Antimicrobial therapy for patients with severe sepsis and septic

shock: An evidence-based review

Pierre-Yves Bochud, MD; Marc Bonten, MD; Oscar Marchetti, MD; Thierry Calandra, MD, PhD

Objective: In 2003, critical care and infectious disease experts
representing 11 international organizations developed manage-
ment guidelines for antimicrobial therapy for patients with severe
sepsis and septic shock that would be of practical use for the
bedside clinician, under the auspices of the Surviving Sepsis
Campaign, an international effort to increase awareness and
improve outcome in severe sepsis.
Design: The process included a modified Delphi method, a con-
sensus conference, several subsequent smaller meetings of sub-
groups and key individuals, teleconferences, and electronic-based
discussion among subgroups and among the entire committee.
Methods:The modified Delphi methodology used for grading
recommendations built on a 2001 publication sponsored by the
International Sepsis Forum. We undertook a systematic review of
the literature graded along five levels to create recommendation
grades from A to E, with A being the highest grade. Pediatric
considerations to contrast adult and pediatric management are in
the article by Parker et al. on p. S591.
Conclusion: Since the prompt institution of therapy that is active
against the causative pathogen is one of the most important predic-
tors of outcome, clinicians must establish a system for rapid admin-
istration of a rationally chosen drug or combination of drugs when
sepsis or septic shock is suspected. The expanding number of
antibacterial, antifungal, and antiviral agents available provides op-
portunities for effective empiric and specific therapy. However, to
minimize the promotion of antimicrobial resistance and cost and to
maximize efficacy, detailed knowledge of the likely pathogens and
the properties of the available drugs is necessary for the intensivist.
(Crit Care Med 2004; 32[Suppl.]:S495–S512)

I n 2001, the International Sepsis
Forum published guidelines on
the management of patients with
severe sepsis and septic shock, in-
cluding an evidence-based review on an-
tibiotic therapy (1). The recommenda-
tions on antibiotic therapy for sepsis
comprised sections on the epidemiology
of sepsis and the impact of appropriate
antimicrobial therapy on outcome of
Gram-negative, Gram-positive, and fun-
gal sepsis and an evidence-based review of
monotherapy vs. combination therapy for
From The Institute for Systems Biology, Seattle,
WA (PYB); the Department of Internal Medicine, Divi-
sion of Acute Internal Medicine and Infectious Dis-
eases, University Medical Center Utrecht, Utrecht, The
Netherlands (MB); Infectious Diseases Service, Depart-
ment of Internal Medicine, Centre Hospitalier Univer-
sitaire Vaudois, Lausanne, Switzerland (OM, TC).
Supported by grants from the Swiss Foundation for
Medical and Biological Grants to PYB (1121), from the
Swiss National Science Foundation to PYB (81LA-
65462) and to TC (3100 – 066972), the Bristol-Myers
Squibb Foundation, the Santos-Suarez Foundation for
Medical Research, and the Leenaards Foundation. TC
is a recipient of a career award from the Leenaards
Foundation.
Copyright © 2004 by the Society of Critical Care
Medicine and Lippincott Williams & Wilkins
DOI: 10.1097/01.CCM.0000143118.41100.14
the empirical management of patients
with severe sepsis or septic shock.
Under the auspices of the Surviving
Sepsis Campaign, a joint venture of the
International Sepsis Forum, the Society
of Critical Care Medicine, and the Euro-
pean Society of Intensive Care Medicine,
the International Sepsis Forum guide-
lines have been updated and extended.
The revised guidelines extend, but do not
entirely replace, the previously published
article (1), as several parts of the initial
document are still up-to-date. As for the
original publication, the revised guide-
lines were developed using an evidence-
based review of the literature. The pro-
cesses used by the panel of experts to
select and review the relevant literature
are outlined in the Material and Methods
section. The panel elected to focus on the
following topics: 1) impact of appropriate
antibiotic therapy on the outcome of sep-
sis; 2) impact of pharmacokinetic and
pharmacodynamic principles on treat-
ment outcome; 3) need for prompt initi-
ation of antimicrobial therapy in patients
with severe sepsis and septic shock; 4)
analyses of the efficacy and toxicity of
monotherapy vs. combination therapy as
empirical therapy; 5) indications for em-
pirical use of anti-Gram-positive antibiot-
ics; 6) need for empirical antifungal ther-
apy in patients with severe sepsis; 7)
analyses of the efficacy and safety of con-
ventional vs. lipid formulations of am-
photericin B and of azoles and echinocan-
dins vs. amphotericin B for the treatment
of fungal sepsis; and 8) impact of antibi-
otic cycling on morbidity and mortality of
sepsis.
Materials and Methods
Selection of Panel, Topics of Interest,
and Review Process. The panel was com-
posed of three board-certified infectious
diseases specialists, with expertise in the
fields of antimicrobial therapy, sepsis, in-
tensive care medicine, and infection con-
trol, who have been selected by the co-
chairs of the “Diagnosis and Management
of Infection in the Severe Sepsis” track of
Surviving Sepsis Campaign. The panel-
ists, chairpersons, and directors of the
Surviving Sepsis Campaign discussed the
specific scopes of the guidelines during a
2-day meeting (June 27–28, 2003). Iden-
tified topics of interest were distributed
among the panelists based on individual
expertise. Each panelist either alone or in
collaboration with colleagues wrote sev-
eral sections of the guidelines, which

Crit Care Med 2004 Vol. 32, No. 11 (Suppl.) S495

were assembled by the panel chair and
reviewed and critiqued by the panelists
and track co-chairs.
Data Source. Medline was used to
search articles published between 1966
and July 1, 2003. Medline searches and
selections of articles were performed by
each of the authors. Medical Subject
Heading (MeSH; http://www.nlm.nih.gov/
mesh/meshhome.html) terms used in the
Medline search were sepsis (comprising
the terms septicemia, sepsis syndrome,
septic shock, bacteremia, fungemia, par-
asitemia, and viremia), pneumonia (com-
prising the terms bronchopneumonia,
pleuropneumonia, aspiration pneumonia,
bacterial pneumonia, viral pneumonia,
and Pneumocystis carinii pneumonia).
Since no ad hoc MeSH term exists for
intraabdominal infections, we used the
words and MeSH terms intraabdominal,
abdomen, abdominal infection, peritoni-
tis, appendicitis, abdominal abscess to
search for articles on intraabdominal in-
fections. The Medline search was then
narrowed by using the MeSH terms anti-
infective agents (comprising the term
antibiotics, which was exploded to in-
clude all classes of antibiotics and all
antibiotic names), clinical trials, or
randomized controlled trials, further
limiting the search to human studies
and English literature. The MeSH key-
words agranulocytosis, antibiotic pro-
phylaxis, and ambulatory care were
used to exclude studies on neutropenic
patients, antibiotic prophylaxis, and
outpatient therapy. Additional articles
were retrieved from references of arti-
cles identified by the Medline search
and of guidelines and review articles on
the following topics: sepsis, communi-
ty-acquired pneumonia, ventilator-
associated pneumonia, and intraab-
dominal infections. Epidemiologic data
were extracted from articles identified
by a Medline search using the keywords
epidemiology and sepsis and by a sys-
tematic review of 27 clinical trials of
antiinflammatory or mediator-targeted
therapies in patients with severe sepsis
and septic shock (2– 6).
For the section on empirical therapy
for Gram-positive severe sepsis or septic
shock, a Medline search was performed
using the terms empirical combined with
either vancomycin or teicoplanin, which
revealed one and 33 hits, respectively.
Extension of the search to the terms em-
pirical, glycopeptide, and clinical trials
yielded 20 hits. However, none of these
articles included nonneutropenic pa-
S496
tients with Gram-positive infections and
severe sepsis. A Medline search using the
terms clinical trial and either linezolid or
quinupristin/dalfopristin and Synercid
yielded 100 and 41 hits, respectively.
Selection of Articles. Abstracts of all
articles satisfying the selection criteria
were reviewed to exclude irrelevant stud-
ies. Review articles and articles on topics
such as antibiotic prophylaxis, pharma-
cology, microbiology, oncology, hematol-
ogy, immunology, mediators of inflam-
mation, allergy, catheter management,
animal studies, chronic infections, or
specific infections (acquired immunode-
ficiency syndrome, endocarditis, chronic
salmonellosis, viral infections in organ
transplant patients, hemorrhagic fever,
viral hepatitis, parasitic infections, and
malaria) were excluded if they did not
fulfill the inclusion criteria. For the com-
parison of monotherapy vs. combination
therapy, articles on each of the three clin-
ical entities reviewed (i.e., sepsis, pneu-
monia, and intraabdominal infections)
were selected only if there was unequiv-
ocal evidence that patients had either
clinically or microbiologically docu-
mented infections and if the study met at
least one of the following criteria: 1) a
definition of sepsis or severe sepsis con-
sistent with the definition of the Consen-
sus Conference of the American College
of Chest Physicians and the Society of
Critical Care Medicine (7); 2) sepsis with
at least one organ dysfunction or sign of
hypoperfusion present in ⬎50% of the
patients; or 3) an overall mortality
⬎10%. This cutoff for mortality was cho-
sen to ensure that articles included a
substantial proportion of patients with
severe sepsis and not simply sepsis. To
ensure that articles on antimicrobial
therapy had been properly selected by one
of the authors (PYB), a random sample of
20% of the articles identified by the Med-
line search were examined by the panel
chair (TC). Overall agreement between
the two reviewers was 94% for articles on
intraabdominal infections (␬ statistic of
0.71) and 96% for articles on communi-
ty-acquired pneumonia and ventilator-
associated pneumonia (␬ statistic of 0.81)
corresponding to “substantial” and “al-
most perfect” agreements, respectively,
according to the criteria proposed by
Sackett et al. (8). Consensus was reached
between the two authors on the articles
for which there was disagreement regard-
ing the inclusion criteria.
Epidemiological Features of
Severe Sepsis and Septic
Shock
Data on secular epidemiologic trends
of the etiological agents and most fre-
quent sites of sepsis have not changed
since the 2001 review and will, therefore,
not be reviewed in detail again (1).
Microorganisms and Sites of Infec-
tion. By the mid-1980s, the frequency of
Gram-positive sepsis (mainly caused by
Staphylococcus aureus, coagulase-nega-
tive staphylococci, enterococci, and
streptococci) has equaled that of Gram-
negative sepsis (mainly caused by Enter-
obacteriaceae, especially Escherichia coli
and Klebsiella pneumoniae, and by
Pseudomonas aeruginosa). Recent epide-
miologic data in the United States and in
Europe indicated that Gram-positive bac-
teria have now surpassed Gram-negative
bacteria as etiological agents of sepsis (9,
10), confirming a trend observed in many
recent sepsis studies (4, 5, 11–17). Fungi
account for about 5% of all cases of sep-
sis, severe sepsis, and septic shock (1).
Most cases of fungal sepsis are caused by
Candida species, which are the fourth
most common bloodstream pathogens in
all recent U.S. studies of nosocomial
bloodstream infections and are associated
with the highest mortality (40%) of all
bloodstream pathogens (18, 19). The in-
cidence of fungal sepsis has increased
three-fold between 1979 and 2000 (10).
However, due to the paucity of published
data on national surveys of fungal infec-
tions, it is unclear whether the incidence
of fungal sepsis has increased in a similar
fashion in other parts of the world. Dutch
investigators reported a two-fold increase
in the incidence of candidemia in the
early 1990s (20). In contrast, a Norwegian
survey also conducted in the early 1990s
reported stable incidence of candidemia
and Candida species distribution (21).
Likewise, a recent survey of the incidence
of candidemia in Switzerland has shown
that the incidence of candidemia re-
mained stable between 1991 and 2000
(median, 0.5 episodes/10,000 patient-
days) (22). By decreasing order of fre-
quency, the predominant sites of infec-
tions in patients with severe sepsis and
septic shock are the lungs, the blood-
stream (i.e., without another identifiable
source of infection), the abdomen, the
urinary tract, and the skin and soft tis-
sues.
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
Need for Prompt Initiation of
Antimicrobial Therapy in
Patients with Severe Sepsis and
Septic Shock
Rationale: Establishing vascular access
and initiating aggressive fluid resuscita-
tion is the first priority when managing
patients with severe sepsis or septic
shock. However, prompt infusion of anti-
microbial agents is also a logical strategy
and may require additional vascular ac-
cess ports. Establishing a supply of pre-
mixed antibiotics in an emergency de-
partment or critical care unit for such
urgent situations is an appropriate strat-
egy for enhancing the likelihood that an-
timicrobial agents will be infused
promptly. Staff should be cognizant that
some agents require more lengthy infu-
sion time, whereas others can be rapidly
infused or even administered as a bolus.
Impact of Appropriate Antibiotic
Therapy on Outcome of Sepsis
Recommendations: Initial empirical anti-
infective therapy should include one or
more drugs that have activity against the
likely pathogens (bacterial or fungal) and
that penetrate into the presumed source
of the sepsis. The choice of drugs should
be guided by the susceptibility patterns of
microorganisms in the community and
in the hospital.
Grade D
Rationale: Early administration of appro-
priate antibiotics reduces mortality in pa-
tients with Gram-positive and Gram-
negative bacteremias. Retrospective studies
conducted in the 1960s and in the 1970s
have shown that appropriate antimicrobial
therapy, defined as the use of at least one
antibiotic active in vitro against the caus-
ative bacteria, reduced the mortality of
Gram-negative bacteremia when compared
with patients receiving inappropriate ther-
apy (23–26). In a landmark study of 173
patients with Gram-negative bacteremia,
who were classified in three categories
based on the severity of the underlying dis-
ease categories (i.e., rapidly fatal, ultimately
fatal, and nonfatal), McCabe et al. (23) ob-
served that appropriate antibiotic therapy
reduced mortality from 48% to 22%. Four
subsequent studies that included larger
numbers of patients yielded similar results
(24 –27). In a recent prospective study of
2,124 patients with Gram-negative bactere-
mia, mortality was 34% in 670 patients
who received inappropriate antibiotics and
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
18% in 1,454 patients who received appro-
priate antibiotics (p ⬍ .0001) (28). Smaller
recent studies showed that the appropriate-
ness of the antibiotic regimen favorably in-
fluenced the outcome of patients infected
with specific Gram-negative bacteria, such
as Enterobacter species (29), P. aeruginosa
(30), and ceftazidime-resistant K. pneu-
moniae or E. coli (31).
Fewer data have been published on the
impact of appropriate antibiotic therapy
in patients with Gram-positive sepsis
(32). Several studies evaluated the impact
of appropriate antimicrobials in patients
with severe infections due to Gram-
negative and Gram-positive bacteria (33–
44). In all but one study (42), appropriate
antibiotic therapy was associated with a
better outcome.
Empirical Antimicrobial Therapy
for Patients with Severe Sepsis
and Septic Shock
The evidence-based review of the lit-
erature on empirical antimicrobial ther-
apy for patients with severe sepsis and
septic shock has focused on three clinical
entities: sepsis (i.e., primary or secondary
bloodstream infections), pneumonia
(both community-acquired and hospital-
acquired), and intraabdominal infections.
Management guidelines and review arti-
cles have been published recently on
these topics (45–53), which should be
used as a supplement to the present arti-
cle, which focuses on clinical trials of
antimicrobial agents for patients with se-
vere sepsis or septic shock.
A total of 2,065 articles (sepsis, 988;
community-acquired pneumonia and
ventilator-associated pneumonia, 509; in-
traabdominal infections, 393; Gram-
positive infections, 175) were identified
using the Medline search strategy de-
scribed in Material and Methods, of which
1,464 did not meet our inclusion criteria
based on the content of the abstract (Fig.
1). A full text review was performed on
the remaining 614 articles, of which 79
(13%) (sepsis, 37; community-acquired
pneumonia and ventilator-associated
pneumonia, 35; and intraabdominal in-
fections, 7) met the inclusion criteria of
severe sepsis or septic shock. Thus, only a
small proportion of the screened clinical
trials of antibiotic therapy have included
critically ill septic patients with high
mortality. The overall mortality of the
patients in the selected studies was 18%.
Most studies performed before 1990 com-
prised a limited number of patients (usu-
ally ⬍100 per trial), a majority of these
patients with Gram-negative sepsis, and
did not include intent-to-treat analyses.
Studies performed in the 1990s were
larger, included patients with both Gram-
positive and Gram-negative sepsis, and in
most instances, included an intent-to-treat
analysis. Given the relatively small number
of studies available and the fact that the
sepsis subset of studies included patients
with various types of infections, including
community-acquired and ventilator-associ-
ated pneumonias and intraabdominal in-
fections, all studies were pooled.
Among the 79 studies identified, 67
met the inclusion criteria for severe sep-
sis and/or septic shock due to bacterial
infections and were grouped in four main
categories: 1) studies comparing different
combination therapies, usually a ␤-lac-
tam antibiotic or a fluoroquinolone plus
an aminoglycoside (n ⫽ 9; sepsis, 4;
pneumonia, 5) (54 – 62); 2) studies com-
paring a monotherapy with a combina-
tion therapy with an aminoglycoside (n ⫽
24; sepsis, 11; pneumonia, 7; intraab-
dominal infections, 6) (63– 86); 3) studies
comparing single-agent therapies (n ⫽
26; sepsis, 14; pneumonia, 11; intraab-
dominal infections, 1) (56, 87–111); and
4) miscellaneous studies (n ⫽ 8; pneu-
monia, 7; sepsis, 1) (112–119).
Monotherapy Vs. Combination
Antibiotic Therapy
Recommendation: Monotherapy is as ef-
ficacious as combination therapy with a
␤-lactam and an aminoglycoside as em-
pirical therapy of patients with severe
sepsis or septic shock.
Carbapenem: Grade B. Third- or
fourth-generation cephalosporins: Grade
B. Extended-spectrum carboxypenicillins
or ureidopenicillins combined with ␤-lac-
tamase inhibitors: Grade E
Rationale. The use of antibiotic combina-
tions for the treatment of severe infec-
tions relies on the following rationale: 1)
a combination of two antibiotics broad-
ens the antibacterial spectrum, which
may be important as treatment is usually
initiated empirically in critically ill septic
patient; 2) combination of antibiotics
may exert additive or synergistic effects
against the infecting pathogen, resulting
in enhanced antibacterial activity and
possibly better clinical response (120 –
122); 3) the use of a combination of an-
tibiotics may reduce the emergence of
resistant bacteria (123) or of superinfec-
tions (124).
S497

