independent prognostic factors were malignancy (OR: 3.64, 95% CI: 1.08e12.31) and clindamycin resistance (OR: 8.06, 95% CI: 2.33e27.94). Table 5 shows the antimicrobial agents that were administered to all patients with anaerobic bacteremia. In the survival group, the antibiotic therapy of two patients was changed from ceftriaxone to carbapenem and piperacillin/tazobactam, respectively, according to the positive anaerobic blood culture results. Fig. 1 shows the KaplaneMeier survival analysis of mortality at 30 days of follow up according to clindamycin susceptibility. The 30-day survival rates were 83% for patients with clindamycinsensitive isolates and 38.1% for patients with clindamycinresistant isolates. A log-rank test showed that clindamycin susceptibility affected the mortality rate (P < 0.001).

T. Umemura et al. / Anaerobe 39 (2016) 45e50

47

Table 1 Anaerobic bacteria detected in the blood cultures of 74 patients and the results of their antibiotic susceptibility testing. Anaerobic bacteria

Bacteroides sp. Bacteroides fragilis Bacteroides thetaiotaomicron Bacteroides vulgatus Bacteroides caccae Bacteroides distasonis Not identified Clostridium perfringens Prevotella sp. Prevotella buccalis Prevotella oris Not identified Peptoniphilus sp. Peptoniphilus asaccharolyticus Peptoniphilus indolicus Parvimonas micra Ruminococcus gnavus Fusobacterium necrophorum Anaerobic Gram positive coccus Anaerobic Gram positive bacillus Anaerobic Gram negative bacillus

n

26 13 5a 3 1 2a 2 6 4 1 1 2 4 3 1 3 2 1 2 13 14

Clindamycin

Amoxicillin/clavulanic acid

Cefmetazole

Imipenem

Moxifloxacin

Susceptible

Resistant

Susceptible

Resistant

Susceptible

Resistant

Susceptible

Resistant

Susceptible

Resistant

9 5 1 1 1 0 1 6 3 1 1 1 4 3 1 3 2 1 1 13 11

16 8 4 2 0 1 1 0 1 0 0 1 0 0 0 0 0 0 1 0 3

10 7 3 0 NT NT 0 4 NT NT NT NT NT NT NT 3 2 1 1 2 1

8 4 2 1 NT NT 2 0 NT NT NT NT NT NT NT 0 0 0 0 0 0

11 10 1 0 NT NT 0 4 NT NT NT NT NT NT NT 3 2 1 1 1 1

7 1 4 1 NT NT 2 0 NT NT NT NT NT NT NT 0 0 0 0 1 0

25 13 5 3 1 1 2 6 4 1 1 2 4 3 1 3 2 1 2 13 14

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1

10 8 2 0 NT NT 0 3 NT NT NT NT NT NT NT 2 2 1 1 1 1

8 3 3 1 NT NT 2 1 NT NT NT NT NT NT NT 1 0 0 0 1 0

NT: not tested. a One patient had multibacterial bacteremia involving Bacteroides distasonis and Bacteroides thetaiotaomicron.

Table 2 Possible source of the anaerobic bacteremia according to the patients other clinical data. Source of bacteremia

Surviving patients (n ¼ 51)

Deceased patients (n ¼ 23)

p value*

Gastrointestinal tract Intraabdominal Hepatobiliary Urinary tract Skin and soft tissue Respiratory tract Others Unknown

17 1 2 9 3 2 1 16

7 4 1 3 2 2 1 3

0.805 0.030 1.000 0.743 0.643 0.584 e e

*chi-square test.

Table 3 The baseline characteristics and univariate logistic regression analysis of patient histories. Factors

Surviving patients (n ¼ 51)

Deceased patients (n ¼ 23)

Odds ratio (95% CI)

Age (mean ± SD) Aged (65 years) Male White blood cell count (mean ± SD) C-reactive protein (mean ± SD) Diabetes mellitus Malignancy Solid Humoral ICU admission history Use of immunosuppressive drugs Clear source of bacteremia Multibacterial bacteremia B. distasonis and B. thetaiotaomicron Combined with aerobic bacteria Bacteroides bacteremia Inappropriate therapy No anti-anaerobic activity Inactive against the organisms Imipenem resistance Clindamycin resistance

64.24 ± 21.62 34 30 11.12 ± 6.74 9.28 ± 9.40 9 18 (16) (2) 5 8 35 9 (0) (9) 15 16 (16) (0) 1 7

69.91 ± 12.33 17 14 9.55 ± 5.35 10.20 ± 9.25 7 15 (15) (0) 6 5 19 3 (1) (2) 10 11 (10) (1) 0 14

e 1.42 1.09 e e 2.04 3.44

(0.65e6.41) (1.22e9.65)

0.244 0.597 0.868 0.328 0.695 0.220 0.017

3.24 1.49 2.17 0.70

(0.88e12.04) (0.43e5.19) (0.64e7.43) (0.17e2.87)

0.060 0.527 0.210 0.743

SD: standard deviation. CI: Confidence interval.

