Prognostic factors of mortality in patients with

community-acquired bloodstream infection

with severe sepsis and septic shock
,
Arturo Artero MD,PhD
a,

, Rafael Zaragoza MD
b
, Juan J. Camarena MD,PhD
c
,
Susana Sancho MD,PhD
b
, Rosa Gonzlez
c
, Jos M. Nogueira MD,PhD
c
a
Department of Internal Medicine, Hospital Universitario Dr Peset Av Gaspar Aguilar 90, 46017 Valencia, Spain
b
Intensive Care Unit, Hospital Universitario Dr Peset Valencia, 46017 Valencia, Spain
c
Department of Microbiology, Hospital Universitario Dr Peset Valencia, 46017 Valencia, Spain
Keywords:
Bacteremia;
Sepsis syndrome;
Septic shock;
Intensive care unit;
Mortality;
Inadequate empirical
antimicrobial treatment
Abstract
Purpose: The purpose of the study was to determine the independent risk factors on mortality in patients
with community-acquired severe sepsis and septic shock.
Methods: A single-site prospective cohort study was carried out in a medical-surgical intensive care unit
in an academic tertiary care center. One hundred twelve patients with community-acquired bloodstream
infection with severe sepsis and septic shock were identified. Clinical, microbiologic, and laboratory
parameters were compared between hospital survivors and hospital deaths.
Results: One-hundred twelve patients were included. The global mortality rate was 41.9%, 44.5% in septic
shock and 34.4% in severe sepsis. One or more comorbidities were present in 66% of patients. The most
commonly identified bloodstream pathogens were Escherichia coli (25%) and Staphylococcus aureus
(21.4%). The proportion of patients receiving inadequate antimicrobial treatment was 8.9%. By univariate
analysis, age, Acute Physiology and Chronic Health Evaluation II score, at least 3 organ dysfunctions, and
albumin, but neither microbiologic characteristics nor site of infection, differed significantly between
survivors and nonsurvivors. Acute Physiology and Chronic Health Evaluation II (odds ratio, 1.13; 95%
confidence interval, 1.06-1.21) and albumin (odds ratio, 0.34; 95% confidence interval, 0.15-0.76) were
independent risk factors associated with global mortality in logistic regression analysis.
Conclusion: In addition to the severity of illness, hypoalbuminemia was identified as the most important
prognostic factor in community-acquired bloodstream infection with severe sepsis and septic shock.
2010 Published by Elsevier Inc.
1. Introduction
Sepsis is an important cause of morbidity and mortality,
which accounts for about 2% of hospital admissions [1]. A
small proportion of these cases progress to severe sepsis and
septic shock, which account for 10% of admissions to
intensive care units (ICUs) [2]. Bacteremia is the essential

Institution at which the work was performed: Hospital Universitario

Dr Peset Av Gaspar Aguilar 90, 46017 Valencia, Spain.

No conflicts of interest to disclose.

Corresponding author. Tel.: +34 961622457; fax: +34 961622301.

E-mail addresses: arturo.artero@uv.es (A. Artero), zaragozar@ono.com
(R. Zaragoza), juan.camarena@uv.es (J.J. Camarena), Sancho_sus@gva.es
(S. Sancho), rosa.gonzalez@uv.es (R. Gonzlez), jose.m.nogueira@uv.es
(J.M. Nogueira).
0883-9441/$ see front matter 2010 Published by Elsevier Inc.
doi:10.1016/j.jcrc.2009.12.004
Journal of Critical Care (2010) 25, 276281
and fundamental pathophysiologic determinant of sepsis
caused by bacteria, and every effort should be made to
demonstrate bacteremia in patients with sepsis [3]. However,
blood cultures were found positive in only 25% of severe
sepsis and in 69% of septic shock [4].