Figure 1. Study flow. Studies were searched in Medline as indicated in Material and Methods. Abstracts
were reviewed, and studies that satisfied the inclusion criteria were selected for a full text review.
MRGP, multiple-resistant Gram-positive bacteria.
In the late 1960s and early 1970s, a
series of retrospectives studies examined
the potential benefits of combination
therapy (27, 125). Overall, combination
therapy was not found to be superior to
single-agent therapy, but some benefits
of antibiotic combinations were noted in
subsets of patients, such as those with
rapidly or ultimately fatal diseases. How-
ever, the impact of these results for to-
day’s practice is limited given the retro-
spective nature of these studies and the
facts that it utilized antibiotics that
would no longer be considered appropri-
ate today and that it did not include mul-
tivariate analyses. Subsequent studies
evaluated the efficacy of different antibi-
otic combinations, usually a ␤-lactam
and an aminoglycoside, for the treatment
of Gram-negative infections (54 –59, 61,
62, 126). By and large, there were no
differences between the various antibiotic
regimens in terms of clinical success, mi-
crobiological eradication, and mortality
rates. However, in one study of 204 pa-
tients with ventilator-associated pneumo-
nia, the clinical cure rate of piperacillin/
tazobactam plus amikacin was superior
to that of ceftazidime plus amikacin (26/
51, 51%, vs. 23/64, 36%; p ⫽ .009), but
mortality was similar in the two treat-
ment groups (8/51, 16%, vs. 13/64, 20%;
p ⫽ .4) (60).
Since the 1980s, the advent of broad-
spectrum and highly bactericidal antibi-
S498
otics, such as the extended-spectrum
penicillins, third-generation or fourth-
generation cephalosporins, or the carbap-
enems, has substantially reduced the
need for aminoglycoside-containing anti-
biotic combinations. During the last two
decades, most antibiotic studies have,
therefore, compared the efficacy and tox-
icity of single-agent antibiotic therapies
(i.e., monotherapies) with that of ␤-lac-
tams paired with aminoglycosides (i.e.,
combination therapies). Twenty-four
studies of empirical monotherapy vs.
combination therapy were identified by
the Medline search. The results obtained
with carbapenems, third-generation or
fourth-generation cephalosporins, and
extended-spectrum penicillins with activ-
ity against Pseudomonas species are
shown in Table 1.
Carbapenems. In five prospective, ran-
domized, controlled studies, mono-
therapy with imipenem-cilastatin or
meropenem was shown to be as effective
as a combination of a ␤-lactam (cefu-
roxime, ceftazidime, or imipenem) and
an aminoglycoside (gentamicin, amika-
cin, or netilmicin) (Table 1) (64 – 66, 74,
75). It is worth mentioning that the same
␤-lactam antibiotic (imipenem) was used
in both treatment arms in only one of
these studies, in which the success rates,
the occurrence of superinfections, and of
P. aeruginosa resistant to the carbap-
enem were similar in the imipenem and
in the imipenem plus netilmicin treat-
ment groups (65). However, nephrotoxic-
ity was significantly more frequent in the
aminoglycoside-containing treatment
arm. In three studies of severe sepsis or
intraabdominal sepsis, imipenem was
found to be at least as efficacious as and
less nephrotoxic (one study) than a com-
bination of gentamicin or tobramycin
and clindamycin or metronidazole (63,
81, 82).
Third-Generation and Fourth-Gener-
ation Cephalosporins. In ten prospective,
randomized, controlled studies, mono-
therapy with a third- or fourth-genera-
tion cephalosporin (cefotaxime, moxalac-
tam, cefoperazone, or ceftazidime) was as
effective as combination therapy with a
␤-lactam (usually an extended-spectrum
penicillin or a cephalosporin) and an ami-
noglycoside (67–71, 77, 78) or clindamy-
cin and an aminoglycoside (76, 83, 84). In
four of the ten studies, clinical success
rates were significantly higher in the
monotherapy treatment groups (69, 76 –
78). However, it is difficult to draw firm
conclusions from these observations as
different ␤-lactams were used in the
monotherapy and combination therapy
treatment groups in all but one study.
Mortality rates were similar in the mono-
therapy and combination therapy treat-
ment groups in nine of the ten studies. A
trend toward a lower mortality in the
monotherapy treatment arm was ob-
served in one trial (77). As expected,
nephrotoxicity was higher in the amin-
oglycoside-containing regimen in seven
of the ten studies, reaching statistical sig-
nificance in five.
Extended-Spectrum Penicillins with
Anti-Pseudomonas Activity. In four pro-
spective, randomized, controlled studies,
extended-spectrum carboxypenicillins (ti-
carcillin) or ureidopenicillins (piperacil-
lin) used either alone or in combination
with a ␤-lactamase inhibitor (clavulanic
acid or tazobactam) were found to be as
effective as amoxicillin-clavulanate, pip-
eracillin, piperacillin-tazobactam, or clin-
damycin combined with an aminoglyco-
side (gentamicin, amikacin, or
netilmicin) as empirical antibiotic ther-
apy for patients with intraabdominal in-
fections, pneumonia, or neonatal sepsis
(72, 79, 85, 86) (Table 1). However, as
noted for the carbapenem or cephalospo-
rin studies, in three of these four studies,
different ␤-lactam antibiotics were used
in the monotherapy and combination
therapy treatment groups. In a study of
206 patients with intraabdominal sepsis,
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
clinical success, and mortality and neph-
rotoxicity rates were similar in patients
treated with piperacillin-tazobactam or
with piperacillin-tazobactam plus amika-
cin (86).
Fluoroquinolones. One study showed
that monotherapy with ciprofloxacin was
as effective as therapy with a ␤-lactam
and/or an aminoglycoside for the treat-
ment of patients with documented Gram-
negative sepsis (80). However, given that
first-generation fluoroquinolones exhibit
suboptimal activities against Gram-
positive bacteria, this class of antibiotics
should not be used as empirical single-
agent therapy in patients with severe sep-
sis.
Comments. The evidence-based re-
view of the literature tends to suggest
that monotherapy with a broad-spectrum
␤-lactam antibiotic is as efficacious and
less nephrotoxic than a combination of a
␤-lactam and an aminoglycoside as em-
pirical therapy for critically ill patients
presenting with severe sepsis or septic
shock. However, monotherapy should not
be regarded as a universal remedy to be
used indiscriminately for several reasons.
Compared with the abundant literature
on empirical therapy for febrile neutro-
penia, few studies have been conducted in
patients with severe sepsis or septic
shock. Moreover, many of the clinical tri-
als reviewed here included ⬍200 pa-
tients, and the statistical power was,
therefore, limited. Furthermore, rarely
was the same ␤-lactam antibiotic used in
the control and experimental treatment
groups, thus limiting the conclusions
that could be drawn about the apparent
equivalent efficacy of the two regimens.
There is clearly a need for large, prospec-
tive, randomized trials to assess the effi-
cacy and toxicity of new antibiotics in
critically ill septic patients.
One should keep in mind that the
choice of empirical antibiotic therapy
depends on several factors related to
the patient’s history (including drug in-
tolerance), underlying diseases, and
susceptibility patterns of microorgan-
isms of the hospital environment and
patient’s community. The initial selec-
tion of an empirical antimicrobial reg-
imen should be broad enough to cover
all likely pathogens. Despite a lack of
clearcut advantage, clinicians may still
prefer to rely initially on a ␤-lactam and
aminoglycoside combination, especially
in the context of high level of antibiotic
resistance or when treating patients
with suspected Pseudomonas infection,
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
even though the benefit of combination
therapy for the latter indication re-
mains controversial. Yet, the benefit
from additive or synergistic effects of
antibiotic combination and the possible
prevention of emerging resistant bacte-
ria must be weighed against the risk of
increased toxicity. Indeed, aminoglyco-
side-containing regimens have been
shown repeatedly to increase the risk of
nephrotoxicity or ototoxicity.
Single-Agent Antibiotic
Therapies
Recommendation: Third-generation and
fourth-generation cephalosporins, car-
bapenems, and extended-spectrum car-
boxypenicillins or ureidopenicillins com-
bined with ␤-lactamase inhibitors are
equally effective as empirical antibiotic
therapy in patients with severe sepsis.
Third- or fourth-generation cephalo-
sporins: Grade A. Carbapenem: Grade B.
Extended-spectrum carboxypenicillins or
ureidopenicillins combined with ␤-lacta-
mase inhibitors: Grade C.
Carbapenems, Third-Generation, or
Fourth-Generation Cephalosporins. Six
prospective, randomized, clinical trials
(of which five included between 150 and
400 patients) were identified that investi-
gated the efficacy and safety of a carbap-
enem (imipenem or meropenem), a
third-generation cephalosporin (ceftazi-
dime or cefotaxime combined with met-
ronidazole), a ureidopenicillin with a
␤-lactamase inhibitor (piperacillin-ta-
zobactam), or another carbapenem (Ta-
ble 2) (92, 95–97, 99, 107, 111). In all but
one study, virtually identical clinical re-
sponse rates and overall mortality were
observed in patients treated with either
imipenem or meropenem or control
␤-lactam antibiotics. In a smaller clinical
trial of 94 patients with intraabdominal
sepsis, the clinical response rate of pa-
tients treated with meropenem was supe-
rior to that of those treated with cefo-
taxime plus metronidazole (111).
In summary, as one might have antic-
ipated given the fairly comparable anti-
microbial spectrum of activity of most
third-generation and fourth-generation
cephalosporins, the studies conducted
with these agents for the treatment of
patients with severe sepsis and commu-
nity-acquired or nosocomial (including
ventilator-associated) pneumonia or mis-
cellaneous types of infections have
yielded very comparable results (Table 2)
(91,93,94,98,101–104,108). Response
rates for the various cephalosporins were
in the range of 65% to 85% in most
studies, except for slightly lower response
rates for ceftriaxone and cefotaxime in
two smaller studies (101, 104).
Aminoglycosides, Monobactams, or
Fluoroquinolones. In the late 1970s and
early 1980s, three studies compared the
efficacy and safety of different aminogly-
cosides (gentamicin vs. tobramycin, gen-
tamicin vs. amikacin, and netilmicin vs.
amikacin) as single-agent therapy for
Gram-negative sepsis (87– 89). Clinical
responses rates, overall mortality, and
nephrotoxicity were similar among the
various aminoglycosides compared in
these studies. A high incidence of amin-
oglycoside-induced nephrotoxicity and
ototoxicity were noted in these studies,
which is a major concern in critically ill
septic patients already at high risk of de-
veloping organ failures due to the circu-
latory collapse associated with severe sep-
sis or septic shock. In two small studies,
aztreonam was found to be as effective
and less toxic than aminoglycosides as
empirical monotherapy of patients with
severe Gram-negative sepsis (90, 100).
However, today, aminoglycosides or a
monobactam antibiotic such as aztreo-
nam should not be used as single-agent
empirical therapy of severe sepsis because
of their narrow antibacterial activity and
risk of toxicity (aminoglycosides).
More recently, five studies, of which
four enrolled only patients with severe
community-acquired or nosocomial
pneumonias, showed that fluoroquinolo-
nes (ciprofloxacin, four studies; levofloxa-
cin, one study) were as efficacious as
␤-lactam antibiotics (imipenem in four
studies; various ␤-lactams with or with-
out an aminoglycoside in one study) for
the treatment of patients with severe sep-
sis caused predominantly by Gram-
negative bacteria (Table 2) (56, 105, 106,
109, 110). Of note, a large number of
patients (ranging between 250 and 400)
were randomized in three of the four
pneumonia studies. However, given the
limited activity of first-generation fluoro-
quinolones (i.e., norfloxacin and cipro-
floxacin) against Gram-positive bacteria
and the possibility of resistance among
Gram-negative bacteria, these antibiotics
cannot be recommended as empirical
monotherapy for patients with severe
sepsis or septic shock of unknown etiol-
ogy, including patients with community-
acquired or healthcare-associated pneu-
monia. The most recent, so-called
respiratory fluoroquinolones (i.e., levo-
S499
Table 1. Monotherapy versus combination antibiotic therapy as empirical treatment of severe sepsis and septic shock

Authors and Yr of Type of Experimental Control Success (%)