(0.47e4.25) (0.40e3.00)

1.85 (0.67e5.13) 2.01 (0.73e5.50)

0.98 (0.94e1.02) 9.78 (3.08e31.08)

p value

0.236 0.174

1.000 <0.001

48

T. Umemura et al. / Anaerobe 39 (2016) 45e50

Table 4 Prognostic factors associated with 30-day mortality of patients with anaerobic bacteremia; multivariate logistic regression analysis. Factors

Surviving patients (n ¼ 51)

Deceased patients (n ¼ 23)

Odds ratio (95% CI)

p value

Malignancy ICU admission Inappropriate therapy Clindamycin resistance

18 5 16 7

15 6 11 14

3.64 2.45 1.95 7.93

0.037 0.266 0.287 0.001

(1.08e12.31) (0.50e12.52) (0.57e6.64) (2.33e27.94)

CI: Confidence interval.

Table 5 Antimicrobial agents administered to all patients with anaerobic bacteremia. Antimicrobial agents Monotherapy Carbapenems Tazobactam/piperacillin Sulbactam/ampicillin 4-th generation cephalosporins 3-th generation cephalosporins 2nd generation cephalosporins 1st generation cephalosporins Clindamycin Minocycline Combination therapy Carbapenems þ clindamycin Cephalosporins þ clindamycin Sulbactam/ampicillin þ clindamycin Pazufloxacin þ clindamycin no antibiotics

Surviving patients (n ¼ 51)

Deceased patients (n ¼ 23)

15* 8* 5 3 1 1 2 1 0

10 0 2 3 0 1 1 0 1

2 3 1 1 8

0 1 0 0 4

*Two patients changed from ceftriaxone after when anaerobic bacteria positive culture returned.

Fig. 1. The results of KaplaneMeier survival graph of mortality at 30 days of follow up according to clindamycin susceptibility and the results of log-rank test.

4. Discussion Antibiotic therapy has been shown to be necessary for the treatment of anaerobic bacteremia [12,16], and some epidemiology studies, including case-control studies, have reported the risk factors for anaerobic bacteremia [8,17]. However, few reports have examined patient prognostic risk factors for anaerobic bacteremia. Therefore, we conducted a retrospective case-control study to identify predictors of death from anaerobic bacteremia, including

drug susceptibility profiles. In this study, Bacteroides species had a high rate of clindamycin resistance (Table 1). In several countries, surveillance programs for the antimicrobial susceptibility of clinical isolates of anaerobic bacteria exist. Consequently, the number of reported cases of Bacteroides species resistant to antimicrobial agents has been increasing. Especially, non-fragilis Bacteroides species have been shown to have a greater likelihood of resistance than B.acteroides fragilis [17e21]. In the present study, Bacteroides bacteremia was not statistically associated with 30-day mortality;

