The influence of inadequate empirical antimicrobial treatment on patients with bloodstream infections in an intensive care unit

ORIGINAL ARTICLE The influence of inadequate empirical antimicrobial treatment on patients with bloodstream infections in an intensive care unit R. Zaragoza1,
, A. Artero2, J. J. Camarena3, S. Sancho1, R. Gonza´lez3 and J. M. Nogueira3 1

Intensive Care Unit, Hospital Universitario Dr Peset, Avda Gaspar Aguilar, 90 46017 Valencia, Internal Medicine Department and 3Microbiology Department, Hospital Universitario Dr Peset, Universidad de Valencia, Spain

2

Objective To determine the occurrence of inadequate antimicrobial therapy among critically ill patients with bacteremia and the factors associated with it, to identify the microorganisms that received inadequate antimicrobial treatment, and to determine the relationship between inadequate treatment and patients outcome.

From June 1995 to January 1999 we collected data on all clinically significant ICU-bacteremias in our teaching hospital. Clinical and microbiological characteristics were recorded and the adequacy of empirical antimicrobial treatment in each case was determined. We defined inappropriate empirical antimicrobial treatment as applying to infection that was not being effectively treated at the time the causative microorganism and its antibiotic susceptibility were known. Multivariate analysis was used to determine the variables associated with inappropriate empirical antimicrobial treatment and to evaluate the influence of this on the related mortality to bacteremia, using the SPSS package (9.0).

Methods

Among 166 intensive care unit patients with bacteremia, 39 (23.5%) received inadequate antimicrobial treatment. In this last group the mean age of patients was 64.1  13.2 years, and 64% were men. Bacteremia was hospital-acquired in 92% of these cases. Eleven percent developed septic shock and 37.7% severe sepsis, and ultimately fatal underlying disease was present in 28.2% of patients given inadequate empirical antimicrobial treatment. The main sources of bacteremias in this group were: a vascular catheter (15.3%), respiratory (7.6%) or unknown (53.8%). The microorganisms most frequently isolated in the group with inadequate empirical antimicrobial treatment were: coagulase-negative staphylococci (29.5%), Acinetobacter baumannii (27.3%), Enterococcus faecalis, Pseudomonas aeruginosa, Enterobacter cloacae, Proteus mirabilis, Escherichia coli, and Candida species (4.5% each). The frequency of coagulase-negative staphylococci in the cases with inappropriate treatment was higher than in the group with appropriate treatment (OR 2.62; 95% CI: 1.10–6.21; P ¼ 0.015). The global mortality rate was 56% and the related mortality was 30% in the group with inadequate empirical antimicrobial treatment. The only factor associated with inappropriate empirical antibiotic treatment was the absence of abdominal or respiratory focus (P ¼ 0.04; OR ¼ 0.35; 95% CI: 0.12-0.97). Septic shock was related to attributable mortality (P ¼ 0.03; OR ¼ 3.19; 95% CI: 1.08-9.40), but not inappropriate empirical antibiotic treatment (P ¼ 0.24; OR ¼ 1.71; 95% CI: 0.66-4.78). Results

Almost a quarter of critically ill patients with bloodstream infections were given inadequate empirical antibiotic treatment, but mortality was not higher in the group with inadequate treatment than in the group with adequate treatment. This fact was probably due to microbiological factors and clinical features, such as the type of microorganism most frequently isolated and the source of the bacteremia.

Conclusion




Corresponding author and reprint requests: Intensive Care Unit, Hospital Universitario Dr Peset, Avda Gaspar Aguilar, 90 46017 Valencia, Spain Tel: þ34 96 3862526 Fax: þ34 96 3861933 E-mail: [email protected]

ß 2003 Copyright by the European Society of Clinical Microbiology and Infectious Diseases