Sepsis incidence is increasing at an estimated annual rate
of 1.5%, and the growing numbers of sepsis patients who
have organ dysfunction indicate that sepsis severity is also
increasing [5]. Most bacteremias in ICUs are hospital-
acquired infections, where the presence of patients with
severe conditions and the need for multiple intravascular and
other devices make such patients vulnerable to bloodstream
infections. Only about 10% to 40% of bloodstream infections
in ICUs are considered community acquired [6-8].
Despite advances in the care of critically ill patients, severe
sepsis and septic shock continue to be associated with a
dismal prognosis [9]. However, recently and based on the
Surviving Sepsis Campaign guidelines for management of
severe sepsis and septic shock [10], a slightly better prognosis
for patients with these conditions has been observed [1,11].
The outcome for critically ill patients with bacteremic sepsis
has been associated with age, sex, severity of illness at onset of
sepsis, comorbidities, nosocomial origin of the infection, and
etiology [1]. Several studies have examined the institution of
appropriate empirical antimicrobials with respect to mortality
in sepsis with diverse results [7,8,12-18]. We have previously
found no association between inadequate empirical antimicro-
bial treatment and the outcome in critically ill patients with
sepsis [8]. However, most of our knowledge about prognostic
factors in severe sepsis and septic shock are based on studies in
which there is a mixture of community- and hospital-acquired
infections or on studies in which only a proportion of patients
with sepsis had severe sepsis or septic shock [7,8,13-18]. Thus,
the objective of this study is to determine the influence of
clinical and microbiological characteristics, and empirical
antimicrobial treatment on the outcomes of patients with
community-acquired bloodstreaminfections with severe sepsis
and septic shock.
2. Patients and methods
2.1. Setting and patients
The study was conducted at a university-affiliated urban
teaching hospital with 600 beds: Hospital Universitario Dr
Peset in Valencia. The period of our study focused on the years
1998 to 2008, inclusive. Adult patients with community-
acquired bloodstreaminfections were collected, and those cases
with severe sepsis or septic shock were prospectively analyzed.
2.2. Study design and data collection
The study design has been previously described [8]. We
collected the following data on the first 24 hours of the
admission at the ICU: demographic characteristics, comorbid
medical conditions, factors predisposing to infection, source
of infection, Acute Physiology and Chronic Health Evalu-
ation (APACHE) II score, severe sepsis, septic shock, organ
dysfunction, and biochemical and hematologic alterations.
During the admission at the intensive care, we also collected
the microorganism isolates, adequacy of their empirical
antimicrobial treatment, and hospital mortality.
2.3. Definitions
Severe sepsis and septic shock were defined following the
criteria of the American College of Chest Physicians/Society of
Critical Care Medicine Consensus Conference [19]. Blood-
streaminfection was defined as the identification of a pathogen
in a blood culture specimen. The bloodstream infections
caused by coagulase-negative staphylococci and other bacteria
of saprophytic skin flora were excluded. Community-acquired
bloodstream infections were required to be established within
48 hours of hospital and ICU admission, and the bloodstream
infection should not be associated with any procedure
performed after hospital or ICU admission. Patients residing
at a nursing home, skilled-care facility, or rehabilitation center
who had a bloodstream infection requiring hospital admission
were classified as having community-acquired infections. The
source of infection was classified as one of the following: lower
respiratory tract, intraabdominal, genitourinary tract, menin-
geal, other, or unknown [13]. A modified McCabe classifica-
tion [20] was used to place the patients' underlying disease in
the following categories: rapidly fatal, ultimately fatal, and
nonfatal. Inappropriate empirical antimicrobial treatment was
considered as the occurrence of bloodstreaminfection that was
not effectively treated at the time when the causative
microorganism and its antibiotic susceptibility were known,
including the absence of antimicrobial agents directed at a
specific class of microorganisms and the administration of an
antimicrobial agent to which the microorganism responsible
for the infection was resistant [8].