T. Umemura et al. / Anaerobe 39 (2016) 45e50

however, 10 of the 25 patients (40%) who had Bacteroides bacteremia has died (Table 3), and this mortality rate was similar to that in a recent report [17]. On the other hand, there was only one unidentified Gram-negative anaerobic bacillus that was carbapenem resistant. Few reports on carbapenem-resistant anaerobic strains are available, but such resistance has been reportedly increasing over the last few years [22,23], and additional research into carbapenem-resistant strains is necessary. Some reports showed that multibacterial anaerobic sepsis was associated with an increased mortality rate [24,25]. In our study, however, only one of the patients who died within 30 days had mixed anaerobic bacteremia involving different Bacteroides species. Nine other survivors and 2 patients who died had mixed blood culture results, in which an anaerobic species and an aerobic species were simultaneously isolated (Table 3). Regarding primary disease was not associated with the 30-day mortality rate (Table 2). On the other hand, regarding the source of bacteremia, an intra-abdominal focus was associated with a high mortality rate (Table 3). Nguyen et al. reported that for Bacteroides bacteremia with an abdominal portal of entry, inactive therapy was associated with a 14-day mortality rate of 50%, whereas the mortality rate was only 14% in cases where active therapy had been instigated [11]. While anaerobes comprise most of the endogenous microflora and can cause bacteremia when mucosal or visceral damage facilitates their entry into the bloodstream [14], most data regarding anaerobic bacteremia comes from studies of surgical patients. In this study, few anaerobic bacteremia patients were identified as having an intra-abdominal portal of entry, so these patients were excluded from the multivariate logistic regression analysis. Future studies should include a larger number of cases with intra-abdominal foci. Blood stream infections due to aerobic and anaerobic bacteria were not associated with mortality (Table 3). The frequency of the presence of aerobic and anaerobic bacteria in the same bottle presented identical ratio as seen in past epidemiological reports [17]. In terms of biomarkers, the C-reactive protein level provides greater diagnostic information than an elevated temperature for the diagnosis of infection [26,27], with procalcitonin being superior to C-reactive protein [28]. In this study, the white blood cell count and the C-reactive protein level were not associated with the 30day mortality rate. Procalcitonin was not included in this analysis because it was not measured in all the cases. Bacteremia, including sepsis, can result in a high mortality rate, and risk prediction is necessary during the monitoring and management of patients with sepsis [29]. Age, septic shock, renal failure, and metabolic acidosis upon admission are factors associated with an increased risk of mortality in cases of sepsis [30]. In the present study, a multivariate logistic regression analysis revealed that malignancy was associated with the 30-day mortality rate (Table 4). Some studies have reported that malignancy is a risk factor for anaerobic bacteremia [2,5e10]. Patients with advanced malignant processes, whose immune systems have been compromised by treatment for malignancy (such as chemotherapy, radiation, and surgery), or malnutrition are thought to have an increased risk of systemic infection [2]. Infection is one of the foremost causes of morbidity and mortality in children with cancer [31]. Nielsen et al. and Laupland et al. reported that malignancy was one of the most common causes of death in patients with bacteremia [32,33]. Zahar et al. reported that in malignancy patients with anaerobic bacteremia, the mortality rate of patients with adequate antimicrobial therapy from the outset was 14%, compared with 63% for patients who were not treated adequately at any time [34]. In our study, the 30-day mortality rate of patients with malignancies and anaerobic bacteremia was 45.5% (15/33 cases). Therefore, when a patient with a malignancy is suspected of having anaerobic

49

bacteremia, the rapid initiation of treatment is important. In addition, our study revealed that clindamycin resistance of anaerobic blood culture isolates was associated with the 30-day mortality rate (Table 4, Fig. 1). The results of systematic reviews and meta-analyses show that appropriate empirical antibiotic treatment is associated with a significant reduction in all-cause mortality [35e37]. In this study, inappropriate therapy was associated with subsequent death, but the correlation was not statistically significant (Table 4). Nguyen et al. reported that for patients with drug-resistant Bacteroides bacteremia, inactive therapy was associated with mortality [11]. However, there were only few cases when clindamycin was used as first line therapy. Patients who survived in most cases received carbapenems or piperacillin/tazobactam (Table 5). The reasons underlying this result remain obscure and warrant further study. Therefore, our results and those of previous studies do not recommend treatment the patients who have anaerobic bacteremia with clindamycin, instead carbapenems and metronidazole are suggested as first line therapy in this situation. An updated Japanese guideline is needed [4]. Although injectable metronidazole has been approved in Japan since September 2014, most anaerobic bacteria are sensitive to this drug at present [18,19,21]. Our study had several limitations. First, our results were based on a retrospective review of routine microbiology data. Second, although our study was a 10-year long study, it only involved one institution; hence, the number of cases with anaerobic bacteremia was not sufficient. A technical bias can occur in routine hospital work. Furthermore, drug susceptibility testing could not be performed for all the strains; consequently, the prognostic factor analysis only included clindamycin and imipenem susceptibility data. Additional long-term prospective studies at multiple institutions are required to investigate a larger number of patients and to identify prognostic factors of death caused by anaerobic bacteremia using multivariate logistic regression analyses. In conclusion, through a multivariate logistic regression analysis and log-rank tests, we demonstrated an association between the outcomes of patients with anaerobic bacteremia and the presence of malignancy and clindamycin resistance. The results of this study are important for the early and appropriate management of patients with anaerobic bacteremia. Conflict of interests The authors state that they have no conflict of interests with the subject matter discussed in this article. Acknowledgments We thank all the clinicians and pharmacists who assisted with the provision of data for this project. References [1] T. Lazarovitch, S. Freimann, G. Shapira, H. Blank, Decrease in anaerobe-related bacteremias and increase in Bacteroides species isolation rate from 1998 to 2007: a retrospective study, Anaerobe 16 (2010) 201e205. [2] B. Lassman, D.R. Gustafson, C. Wood, J.E. Rosenblatt, Reemergence of anaerobic bacteremia, Clin. Infect. Dis. 44 (2007) 895e900. [3] Giamrellou, Anaerobic infection therapy, Int. J. Antimicrob. Agents 16 (2000) 341e346. [4] Japanese Society of Chemotherapy Committee on guidelines for treatment of anaerobic infections, Japanese Association for Anaerobic Infections Research, Chapter 2e2. Anaerobic infections (individual fields): blood stream infections with anaerobic bacteremia, J. Infect. Chemother. 17 (Suppl. 1) (2011) 47e49. [5] L. Blairon, Y. DeGheldre, B. Delaere, A. Sonet, A. Bosly, Y.A. Glupcznski, 62month retrospective epidemiological survey of anaerobic bacteremia in a university hospital, Clin. Microbiol. Infect. 12 (2006) 527e532. [6] A.A. Venugopal, S. Szpunar, L.B. Johnson, Risk and prognostic factors among