Publication Infection Therapy Therapy (Exp. vs. Control) p

Carbapenems vs. ␤-lactams or antianaerobes combined with an aminoglycoside

Solomkin et al. 1985 (63) SEPSIS Imipenem Gentamicin ⫹ clindamycin 28/37 (76) vs. 27/37 (73) 1.000
Poenaru et al. 1990 (81) ABDOM Imipenem Tobramycin ⫹ clindamycin — —
or metronidazole
Solomkin et al. 1990 (82) ABDOM Imipenem Tobramycin ⫹ clindamycin 67/81 (83) vs. 57/81 (70) .094
Mouton et al. 1990 (64) SEPSIS Imipenem Cefotaxime ⫹ amikacin 58/70 (83) vs. 54/70 (77) .527
a
Cometta et al. 1994 (65) SEPSIS Imipenem Imipenem ⫹ netilmycin 113/142 (80) vs. 119/138 (86) .155
Sieger et al. 1997 (74) HAP Meropenem Ceftazidime ⫹ amikacin 76/106 (72) vs. 62/105 (59) .061
Jaspers et al. 1998 (66) SEPSIS Meropenem Cefuroxime ⫹ gentamicin 27/39 (69) vs. 25/40 (63) .637
⫹/⫺ metronidazole
Alvarez-Lerma et al. 2001 (75) HAP Meropenem Ceftazidime ⫹ amikacin 47/69 (68) vs. 39/71 (55) .121
Third- or fourth-generation cephalosporins vs. ␤-lactams or antianaerobes combined with an aminoglycoside
Arich et al. 1987 (67) SEPSIS Cefotaxime Cefazolin ⫹ tobramycin 22/25 (88) vs. 17/22 (77) .446
Schentag et al. 1983 (83) ABDOM Moxalactam Tobramycin ⫹ clindamycin 37/49 (76) vs. 36/49 (73) 1.000
a
Oblinger et al. 1982 (68) SEPSIS Moxalactam Conventional therapy 33/38 (87) vs. 32/40 (80) .547
Mangi et al. 1988 (76) HAP Cefoperazone Clindamycin or cefazolin 41/46 (89) vs. 44/61 (72) .052
and gentamicin
Greenberg et al. 1994 (84) ABDOM Cefoperazone- Clindamycin ⫹ gentamicin 33/47 (70) vs. 15/29 (52) .143
sulbactam
b
Fernandez-Guerrero et al. 1991 (77) HAP Cefotaxime Combination therapy 217/275 (79) vs. 193/273 (71) .030
Croce et al. 1993 (78) HAP Ceftazidime or Ceftazidime or cefoperazone 22/39 (56) vs. 22/70 (31) .015
cephoperazone ⫹ gentamicin
Rubinstein et al. 1995 (69) SEPSIS Ceftazidime Ceftriaxone ⫹ tobramycin 227/306 (74) vs. 179/274 (65) .023
d
Extermann et al. 1995 (70) SEPSIS Ceftazidime “Best guess” combination 38/41 (93) vs. 28/30 (93) 1.000
McCormick et al. 1997 (71) SEPSIS Ceftazidime Mezlocilin ⫹ netilmicin 50/65 (77) vs. 48/63 (76) 1.000
Antipseudomonas penicillins vs. ␤-lactams or anti-anaerobes combined with an aminoglycoside
Fink 1991 (85) ABDOM Ticarcillin-clavulanate Gentamicin ⫹ clindamycin 15/20 (75) vs. 16/25 (64) .525
Hammerberg et al. 1989 (72) SEPSIS Piperacillin Ampicillin ⫹ amikacin — —
Speich et al. 1998 (79) CAP ⫹ HAP Piperacillin- Amoxiclav ⫹ netilmicin or 37/41 (90) vs. 36/43 (84) .521
tazobactam gentamicin
Dupont et al. 2000 (86) ABDOM Piperacillin- Piperacillin-tazobactam ⫹ 44/99 (44) vs. 55/105 (52) .266
tazobactam amikacin
Miscellaneous monotherapies versus ␤-lactams combined with an aminoglycoside
Korvick et al. 1992 (73) SEPSIS f
␤-lactam or ␤-lactam ⫹ aminoglycoside — —
aminoglycoside
alone
Manhold et al. 1998 (80) HAP Ciprofloxacin Ceftazidime ⫹ gentamicin — —

p, two-tailed Fisher’s exact test; ABDOM, intraabdominal infections; CAP, community-acquired pneumonia; HAP, hospital-acquired pneumonia
(ventilator-associated or not).
a
Conventional therapy included a ␤-lactam (i.e., penicillin, ampicillin, ticarcillin, nafcillin, cefamandole, cefoxitin) given either alone or in combination
with an aminoglycoside (tobramycin or amikacin); bcombination therapy comprised a ␤-lactam (a cephalosporin or a “broad-spectrum” penicillin) with an
aminoglycoside (gentamicin, tobramycin, or amikacin); cinfection-related mortality; dbest guess combination usually include a ␤-lactam with an
aminoglycoside (not specified); f␤-lactam included a cephalosporin, imipenem, or an extended-spectrum penicillin; aminoglycosides included gentamicin,
tobramycin, amikacin, or kanamycin.

floxacin, gatifloxacin, moxifloxacin, or
gemifloxacin) exhibit enhanced in vitro
activities against Gram-positive bacteria.
Numerous studies have shown that these
agents are highly efficacious as single-
agent therapy of community-acquired
pneumonia in the outpatient or inpatient
setting (48, 50, 127). When used for the
treatment of severe community-acquired
or nosocomial infections in intensive care
unit (ICU) patients in whom Pseudomo-
nas infection is a concern, ciprofloxacin
or the respiratory fluoroquinolones
should preferentially be combined with
an anti-Pseudomonas antibiotic (i.e., a
carbapenem, a third- or fourth-genera-
tion cephalosporin, or an extended-
spectrum penicillin) with or without an
aminoglycoside.
Empirical Use of Anti-Gram-
Positive Antibiotics in Patients
with Severe Sepsis
Recommendation: Empirical use of gly-
copeptide antibiotics (vancomycin, teico-
planin), oxazolidinones (linezolid), or
streptogramins (quinupristin/dalfopris-
tin) in patients with severe sepsis or sep-
tic shock is justified in patients with hy-
persensitivity to ␤-lactams or in
institutions with resistant Gram-positive
bacteria (i.e., methicillin-resistant staph-
ylococci, penicillin-resistant pneumo-
cocci, or ampicillin-resistant entero-
cocci) in the community or in the
hospital.
Grade E
Rationale: Antimicrobial coverage of both
Gram-positive and Gram-negative bacte-
ria is mandatory in the empirical treat-
ment of critically ill patients with severe
sepsis. Naturally, many broad-spectrum
antibiotics reliably cover both pathogens
and, therefore, can be used for empirical
therapy. The choice of empirical treat-
ment, however, should depend on the lo-
cal epidemiology of pathogens associated
with infections and their antimicrobial

S500 Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)

Table 1. Continued.

Mortality (%) (Exp. vs. Control) p Nephrotoxicity (%) (Exp. vs. Control) p

6/37 (16) vs. 4/37 (11) .736 1/37 (3) vs. 10/27 (27) .007
4/52 (8) vs. 9/52 (17) .235 — —

11/81 (14) vs. 14/81 (17) .664 — —

7/70 (10) vs. 7/70 (10) 1.000 1/105 (1) vs. 4/102 (4) .369
18/142 (13) vs. 13/138 (9) .448 0/158 (0) vs. 6/149 (4) .014
13/104 (13) vs. 23/107 (21) .100 0/104 (0) vs. 2/105 (2) .498
3/39 (8) vs. 4/40 (10) 1.000 2/39 (5) vs. 5/35 (13) .432

16/69 (23) vs. 20/71 (28) .564 0/69 (0) vs. 2/69 (3) .497

8/25 (32) vs. 5/22 (23) .530 0/25 (0) vs. 3/19 (14) .095
7/49 (14) vs. 6/49 (12) 1.000 6/49 (12) vs. 14/35 (29) .078
8/33 (24) vs. 9/32 (28) .783 3/41 (7) vs. 11/36 (23) .046
9/61 (15) vs. 13/71 (18) .645 6/61 (10) vs. 12/66 (15) .447

6/47 (13) vs. 3/29 (10) 1.000 0/47 (0) vs. 1/28 (3) .382

36/275 (13) vs. 52/273 (19) .063 0/275 (0) vs. 7/266 (3) .007
c
2/39 (5) vs. 7/70 (10) .485 5/39 (13) vs. 25/45 (36) .013

31/306 (10) vs. 33/274 (12) .508 0/306 (0) vs. 9/265 (3) .001
6/41 (15) vs. 4/30 (13) 1.000 — —
13/65 (20) vs. 9/63 (14) .484 2/65 (3) vs. 8/55 (13) .053

3/20 (15) vs. 5/25 (20) .716 1/20 (5) vs. 0/25 (0) .444
17/200 (9) vs. 27/196 (14) .110 50/200 (25) vs. 43/153 (22) .480
1/41 (2) vs. 6/43 (14) .110 0/44 (0) vs. 2/43 (4) .494

19/99 (19) vs. 22/105 (21) .862 3/99 (3) vs. 3/102 (3) 1.000

24/118 (20) vs. 20/112 (18) .738 — —

13/28 (46) vs. 6/23 (26) .158 — —

susceptibility. It is, therefore, essential to
get frequent feedbacks from the microbi-
ology laboratory on antibiotic susceptibil-
ities trends. Major changes in the relative
importance of Gram-positive and Gram-
negative infections among critically ill
patients have occurred during the last 10
yrs. While Gram-negative bacteria pre-
dominated until the middle 1980s, Gram-
positive bacteria now account for at least
one-half of the infections occurring in
patients with severe sepsis (128, 129).
Moreover, methicillin-resistant S. aureus
(MRSA) and methicillin-resistant Staph-
ylococcus epidermidis are responsible for
a majority of staphylococcal infections in
some institutions. The frequency of pen-
icillin-resistant S. pneumoniae is also in-
creasing in many areas of the world. Yet,
the relevance of reduced susceptibility of
Streptococcus pneumoniae to penicillin
on clinical outcome of patients treated
with ␤-lactam antibiotics is uncertain,
especially for respiratory tract infections
(130). Clinical failures, however, have
been reported in patients with bacterial
meningitis caused by a penicillin-resis-
tant S. pneumoniae treated with ceftriax-
one (131). Does this epidemiologic con-
text justify the empirical use of
glycopeptides (vancomycin or teicopla-
nin), oxazolidinones (linezolid), or
streptrogramins (quinupristin/dalfopris-
tin) on a routine basis in all patients with
severe sepsis and septic shock?
Glycopeptides. To the best of our
knowledge, randomized trials comparing
empirical treatment with or without
Gram-positive coverage (including glyco-
peptides) have never been performed in
adult nonneutropenic patients and such
an approach would probably be judged as
unethical. None of the articles retrieved
by the Medline search included a compar-
ison of empirical treatment with or with-
out Gram-positive patients in nonneutro-
penic adults with severe sepsis. Although
the indiscriminate use of glycopeptides
for presumed Gram-positive infections in
patients with severe sepsis or septic shock

Crit Care Med 2004 Vol. 32, No. 11 (Suppl.) S501

Table 2. Comparison of monotherapies as empirical treatment of severe sepsis and septic shock

Authors and Yr of Type of Experimental Success (%) Mortality (%)

Publication Infection Therapy Control Therapy (Exp. vs. Control) p (Exp. vs. Control) p

Carbapenems versus third-generation cephalosporins or carbapenems

Norrby et al. 1993 (92) SEPSIS Imipenem Ceftazidime 27 /197 (64) vs. 124/196 (63) .834 20/197 (10) vs. 31/196 (16) .101
Kempf et al. 1996 (111) ABDOM Meropenem Cefotaxime ⫹ 41 /43 (95) vs. 30/40 (75) .012 3/43 (7) vs. 5/40 (13) .473
metronidazole
Colardyn and Faulkner 1996 (95) SEPSIS Meropenem Imipenem 68 /90 (76) vs. 67/87 (77) .861 24/106 (23) vs. 17/98 (17) .385
Mehtar et al. 1997 (96) SEPSIS Meropenem Ceftazidime ⫹ 46 /48 (96) vs. 40/43 (93) .664 10/68 (15) vs. 11/63 (17) .812
metronidazole
Garau et al. 1997 (97) SEPSIS Meropenem Imipenem 55 /66 (83) vs. 49/67 (73) .208 22/76 (29) vs. 26/75 (35) .488
a
Jaccard et al. 1998 (107) HAP Imipenem Piperacillin-tazobactam 56 /79 (71) vs. 62/75 (83) .091 6/79 (8) vs. 7/75 (9) .777
Verwaest 2000 (99) SEPSIS Meropenem Imipenem 67 /87 (77) vs. 62/91 (68) .240 22/107 (21) vs. 14/105 (13) .201
Third- or fourth-generation cephalosporins vs. a third-generation cephalosporin
Mangi et al. 1988 (91) SEPSIS Cefoperazone Ceftazidime 48 /62 (77) vs. 54/63 (86) .256 23/62 (37) vs. 24/63 (38) 1.000
Reeves et al. 1989 (101) HAP Ceftriaxone Cefotaxime 12 /25 (48) vs. 19/26 (73) .089 2/25 (8) vs. 4/26 (15) .668
Mangi et al. 1992 (102) HAP Ceftriaxone Cefoperazone 35 /50 (70) vs. 48/60 (80) .269 12/50 (24) vs. 10/60 (17) .351
Thomas et al. 1992 (103) HAP Cefotaxime Ceftriaxone 8 /12 (67) vs. 10/15 (67) 1.000 1/12 (8) vs. 5/15 (33) .182
Norrby and Geddes 1993 (93) SEPSIS Cefpirome Ceftazidime 131 /176 (74) vs. 34/50 (68) .372 28/282 (10) vs. 10/80 (13) .536
Barckow et al. 1993 (104) CAP ⫹ HAP Cefepime Cefotaxime 27 /37 (73) vs. 10/18 (56) .231 13/37 (35) vs. 4/18 (22) .372
Schrank et al. 1995 (94) SEPSIS Cefepime Ceftazidime 11 /13 (85) vs. 12/15 (80) 1.000 1/13 (8) vs. 2/15 (13) 1.000
Norrby et al. 1998 (98) SEPSIS Cefpirome Ceftazidime 123 /188 (65) vs. 132/188 (70) .377 15/188 (8) vs. 23/188 (12) .231
Grossman et al. 1999 (108) CAP Cefepime Ceftriaxone 53 /67 (79) vs. 46/61 (75) .676 7/76 (9) vs. 7/75 (9) 1.000
Fluoroquinolones vs. ␤-lactams
b
Fink et al. 1994 (105) CAP ⫹ HAP Ciprofloxacin Imipenem 92 /202 (46) vs. 90/200 (45) .921 43/202 (21) vs. 38/200 (19) .620
Siami et al. 1995 (106) CAP ⫹ HAP Ciprofloxacin Imipenem 17 /24 (71) vs. 14 /21 (67) 1.000 2/24 (8) vs. 3/21 (14) .652
Krumpe et al. 1999 (56) SEPSIS Ciprofloxacin Several ␤-lactams 138 /166 (83) vs. 74/87 (85) .858 26/207 (13) vs. 13/99 (13) .857
Torres et al. 2000 (109) HAP Ciprofloxacin Imipenem 40 /57 (70) vs. 34/52 (65) .683 8/41 (20) vs. 4/34 (12) .529
West et al. 2003 (110) HAP Levofloxacin Imipenem followed by 135 /204 (66) vs. 143/206 (69) .526 38/220 (17) vs. 32/218 (15) .515
(IV then oral) oral ciprofloxacin

p, two-tailed Fisher’s exact test; CAP, community-acquired pneumonia; HAP, hospital-acquired pneumonia (ventilator-associated or not).
a
Infection-related mortality; bsame study as Snydam et al. 1994.