2.4. Statistical analysis
Categorical variables were compared using the
2
test or
the Fisher exact test (when necessary). Continuous variables
were expressed as means SD, and Student t test was used
for comparing means. Variables with a P .05, and a
plausible biological relationship to the dependent outcome
variable, were further analyzed with a nonconditional
multivariate analysis to determine the independent factors
associated with the presence or absence of hospital mortality.
3. Results
During the study period, 146 patients were admitted to the
adult ICU with community-acquired bloodstream infections,
277 Mortality in community-acquired bloodstream infections
which represent 24.5% of the total of 594 bloodstream
infections admitted to ICU during the study period. Among
the community cases, severe sepsis or septic shock was
reported in 112 patients. The patients' mean age was 63.5
years (range, 16-85). Sixty percent of patients were male.
Table 1 shows the clinical and epidemiologic characteristics
of community-acquired bloodstream infections and their
relationship to hospital mortality. The global mortality rate
was 41.9%, 45.5%in septic shock and 34.4%in severe sepsis
(P = .343). Four variables were related to hospital mortality
Table 1 Epidemiologic and clinical characteristics of bloodstream infections and their relationship to hospital mortality
Total
(n = 112)
Hospital
survivors
(n = 65)
Hospital
nonsurvivors
(n = 47)
OR (95% CI) P
Mean age in years (SD) 63.5 (15.8) 61.0 (16.6) 67.1 (14.0) 1.02 (1.00-1.05) .047
Male sex 67 (59.8%) 40 (61.5%) 27 (57.4%) 1.18 (0.55-2.54) .663
One or more comorbidities 74 (66.0%) 39 (60.0%) 35 (74.4%) 1.94 (0.85-4.42) .111
Diabetes mellitus 32 (28.5%) 20 (30.7%) 12 (25.5%) 1.29 (0.55-3.00) .545
Chronic pulmonary disease 19 (16.9%) 7 (10.7%) 12 (25.5%) 2.84 (1.02-7.89) .045
Heart failure 19 (16.9%) 9 (13.8%) 10 (21.2%) 1.68 (0.62-4.53) .304
Alcoholism 14 (12.5%) 7 (10.7%) 7 (14.8%) 1.45 (0.47-4.45) .516
Neoplasm 10 (8.9%) 3 (4.6%) 7 (14.8%) 3.61 (0.88-14.80) .074
Liver cirrhosis 9 (8.0%) 5 (7.6%) 4 (8.5%) 1.11 (0.28-4.40) .875
Chronic renal failure 9 (8.0%) 5 (7.6%) 4 (8.5%) 1.11 (0.28-4.40) .875
Ultimately or rapidly fatal disease
(McCabe classification)
14 (12.5%) 6 (9.2%) 8 (17.0%) 2.01 (0.64-6.26) .219
Mean APACHE II (SD) 22.0 (8.0) 18.7 (7.1) 26.5 (7.0) 1.16 (1.08-1.23) b.001
Septic shock 83 (74.1%) 46 (70.7%) 37 (78.7%) 1.52 (0.63-3.68) .343
Glasgow scale 14 37 (33.0%) 18 (27.6%) 19 (40.4%) 1.77 (0.79-3.92) .157
3 Organ dysfunctions 56 (50.0%) 19 (29.2%) 37 (78.7%) 3.70 (2.04-6.68) b.001
Source
Respiratory 27 (24.1%) 15 (23.0%) 12 (25.5%) 1.14 (0.47-2.73) .764
Abdominal 19 (16.9%) 12 (18.4%) 7 (14.8%) 0.77 (0.27.2.14) .620
Genitourinary 17 (15.1%) 10 (15.3%) 7 (14.8%) 0.96 (0.33-2.74) .943
Meningeal 13 (11.6%) 9 (13.8%) 4 (8.5%) 0.57 (0.16-2.00) .389
Others 21 (18.7%) 14 (21.5%) 7 (14.8%) 0.63 (0.23-1.72) .376
Unknown 15 (13.3%) 5 (7.6%) 10 (21.2%) 3.24 (1.02-10.23) .045
Laboratory data
Leukopenia b4 10
9
/L 25 (22.3%) 12 (18.4%) 13 (27.6%) 1.68 (0.69-4.13) .249
Thrombocytopenia b100 10
9
/L 48 (42.8%) 25 (38.4%) 23 (48.9%) 1.53 (0.71-3.27) .269
Albumin b20g/L 27 (31.3%) 10 (21.2%) 17 (43.5%) 2.85 (1.11-7.33) .026
Antimicrobial monotherapy 56 (50.0%) 31 (47.6%) 25 (53.1%) 1.24 (0.58-2.64) .566
Inadequate empirical
antimicrobial treatment
9 (8.0%) 5 (7.6%) 4 (8.5%) 1.11 (0.28-4.40) 1.000
OR indicates odds ratio; CI, confidence interval.