Zaragoza et al Inadequate treatment of ICU-Bacteremia 413

Keywords Bacteremia, intensive care, staphylococci, attributable mortality

inadequate

treatment,

coagulase-negative

Accepted 12 September 2002

Clin Microbiol Infect 2003; 9: 412–418

INTRODUCTION

MATERIALS AND METHODS

Bloodstream infections are probably the most serious infections of hospitalized patients requiring intensive care, especially those with severe sepsis and septic shock [1–4]. Some studies have demonstrated an important relationship between hospital mortality and inadequate empirical antimicrobial treatment of infections in intensive care unit (ICU) patients, especially those with bloodstream infections and ventilator-associated pneumonia [5–8]. Antimicrobial resistance has emerged as an important determinant of outcome for patients in the ICU. This is largely due to the administration of inadequate empirical antimicrobial treatment, which is most often related to bacterial antibiotic resistance. The patterns of such resistance clearly vary among hospitals [9]. Inadequate empirical antimicrobial treatment has also resulted in greater patient morbidity, higher mortality rates, and increased healthcare costs [10,11]. Some microorganisms, such as vancomycinresistant enterococci, methicillin-resistant Staphylococcus aureus (MRSA), and coagulase-negative staphylococci (CNS), have been classically associated with inadequate empirical antimicrobial treatment [5,12–16]. However, their influence on mortality is variable and dependent on the type of microorganism and the place where the study was done [17–20]. In a preliminary study, we did not find significant differences in attributable mortality rates between susceptible and resistant pathogens in ICU bacteremias [21]. We have performed a prospective cohort study with three main objectives: first, to determine the occurrence of inadequate empirical antimicrobial treatment among critically ill patients with bacteremia, and the factors associated with it; second, to identify the microorganisms that received inadequate antimicrobial treatment; and third, to determine the relationship between inappropriate empirical antimicrobial treatment and patient outcome.

Study location and patients The study was carried out in the Hospital Universitario Dr Peset (600 beds), a universityaffiliated urban teaching hospital in Valencia, Spain. During a 43-month period (June 1995 to January 1999), all patients admitted to the medical–surgical ICU (12 beds) were potentially eligible for this investigation. The requirement for antibiotic treatment and the selection of specific antimicrobial agents were determined by the patients’ treating physicians. Study design and data collection A prospective cohort study design was used, segregating patients with a bloodstream infection according to hospital survival and the adequacy of their empirical antimicrobial treatment. Hospitalattributable mortality was the main outcome variable evaluated. The other studied variables were: age, sex, underlying disease, nosocomial origin of bacteremia, source of infection, severe sepsis, septic shock, previous use of antibiotics, microorganism isolates, and global hospital mortality. Definitions Bloodstream infection was considered definitive, and the possibility of a contaminated blood culture was excluded, by the presence of clinical or laboratory evidence of infection (fever, hypothermia, localized infection, inadequate organ perfusion, septic shock, leukocytosis, or laboratory findings compatible with disseminated intravascular coagulopathy) [6]. Organisms that are commonly recovered from the environment or the skin (mainly CNS and aerobic Gram-positive rods) were judged to be contaminants unless the clinical findings, the results of cultures of material from other body sites and the number of positive sets (two or more) indicated a high probability of bloodstream infection [6].

ß 2003 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 9, 412–418

414 Clinical Microbiology and Infection, Volume 9 Number 5, May 2003

For the purposes of this investigation, inappropriate empirical antimicrobial treatment of a bloodstream infection was defined, following Ibrahim et al. [5] and Kollef et al. [7], as the occurrence of infection (i.e. a positive blood culture result) that was not effectively treated at the time when the causative microorganism and its antibiotic susceptibility were known. Inappropriate empirical antimicrobial treatment included the absence of antimicrobial agents directed at a specific class of microorganisms (e.g. absence of therapy for fungemia caused by Candida spp.) and the administration of an antimicrobial agent to which the microorganism responsible for the infection was resistant (e.g. empirical treatment with oxacillin for bacteremia subsequently attributed to MRSA on the basis of blood culture results). It was also necessary to consider the antibiotic treatment as appropriate if at least one effective drug was included in the empirical treatment, and the dose, pattern of administration and duration of treatment were according to current medical standards [22–24]. All blood cultures for establishing the presence of a bloodstream infection were required to be obtained percutaneously, using sterile technique, and not drawn from indwelling vascular catheters. Bacteremia-related mortality was diagnosed when a patient died during treatment for a community-acquired or nosocomial bloodstream infection, which could be easily explained by the infection, and the death could not be directly attributed to any other cause [5,23]. McCabe’s classification [25] was used to place the patients’ underlying disease in the following categories: rapidly fatal, ultimately fatal and non-fatal. Community-acquired bloodstream infections were required to be established within 48 h of hospital admission. Nosocomial bloodstream infections were required to be established after 48 h of hospitalization [22,26]. Nosocomial bloodstream infections, as well as other nosocomial infections (urinary tract, wound infection), were defined according to criteria established by the Centers for Disease Control and Prevention [27]. The source of infection was designated as one of the following: respiratory, surgical wound, vascular catheter related, intra-abdominal, urinary tract, unknown and others [22]. Following Rello et al. and Valle´ s et al. [22,24], abdominal and respiratory sources of bacteremia were considered as risk foci of infection, because these foci were associated with a higher mortality than other sources of