should be avoided, their use is appropri-
ate in severely ill patients in the following
circumstances: 1) endemic levels of
MRSA; 2) documented hypersensitivity to
␤-lactam antibiotics; and 3) treatment of
bacterial meningitis in areas with high
levels of penicillin-resistance of S. pneu-
moniae. Naturally, “endemic” and “high”
levels of resistance are subjective, and no
studies have been performed to deter-
mine the level of resistance above which
the empirical use of antibiotics active
against these resistant pathogens is war-
ranted. There probably is an inverse rela-
tionship between the severity of sepsis
and the willingness to accept a risk not to
cover an antibiotic-resistant pathogen
empirically. However, the possible clini-
cal benefit associated with the empirical
use of glycopeptides should be weighed
against the risks of selecting resistant mi-
croorganisms and of increased toxicity,
especially when vancomycin is used in
combination with an aminoglycoside or
other nephrotoxic agents. To further re-
duce the risk of emergence of vancomy-
cin-resistant staphylococci, empirical
vancomycin therapy should be rapidly
discontinued in patients in whom Gram-
positive infections have been ruled out.
Finally, it is rarely, if ever, appropriate to
use vancomycin alone as empirical ther-
apy since most cases require additional
Gram-negative coverage, at least until
microbiological results are available.
Empirical use of either vancomycin or
teicoplanin has been common practice in
the management of neutropenic patients
with fever, in whom coagulase-negative
staphylococci and viridans streptococci
are predominant pathogens. Although
some studies have suggested that the use
of glycopeptides at the initiation of em-
pirical therapy might be beneficial, a re-
cent prospective, randomized, double-
blind trial failed to demonstrate clinical
benefits (132). Prolonged courses of van-
comycin therapy have been associated
with the development of vancomycin-
resistance S. aureus, either through re-
duced cell wall permeability (133) or
through acquisition of the vanA gene
from vancomycin-resistant enterococci
(VRE) (134).
Linezolid and Quinupristin/Dalfopris-
tin. Linezolid and quinupristin/dalfopris-
tin are recently introduced antibiotics for
the treatment of Gram-positive infec-
tions. Linezolid is a new oxazolidinone
antibiotic with activity against staphylo-
cocci (including MRSA), and Enterococ-
cus faecium and Enterococcus faecalis
(including strains resistant to vancomy-
cin) (135). Indications for the empirical
use of linezolid are similar to that of
glycopeptides plus two other indications,
namely the need for empirical coverage of
VRE and documented hypersensitivity to
glycopeptides. The use of linezolid is ap-
proved for infections with VRE (including
bacteremia), community-acquired and
nosocomial pneumonia caused by S.
pneumoniae or S. aureus (including
MRSA), and skin and soft tissue infec-
tions caused by Gram-positive bacteria.
Four randomized trials comparing the
efficacy of linezolid with alternative treat-
ment in adult nonneutropenic patients
have been identified (Table 3). In an
open-labeled randomized trial, San Pedro
and coworkers (136) compared linezolid
(sequential intravenous-to-oral therapy)
with ceftriaxone IV followed by oral cef-
podoxime for patients hospitalized with
community-acquired pneumonia. Epi-
sodes of community-acquired pneumonia
were primarily caused by S. pneumoniae,
and overall clinical cure rates were
higher for linezolid-treated patients, es-
pecially for patients with S. pneumoniae
bacteremia (93.1% vs. 68.2%; p ⫽ .021).
In two double-blind, randomized trials, a
combination of linezolid and aztreonam
was compared with vancomycin and az-
treonam for the treatment of patients
with hospital-acquired pneumonia (119,
137). Although equivalence of activity
was found in both trials, a subsequent
retrospective analysis of the combined

S502 Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)