Table 2 Microorganisms associated with community-acquired bloodstream infections
Microorganisms Septic shock
n = 83
Severe sepsis
n = 29
OR (95% CI) P Hospital
survivors
n = 65
Hospital
nonsurvivors
n = 47
OR (95% CI) P
Gram positive 36 19 0.40 (0.16-0.97) .040 34 21 0-73 (0.34-1.56) .426
Staphylococcus aureus 14 9 0.45 (0.17-1.19) .104 14 9 0.86 (0.33-2.20) .757
Streptococcus pneumoniae 15 4 1.37 (0.41-4.55) .597 13 6 1.70 (0.59-4.88) .314
Others 7 6 0.35 (0.10-1.15) .076 7 6 1.21 (0.37-3.87) .745
Gram negative 39 9 1.96 (0.80-4.83) .135 27 21 1.13 (0.53-2.42) .740
E coli 26 2 6.15 (1.36-27.8) .009 17 11 1.15 (0.48-2.77) .740
Pseudomonas spp 3 3 0.32 (0.06-1.70) .166 3 3 1.40 (0.27-7.31) .682
Others 10 4 0.85 (0.24-2.97) .755 7 7 1.45 (0.47-4.45) .515
Polymicrobial and fungi 8 1 2.98 (0.35-24.9) .442 4 5 1.81 (0.46-7.16) .389
278 A. Artero et al.
by univariate analysis: age (P= .044), APACHE II (Pb .001),
at least 3 organ dysfunctions (P b .001), and albumin less than
20g/L (P = .026). Only 8.% of the patients received
inadequate antimicrobial treatment, with no relation to
mortality. After excluding the deaths that occurred within
48 hours of admission (n = 15), there was no relationship
between inadequate empirical antimicrobial treatment and
mortality (P = 1.000). The etiology of cases that received
inadequate empirical antimicrobial treatment was as follows:
Escherichia coli (n = 3), Staphylococcus aureus (n = 2),
Candida glabrata (n = 1), Cryptococcus neoformans (n = 1),
and polymicrobial (n = 2). The site of infection of these cases
was as follows: urinary (n = 2), abdominal (n = 2), respiratory
(n = 2), endocarditis (n = 1), and unknown (n = 2).
The most frequently isolated microorganisms are shown
in Table 2. Gram-positive microorganisms were isolated in
49.1% of blood cultures; gram-negative, 42.8%; and
polymicrobial and fungi, 8%. Septic shock was slightly
more frequent in bloodstream infections caused by gram-
positive microorganisms. Neither gram-positive nor gram-
negative organisms were related to mortality. Escherichia
coli was the most frequent microorganisms isolated (58.3%
of gram-negative bacilli). The quinolone resistance rate to
E coli was 35.7%, and the prevalence of extended-spectrum
-lactamases was 14.2%. Escherichia coli was associated
with septic shock (P = .009) but not with mortality. None
of the other microorganisms were associated with either
shock or mortality. Pseudomonas aeruginosa was isolated
in 5.3% of cases, with appropriate empirical antimicrobial
treatment in all cases and with a mortality rate of 50%.