50

[7]

[8]

[9]

[10]

[11]

[12] [13]

[14] [15]

[16]

[17]

[18]

[19] [20]

[21]

[22]

T. Umemura et al. / Anaerobe 39 (2016) 45e50 patients with bacteremia due to Eggerthella lenta, Anaerobe 18 (2012) 475e478. F.R. Cockerill 3rd, J.G. Hughes, E.A. Vetter, R.A. Mueller, A.L. Weaver, Ilstrup DM. Analysis of 281,797 consecutive blood cultures performed over an eightyear period: trends in microorganisms isolated and the value of anaerobic culture of blood, Clin. Infect. Dis. 24 (1997) 403e418. T. Umemura, Y. Yamagishi, Y. Hamada, H. Suematsu, H. Mikamo, Investigation of the risk factors of anaerobic bacteremia in a caseecontrol study, Jpn. J. Antibiot. 67 (2014) 133e143. R.L. Lark, S.A. McNeil, K. VanderHyde, Z. Noorani, J. Uberti, C. Chenoweth, Risk factors for anaerobic bloodstream infections in bone marrow transplant recipients, Clin. Infect. Dis. 33 (2001) 338e343. N. Coleman, G. Speirs, J. Khan, V. Broadbent, D.G. Wight, R.E. Warren, Neutropenic enterocolotis associated with Clostridium tertium, J. Clin. Pathol. 46 (1993) 180e183. M.H. Nguyen, V.L. Yu, A.J. Morris, L. McDermott, M.W. Wagener, L. Harrell, et al., Antimicrobial resistance and clinical outcome of Bacteroides bacteremia: findings of a multicenter prospective observational trial, Clin. Infect. Dis. 30 (2000) 870e876. J.H. Salonen, E. Eerola, O. Meurman, Clinical significance and outcome of anaerobic bacteremia, Clin. Infect. Dis. 26 (1998) 1413e1417. R. Robert, A. Deraignac, G. Le Moal, S. Ragot, G. Grollier, Prognostic factors and impact of antibiotherapy in 117 cases of anaerobic bacteremia, Eur. J. Clin. Microbiol. Infect. Dis. 27 (2008) 671e678. J.R. Wilson, A.P. Limaye, Risk factors for mortality in patients with anaerobic bacteremia, Eur. J. Clin. Microbiol. Infect. Dis. 23 (2004) 310e316. O. Gasch, M. Camoez, M.A. Dominguez, B. Padilla, V. Pintado, B. Almirante, et al., Predictive factors for early mortality among patients with methicillinresistant Staphylococcus aureus bacteremia, J. Antimicrob. Chemother. 68 (2013) 1423e1430. Y.M. Chen, H.C. Lee, C.M. Chang, Y.C. Chuang, W.C. Ko, Clostridium bacteremia: emphasis on the poor prognosis in cirrhotic patients, J. Microbiol. Immunol. Infect. 34 (2001) 113e118. J.T. Ingo, M.D. Parkins, D.B. Gregson, J.D. Pitout, T. Ross, D.L. Church, et al., Population -based assessment of the incidence, risk factors, and outcomes of anaerobic bloodstream infections, Infection 41 (2013) 41e48. A. Marchand-Austin, P. Rawte, B. Toye, F.B. Jamieson, D.J. Farrell, S.N. Patel, Antimicrobial susceptibility of clinical isolates of anaerobic bacteria in Ontario, 2010-2011, Anaerobe 28 (2014) 120e125. A.C. Veloo, A.J. van Winkelhoff, Antibiotic susceptibility profiles of anaerobic pathogens in The Netherlands, Anaerobe 28 (2014) 1e6. B.P. Gomes, R.C. Jacinto, F. Montagner, E.L. Sousa, C.C. Ferraz, Analysis of the antimicrobial susceptibility of anaerobic bacteria isolated from endodontic infections in Brazil during a period of nine years, J. Endod. 37 (2011) 1058e1062. W. Jamal, G. Al Hashem, V.O. Rotimi, Antimicrobial resistance among anaerobes isolated from clinical specimens in Kuwait hospitals: comparative analysis of 11-year data, Anaerobe 28 (2014) 1e6. ~ o, P. Areses, M.D. Pen ~ alver, S. Cortizo, F. Pardo, M.L. del Molino, et al., M. Trevin