Table 3. Clinical studies of linezolid in patients with Gram-positive infections
Authors and Yr of Type of
Publication Infection
Rubinstein et al. 2001 (119) HAP
Stevens et al. 2002 (139) Known/suspected MRSA Linezolid
infections
Documented MRSA
infections
San Pedro et al. 2002 (136) CAP due to S.
pneumoniae
Wunderink et al. 2003 (137) HAP
Wunderink et al. 2003 (138) Documented SA HAP
(n ⫽ 339)
Documented MRSA HAP
(n ⫽ 160)
SA HAP diagnosed by
invasive procedure
(n ⫽ 223)
MRSA HAP diagnosed by
invasive procedure
(n ⫽ 95)
HAP, hospital-acquired pneumonia (ventilator-associated or not); MRSA, methicillin-resistant Staphylococcus aureus; SA, Staphylococcus aureus.
a
In the clinically evaluable population; bIntention-to-treat analysis.
dataset revealed that initial therapy with
linezolid was associated with significantly
better survival and clinical cure rates in
patients with nosocomial pneumonia due
to MRSA (138). Finally, in an open-
labeled randomized trial, Stevens and co-
workers (139) compared the efficacy of
linezolid with that of vancomycin for the
treatment of patients with suspected
MRSA. Linezolid and vancomycin had
similar efficacy, both in an intention-to-
treat analysis and in a subgroup of pa-
tients with documented MRSA infections.
Most patients, however, had skin or soft
tissue infections and probably did not ful-
fill the criteria of severe sepsis. No clini-
cal data are available on the use of lin-
ezolid for bacterial meningitis in areas of
high levels of penicillin-resistant S. pneu-
moniae.
Resistance to linezolid is based on spe-
cific point mutations in the 23S ribo-
somal RNA of the 50S subunit of the
ribosome preventing binding of linezolid
(135). Resistance development of VRE
and S. aureus during therapy was associ-
ated with the presence of indwelling pros-
thetic devices and prolonged courses of
antimicrobial therapy (140, 141). More-
over, nosocomial spread of linezolid-
resistant VRE has been reported as well
(142). Therefore, empirical linezolid ther-
apy should be rapidly discontinued in pa-
tients in whom Gram-positive infections
have been ruled out.
Quinupristin/dalfopristin is an in-
jectable streptogramin available for the
treatment of Gram-positive infections
caused by S. aureus (including MRSA)
and E. faecium (including VRE), but it
has no activity against E. faecalis. Two
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
Experimental Control Success (%)
Therapy Therapy (Exp. vs. Control)
Linezolid ⫹ Vancomycin ⫹ 71/107 (66.4)a vs. 62/91 (68.1)a
aztreonam aztreonam 86/161 (53)b vs. 74/142 (52)b
Vancomcyin 109/192 (57)b vs. 93/169 (55)b
41/56 (73) vs. 38/52 (73)
Linezolid Ceftriaxon/ 316/381 (83) vs. 280/366 (76)
cefpodoxim
Linezolid ⫹ Vancomycin ⫹ 135/256 (53)b vs. 128/245 (52)b
aztreonam aztreonam 114 /168 (68)a vs. 111/171 (65)a
Linezolid ⫹ Linezolid ⫹ 70/136 (51.5)a vs. 59/136 (43.4)a
aztreonam aztreonam
36/61 (59.0)a vs. 22/62 (35.5)a
47/92 (51)a vs. 39/90 (43)a
19/33 (58)a vs. 13/39 (33)a
randomized clinical trials of quinupris-
tin/dalfopristin have been identified
(Table 4). Fagon and coworkers (118)
compared quinupristin/dalfopristin
with vancomycin (aztreonam was al-
lowed in both treatment arms) in pa-
tients with Gram-positive nosocomial
pneumonia. The second study was a
combined report of two nearly identical
open-label randomized trials in which
patients with Gram-positive skin and
skin structure infections were treated
with quinupristin/dalfopristin or with
one of three comparator antibiotics (ce-
fazolin, vancomycin, or oxacillin) (143).
In both studies, clinical cure rates of
quinupristin/dalfopristin and of control
antibiotics were similar. However, only
a few patients fulfilled the criteria of
severe sepsis in the latter study (143). A
formal head-to-head comparison of lin-
ezolid and quinupristin/dalfopristin has
not been performed. Therefore, empir-
ical therapy with quinupristin/dalfo-
pristin should be limited to severely ill
patients with a high likelihood of infec-
tion caused by vancomycin-resistant E.
faecium or MRSA and to patients with
documented or presumed hypersensi-
tivity to linezolid. The indiscriminate
use of linezolid or of quinupristin/
dalfopristin for presumed Gram-posi-
tive infections in patients with severe
sepsis or septic shock should be
avoided. The possible clinical benefit
associated with the empirical use of
these agents should be weighed against
the risks of selection of resistant micro-
organisms, toxicity, and increased
treatment costs. Finally, it is rarely, if
ever, appropriate to use linezolid or
Survival (%) p
p (Exp. vs. Control)
NS 36/203 (18)b vs. 49/193 (25) .067
.908
NS b
40/240 (17) vs. 30/220 (14) b
NS
NS
.04 366/381 (4) vs. 347/366 (5) NS
NS 257/321 (80)b vs. 241/302 (80)b NS
145/157 (92)a vs. 150 /166 (90)a
NS 131/168 (78) vs. 121/171 (71) NS
⬍.01 60/75 (80) vs. 54/85 (64) .025
NS 86/109 (79) vs. 82/114 (72) NS
.04 34/40 (85) vs. 37/55 (67) .05
quinupristin/dalfopristin alone as em-
pirical therapy of severe sepsis or septic
shock, since Gram-negative coverage is
needed, at least until microbiological
results become available.
Modification of Empirical
Antimicrobial Therapy Based on
Culture Results
Recommendation: Modification of empir-
ical antimicrobial therapy with the aim to
restrict the number of antibiotics and
narrow the spectrum of antimicrobial
therapy is an important and responsible
strategy for minimizing the development
of resistant pathogens and for containing
of costs.
Grade E
To the best of our knowledge no stud-
ies have been performed in which pa-
tients were randomized either to con-
tinue to receive empirical broad-
spectrum antibiotic therapy or to be
switched to a narrow-spectrum antibiotic
regimen selected on the basis of suscep-
tibility data of the causative pathogen. In
a before-after study design, Ibrahim and
coworkers (144) assessed the effects of a
clinical guideline on the appropriateness
of empirical therapy in patients with ven-
tilator-associated pneumonia. The clini-
cal guideline consisted of empirical ther-
apy with imipenem, ciprofloxacin, and
vancomycin, followed by narrowing of
antibiotic therapy when a microbiological
cause of infection was identified and dis-
continuation of antibiotic therapy after 7
days when the clinical condition allowed.
Implementation of the clinical guideline
S503
Table 4. Clinical studies of quinupristin/dalfopristin in patients with Gram-positive infections
Authors and Yr of Type of
Publication Infection
Fagon et al. 2000 (118) HAP
Nichols et al. 1999 (143) Complicated Gram-
positive skin and
structure infections
HAP, hospital-acquired pneumonia (ventilator-associated or not).
a
In the clinically evaluable population.
was associated with higher appropriate-
ness of empirical therapy and a reduction
in the mean duration of therapy. How-
ever, the number of patients in whom
antibiotics were narrowed was not re-
ported.
In fact, there are arguments both in
favor and against such an approach. On
the one hand, narrowing of the antimi-
crobial spectrum may reduce the risk of
selecting resistant microorganisms and
treatment costs. On the other hand, mi-
crobiological documentation of the etiol-
ogy of sepsis is lacking in a significant
proportion of sepsis cases, making it dif-
ficult, if at all possible, to narrow the
antibiotic coverage based on clinical cri-
teria only.
Recommendations: The antimicrobial
regimen should always be reassessed after
48 –72 hrs on the basis of microbiological
and clinical data with the aim of using a
narrow-spectrum antibiotic to recant the
development of resistance, to reduce tox-
icity, and to reduce costs. Once a caus-
ative pathogen is identified, there is no
evidence that combination therapy is
more effective monotherapy. The dura-
tion of therapy should typically be 7–10
days and guided by clinical response.
Grade E
Recommendations: If the presenting clini-
cal syndrome is determined to be due to a
noninfectious cause, antimicrobial therapy
should be stopped promptly to minimize
the development of resistant pathogens and
superinfection with other pathogenic or-
ganisms.
Grade E
Antibiotic Cycling
Recommendations: At the present time,
there is insufficient evidence to recom-
mend the use of antibiotic cycling as a
strategy to reduce the development of
antibiotic resistance.
S504
Experimental Control Success (%)
Therapy Therapy (Exp. vs. Control)
Quinup/Dalfop ⫹ Vancomycin ⫹ 49/87 (56)a vs. 49/84 (58)a
aztreonam aztreonam
b b
65/150 (43) vs. 67/148 (45)
Quinup/Dalfop Cefazolin or 68.2%a vs. 70.7%
vancomycin
or oxacillin
Grade C
Rationale: Antibiotic cycling has been
proposed as a strategy to minimize the
development of antibiotic resistance and
thus to improve the patient’s outcome.
The rationale is that the cyclic exposure
to different classes of antibiotics should
prevent emergence of resistance by ex-
posing the microbial flora to homoge-
neous selective antibiotic pressure during
a limited period of time. This concept is
based on several assumptions. First, it is
assumed that antibiotic-resistant micro-
organisms have a growth disadvantage
when the selective antibiotic pressure is
withdrawn. Therefore, development of re-
sistance during exposure to a given anti-
biotic will be counterbalanced during pe-
riods of nonexposure. Second, should
resistance develop during one period, ex-
posure to another class of antibiotics in
the following cycle will eliminate the re-
sistant microorganisms. For this to oc-
cur, it presupposes the absence of cross-
resistance (i.e., that the mechanism of
resistance be different among different
classes of antibiotics). However, besides
these theoretical considerations, antibi-
otics are just one of many factors that
have an impact on the development of
resistance. For example, changes in the
numbers of patients introducing resistant
microorganisms into the unit or changes
in compliance with hygienic measures
will also affect the emergence of antibi-
otic resistance. The effectiveness of infec-
tion control programs can be influenced
by changes in the workload of healthcare
workers or understaffing, which both will
lead to more contacts between patients
and healthcare workers and less adher-
ence to hand hygiene and thus to more
transmission of pathogens (145–147).
Whether the theoretical benefits of anti-
biotic cycling hold true in daily practice
can only be tested in studies in which
there is careful control of confounding
variables.
Survival (%)
p (Exp. vs. Control) p
NS NS
b b
NS 38/150 (25) vs. 116/148 (22)
NS
The “real-life” experience with antibi-
otic cycling is sparse, and interpretation
of findings is seriously hampered by nu-
merous methodologic problems in study
design. These included a nonstructured
antibiotic cycling scheme (148), evalua-
tion of changes in empirical therapy in a
before-after design (149), use of multiple
antibiotics for cycling and different cy-
cling intervals (148, 150), presence of im-
portant confounding factors, such as re-
duction in the antibiotic use (151), or
changes in infection control strategies
(152).
Clinical Data. The impact of antibiotic
cycling on a patient’s outcome has been
analyzed in three studies. Kollef and co-
workers (149) compared the effects of re-
placing ceftazidime by ciprofloxacin in
the empirical treatment of suspected
Gram-negative infections in two 6-month
periods. The incidence of ventilator-
associated pneumonia (VAP) decreased
from 11.6% to 6.7% (p ⫽ .028), as did the
incidences of VAP and bacteremia due to
antibiotic-resistant Gram-negative bacte-
ria (4% to 0.9%, p ⫽ .013, and 1.7% to
0.3%, p ⫽ .12, respectively). However,
outcome parameters such as survival and
length of stay remained unaffected. In
another trial, Raymond and coworkers
(152) compared a 1-yr period of nonpro-
tocol-driven antibiotic use with a subse-
quent 1-yr period of rotating empirical
antibiotic assignment. Different classes of
antibiotics were used for different indica-
tions (pneumonia vs. peritonitis or sepsis
of unknown origin) with rotation of an-
tibiotic combinations after 4 months. In-
fection rates with resistant Gram-positive
and Gram-negative bacteria significantly
decreased, survival in the ICU increased,
and antibiotic rotation was identified as
an independent predictor for survival in
logistic regression analysis. Yet, other
confounding factors were changed as
well. An antibiotic surveillance team was
instituted in the second half of the first
study period, and alcohol hand dispensers
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
were distributed at the start of the second
period. In the third study, Gruson and
coworkers (151) investigated whether a
new program of antibiotic use had an
impact on the incidence of VAP caused by
antibiotic-resistant bacteria in a before-
after study. In the second part of the
study, ceftazidime and ciprofloxacin use
were restricted, antibiotics were rotated
without favoring any one antibiotic, and
each antibiotic prescription was deter-
mined by one of the two investigators. In
addition to these measures, a policy of
shorter duration of aminoglycoside ther-
apy was implemented and compliance
with protocol was stimulated by daily
meetings between of one of the two in-
vestigators with critical care physicians,
weekly meetings with nurses to identify
nosocomial problems, monthly evalua-
tion of antibiotic consumption, and
three-monthly evaluation of changes in
resistance patterns. The rotation protocol
included the variable use of a ␤-lactam
antibiotic (cefepime or piperacillin-
tazobactam or imipenem or ticarcillin
during periods of at least 1 month) and
an aminoglycoside (amikacin or tobramy-
cin, netilmicin or isepamicin) for late-
onset VAP. For early-onset VAP, prescrip-
tion was rotated on a monthly basis using
either amoxicillin-clavulanic acid, cefo-
taxime, ceftriaxone, or cefpirome in com-
bination with an aminoglycoside or fos-
fomycin. Importantly, implementation of
all these measures reduced overall anti-
biotic use by ⬎50%. In addition, antibi-
otic resistance levels of isolated patho-
gens decreased, both for antibiotic use
that had been reduced (i.e., ciprofloxacin
and ceftazidime) and for antibiotics use
that had been increased (i.e., cefepime
and piperacillin-tazobactam). Incidences
of microbiologically confirmed VAP de-
creased from 22.1% to 15.7% (p ⬍ .01)
among patients requiring mechanical
ventilation for ⬎48 hrs, but ICU mortal-
ity did not change (40.6% to 37.2%; p ⫽
NS).
In summary, none of these studies
have taken into account the most impor-
tant factors contributing to the develop-
ment of dynamics of antibiotic resistance
in the hospital settings (i.e., relative im-
portance of introduction, cross-transmis-
sion, and endogenous acquisition by se-
lective antibiotic pressure), precluding a
reliable assessment of the role played by
antibiotic cycling. Furthermore, the op-
timal choice of antibiotics and cycle in-
tervals remain to be determined (153,
154).
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
Empirical Use of Anti-Fungal
Agents in Patients with Severe
Sepsis or Septic Shock
Recommendation: Empirical antifungal
therapy should not be used on a routine
basis in patients with severe sepsis or
septic shock, but it may be justified in
selected subsets of septic patients at high
risk for invasive candidiasis (155).
Grade E
Rationale: Since the 1980s, fungi have
emerged worldwide as an increasingly
frequent cause of nosocomial infections
in critically ill patients and are associated
with significant morbidity and mortality
(156 –159). Candida was the fourth most
frequent cause of bloodstream infections
in U.S. hospitals in the 1990s (18). About
one third of all episodes of candidemia
occur in medical, surgical, or pediatric
ICUs (22, 160). Fungi were isolated from
17% of the patients with ICU-acquired
infections in a 1-day point prevalence
study conducted in 1,417 ICUs in western
European countries (129). However, it is
unclear whether all these fungal isolates
were the etiological agent of infections or
simply colonizing microorganisms. Clin-
ical manifestations of candidiasis are usu-
ally not specific, and standard culture
techniques lack sensitivity. Detection of
Candida antigens (mannan and ␤-glu-
can), metabolites (arabinitol and eno-
lase), or antibodies (anti-mannan) and
amplification of fungal DNA by polymer-
ase chain reaction have been or are cur-
rently under investigation. The facts that
Candida infections are difficult to diag-
nose and associated with severe morbid-
ity and high mortality are arguments in
favor of the empirical or preemptive use
of antifungal agents in critically ill ICU
patients at high risk of candidiasis. The
likelihood of fungal sepsis is increased in
patients who have been treated with sev-
eral broad-spectrum antibiotics, who are
colonized with Candida at multiple sites,
who have damaged physiologic barriers
(i.e., recurrent gastrointestinal perfora-
tions or anastomotic leakages, acute ne-
crotizing pancreatitis, chemotherapy-
induced mucositis, vascular access
devices, and total parenteral nutrition),
or who are immunosuppressed (i.e., can-
cer patients with neutropenia, or hema-
topoietic stem cells or solid organ trans-
plant recipients) (155, 161). However,
recent epidemiologic studies and multi-
center trials have shown that fungi ac-
count for only 5% of all cases of severe
sepsis or septic shock, which does not
justify the use of antifungal therapy on a
routine basis, but only in selected subsets
of septic patients at high risk for invasive
candidiasis.
Treatment of Candidemia
Recommendation: Azoles (fluconazole)
and echinocandins (caspofungin) are as
efficacious as and less toxic than ampho-