Staphylococcus aureus and Streptococcus pneumoniae
were the most frequent gram-positive microorganisms
isolated (41.8% and 34.5% of gram-positive microorgan-
isms, respectively). The rate of methicillin-resistant Sta-
phylococcus aureus was 17.3%, and there was no case
producing Panton-Valentine leukocidin.
Four risk factors were independently associated to global
mortality by univariate analysis: age, APACHE II, number
of organ dysfunction, and albumin. Table 3 shows
independent risk factors associated with hospital mortality
by multivariate analysis.
4. Discussion
In our series, we found that a simple laboratory finding
such as abnormal serum albumin and the well-established
prognostic index APACHE II were independently associated
with the mortality. These results increase our knowledge
about prognostic factors in the subgroup of critically ill
patients with community-acquired bloodstream infections
with severe sepsis and septic shock.
We found a 41.9% hospital mortality rate, which is almost
identical to the 43.1% found by Valls et al [13] in
community-acquired bloodstream infection in critically ill
adult patients, despite the fact that in this study the patients
were older (65.1% vs 55.2% of patients aged 60 years) and
less severe forms of sepsis were not included.
The analysis of the microbiological characteristics of
cases showed that 3 microorganismsE coli, Staphylococ-
cus aureus, and Streptococcus pneumoniaewere respon-
sible for the majority of community-acquired bloodstream
infections (62.5%), according with previous publications
[13,21]. Only E coli was associated with septic shock, but
not with mortality. Gram-positive microorganisms had a
lower rate of septic shock (65.4% vs 82.4%), whereas gram-
negative and gram-positive bacteria had similar outcomes,
according with other studies on severe sepsis [22].
It should be noted that despite the fact that abdominal or
respiratory foci of infections have been found to be related to
mortality in studies on sepsis and septic shock, these studies
included patients with hospital-acquired infections [23-25].
However, in our study with only community-acquired
infections, we were unable to define any influence of the
source of infection on prognosis, as has been described by
Valls and the Spanish collaborative group for infections in
ICUs in community-acquired bloodstream infections in
critically adult ill patients [13].
We found APACHE II and serum albumin to be
associated with mortality. The APACHE II has been widely
accepted as one of the most important prognostic factors in
severe bloodstream infections [21]; but despite the fact that
low serum albumin appears to reflect a poor prognosis in a
broad range of clinical settings [26,27], it is not usually
included as an independent prognostic factor in severe sepsis
or septic shock. The third version of APACHE score
includes albumin, but its performance was not superior to
that of its predecessor for a cohort of critically ill patients
[28]. We want to emphasize that because our patients had
community-acquired infections, hypoalbuminemia was
probably related to infection, whereas hypoalbuminemia in
hospital-acquired sepsis can also be caused by previous
illnesses; and it should be included in studies of prediction of
outcome in community-acquired infections.
In sepsis, empirical antimicrobial therapy, based on
coverage of predictable pathogens determined by the focus
of infection, should be administered as early as possible [3].
However, our data showed no association between inade-
quate empirical antimicrobial treatment and mortality. This
might be derived from different reasons: the low rate of
inadequate empirical antimicrobial treatment (8.9%) in our
study as a consequence of the prescription of combined
antibiotics and broad-spectrum monotherapy for these severe
Table 3 Independent risk factors associated with global
mortality to bloodstream infections by logistic regression
Risk factor Global mortality
P value OR (95% CI)
APACHE II b.001 1.13 (1.06-1.21)
Albumin (10 g/L) .009 0.34 (0.15-0.76)
279 Mortality in community-acquired bloodstream infections
forms of sepsis; the low rate of bacteria resistant to
antibiotics in community-acquired infections, particularly
methicillin-resistant Staphylococcus aureus; and finally, the
exclusion of bloodstream infections caused by coagulase-
negative staphylococci. We previously found that coagulase-
negative staphylococcus was the most frequent organism in
critically ill patients with nosocomial-acquired infections
related to inadequate empirical antimicrobial treatment [8].