bacteremia. The definitions used for the systemic inflammatory response syndrome, sepsis, severe sepsis and septic shock were those proposed by the American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference [28]. Because Enterobacteriaceae other than Escherichia coli, Pseudomonas aeruginosa, Acinetobacter baumannii and Candida spp. have been reported to cause high mortality, they were considered to be high-risk microorganisms [7,24]. Data analysis The chi-square test was used for contingency tables. Student’s t-test was used for comparing means. Stepwise logistic regression analysis was used to determine independent predictors of related bacteremia in-hospital mortality and independent predictors of inadequate empirical antimicrobial treatment of bacteremia. We estimated multivariable-adjusted odds ratios with 95% confidence intervals. SPSS 9.0 software was used for data analysis. RESULTS During the study period, 3225 consecutive eligible patients were evaluated. Among these, 166 episodes of ICU bacteremia were detected. In 39 episodes, inappropriate empirical antimicrobial treatment was given (23.5% of the cases of bacteremia). The characteristics of the 166 episodes of ICU bacteremia according to the adequacy of empirical antibiotic treatment are shown in Table 1. None of these variables was statistically different between the two groups with appropriate and inappropriate empirical antibiotic treatment, except that the presence of an abdominal or respiratory focus as a source of infection was significantly less frequent in the inadequate empirical treatment group in the univariate analysis (12% versus 33%; P ¼ 0.01). The commonest sources of inappropriately treated bacteremia were: unknown (53.8%), vascular catheter (15.3%), and respiratory (7.6%). The systemic response to inadequately treated bacteremia was classified as sepsis in 16 episodes (41%), severe sepsis in 15 episodes (38.5%), and septic shock in four episodes (11%). Forty-four microorganisms were isolated in the 39 episodes of bacteremia with inappropriate antimicrobial therapy. The causative microorganisms

ß 2003 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 9, 412–418

Zaragoza et al Inadequate treatment of ICU-Bacteremia 415

Table 1 Characteristics of 166 cases of intensive care bacteremia according to the adequacy of empirical antimicrobial treatment

Male/female (%) Years of age (mean  SD) Nosocomial origin of the bacteremia (%) Ultimately fatal disease (%) Respiratory or abdominal source of the bacteremia (%) High-risk microorganisms (%) Previous antibiotic treatment (%) Severe sepsis or septic shock (%) Global mortality (%) Related mortality (%)

Adequate treatment N ¼ 127

Inadequate treatment N ¼ 39

P

60.6/39.4 63.2  16.3 79.5 33.8 33.6

64.1/35.9 64.2  13.2 92.3 28.2 12.1

0.697 0.733 0.066 0.640 0.017

34.9 62.2 58.3 50.3 22.8

42.1 74.3 48.7 56.4 30.8

0.420 0.164 0.293 0.512 0.315

of intensive care cases of bacteremia according to the adequacy of empirical antimicrobial treatment are listed in Table 2. CNS (29.5%) and A. baumannii (27.3%) were the microorganisms most frequently isolated. This high incidence of A. baumannii bacteremia was due to a carbapenem-resistant outbreak in our ICU during the period of the study. Gram-positive microorganisms were isolated in 41% of bacteremias with inappropriate antimicrobial treatment, whereas Gram-negative organisms represented 54%. Statistical analysis of the influence of inadequate empirical antimicrobial treatment on attributable mortality only for patients with bloodstream infections due to Gram-negative bacteremia, with and without A. baumannii, did not show

significant differences (OR 1.07, 95% CI 0.25–4.28, P ¼ 0.574; OR 0.58, 95% CI 0.02–63.4, P ¼ 0540). Fungemia due to Candida spp. represented only 4.5% of the isolates. The frequency of CNS in the cases with inappropriate treatment was higher than in the group with appropriate treatment (OR 2.62, 95% CI 1.10–6.21, P ¼ 0.015). Logistic regression showed that the only risk factor associated with the administration of inadequate treatment was the absence of an abdominal or respiratory focus of bacteremia (OR ¼ 0.35, 95% CI 0.12–0.97, P ¼ 0.04). Multivariate analysis failed to demonstrate that inadequate treatment, risk source of infection, isolation of a risk microorganism and nosocomial origin of bacteremia were associated with related