tericin B deoxycholate for the treatment
of patients with candidemia. Albeit better
tolerated, there is no evidence that lipid
formulations of amphotericin B are supe-
rior to amphotericin B deoxycholate for
the treatment of candidemia.
Azoles: Grade A; Echinocandins: Grade B;
Lipid formulations of amphotericin B:
Grade E
Rationale: Candidemia is associated with
significant morbidity, prolonged hospital
stay, long-term sequelae, and high crude
mortality rates (40% to 60%) (162, 163).
Moreover, the presence of disseminated
infection is an independent prognostic
factor of fatal outcome in patients with
candidiasis (164), and antifungal therapy
has been shown to reduce mortality (165,
166). Therefore, recent management
guidelines have recommended that all pa-
tients with candidemia be treated with
antifungal agents (155, 167).
For decades, amphotericin B deoxy-
cholate, a fungicidal compound targeting
ergosterol in the cell membrane of a
broad spectrum of fungi, has been the
agent of choice for the empirical therapy
of invasive fungal infections. However,
nephrotoxicity and infusion-related ad-
verse events, especially fever and rigors,
are frequent adverse events of conven-
tional amphotericin B therapy that have
limited its use.
In the late 1980s, the advent of azoles
constituted a major progress in the man-
agement of invasive mycoses. This class
of drugs inhibits the synthesis of ergos-
terol. Fluconazole and itraconazole be-
came available in the early 1990s. Differ-
ent clinical studies compared the efficacy
of fluconazole and amphotericin B deoxy-
cholate for the treatment of candidemia.
Azoles. In a multicenter study of 206
nonneutropenic patients with candi-
demia, fluconazole (400 mg/day) was
shown to be as efficacious (success rates
were 72% and 79%, respectively) as and
better tolerated than amphotericin B de-
oxycholate (0.5– 0.6 mg/kg/day) (168).
Similar findings were made in a prospec-
S505
tive observational candidemia study in
which 227 patients were treated with am-
photericin B deoxycholate (median daily
dose, 0.5– 0.7 mg/kg) and 67 patients
with fluconazole (median daily dose, 200
mg) (165) and in several other smaller
comparative studies (169 –171). Noncom-
parative studies yielded similar success
rates for fluconazole (172–178). In sum-
mary, fluconazole (400 mg/day) was
found to be as effective as and better
tolerated than amphotericin B deoxy-
cholate (used at a dose of 0.3–1.2 mg/kg/
day) for the treatment of candidemia or
invasive candidiasis. Whether higher
doses of fluconazole would be associated
with better response rates is unknown. In
a study of 65 patients with Candida albi-
cans bloodstream infections, clinical re-
sponse rates were 60% and 83% for pa-
tients treated with 5 or 10 mg/kg/day of
fluconazole (179). The favorable efficacy-
toxicity profile of fluconazole led to a
rapidly increasing use of this azole for
prophylaxis and treatment of invasive
candidiasis. As anticipated, in the late
1990s an increasing incidence of infec-
tions due to non-C. albicans species with
reduced dose-dependent susceptibility
(Candida glabrata) or intrinsic resistance
(Candida krusei) to azoles was observed
in some U.S. centers (18, 180). The use of
high doses (800 –1200 mg) of fluconazole
has been advocated for the empirical
treatment of candidemia when infections
with azole-resistant Candida species are
suspected, but efficacy data from compar-
ative trials are lacking. However, for most
experts, amphotericin B remained clearly
the first choice for empirical therapy of
invasive candidiasis in the preechinocan-
din area (155, 167). This may have
changed since the advent of the echino-
candins.
Itraconazole, an azole with improved
in vitro activity against fluconazole-
resistant Candida species, has been avail-
able since the early 1990s. However, a
poor bioavailability of the oral formula-
tion of itraconazole has been a major
limitation for its use in critically ill pa-
tients. Recently, an intravenous form of
itraconazole has been developed, but
there are very few data on the efficacy of
IV itraconazole for the treatment of can-
didemia.
Voriconazole, a member of the newest
generation of antifungal triazoles, has
been shown to exhibit excellent in vitro
and in vivo activities against Candida spe-
cies (181). In patients with oropharyngeal
and/or esophageal candidiasis, voricon-
S506
azole was found to be at least as effective
as fluconazole (182). Moreover, voricon-
azole salvage therapy showed an overall
treatment success rate of 55% and 61%
in patients with refractory systemic or
esophageal candidiasis, respectively
(183). A large multicenter study compar-
ing voriconazole with sequential use of
amphotericin B deoxycholate followed by
fluconazole in nonneutropenic patients
with candidemia has recently been com-
pleted, and the first results of that trial
indicate that voriconazole is as effective
as and was better tolerated than ampho-
tericin B/fluconazole. Posaconazole and
ravuconazole are two other triazoles in
clinical development.
Echinocandins. Echinocandins are a
new class of antifungal agents that act by
inhibiting the synthesis of ␤-(1, 3)-D-
glucan, a component of the fungal cell
wall (184). Caspofungin, the first repre-
sentative of this new family of antifungal
agents, is active against Candida (C. al-
bicans and non-C. albicans) and Aspergil-
lus species. In immunosuppressed,
mostly human immunodeficiency virus-
positive patients with oropharyngeal and
esophageal candidiasis, caspofungin was
observed to be as effective as and better
tolerated than amphotericin B deoxy-
cholate (185, 186). In a large multicenter
trial that included 239 patients (of whom
24 were neutropenic) with invasive can-
didiasis (80% of the patients had candi-
demia), caspofungin (70-mg loading
dose, followed by 50 mg daily) was at least
as efficacious as and less toxic than am-
photericin B deoxycholate (0.6 –1.0 mg/
kg/day) (success rates, 73% vs. 62%; dis-
continuation for adverse events, 3% vs.
23%) (187). Of note, the success rates of
caspofungin against C. glabrata and C.
krusei were comparable with those ob-
tained in azole-susceptible Candida spe-
cies. Micafungin and anidulafungin are
two other echinocandins that are cur-
rently in clinical development.
In summary, azoles (fluconazole) and
echinocandins (caspofungin) have been
shown to be as efficacious as and better
tolerated than amphotericin B deoxy-
cholate for the treatment of patients with
candidemia and invasive candidiasis.
Limited information is available on the
efficacy of these agents in neutropenic
patients. Amphotericin B and caspofun-
gin are the antifungal agents of choice for
the treatment of infections caused by flu-
conazole-resistant non-C. albicans spe-
cies. The place of the newest triazole vori-
conazole with improved in vitro activity
against resistant strains awaits the results
of a recently completed large multicenter
study in patients with candidemia. Given
the poor prognosis of fungal sepsis, clini-
cians have shown great interest for the
use of combinations of antifungal agents
of different classes for the treatment of
critically ill patients with invasive myco-
ses. However, up to now, there have been
relatively few in vitro or in vivo studies of
combinations of antifungal agents. While
awaiting the results of prospective, ran-
domized clinical trials demonstrating
that combinations of antifungals are su-
perior to and reasonably not more toxic
than treatment with single agents, the
undiscriminating use of these costly
treatment regimens should be discour-
aged.
Lipid Formulations of Amphotericin
B. Nephrotoxicity and infusion-related
reactions are frequent adverse events of
treatment with amphotericin B deoxy-
cholate, and treatment interruptions be-
cause of toxicity may have a negative im-
pact on efficacy. In noncomparative
studies, continuous infusion of ampho-
tericin B deoxycholate over 24 hrs has
been shown to limit infusion-related re-
actions and nephrotoxicity and did not
seem to affect efficacy (188, 189). Yet, the
influence of this mode of administration
of amphotericin B on treatment efficacy
remains to be demonstrated, as it results
in a marked alteration of drug pharma-
cokinetics. Lipid formulations have been
developed to improve the toxicity and
possibly also the efficacy profile of am-
photericin B. Three lipid formulations
with different pharmacologic properties
are available: liposomal amphotericin B,
amphotericin B lipid complex, and am-
photericin B colloidal dispersion. They
have been primarily used for the treat-
ment of cancer patients with neutropenia
and persistent fever or with invasive as-
pergillosis. In large, multicenter clinical
trials, the lipid formulations were shown
to be as efficacious as and usually better
tolerated than amphotericin B deoxy-
cholate (190 –197). Yet, there were differ-
ences between the lipid preparations in
terms of adverse events. Infusion-related
reactions were significantly less frequent
with liposomal amphotericin B (5% to
20%) than with amphotericin B lipid
complex (40% to 80%) or colloidal dis-
persion of amphotericin B (50% to 80%).
Therefore, liposomal amphotericin B has
been used as the control treatment regi-
men in most recent studies of empirical
antifungal therapy for persistent fever in
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
neutropenic cancer patients (193, 198).
Lipid formulations are considerably more
expensive than conventional amphoteri-
cin B, which has had a negative impact
on their use.
There are few data on the direct com-
parison of the efficacy and toxicity of am-
photericin B deoxycholate and the lipid
formulations for the treatment of pa-
tients with invasive candidiasis. Small
noncomparative studies suggest that lipid
formulations of amphotericin B are as
efficacious as but better tolerated than
conventional amphotericin B (199 –202).
Their high costs, the paucity of clinical
data, and the existence of alternative an-
tifungal therapies (azoles and echinocan-
dins) explain why the use of lipid formu-
lations has been limited in patients with
invasive candidiasis (155).
Summary Recommendations
1. Antibiotic therapy should be started
within the first hour of recognition of
severe sepsis, after appropriate cul-
tures have been obtained.
Recommendation: Grade E
2. Initial empirical antiinfective therapy
should include one or more drugs that
have activity against the likely patho-
gens (bacterial or fungal) and that
penetrate into the presumed source of
sepsis. The choice of drugs should be
guided by the susceptibility patterns of
microorganisms in the community
and in the hospital.
Recommendation: Grade D
3. Monotherapy is as efficacious as com-
bination therapy with a ␤-lactam and
an aminoglycoside as empirical ther-
apy of patients with severe sepsis or
septic shock.
Recommendation: Carbapenem: Grade B.
Third- or fourth-generation cephalospo-
rins: Grade B. Extended-spectrum car-
boxypenicillins or ureidopenicillins com-
bined with ␤-lactamase inhibitors: Grade
E.
4. Third-generation and fourth-genera-
tion cephalosporins, carbapenems,
and extended-spectrum carboxypeni-
cillins or ureidopenicillins combined
with ␤-lactamase inhibitors are
equally effective as empirical antibi-
otic therapy in patients with severe
sepsis.
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
Recommendation: Third- or fourth-
generation cephalosporins: Grade A. Car-
bapenem: Grade B. Extended-spectrum
carboxypenicillins or ureidopenicillins
combined with ␤-lactamase inhibitors:
Grade C
5. Empirical use of glycopeptide antibi-
otics (vancomycin, teicoplanin), ox-
azolidinones (linezolid), or strepto-
gramins (quinupristin/dalfopristin) in
patients with severe sepsis or septic
shock is justified in patients with hy-
persensitivity to ␤-lactams or in insti-
tutions with resistant Gram-positive
bacteria (i.e., methicillin-resistant
staphylococci, penicillin-resistant
pneumococci, or ampicillin-resistant
enterococci) in the community or in
the hospital.
Recommendation: Grade E
6. Modification of empirical antimicro-
bial therapy with the aim to restrict
the number of antibiotics and narrow
the spectrum of antimicrobial therapy
is an important and responsible strat-
egy for minimizing the development
of resistant pathogens and for contain-
ing costs.
Recommendation: Grade E
7. The antimicrobial regimen should al-
ways be reassessed after 48 –72 hrs on
the basis of microbiological and clini-
cal data with the aim to use a narrow-
spectrum antibiotic to prevent the de-
velopment of resistance, to reduce
toxicity, and to reduce costs. Once a
causative pathogen is identified, there
is no evidence that combination ther-
apy is more effective monotherapy.
The duration of therapy should typi-
cally be 7–10 days and guided by clin-
ical response.
Recommendation: Grade E
8. If the presenting clinical syndrome is
determined to be due to a noninfec-
tious cause, antimicrobial therapy
should be stopped promptly to mini-
mize the development of resistant
pathogens and superinfection with
other pathogenic organisms.
Recommendation: Grade E
9. At the present time, there is insuffi-
cient evidence to recommend the use
of antibiotic cycling as a strategy to
reduce the development of antibiotic
resistance.
Recommendation: Grade C
10. Empirical antifungal therapy should
not be used on a routine basis in
patients with severe sepsis or septic
shock, but it may be justified in se-
lected subsets of septic patients at
high risk for invasive candidiasis
(155).
Recommendation: Grade E
11. Azoles (fluconazole) and echinocan-
dins (caspofungin) are as efficacious
as and less toxic than amphotericin B
deoxycholate for the treatment of pa-
tients with candidemia. Albeit better
tolerated, there is no evidence that
lipid formulations of amphotericin B
are superior to amphotericin B de-
oxycholate for the treatment of can-
didemia.
Recommendation: Azoles: Grade A. Echi-
nocandins: Grade B. Lipid formulations
of amphotericin B: Grade E.
REFERENCES
1. Bochud PY, Glauser M, Calandra T: Antibi-
otics in sepsis. Intensive Care Med 2001;
27(14 Suppl):S33–S48
2. Zeni F, Freeman B, Natanson C: Anti-
inflammatory therapies to treat sepsis and
septic shock: A reassessment. Crit Care Med
1997; 25:1095–1100
3. Brun-Buisson C, Doyon F, Carlet J, et al:
Incidence, risk factors, and outcome of se-
vere sepsis and septic shock in adults: A
multicenter prospective study in intensive
care units. JAMA 1995; 274:968 –974
4. Brun-Buisson C, Doyon F, Carlet J: Bacte-
remia and severe sepsis in adults: A multi-
center prospective survey in ICUs and wards
of 24 hospitals. Am J Respir Crit Care Med
1996; 154:617– 624
5. Cohen J, Carlet J, for the INTERSEPT Study
Group: An international, multicenter, pla-
cebo-controlled trial of monoclonal anti-
body to human tumor necrosis factor-␣ in
patients with sepsis. Crit Care Med 1996;
24:1431–1440
6. Sands KE, Bates DW, Lanken PN, et al:
Epidemiology of sepsis syndrome in 8 aca-
demic medical centers. JAMA 1997; 278:
234 –240
7. Bone RC, Sibbald WJ, Sprung CL: The
ACCP-SCCM consensus conference on sep-
sis and organ failure. Chest 1992; 101:
1481–1483
8. Sackett DL, Haynes RB, Guyatt GH, et al:
Clinical Epidemiology. Philadelphia, Lip-
pincott-Raven, 1991
9. Marchetti O, Calandra T: Infections in neu-
tropenic cancer patients. Lancet 2002; 359:
723–725
10. Martin GS, Mannino DM, Eaton S, et al: The
S507
 epidemiology of sepsis in the United States
from 1979 through 2000. N Engl J Med
2003; 348:1546 –1554
11. Abraham E, Wunderink R, Silverman H, et
al: Efficacy and safety of monoclonal anti-
body to human tumor necrosis factor ␣ in
patients with sepsis syndrome: A random-
ized, controlled, double-blind, multicenter
clinical trial. JAMA 1995; 273:934 –941
12. Abraham E, Glauser MP, Butler T, et al: p55
tumor necrosis factor receptor fusion pro-
tein in the treatment of patients with severe
sepsis and septic shock: A randomized con-
trolled multicenter trial. JAMA 1997; 277:
1531–1538
13. Abraham E, Anzueto A, Gutierrez G, et al:
Double-blind randomised controlled trial of
monoclonal antibody to human tumour ne-
crosis factor in treatment of septic shock.
NORASEPT II Study Group. Lancet 1998;
351(9107):929 –933
14. Cohen J, Abraham E: Microbiologic findings
and correlations with serum tumor necrosis
factor-alpha in patients with severe sepsis
and septic shock. J Infect Dis 1999; 180:
116 –121
15. Fisher CJ, Agosti JM, Opal SM, et al: Treat-
ment of septic shock with the tumor necro-
sis factor receptor:Fc fusion protein. N Engl
J Med 1996; 334:1697–1702
16. Fisher CJ Jr, Opal SM, Dhainaut JF, et al:
Influence of an anti-tumor necrosis factor
monoclonal antibody on cytokine levels in
patients with sepsis. Crit Care Med 1993;
21:318 –327
17. Opal SM, Fisher CJ, Dhainaut JF, et al:
Confirmatory interleukin-1 receptor antag-
onist trial in severe sepsis: A phase III, ran-
domized, double-blind, placebo-controlled,
multicenter trial. Crit Care Med 1997; 25:
1115–1124
18. Edmond MB, Wallace SE, McClish DK, et al:
Nosocomial bloodstream infections in
United States hospitals: A three-year analy-
sis. Clin Infect Dis 1999; 29:239 –244
19. Pfaller MA, Jones RN, Messer SA, et al: Na-
tional surveillance of nosocomial blood
stream infection due to species of Candida
other than Candida albicans: Frequency of
occurrence and antifungal susceptibility in
the SCOPE Program. Surveillance and Con-
trol of Pathogens of Epidemiologic. Diagn
Microbiol Infect Dis 1998; 30:121–129
20. Voss A, Kluytmans JA, Koeleman JG, et al:
Occurrence of yeast bloodstream infections
between 1987 and 1995 in five Dutch uni-
versity hospitals. Eur J Clin Microbiol Infect
Dis 1996; 15:909 –912