In life-threatening infections, therapy is usually begun
with a combination of antibiotics; but a single antibiotic may
be sufficient, especially in abdominal and skin and soft tissue
infections. In this series, 50% of the patients received
monotherapy. We did not find any relationship between
monotherapy and mortality or inadequate empirical antimi-
crobial treatment. Nevertheless, more studies with larger
number of cases would be necessary to provide more precise
information on these matters.
This study has several limitations. It was not designed for
identifying health careassociated bloodstream infections,
which often resemble nosocomial infections [29]. Some of
these cases could be included in this category, especially if we
consider that the mean age of our patients was 63.5 years and
the high rate of underlying disorders. We do not knowwhether
our results apply to patients with health careassociated
bloodstream infections. Because of the fact that the lack of
influence of inadequate empirical antimicrobial treatment on
mortality was based on a very limited number of cases, this
should be interpreted as a type 2 error, lacking statistical power.
The design of this study did not allowus to analyze the time to
antibiotic administration or early goal-directed therapy, which
have been associated with a better prognosis [30,31]. Finally,
our study was a single-site investigation of patients admitted to
an ICU; and the risk factors identified could be different in
other centers or settings.
Our results highlight the added importance of serum
albumin, in addition to APACHE II, in predicting the
mortality risk in community-acquired severe sepsis and
septic shock in standard clinical setting.
References
[1] Annane D, Bellissant E, Cavaillon JM. Septic shock. Lancet 2005;365:
63-78.
[2] Annane D, Aegerter P, Jars-Guincestre MC, et al. Current epidemi-
ology of septic shock: the CUB-Rea Network. Am J Respir Crit Care
Med 2003;1687:165-72.
[3] Cunha BA. Sepsis and septic shock: selection of empiric antimicrobial
therapy. Crit Care Clin 2008;24:313-4.
[4] Rangel-Frausto MS, Pittet D, Costigan M, et al. The natural history of
the systemic inflammatory response syndrome (SIRS). JAMA
1995;273:117-23.
[5] Martin GS, Mannino DM, Eaton S, et al. The epidemiology of sepsis in
the United States from 1979 through 2000. N Engl J Med 2003;348:
1546-54.
[6] Crowe M, Ispahani P, Humphreys H, et al. Bacteremia in the adult
intensive care unit of a teaching hospital in Nottingham, UK, 1985-
1996. Eur J Clin Microbiol Infect Dis 1998;17:377-84.
[7] Ibrahim EH, Sherman G, Ward S, et al. The influence of inadequate
antimicrobial treatment of bloodstream infections on patient outcomes
in the ICU setting. Chest 2000;118:146-55.
[8] Zaragoza R, Artero A, Camarena JJ, et al. The influence of
inadequate empirical antimicrobial treatment on patients with
bloodstream infections in an intensive care unit. Clin Microbiol
Infect 2003;9:1-7.
[9] Hugonnet S, Harbarth S, Ferrire K, et al. Bacteremic sepsis in
intensive care: temporal trends in incidence, organ dysfunction, and
prognosis. Crit Care Med 2003;31:390-4.
[10] Dellinger RP, Carlet JM, Masur H, et al. Surviving Sepsis Campaign
guidelines for management of severe sepsis and septic shock. Intensive
Care Med 2004;30:536-55.
[11] Ferrer R, Artigas A, Levy MM, et al. Improvement in process of care
and outcome after a multicenter severe sepsis educational program in
Spain. JAMA 2008;299:2294-303.
[12] Talan DA, Moran GJ, Abrahamian FM. Severe sepsis and septic
shock in the emergency department. Infect Dis Clin North Am
2008;22:1-31.
[13] Valls J, Rello J, Ochagava A, et al. Community-acquired
bloodstream infection in critically ill adult patients: impact of shock
and inappropriate antibiotic therapy on survival. Chest 2003;123:
1615-24.