Table 2 Microorganisms causing cases of intensive care bacteremia according to the adequacy of empiric antimicrobial treatment

Microorganisms

Inadequate empirical antimicrobial treatment No. (%)

Coagulase-negative staphylococci Acinetobacter baumannii Enterococcus spp. Staphylococcus aureus Pseudomonas aeruginosa Enterobacter cloacae Proteus mirabilis Escherichia coli Candida albicans Klebsiella pneumoniae Salmonella spp. Citrobacter freundii Serratia marcescens Streptococcus mutans Others

13 12 2 2 2 2 2 2 2 1 1 1 1 1 0

Total

44 (100)

(29.5)a (27.3) (4.5) (4.5) (4.5) (4.5) (4.5) (4.5) (4.5) (2.3) (2.3) (2.3) (2.3) (2.3) (0)

Adequate empirical antimicrobial treatment No. (%) 21 30 13 15 17 3 2 18 0 4 0 0 0 0 29

(13.8)a (19.7) (8.5) (9.8) (11.1) (2) (1.3) (11.8) (0) (2.6) (0) (0) (0) (0) (19.1)

152 (100)

a

Coagulase-negative staphylococci were the only organisms with significant differences, P ¼ 0.015.

ß 2003 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 9, 412–418

416 Clinical Microbiology and Infection, Volume 9 Number 5, May 2003

mortality. The presence of septic shock at the moment of bacteremia was the only related factor with attributable mortality in our study (OR ¼ 3.19, 95% CI 1.08–9.40, P ¼ 0.03). DISCUSSION It is well known that the presence of bloodstream infection is a risk factor for the administration of inadequate empirical treatment of infections [7]. In our setting, 23.5% of patients with cases of ICU bacteremia received inappropriate empirical antimicrobial treatment. Similar percentages have been described by several authors, varying between 20% and 30% [5,6,24]. Previous studies have identified an important association between the administration of inadequate antimicrobial treatment of bloodstream infections and hospital mortality [5–8,22]. Ibrahim et al. [5] found an important reduction in fatality in patients receiving appropriate treatment (61.9% versus 28.4%, RR 2.18), and Leibovici et al [6] demonstrated that the contribution of inappropriate empirical antimicrobial treatment to fatality was independent of other risk factors in a multivariate logistic analysis. However, our study failed to demonstrate this relationship between attributable mortality and the administration of inappropriate empirical antimicrobial treatment in patients with ICU bacteremia. To our knowledge, this is the first study to find no differences in mortality rates between critically ill patients with a bloodstream infection who received inappropriate empirical antimicrobial treatment and those who were given appropriate treatment. Factors such as the source of the bacteremia and the microorganism isolated may explain our results, in part. The presence of abdominal or respiratory foci as risk sources of infection was significantly less frequent in the inadequate treatment group in the univariate analysis (12% versus 33%, P ¼ 0.01) in our setting. It is well recognized that the mortality rate of cases of ICU bacteremias depends on the source of the bacteremia. Classically, urinary and vascular catheter foci have been associated with better outcome, and pulmonary and intra-abdominal foci with the worst prognosis [22,24,29,30]. Valle´ s and the Spanish collaborative group for Infections in Intensive Care Units [24] demonstrated that the source of bacteremia other than intravascular catheters was a risk factor asso-