[23]

[24] [25]

[26] [27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

Susceptibility trends of Bacteroides fragilis group and characterisation of carbapenemase-producing strains by automated REP-PCR and MALDI TOF, Anaerobe 18 (2012) 37e43. D.R. Snydman, N.V. Jacobus, L.A. McDermott, Y. Golan, E.J. Goldstein, L. Harrell, et al., Update resistance of bacteroides fragilis group and related species with special attention to carbapenems 2006-2009, Anaerobe 17 (2011) 147e151. I. Brook, Recovery of anaerobic bacteria from clinical specimens in 12 years at two military hospitals, J. Clin. Microbiol. 26 (1988) 1181e1188. V.A. Peraino, S.A. Cross, E.J. Goldstein, Incidence and clinical significance of anaerobic bacteremia in a community hospital, Clin. Infect. Dis. 16 (1993) S288eS291. J.C. Marshall, K. Reinhart, International sepsis forum. Biomarkers of sepsis, Crit. Care Med. 37 (2009) 2290e2298. voa, L. Coelho, E. Almeida, A. Fernandes, R. Mealha, P. Moreira, et al., CP. Po reactive protein as a marker of infection in critically ill patients, Clin. Microbiol. Infect. 11 (2005) 101e108. L. Simon, F. Gauvin, D.K. Amre, P. Saint-Louis, J. Lacroix, Serum procalcitonin and C-reactive protein levels as markers of bacterial infection: a systematic review and meta-analysis, Clin. Infect. Dis. 39 (2004) 206e217. H.W. Kim, J.H. Yoon, S.J. Jin, S.B. Kim, N.S. Ku, S.J. Jeong, et al., Delta neutrophil index as a prognostic marker of early mortality in gram negative bacteremia, Infect. Chemother. 46 (2014) 94e102. C.A. Gogos, A. Lekkou, O. Papageorgiou, D. Siagris, A. Skoutelis, H.P. Bassaris, Clinical prognostic markers in patients with severe sepsis: a prospective analysis of 139 consecutive cases, J. Infect. 47 (2003) 300e306. P. Quarello, P. Saracco, M. Giacchino, D. Caselli, I. Caviglia, D. Longoni, et al., Epidemiology of infections in children with acquired aplastic anemia: a retrospective multicenter study in Italy, Eur. J. Haematol. 88 (2012) 526e534. S.L. Nielsen, A.T. Lassen, K.O. Gradel, T.G. Jensen, H.J. Kolmos, J. Hallas, et al., Bacteremia is associated with excess long-term mortality: a 12-year population-based cohort study, J. Infect. 70 (2014) 111e126. K.B. Laupland, L.W. Svenson, D.B. Gregson, D.L. Church, Long-term mortality associated with community-onset bloodstream infection, Infection 39 (2011) 405e410. J.R. Zahar, H. Farhat, E. Chachaty, P. Meshaka, S. Antoun, Nitenberg, Incidence and clinical significance of anaerobic bacteraemia in cancer patients: a 6-year retrospective study, Clin. Microbiol. Infect. 11 (2005) 724e729. M. Paul, V. Shani, E. Muchtar, G. Kariv, E. Robenshtok, L. Leibovici, Systematic review and meta-analysis of the efficacy of appropriate empiric antibiotic therapy for sepsis, Antimicrob. Agents Chemother. 54 (2010) 4851e4863. A. Esparcia, A. Artero, J.M. Eiros, M. Balaguer, M. Madrazo, J. Alberola, et al., Influence of inadequate antimicrobial therapy on prognosis in elderly patients with severe urinary tract infections, Eur. J. Intern Med. 25 (2014) 523e527. ndez-Hidalgo, C. Pigrau, C. Ferrer, D. Rodríguez-Pardo, B. Almirante, N. Ferna A.M. Planes, et al., Impact of prompt catheter withdrawal and adequate antimicrobial therapy on the prognosis of hospital-acquired parenteral nutrition catheter-related bacteraemia, Clin. Microbiol. Infect. 20 (2014) 1205e1210.