[14] Kollef MH, Sherman G, Ward S, et al. Inadequate antimicrobial
treatment of infections. A risk factor of hospital mortality among
critically ill patients. Chest 1999;115:462-74.
[15] Garnacho-Montero J, Ortiz-Leyva C, Herrera-Melero I, et al. Mortality
and morbidity attributable to inadequate empirical antimicrobial
therapy in patients admitted to the ICU with sepsis: a matched cohort
study. J Antimicrob Chemother 2008;61:436-41.
[16] Leibovici L, Shraga I, Drucker M, et al. The benefit of appropriate
empirical antibiotic treatment in patients with bloodstream infection. J
Intern Med 1998;244:379-86.
[17] Kollef MH, Sherman G, Ward S, et al. Inadequate antimicrobial
treatment of infections: a risk factor for hospital mortality among
critically ill patients. Chest 1999;115:462-74.
[18] Garnacho-Montero J, Garcia-Garmendia JL, Barrero-Almodovar A,
et al. Impact of adequate empirical antibiotic therapy on the outcome of
patients admitted to the intensive care unit with sepsis. Crit Care Med
2003;31:2742-51.
[19] American College of Chest Physicians/Society of Critical Care
Medicine Consensus Committee. Definition for sepsis and organ
failures and guidelines for the use of innovative therapies in sepsis.
Chest 1992;101:1644-55.
[20] McCabe WR, Jackson GG. Gram negative bacteremia. Arch Intern
Med 1962;110:92-100.
[21] Laupland KB, Gregson DB, Zygun DA, et al. Severe bloodstream
infections: a population-based assessment. Crit Care Med 2004;32:
992-7.
[22] Bone RC, Fisher Jr CJ, Clemmer TP, et al. A controlled clinical trial of
high-dose methylprednisolone in the treatment of severe sepsis and
septic shock. N Engl J Med 1987;317:653-8.
[23] Rello J, Ricart M, Mirelis B, et al. Nosocomial bacteremia in a
medical-surgical intensive care unit: epidemiologic characteristics and
factors influencing mortality in 111 episodes. Intensive Care Med
1994;20:94-8.
[24] Pittet D, Thievent B, Wenzel RP, et al. Bedside prediction of mortality
from bacteremic sepsis. A dynamic analysis of ICU patients. Am J
Respir Crit Care Med 1996;153:684-93.
[25] Valls J, Len C, Alvarez-Lerma F. Nosocomial bacteremia in
critically ill patients: a multicenter study evaluating epidemiology
and prognosis. Spanish Collaborative Group for Infections in
Intensive Care Units of Sociedad Espaola de Medicina Intensiva
y Unidades Coronarias (SEMIUC). Clin Infect Dis 1997;24:
387-95.
[26] Wallace SM, Walton BI, Kharbanda RK, et al. Mortality from infective
endocarditis: clinical predictors of outcome. Heart 2002;88:53-60.
280 A. Artero et al.
[27] Goldwasser P, Feldman J. Association of serum albumin and mortality
risk. J Clin Epidedemiol 1997;50:693-703.
[28] Beck DH, Taylor BL, Millar B, et al. Prediction of outcome from
intensive care: a prospective cohort study comparing Acute Physiology
and Chronic Health Evaluation II and III prognostic systems in a
United Kingdom intensive care unit. Crit Care Med 1997;25:9-15.
[29] Friedman ND, Kaye KS, Stout JE, et al. Health care-associated
bloodstream infections in adults: a reason to change the accepted
definition of community-acquired infections. Ann Intern Med
2002;137:791-7.
[30] Kumar A, Roberts D, Wood KE, et al. Duration of hypotension before
initiation of effective antimicrobial therapy is the critical determinant
of survival in human septic shock. Crit Care Med 2006;34:1589-96.
[31] Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in
the treatment of severe sepsis and septic shock. N Engl J Med
2001;345:1368-77.
281 Mortality in community-acquired bloodstream infections