ciated with bacteremia-related mortality. Rello et al. [22] found an important relationship between an intra-abdominal origin of infection and hospital mortality, and Leibovici et al. [6], in a stratified analysis, demonstrated that the highest benefit of appropriate antibiotic treatment was apparent in intra-abdominal infections and infections of skin and soft tissues, whereas all the other sources of bacteremia were comparable. The low prevalence of risk foci of infection in the inadequate treatment group, as described above, may be, responsible for our results in part. The reason why patients with respiratory or abdominal foci of infection were more frequently treated with inadequate empirical antimicrobial treatment than patients with other foci of infection was probably the fact that this type of infection followed a standardized treatment regimen in our ICU. Microbiological factors can also influence the outcome of sepsis. The most commonly identified bloodstream pathogens associated with inadequate treatment described in several reports were vancomycin-resistant enterococci, MRSA, CNS, P. aeruginosa, Enterobacter, Acinetobacter, and Candida spp. [5,13–16]. However, to date, the evidence with MRSA and vancomycin-resistant enterococci is inconsistent with regard to the effect on mortality rates [17–19], although the use of vancomycin was associated with increased mortality [20]. Similar findings were previously reported by our group, which did not find significant differences in attributable mortality rates between susceptible and resistant pathogens in cases of ICU bacteremia [21]. Furthermore, Rangel-Frausto [31] found that bacteremia caused by Candida and Enterococus spp. was associated with the highest attributable mortality (30–40%), and CNS were associated with the lowest attributable mortality (15–20%). We did not find differences in the percentage of high-risk microorganisms between the groups with appropriate and inappropriate treatment, but the frequency of CNS (33.3%) in the cases with inappropriate treatment was higher than in the group with appropriate treatment (16.3%) (OR 2.62, 95% CI 1.10–6.21, P ¼ 0.015). This fact may also contribute to the absence of benefit of the administration of adequate treatment in our study, due to the low mortality of cases of CNS bacteremia in critically ill patients [24,31]. A bloodstream infection attributed to Candida spp., prior administration of antibiotics during the same hospitalization, decreasing serum albumin

ß 2003 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 9, 412–418

Zaragoza et al Inadequate treatment of ICU-Bacteremia 417

concentrations and increasing central catheter duration were factors described by Ibrahim et al. [5] as being independently associated with the administration of inadequate antimicrobial treatment. No factors other than the absence of an abdominal or a respiratory source of infection were associated with the administration of inadequate treatment in our study. Finally, other factors that have been independently associated with a higher directly related mortality to bacteremia as well as inappropriate treatment, are severe sepsis or septic shock, development of multi-organ failure or number of organ dysfunctions, adult respiratory distress syndrome, acute renal failure, APACHE II score at the onset of sepsis, Gramnegative bacteremia or candidal bacteremia, and abdominal or respiratory foci [22–24]. Septic shock was the only factor associated with attributable mortality in our study. Inappropriate treatment, risk focus of infection and risk microorganism were not related to attributable mortality in this study, probably due to the limited number of cases in our setting. CONCLUSIONS Although the administration of inappropriate empirical antimicrobial treatment to patients with ICU bacteremia was high in our study, we did not find an excess of attributable mortality in this group, probably due to microbiological factors and clinical features such as the type of microorganism most frequently isolated and the source of bacteremia involved. REFERENCES 1. Balk RA. Septic shock. Pathophysiol Curr Opin Anesthes 1994; 7: 136–40. 2. Bone RC. The pathogenesis of sepsis. Ann Intern Med 1991; 115: 457–69. 3. Bone RC. Gram-negative sepsis: background, clinical features and intervention. Chest 1991; 100: 802– 8. 4. Zaragoza R, Artero A, Borras S et al. Intensive care unit bacteremias and septic shock. Analysis of clinical and microbiological features and risk factors. Intensive Care Med 2001; 27(suppl 2): S236. 5. 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.

6. Leibovici L, Shraga I, Druckner M et al. The benefit of appropriate empirical antibiotic treatment in patients with bloodstream infection. J Intern Med 1998; 244: 379–86. 7. Kollef MH, Sherman G, Ward S et al. Inadequate antimicrobial treatment of infections: a risk factor for hospital mortality among critically ill patients. Chest 1999; 115: 462–74. 8. Bishara J, Leibovici L, Ashkenazi S, Samra Z, Pitlik S. Seven-year study of bacteraemic pneumonia in a single institution. Eur J Clin Microbiol Infect Dis 2000; 19: 926–31. 9. Rello J, Sa-Borges M, Correa H et al. Variations in etiology of ventilator-associated pneumonia across four treatment sites: implications for antimicrobial prescribing practices. Am J Respir Crit Care Med 1999; 160: 608–13. 10. Kollef MH. Is there a role for antibiotic cycling in the intensive care unit? Crit Care Med 2001; 29(4 Suppl): N135–42. 11. Niederman MS. Impact of antibiotic resistance on clinical outcomes and the cost of care. Crit Care Med 2001; 29(suppl 4): S114–20. 12. Kollef MH, Ward S, Sherman G et al. Inadequate treatment of nosocomial infections is associated with certain empiric antibiotic choices. Crit Care Med 2000; 28: 3456–64. 13. Chow JW, Fine MJ, Shlaes DM et al. Enterobacter bacteremia: clinical features and emergence of antibiotic resistance during therapy. Ann Intern Med 1991; 115: 585–90. 14. Vidal F, Mensa J, Almela M et al. Epidemiology and outcome of Pseudomonas aeruginosa bacteremia, with special emphasis on the influence of antibiotic treatment: analysis of 189 episodes. Arch Intern Med 1996; 156: 2121–6. 15. Garcia-Garmendia JL, Ortiz-Leba C, GarnachoMontero J et al. Mortality and the increase in length of stay attributable to the acquisition of Acinetobacter in critically ill patients. Crit Care Med 1999; 27: 1794–9. 16. Carmeli Y, Troillet N, Karchmer AW et al. Health and economic outcomes of antibiotic resistance in Pseudomonas aeruginosa. Arch Intern Med 1999; 159: 1127–32. 17. Mainous MR, Lipsett PA, O’brien M. Enterococcal bacteremia in the surgical intensive care unit: does vancomycin resistance affect mortality? The Johns Hopkins SICU Study Group. Arch Surg 1997; 132: 76–81. 18. Lucas GM, Lechtzin N, Puryear DW et al. Vancomycin-resistant and vancomycin-susceptible enterococcal bacteremia: comparison of clinical features and outcomes. Clin Infect Dis 1998; 26: 1127–33. 19. Steinberg JP, Clark CC, Hackman BO. Nosocomial and community-acquired Staphylococcus aureus