21. Sandven P, Bevanger L, Digranes A, et al:
Constant low rate of fungemia in Norway,
1991 to 1996. J Clin Microbiol 1998; 36:
3455–3459
22. Marchetti O, Bille J, Fluckiger U, et al: Ep-
idemiology of candidemia in Swiss tertiary
care hospitals: Secular trends, 1991–2000.
Clin Infect Dis 2004; 38:311–320
23. Mc Cabe WR, Jackson GG: Gram negative
S508
bacteremia. Arch Intern Med 1962; 110:
92–100
24. Freid MA, Vosti KL: The importance of un-
derlying disease in patients with gram-
negative bacteremia. Arch Intern Med 1968;
121:418 – 423
25. Bryant RE, Hood AF, Hood CE, et al: Fac-
tors affecting mortality of gram-negative
rod bacteremia. Arch Intern Med 1971; 127:
120 –128
26. Young LS, Martin WJ, Meyer RD, et al:
Gram-negative rod bacteremia: Microbio-
logic, immunologic, and therapeutic con-
siderations. Ann Intern Med 1977; 86:
456 – 471
27. Kreger BE, Craven DE, McCabe WR: Gram-
negative bacteremia. IV. Re-evaluation of
clinical features and treatment in 612 pa-
tients. Am J Med 1980; 68:344 –355
28. Leibovici L, Paul M, Poznanski O, et al:
Monotherapy versus beta-lactam-aminogly-
coside combination treatment for gram-
negative bacteremia: A prospective, obser-
vational study. Antimicrob Agents
Chemother 1997; 41:1127–1133
29. Chow JW, Fine MJ, Shlaes DM, et al: Enter-
obacter bacteremia: Clinical features and
emergence of antibiotic resistance during
therapy. Ann Intern Med 1991; 115:585–590
30. Vidal F, Mensa J, Almela M, et al: Epidemi-
ology and outcome of Pseudomonas aerugi-
nosa bacteremia, with special emphasis on
the influence of antibiotic treatment: Anal-
ysis of 189 episodes. Arch Intern Med 1996;
156:2121–2126
31. Schiappa DA, Hayden MK, Matushek MG, et
al: Ceftazidime-resistant Klebsiella pneu-
moniae and Escherichia coli bloodstream
infection: A case-control and molecular ep-
idemiologic investigation. J Infect Dis 1996;
174:529 –536
32. Caballero-Granado FJ, Cisneros JM, Luque
R, et al: Comparative study of bacteremias
caused by Enterococcus spp. with and with-
out high-level resistance to gentamicin.
J Clin Microbiol 1998; 36:520 –525
33. Ispahani P, Pearson NJ, Greenwood D: An
analysis of community and hospital-ac-
quired bacteraemia in a large teaching hos-
pital in the United Kingdom. Q J Med 1987;
63:427– 440
34. Leibovici L, Drucker M, Konigsberger H, et
al: Septic shock in bacteremic patients: Risk
factors, features and prognosis. Scand J In-
fect Dis 1997; 29:71–75
35. Weinstein MP, Towns ML, Quartey SM, et
al: The clinical significance of positive blood
cultures in the 1990s: A prospective com-
prehensive evaluation of the microbiology,
epidemiology, and outcome of bacteremia
and fungemia in adults. Clin Infect Dis
1997; 24:584 – 602
36. Leibovici L, Shraga I, Drucker M, et al: The
benefit of appropriate empirical antibiotic
treatment in patients with bloodstream in-
fection. J Intern Med 1998; 244:379 –386
37. Kollef MH, Sherman G, Ward S, et al: Inad-
equate antimicrobial treatment of infec-
tions: A risk factor for hospital mortality
among critically ill patients. Chest 1999;
115:462– 474
38. Ibrahim EH, Sherman G, Ward S, et al: The
influence of inadequate antimicrobial treat-
ment of bloodstream infections on patient
outcomes in the ICU setting. Chest 2000;
118:146 –155
39. Harbarth S, Ferriere K, Hugonnet S, et al:
Epidemiology and prognostic determinants
of bloodstream infections in surgical inten-
sive care. Arch Surg 2002; 137:1353–1359
40. Hanon FX, Monnet DL, Sorensen TL, et al:
Survival of patients with bacteraemia in re-
lation to initial empirical antimicrobial
treatment. Scand J Infect Dis 2002; 34:
520 –528
41. Harbarth S, Garbino J, Pugin J, et al: Inap-
propriate initial antimicrobial therapy and
its effect on survival in a clinical trial of
immunomodulating therapy for severe sep-
sis. Am J Med 2003; 115:529 –535
42. Zaragoza R, Artero A, Camarena JJ, et al:
The influence of inadequate empirical anti-
microbial treatment on patients with blood-
stream infections in an intensive care unit.
Clin Microbiol Infect 2003; 9:412– 418
43. Leone M, Bourgoin A, Cambon S, et al:
Empirical antimicrobial therapy of septic
shock patients: Adequacy and impact on the
outcome. Crit Care Med 2003; 31:462– 467
44. Garnacho-Montero J, Garcia-Garmendia JL,
Barrero-Almodovar A, et al: Impact of ade-
quate empirical antibiotic therapy on the
outcome of patients admitted to the inten-
sive care unit with sepsis. Crit Care Med
2003; 31:2742–2751
45. Mandell LA, Campbell GD Jr: Nosocomial
pneumonia guidelines: An international
perspective. Chest 1998; 113(3 Suppl):
188S–193S.
46. Mandell LA, Marrie TJ, Grossman RF, et al:
Canadian guidelines for the initial manage-
ment of community-acquired pneumonia:
An evidence-based update by the Canadian
Infectious Diseases Society and the Cana-
dian Thoracic Society. Clin Infect Dis 2000;
31:383– 421
47. BTS guidelines for the management of
community acquired pneumonia in adults.
Thorax 2001; 56(Suppl 4):IV1–IV64
48. Niederman MS, Mandell LA, Anzueto A, et
al: Guidelines for the management of adults
with community-acquired pneumonia: Di-
agnosis, assessment of severity, antimicro-
bial therapy, and prevention. Am J Respir
Crit Care Med 2001; 163:1730 –1754
49. Memish ZA, Shibl AM, Ahmed QA: Guide-
lines for the management of community-
acquired pneumonia in Saudi Arabia: A
model for the Middle East region. Int J
Antimicrob Agents 2002; 20(Suppl 1):
S1–S12
50. Mandell LA, Bartlett JG, Dowell SF, et al:
Update of practice guidelines for the man-
agement of community-acquired pneumo-
nia in immunocompetent adults. Clin In-
fect Dis 2003; 37:1405–1433
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
 51. Mazuski JE, Sawyer RG, Nathens AB, et al:
The surgical infection society guidelines on
antimicrobial therapy for intra-abdominal
infections: An executive summary. Surg In-
fect 2002; 3:161–173
52. Mazuski JE: The surgical infection society
guidelines on antimicrobial therapy for in-
tra-abdominal infections: Evidence for the
recommendations. Surg Infect 2002;
3:175–233
53. Solomkin JS, Mazuski JE, Baron EJ, et al:
Guidelines for the selection of anti-infective
agents for complicated intra-abdominal in-
fections. Clin Infect Dis 2003; 37:997–1005
54. Klastersky J, Meunier-Carpentier F, Prevost
JM: Significance of antimicrobial synergism
for the outcome of gram negative sepsis.
Am J Med Sci 1977; 273:157–167
55. Hanson B, Coppens L, Klastersky J: Com-
parative studies of ticarcillin and mezlocil-
lin plus sisomicin in Gram-negative bacil-
lary bacteraemia and bronchopneumonia.
J Antimicrob Chemother 1982; 10:335–341
56. Krumpe PE, Cohn S, Garreltes J, et al: In-
travenous and oral mono- or combination-
therapy in the treatment of severe infec-
tions: Ciprofloxacin versus standard
antibiotic therapy. J Antimicrob Chemother
1999; 43(Suppl A):117–128
57. Beaucaire G: Evaluation of the efficacy and
safety of isepamicin compared with amika-
cin in the treatment of nosocomial pneu-
monia and septicaemia. J Chemother 1995;
7(Suppl 2):165–173
58. Colardyn F: The efficacy and safety of ise-
pamicin and ceftazidime compared with
amikacin and ceftazidime in acute lower
respiratory tract infection. J Chemother
1995; 7(Suppl 2):129 –135
59. Wolff M: Comparison of strategies using
cefpirome and ceftazidime for empiric
treatment of pneumonia in intensive care
patients. Antimicrob Agents Chemother
1998; 42:28 –36
60. Brun-Buisson C, Sollet JP, Schweich H, et
al: Treatment of ventilator-associated pneu-
monia with piperacillin-tazobactam/amika-
cin versus ceftazidime/amikacin: A multi-
center, randomized controlled trial. Clin
Infect Dis 1998; 26:346 –354
61. Alvarez-Lerma F, Insausti-Ordenana J,
Jorda-Marcos R, et al: Efficacy and tolera-
bility of piperacillin/tazobactam versus
ceftazidime in association with amikacin for
treating nosocomial pneumonia in inten-
sive care patients: A prospective random-
ized multicenter trial. Intensive Care Med
2001; 27:493–502
62. Beaucaire G, Nicolas MH, Martin C, et al:
Phare study: Comparative study of com-
bined cefepime-amikacin versus ceftazi-
dime combined with amikacin in the treat-
ment of nosocomial pneumonias in
ventilated patients. Ann Fr Anesth Reanim
1999; 18:186 –195
63. Solomkin JS, Fant WK, Rivera JO, et al:
Randomized trial of imipenem/cilastatin
versus gentamicin and clindamycin in
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
mixed flora infections. Am J Med 1985; 78:
85–91
64. Mouton Y, Deboscker Y, Bazin C, et al:
Prospective, randomized, controlled study
of imipenem-cilastatin versus cefotaxime-
amikacin in the treatment of lower respira-
tory tract infection and septicemia at inten-
sive care units. Presse Medicale 1990; 19:
607– 612
65. Cometta A, Baumgartner JD, Lew D, et al:
Prospective randomized comparison of imi-
penem monotherapy with imipenem plus
netilmicin for treatment of severe infec-
tions in nonneutropenic patients. Antimi-
crob Agents Chemother 1994; 38:
1309 –1313
66. Jaspers CA, Kieft H, Speelberg B, et al:
Meropenem versus cefuroxime plus genta-
micin for treatment of serious infections in
elderly patients. Antimicrob Agents Che-
mother 1998; 42:1233–1238
67. Arich C, Gouby A, Bengler C, et al: Com-
parison of the efficacy of cefotaxime alone
and the combination cefazolin-tobramycin
in the treatment of enterobacterial septice-
mia. Pathol Biol 1987; 35:613– 615
68. Oblinger MJ, Bowers JT, Sande MA, et al:
Moxalactam therapy vs. standard antimicro-
bial therapy for selected serious infections.
Rev Infect Dis 1982; 4(Suppl):S639 –S649
69. Rubinstein E, Lode H, Grassi C: Ceftazidime
monotherapy vs. ceftriaxone/tobramycin for
serious hospital-acquired gram-negative in-
fections. Clin Infect Dis 1995; 20:
1217–1228
70. Extermann M, Regamey C, Humair L, et al:
Initial treatment of sepsis in non-neutro-
penic patients: Ceftazidime alone versus
‘best guess’ combined antibiotic therapy.
Chemotherapy 1995; 41:306 –315
71. McCormick PA, Greenslade L, Kibbler CC,
et al: A prospective randomized trial of
ceftazidime versus netilmicin plus me-
zlocillin in the empirical therapy of pre-
sumed sepsis in cirrhotic patients. Hepatol-
ogy 1997; 25:833– 836
72. Hammerberg O, Kurnitzki C, Watts J, et al:
Randomized trial using piperacillin versus
ampicillin and amikacin for treatment of
premature neonates with risk factors for
sepsis. Eur J Clin Microbiol Infect Dis 1989;
8:241–244
73. Korvick JA, Bryan CS, Farber B, et al: Pro-
spective observational study of Klebsiella
bacteremia in 230 patients: Outcome for
antibiotic combinations versus mono-
therapy. Antimicrob Agents Chemother
1992; 36:2639 –2644
74. Sieger B, Berman SJ, Geckler RW, et al:
Empiric treatment of hospital-acquired
lower respiratory tract infections with
meropenem or ceftazidime with tobramy-
cin: A randomized study. Crit Care Med
1997; 25:1663–1670
75. Alvarez-Lerma F: Efficacy of meropenem as
monotherapy in the treatment of ventilator-
associated pneumonia. J Chemother 2001;
13:70 – 81
76. Mangi RJ, Greco T, Ryan J, et al: Cefopera-
zone versus combination antibiotic therapy
of hospital-acquired pneumonia. Am J Med
1988; 84:68 –74
77. Fernandez-Guerrero M, Gudiol F, Rodri-
guez-Torres A, et al: Nosocomial pneumo-
nia: Comparative multicentre trial between
monotherapy with cefotaxime and treat-
ment with antibiotic combinations. Infec-
tion 1991; 19(Suppl 6):S320 –S325
78. Croce MA, Fabian TC, Stewart RM, et al:
Empiric monotherapy versus combination
therapy of nosocomial pneumonia in
trauma patients. J Trauma 1993; 35:
303–309
79. Speich R, Imhof E, Vogt M, et al: Efficacy,
safety, and tolerance of piperacillin/tazobac-
tam compared to co-amoxiclav plus an ami-
noglycoside in the treatment of severe
pneumonia. Eur J Clin Microbiol Infect Dis
1998; 17:313–317
80. Manhold C, von Rolbicki U, Brase R, et al:
Outbreaks of Staphylococcus aureus infec-
tions during treatment of late onset pneu-
monia with ciprofloxacin in a prospective,
randomized study. Intensive Care Med
1998; 24:1327–1330
81. Poenaru D, De Santis M, Christou NV: Imi-
penem versus tobramycin–antianaerobe an-
tibiotic therapy in intra-abdominal infec-
tions. Can J Surg 1990; 33:415– 422
82. Solomkin JS, Dellinger EP, Christou NV, et
al: Results of a multicenter trial comparing
imipenem/cilastatin to tobramycin/clinda-
mycin for intra-abdominal infections. Ann
Surg 1990; 212:581–591
83. Schentag JJ, Wels PB, Reitberg DP, et al: A
randomized clinical trial of moxalactam
alone versus tobramycin plus clindamycin
in abdominal sepsis. Ann Surg 1983; 198:
35– 41
84. Greenberg RN, Cayavec P, Danko LS, et al:
Comparison of cefoperazone plus sulbactam
with clindamycin plus gentamicin as treat-
ment for intra-abdominal infections. J An-
timicrob Chemother 1994; 34:391– 401
85. Fink MP: Antibiotic therapy of intra-
abdominal sepsis in the elderly: Experience
with ticarcillin and clavulanic acid. Surg
Gynecol Obstet 1991; 172(Suppl):36 – 41
86. Dupont H, Carbon C, Carlet J: Monotherapy
with a broad-spectrum beta-lactam is as ef-
fective as its combination with an amino-
glycoside in treatment of severe generalized
peritonitis: a multicenter randomized con-
trolled trial. Antimicrob Agents Chemother
2000; 44:2028 –2033
87. Stone HH, Kolb LD, Geheber CE, et al:
Treatment of surgical infections with to-
bramcin. Am Surg 1975; 41:301–308
88. Smith CR, Baughman KL, Edwards CQ, et
al: Controlled comparison of amikacin and
gentamicin. N Engl J Med 1977; 296:
349 –353
89. Bock BV, Edelstein PH, Meyer RD: Prospec-
tive comparative study of efficacy and tox-
icity of netilmicin and amikacin. Antimi-
crob Agents Chemother 1980; 17:217–225
S509
 90. Gudiol F, Pallares R, Ariza X, et al: Com-
parative clinical evaluation of aztreonam
versus aminoglycosides in gram-negative
septicaemia. J Antimicrob Chemother 1986;
17:661– 671
91. Mangi RJ, Ryan J, Berenson C, et al: Ce-
foperazone versus ceftazidime monotherapy
of nosocomial pneumonia. Am J Med 1988;
85:44 – 48
92. Norrby SR, Finch RG, Glauser M: Mono-
therapy in serious hospital-acquired infec-
tions: A clinical trial of ceftazidime versus
imipenem/cilastatin. J Antimicrob Che-
mother 1993; 31:927–937
93. Norrby SR, Geddes AM: Efficacy of cefpi-
rome in the treatment of septicaemia.
Scand J Infect Dis 1993; 91:51–59
94. Schrank JH Jr, Kelly JW, McAllister CK:
Randomized comparison of cefepime and
ceftazidime for treatment of hospitalized
patients with gram-negative bacteremia.
Clin Infect Dis 1995; 20:56 –58
95. Colardyn F, Faulkner KL: Intravenous
meropenem versus imipenem/cilastatin in
the treatment of serious bacterial infections
in hospitalized patients. J Antimicrob Che-
mother 1996; 38:523–537
96. Mehtar S, Dewar EP, Leaper DJ, et al: A
multi-centre study to compare meropenem
and cefotaxime and metronidazole in the
treatment of hospitalized patients with se-
rious infections. J Antimicrob Chemother
1997; 39:631– 638
97. Garau J, Blanquer J, Cobo L, et al: Prospec-
tive, randomised, multicentre study of
meropenem versus imipenem/cilastatin as
empiric monotherapy in severe nosocomial
infections. Eur J Clin Microbiol Infect Dis
1997; 16:789 –796
98. Norrby SR, Geddes AM, Shah PM: Random-
ized comparative trial of cefpirome versus
ceftazidime in the empirical treatment of
suspected bacteraemia or sepsis. J Antimi-
crob Chemother 1998; 42:503–509
99. Verwaest C: Meropenem versus imipenem/
cilastatin as empirical monotherapy for se-
rious bacterial infections in the intensive
care unit. Clin Microbiol Infect 2000;
6:294 –302
100. Schentag JJ, Vari AJ, Winslade NE, et al:
Treatment with aztreonam or tobramycin
in critical care patients with nosocomial
gram-negative pneumonia. Am J Med 1985;
78:34 – 41
101. Reeves JH, Russell GM, Cade JF, et al: Com-
parison of ceftriaxone with cefotaxime in
serious chest infections. Chest 1989; 96:
1292–1297
102. Mangi RJ, Peccerillo KM, Ryan J, et al: Ce-
foperazone versus ceftriaxone monotherapy
of nosocomial pneumonia. Diagn Microbiol
Infect Dis 1992; 15:441– 447
103. Thomas P, Daly S, Misan G, et al: Compar-
ison of the efficacy and adverse effect profile
of cefotaxime and ceftriaxone in ICU pa-
tients with susceptible infections. Diagn
Microbiol Infect Dis 1992; 15:89 –97
104. Barckow D, Schwigon CD: Cefepime versus
S510
cefotaxime in the treatment of lower respi-
ratory tract infections. J Antimicrob Che-
mother 1993; 32(Suppl B):187–193
105. Fink MP, Snydman DR, Niederman MS, et
al: Treatment of severe pneumonia in hos-
pitalized patients: Results of a multicenter,
randomized, double-blind trial comparing
intravenous ciprofloxacin with imipenem-
cilastatin. Antimicrob Agents Chemother
1994; 38:547–557
106. Siami GA, Wilkins WT, Bess DT, et al: Com-
parison of ciprofloxacin with imipenem in
the treatment of severe pneumonia in hos-
pitalised geriatric patients. Drugs 1995;
49(Suppl 2):436 – 438
107. Jaccard C, Troillet N, Harbarth S, et al:
Prospective randomized comparison of imi-
penem-cilastatin and piperacillin-tazobac-
tam in nosocomial pneumonia or peritoni-
tis. Antimicrob Agents Chemother 1998; 42:
2966 –2972
108. Grossman RF, Campbell DA, Landis SJ, et
al: Treatment of community-acquired
pneumonia in the elderly: The role of
cefepime, a fourth-generation cephalospo-
rin. J Antimicrob Chemother 1999; 43:
549 –554
109. Torres A, Bauer TT, Leon-Gil C, et al: Treat-
ment of severe nosocomial pneumonia: A
prospective randomised comparison of in-
travenous ciprofloxacin with imipenem/
cilastatin. Thorax 2000; 55:1033–1039
110. West M, Boulanger BR, Fogarty C, et al:
Levofloxacin compared with imipenem/
cilastatin followed by ciprofloxacin in adult
patients with nosocomial pneumonia: A
multicenter, prospective, randomized,
open-label study. Clin Ther 2003; 25:
485–506