ß 2003 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 9, 412–418

418 Clinical Microbiology and Infection, Volume 9 Number 5, May 2003

20.

21.

22.

23.

24.

bacteremias from 1980 to 1993: impact of intravascular devices and methicillin resistance. Clin Infect Dis 1996; 23: 255–9. Gonza´ lez C, Rubio M, Romero-Vivas J et al. Bacteremic pneumonia due to Staphylococcus aureus: a comparison of disease caused by methicillin-resistant and methicillin-susceptible organisms. Clin Infect Dis 1999; 29: 1171–7. Zaragoza R, Vicente R, Artero A et al. Influence of clinical and microbiological factors on mortality of intensive care unit bacteremias. Intensive Care Med 2000; 26(suppl 3): S348. Rello J, Ricart M, Mirelis B et al. Nosocomial bacteremia in a medical–surgical intensive care unit: epidemiologic characteristics and factors influencing mortality in 111 episodes. Intensive Care Med 1994; 20: 94–8. Pittet D, Thievent B, Wenzel RP, Li N, Auckenthaler R, Suter PM. Bedside prediction of mortality from bacteremic sepsis. A dynamic analysis of ICU patients. Am J Respir Crit Care Med 1996; 153: 684– 93. Valle´ s J, Leo´ n 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 Espan˜ ola de Medicina Intensiva y Unidades Coronarias (SEMIUC) Clin Infect Dis 1997; 24: 387–95.

25. McCabe WR, Jackson GG. Gram negative bacteremia. Arch Intern Med 1962; 110: 92–100. 26. Pingleton SK, Fagon JY, Leeper KV Jr. Patient selection for clinical investigation of ventilatorassociated pneumonia: criteria for evaluating diagnostic techniques. Chest 1992; 102(suppl 1): S553–6. 27. Garner JS, Jarvis WR, Emori TB et al. CDC definitions for nosocomial infections. Am J Infect Control 1988; 16: 128–40. 28. American College of Chest Physicians/Society of Critical Care Medical Consensus Conference. Definitions for sepsis and multiple organ failure and guidelines for the use of innovative therapies in sepsis. Chest 1992; 101: 1644–55. 29. Brun-Buisson C, Doyon F, Carlet J. Bacteremia and severe sepsis in adults: a multicenter prospective survey in ICUs and wards of 24 hospitals. French Bacteremia–Sepsis Study Group. Am J Respir Crit Care Med 1996; 154: 617–24. 30. Gatell JM, Trilla A, Latorre X et al. Nosocomial bacteremia in a large Spanish teaching hospital: analysis of factors influencing prognosis. Rev Infect Dis 1988; 10: 203–10. 31. Rangel-Frausto MS. The epidemiology of bacterial sepsis. Infect Dis Clin North Am 1999; 13: 299–312. 32. Rello J, Ochagavia A, Sabanes E et al. Evaluation of outcome of intravenous catheter-related infections in critically ill patients. Am J Respir Crit Care Med 2000; 162: 1027–30.

ß 2003 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 9, 412–418