111. Kempf P, Bauernfeind A, Muller A, et al:
Meropenem monotherapy versus cefo-
taxime plus metronidazole combination
treatment for serious intra-abdominal in-
fections. Infection 1996; 24:473– 479
112. Macfarlane JT, Finch RG, Ward MJ, et al:
Erythromycin compared with a combina-
tion of ampicillin and amoxycillin as initial
therapy for adults with pneumonia includ-
ing Legionnaires’ disease. J Infect 1983;
7:111–117
113. Shann F, Barker J, Poore P: Chloramphen-
icol alone versus chloramphenicol plus pen-
icillin for severe pneumonia in children.
Lancet 1985; 2:684 – 686
114. Gaillat J, Bru JP, Sedallian A: Penicillin
G/ofloxacin versus erythromycin/amoxicil-
lin-clavulanate in the treatment of severe
community-acquired pneumonia. Eur
J Clin Microbiol Infect Dis 1994; 13:
639 – 644
115. Roson B, Carratala J, Tubau F, et al: Use-
fulness of betalactam therapy for commu-
nity-acquired pneumonia in the era of drug-
resistant Streptococcus pneumoniae: A
randomized study of amoxicillin-clavu-
lanate and ceftriaxone. Microb Drug Resist
2001; 7:85–96
116. Linden PK, Pasculle AW, McDevitt D, et al:
Effect of quinupristin/dalfopristin on the
outcome of vancomycin-resistant Entero-
coccus faecium bacteraemia: Comparison
with a control cohort. J Antimicrob Che-
mother 1997; 39(Suppl A):145–151
117. Polk HC Jr, Livingston DH, Fry DE, et al:
Treatment of pneumonia in mechanically
ventilated trauma patients: Results of a pro-
spective trial. Arch Surg 1997; 132:
1086 –1092
118. Fagon J, Patrick H, Haas DW, et al: Treat-
ment of gram-positive nosocomial pneumo-
nia: Prospective randomized comparison of
quinupristin/dalfopristin versus vancomy-
cin. Am J Respir Crit Care Med 2000; 161:
753–762
119. Rubinstein E, Cammarata S, Oliphant T, et
al: Linezolid (PNU-100766) versus vanco-
mycin in the treatment of hospitalized pa-
tients with nosocomial pneumonia: A ran-
domized, double-blind, multicenter study.
Clin Infect Dis 2001; 32:402– 412
120. Calandra T, Glauser MP: Immunocompro-
mised animal models for the study of anti-
biotic combinations. Am J Med 1986; 80:
45–52
121. Klastersky J, Zinner SH: Synergistic com-
binations of antibiotics in gram-negative
bacillary infections. Rev Infect Dis 1982;
4:294 –301
122. De Jongh CA, Joshi JH, Thompson BW, et
al: A double beta-lactam combination ver-
sus an aminoglycoside-containing regimen
as empiric antibiotic therapy for febrile
granulocytopenic cancer patients. Am J
Med 1986; 80:101–111
123. Gribble MJ, Chow AW, Naiman SC, et al:
Prospective randomized trial of piperacillin
monotherapy versus carboxypenicillin-
aminoglycoside combination regimens in
the empirical treatment of serious bacterial
infections. Antimicrob Agents Chemother
1983; 24:388 –393
124. The EORTC International Antimicrobial
Therapy Cooperative Group: Ceftazidime
combined with a short or long course of
amikacin for empirical therapy of gram-
negative bacteremia in cancer patients with
granulocytopenia. N Engl J Med 1987; 317:
1692–1698
125. Anderson ET, Young LS, Hewitt WL: Anti-
microbial synergism in the therapy of
gram-negative rod bacteremia. Chemother-
apy 1978; 24:45–54
126. Nicolau DP, McNabb J, Lacy MK, et al: Con-
tinuous versus intermittent administration
of ceftazidime in intensive care unit pa-
tients with nosocomial pneumonia. Int J
Antimicrob Agents 2001; 17:497–504
127. Bartlett JG, Dowell SF, Mandell LA, et al:
Practice guidelines for the management of
community-acquired pneumonia in adults.
Clin Infect Dis 2000; 31:347–382
128. Richards MJ, Edwards JR, Culver DH, et al:
Nosocomial infections in medical intensive
care units in the United States. Crit Care
Med 1999; 27:887– 892
129. Vincent JL, Bihari DJ, Suter PM, et al: The
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
 prevalence of nosocomial infection in inten-
sive care units in Europe: Results of the
European Prevalence of Infection in Inten-
sive Care (EPIC) Study. JAMA 1995; 274:
639 – 644
130. Yu VL, Chiou CC, Feldman C, et al: An
international prospective study of pneumo-
coccal bacteremia: Correlation with in vitro
resistance, antibiotics administered, and
clinical outcome. Clin Infect Dis 2003; 37:
230 –237
131. John CC: Treatment failure with use of a
third-generation cephalosporin for penicil-
lin-resistant pneumococcal meningitis:
Case report and review. Clin Infect Dis
1994; 18:188 –193
132. Cometta A, Kern WV, De Bock R, et al:
Vancomycin versus placebo for treating per-
sistent fever in patients with neutropenic
cancer receiving piperacillin-tazobactam
monotherapy. Clin Infect Dis 2003; 37:
382–389
133. Smith TL, Pearson ML, Wilcox KR, et al:
Emergence of vancomycin resistance in
Staphylococcus aureus. N Engl J Med 1999;
340:493–501
134. Weigel LM, Clewell DB, Gill SR, et al: Ge-
netic analysis of a high-level vancomycin-
resistant isolate of Staphylococcus aureus.
Science 2003; 302:1569 –1571
135. Moellering RC: Linezolid: The first oxazo-
lidinone antimicrobial. Ann Intern Med
2003; 138:135–142
136. San Pedro GS, Cammarata SK, Oliphant
TH, et al: Linezolid versus ceftriaxone/
cefpodoxime in patients hospitalized for the
treatment of Streptococcus pneumoniae
pneumonia. Scand J Infect Dis 2002; 34:
720 –728
137. Wunderink RG, Cammarata SK, Oliphant
TH, et al: Continuation of a randomized,
double-blind, multicenter study of linezolid
versus vancomycin in the treatment of pa-
tients with nosocomial pneumonia. Clin
Ther 2003; 25:980 –992
138. Wunderink RG, Rello J, Cammarata SK, et
al: Linezolid vs vancomycin: Analysis of two
double-blind studies of patients with methi-
cillin-resistant Staphylococcus aureus nos-
ocomial pneumonia. Chest 2003; 124:
1789 –1797
139. Stevens DL, Herr D, Lampiris H, et al: Lin-
ezolid versus vancomycin for the treatment
of methicillin-resistant Staphylococcus au-
reus infections. Clin Infect Dis 2002; 34:
1481–1490
140. Gonzales RD, Schreckenberger PC, Graham
MB, et al: Infections due to vancomycin-
resistant Enterococcus faecium resistant to
linezolid. Lancet 2001; 357:1179
141. Tsiodras S, Gold HS, Sakoulas G, et al:
Linezolid resistance in a clinical isolate of
Staphylococcus aureus. Lancet 2001; 358:
207–208
142. Herrero IA, Issa NC, Patel R: Nosocomial
spread of linezolid-resistant, vancomycin-
resistant Enterococcus faecium. N Engl
J Med 2002; 346:867– 869
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)
143. Nichols RL, Graham DR, Barriere SL, et al:
Treatment of hospitalized patients with
complicated gram-positive skin and skin
structure infections: Two randomized, mul-
ticentre studies of quinupristin/dalfopristin
versus cefazolin, oxacillin or vancomycin.
J Antimicrob Chemother 1999; 44:263–273
144. Ibrahim EH, Ward S, Sherman G, et al:
Experience with a clinical guideline for the
treatment of ventilator-associated pneumo-
nia. Crit Care Med 2001; 29:1109 –1115
145. Haley RW, Bregman DA: The role of under-
staffing and overcrowding in recurrent out-
breaks of staphylococcal infection in a neo-
natal special-care unit. J Infect Dis 1982;
145:875– 885
146. Fridkin SK, Pear SM, Williamson TH, et al:
The role of understaffing in central venous
catheter-associated bloodstream infections.
Infect Control Hosp Epidemiol 1996; 17:
150 –158
147. Grundmann H, Hori S, Winter B, et al: Risk
factors for the transmission of methicillin-
resistant Staphylococcus aureus in an adult
intensive care unit: Fitting a model to the
data. J Infect Dis 2002; 185:481– 488
148. Gerding DN, Larson TA, Hughes RA, et al:
Aminoglycoside resistance and aminoglyco-
side usage: Ten years of experience in one
hospital. Antimicrob Agents Chemother
1991; 35:1284 –1290
149. Kollef MH, Vlasnik J, Sharpless L, et al:
Scheduled change of antibiotic classes: A
strategy to decrease the incidence of venti-
lator-associated pneumonia. Am J Respir
Crit Care Med 1997; 156:1040 –1048
150. Dominguez EA, Smith TL, Reed E, et al: A
pilot study of antibiotic cycling in a hema-
tology-oncology unit. Infect Control Hosp
Epidemiol 2000; 21(1 Suppl):S4 –S8
151. Gruson D, Hilbert G, Vargas F, et al: Rota-
tion and restricted use of antibiotics in a
medical intensive care unit: Impact on the
incidence of ventilator-associated pneumo-
nia caused by antibiotic-resistant gram-
negative bacteria. Am J Respir Crit Care
Med 2000; 162:837– 843
152. Raymond DP, Pelletier SJ, Crabtree TD, et
al: Impact of a rotating empiric antibiotic
schedule on infectious mortality in an in-
tensive care unit. Crit Care Med 2001; 29:
1101–1108
153. Fridkin SK: Routine cycling of antimicro-
bial agents as an infection-control measure.
Clin Infect Dis 2003; 36:1438 –1444
154. Paterson DL, Rice LB: Empirical antibiotic
choice for the seriously ill patient: Are min-
imization of selection of resistant organ-
isms and maximization of individual out-
come mutually exclusive? Clin Infect Dis
2003; 36:1006 –1012
155. Rex JH, Walsh TJ, Sobel JD, et al: Practice
guidelines for the treatment of candidiasis.
Clin Infect Dis 2000; 30:662– 678
156. Banerjee SN, Emori TG, Culver DH, et al:
Secular trends in nosocomial primary
bloodstream infections in the United States,
1980 –1989. Am J Med 1991; 91:86S– 89S
157. Beck-Sagué CM, Jarvis WR, the National
Nosocomial Infections Surveillance System:
Secular trends in the epidemiology of nos-
ocomial fungal infections in the United
States, 1980 –1990. J Infect Dis 1993; 167:
1247–1251
158. Fisher-Hoch SP, Hutwagner L: Opportunis-
tic candidiasis: An epidemic of the 1980s.
Clin Infect Dis 1995; 21:897–904
159. Fraser VJ, Jones M, Dunkel J, et al: Candi-
demia in a tertiary care hospital: Epidemi-
ology, risk factors, and predictors of mor-
tality. Clin Infect Dis 1992; 15:414 – 421
160. Wenzel RP: Nosocomial candidemia: Risk
factors and attributable mortality. Clin In-
fect Dis 1995; 20:1531–1534
161. Calandra T, Bille J, Schneider R, et al: Clin-
ical significance of Candida isolated from
peritoneum in surgical patients. Lancet
1989; 2:1437–1440
162. Wey SB, Mori M, Pfaller MA, et al: Hospital-
acquired candidemia: The attributable mor-
tality and excess length of stay. Arch Intern
Med 1988; 148:2642–2645
163. Gudlaugsson O, Gillespie S, Lee K, et al:
Attributable mortality of nosocomial candi-
demia, revisited. Clin Infect Dis 2003; 37:
1172–1177
164. Uzun O, Ascioglu S, Anaissie EJ, et al: Risk
factors and predictors of outcome in pa-
tients with cancer and breakthrough candi-
demia. Clin Infect Dis 2001; 32:1713–1717
165. Nguyen MH, Peacock JE Jr, Tanner DC, et
al: Therapeutic approaches in patients with
candidemia: Evaluation in a multicenter,
prospective, observational study. Arch In-
tern Med 1995; 155:2429 –2435
166. Nolla-Salas J, Sitges-Serra A, Leon-Gil C, et
al: Candidemia in non-neutropenic criti-
cally ill patients: Analysis of prognostic fac-
tors and assessment of systemic antifungal
therapy. Intensive Care Med 1997; 23:23–30
167. Edwards JE Jr, Bodey GP, Bowden RA, et al:
International Conference for the Develop-
ment of a Consensus on the Management
and Prevention of Severe Candidal Infec-
tions. Clin Infect Dis 1997; 25:43–59
168. Rex JH, Bennett JE, Sugar AM, et al: A
randomized trial comparing fluconazole
with amphotericin B for the treatment of
candidemia in patients without neutrope-
nia. N Engl J Med 1994; 331:1325–1330
169. Phillips P, Shafran S, Garber G, et al: Mul-
ticenter randomized trial of fluconazole
versus amphotericin B for treatment of can-
didemia in non-neutropenic patients. Eur
J Clin Microbiol Infect Dis 1997; 16:
337–345
170. Anaissie EJ, Vartivarian SE, Abi-Said D, et
al: Fluconazole versus amphotericin B in
the treatment of hematogenous candidiasis:
A matched cohort study. Am J Med 1996;
101:170 –176
171. Driessen M, Ellis JB, Cooper PA, et al: Flu-
conazole vs. amphotericin B for the treat-
ment of neonatal fungal septicemia: A pro-
spective randomized trial. Ped Infect Dis J
1996; 15:1107–1112
S511
172. van’t Wout JW, Mattie H, van Furth R: A
prospective study of the efficacy of flucon-
azole (UK-49,858) against deep-seated fun-
gal infections. J Antimicrob Chemother
1988; 21:665– 672
173. Kujath P, Lerch K: Secondary mycosis in
surgery: treatment with fluconazole. Infec-
tion 1989; 17:111–117
174. Ikemoto H: A clinical study of fluconazole
for the treatment of deep mycoses. Diagn
Microbiol Infect Dis 1989; 12(4 Suppl):
239S–247S
175. Viscoli C, Castagnola E, Fioredda F, et al:
Fluconazole in the treatment of candidiasis
in immunocompromised children. Antimi-
crob Agents Chemother 1991; 35:365–367
176. Nolla-Salas J, Leon C, Torres-Rodriguez JM,
et al: Treatment of candidemia in critically
ill surgical patients with intravenous flu-
conazole. Clin Infect Dis 1992; 14:952–954
177. Fasano C, O’Keeffe J, Gibbs D: Fluconazole
treatment of children with severe fungal
infections not treatable with conventional
agents. Eur J Clin Microbiol Infect Dis
1994; 13:344 –347
178. Presterl E, Graninger W: Efficacy and safety
of fluconazole in the treatment of systemic
fungal infections in pediatric patients. Eur
J Clin Microbiol Infect Dis 1994; 13:
347–351
179. Graninger W, Presteril E, Schneeweiss B, et
al: Treatment of Candida albicans fungae-
mia with fluconazole. J Infect 1993; 26:
133–146
180. Nguyen MH, Peacock JE Jr, Morris AJ, et al:
The changing face of candidemia: Emer-
gence of non-Candida albicans species and
antifungal resistance. Am J Med 1996; 100:
617– 623
181. Johnson LB, Kauffman CA: Voriconazole: A
new triazole antifungal agent. Clin Infect
Dis 2003; 36:630 – 637
182. Ally R, Schurmann D, Kreisel W, et al: A
randomized, double-blind, double-dummy,
multicenter trial of voriconazole and flu-
conazole in the treatment of esophageal
candidiasis in immunocompromised pa-
tients. Clin Infect Dis 2001; 33:1447–1454
183. Perfect JR, Marr KA, Walsh TJ, et al: Vori-
conazole treatment for less-common,
S512
emerging, or refractory fungal infections.
Clin Infect Dis 2003; 36:1122–1131
184. Denning DW: Echinocandin antifungal
drugs. Lancet 2003; 362:1142–1151
185. Villanueva A, Arathoon EG, Gotuzzo E, et
al: A randomized double-blind study of
caspofungin versus amphotericin for the
treatment of candidal esophagitis. Clin In-
fect Dis 2001; 33:1529 –1535
186. Arathoon EG, Gotuzzo E, Noriega LM, et al:
Randomized, double-blind, multicenter
study of caspofungin versus amphotericin B
for treatment of oropharyngeal and esoph-
ageal candidiases. Antimicrob Agents Che-
mother 2002; 46:451– 457
187. Mora-Duarte J, Betts R, Rotstein C, et al:
Comparison of caspofungin and amphoter-
icin B for invasive candidiasis. N Engl J Med
2002; 347:2020 –2029
188. Eriksson U, Seifert B, Schaffner A: Compar-
ison of effects of amphotericin B deoxy-
cholate infused over 4 or 24 hrs: Random-
ised controlled trial. BMJ 2001; 322:
579 –582
189. Imhof A, Walter RB, Schaffner A: Continu-
ous infusion of escalated doses of ampho-
tericin B deoxycholate: An open-label obser-
vational study. Clin Infect Dis 2003; 36:
943–951
190. Prentice HG, Hann IM, Herbrecht R, et al: A
randomized comparison of liposomal versus
conventional amphotericin B for the treat-
ment of pyrexia of unknown origin in neu-
tropenic patients. Br J Haematol 1997; 98:
711–718
191. Clark AD, McKendrick S, Tansey PJ, et al: A
comparative analysis of lipid-complexed and
liposomal amphotericin B preparations in
haematological oncology. Br J Haematol
1998; 103:198 –204
192. White MH, Bowden RA, Sandler ES, et al:
Randomized, double-blind clinical trial of
amphotericin B colloidal dispersion vs. am-
photericin B in the empirical treatment of
fever and neutropenia. Clin Infect Dis 1998;
27:296 –302
193. Walsh TJ, Finberg RW, Arndt C, et al: Lipo-
somal amphotericin B for empirical therapy
in patients with persistent fever and neu-
tropenia. N Engl J Med 1999; 340:764 –771
194. Wingard JR, White MH, Anaissie E, et al: A
randomized, double-blind comparative trial
evaluating the safety of liposomal ampho-
tericin B versus amphotericin B lipid com-
plex in the empirical treatment of febrile
neutropenia. Clin Infect Dis 2000; 31:
1155–1163
195. Ellis M, Spence D, de Pauw B, et al: An
EORTC international multicenter random-
ized trial (EORTC number 19923) compar-
ing two dosages of liposomal amphotericin
B for treatment of invasive aspergillosis.
Clin Infect Dis 1998; 27:1406 –1412
196. Bowden R, Chandrasekar P, White MH, et
al: A double-blind, randomized, controlled
trial of amphotericin B colloidal dispersion
versus amphotericin B for treatment of in-
vasive aspergillosis in immunocompro-
mised patients. Clin Infect Dis 2002; 35:
359 –366
197. Leenders AC, Daenen S, Jansen RL, et al:
Liposomal amphotericin B compared with
amphotericin B deoxycholate in the treat-
ment of documented and suspected neutro-
penia-associated invasive fungal infections.
Br J Haematol 1998; 103:205–212
198. Walsh TJ, Pappas P, Winston DJ, et al: Vori-
conazole compared with liposomal ampho-
tericin B for empirical antifungal therapy in
patients with neutropenia and persistent fe-
ver. N Engl J Med 2002; 346:225–234
199. Bowden RA, Cays M, Gooley T, et al: Phase
I study of amphotericin B colloidal disper-
sion for the treatment of invasive fungal
infections after marrow transplant. J Infect
Dis 1996; 173:1208 –1215
200. Noskin GA, Pietrelli L, Coffey G, et al: Am-
photericin B colloidal dispersion for treat-
ment of candidemia in immunocompro-
mised patients. Clin Infect Dis 1998; 26:
461– 467
201. Juster-Reicher A, Flidel-Rimon O, Amitay
M, et al: High-dose liposomal amphotericin
B in the therapy of systemic candidiasis in
neonates. Eur J Clin Microbiol Infect Dis
2003; 22:603– 607
202. Lopez Sastre JB, Coto Cotallo GD, Fernan-
dez CB: Neonatal invasive candidiasis: A
prospective multicenter study of 118 cases.
Am J Perinatol 2003; 20:153–163
Crit Care Med 2004 Vol. 32, No. 11 (Suppl.)