Albumin Versus Other Fluids For Fluid Resuscitation in Patients With Sepsis: A Meta-Analysis

Download as pdf or txt
Download as pdf or txt
You are on page 1of 22

RESEARCH ARTICLE

Albumin versus Other Fluids for Fluid


Resuscitation in Patients with Sepsis: A
Meta-Analysis
Libing Jiang, Shouyin Jiang, Mao Zhang*, Zhongjun Zheng, Yuefeng Ma*
Department of Emergency Medicine, Second Affiliated Hospital, School of Medicine & Institute of Emergency
Medicine, Zhejiang University, Hangzhou, China
*[email protected] (YFM); [email protected] (MZ)

Abstract

OPEN ACCESS
Citation: Jiang L, Jiang S, Zhang M, Zheng Z, Ma
Y (2014) Albumin versus Other Fluids for Fluid
Resuscitation in Patients with Sepsis: A MetaAnalysis. PLoS ONE 9(12): e114666. doi:10.1371/
journal.pone.0114666
Editor: James D. Chalmers, University of Dundee,
United Kingdom
Received: June 12, 2014
Accepted: November 12, 2014
Published: December 4, 2014
Copyright: 2014 Jiang et al. This is an openaccess article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original author
and source are credited.
Data Availability: The authors confirm that all data
underlying the findings are fully available without
restriction. All relevant data are within the paper
and its Supporting Information files.
Funding: The authors have no support or funding
to report.
Competing Interests: The authors have declared
that no competing interests exist.

Background: Early fluid resuscitation is vital to patients with sepsis. However, the
choice of fluid has been a hot topic of discussion. The objective of this study was to
evaluate whether the use of albumin-containing fluids for resuscitation in patients
with sepsis was associated with a decreased mortality rate.
Methods: We systematically searched PubMed, EMBASE and Cochrane library for
eligible randomized controlled trials (RCTs) up to March 2014. The selection of
eligible studies, assessment of methodological quality, and extraction of all relevant
data were conducted by two authors independently.
Results: In total, 15 RCTs were eligible for analysis. After pooling the data, we
found there was no significant effect of albumin-containing fluids on mortality in
patients with sepsis of any severity (RR: 0.94, 95% CI: 0.87, 1.02 and RD: 0.01,
95% CI: 0.03, 0.01). The results were robust to subgroup analyses, sensitivity
analyses and trial sequential analyses.
Conclusion: The present meta-analysis did not demonstrate significant advantage
of using albumin-containing fluids for resuscitation in patients with sepsis of any
severity. Given the cost-effectiveness of using albumin, crystalloids should be the
first choice for fluid resuscitation in septic patients.

Introduction
Sepsis is a common serious health problem. It is estimated that the annual
number of patients with severe sepsis exceeds 750,000 in the United States and 19
million worldwide, with a short-term mortality of 20% to 30%, reaching up to
50% when shock is present [14]. Meanwhile, the total number of deaths from

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

1 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

sepsis continues to increase [1]. It has been reported that sepsis is the leading
cause of death among hospitalized patients in non-coronary intensive care units
[5, 6].
Early fluid resuscitation is one of the key interventions for patients with sepsis
which has been widely accepted by clinicians. However, the optimal choice of
fluid remains inconclusive [710]. Albumin has been used as one type of
resuscitation fluids since the Second World War [11]. However, until recently, the
pragmatic value of albumin in sepsis is still under debate [1215]. In 2011, a large
meta-analysis which included 17 studies demonstrated that albumin use in
patients with sepsis was associated with a decrease in mortality [16]. However, this
study has some flaws. Firstly, it is vulnerable to bias because the most influential
trial included was the pre-defined subgroup of patients with severe sepsis in the
SAFE (the saline versus albumin fluid evaluation) study [13]. Secondly, six studies
by Dr. Joachim Boldt (whose studies are suspected of lacking of integrity) were
included in this meta-analysis [9]. As several large studies regarding which fluid
should be used for resuscitation have been published recently, the purpose of this
study was to further evaluate whether the use of albumin-containing fluids was
associated with a decreased mortality rate in patients with sepsis.

Methods
This meta-analysis was performed according to the Preferred Reporting Items for
Systematic Reviews and Meta-Analyses (PRISMA) statement (Checklist S1) [17].

Eligibility Criteria
Patients: patients with sepsis of any severity (including sepsis, severe sepsis and
septic shock).
Intervention: fluid resuscitation.
Comparison: fluid resuscitation with albumin-containing fluids (of any
concentration) vs. other resuscitation fluids (including any colloid or crystalloid).
Outcome: all-cause mortality at the longest follow-up available (including 48 h
mortality, ICU mortality, hospital mortality, 28/30 days mortality, 90 days
mortality, whichever was longest.).
Study Design: randomized controlled trials (RCTs).

Literature Search and Study Selection


We systematically searched PubMed, EMBASE and Cochrane library up to March
2014. The following free text words or Medical Subject Headings were used:
sepsis, septic, systemic inflammatory response syndrome, SIRS, septicemia, fluid
therapy, resuscitation, plasma substitute, albumin and serum albumin. In
addition, we also screened reference lists of all eligible studies and relevant reviews
to obtain additional trials. There was no language restriction. The search strategy
is showed in Text S2. Two investigators (JLB and MYF) independently screened

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

2 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

the titles and abstracts of all records identified from the literature search. After
excluding obviously non-relevant publications, potentially eligible articles were
further screened in full text according to our pre-defined inclusion criteria.
Discrepancies were resolved by consensus.

Data Extraction and Quality Assessment


Data on the following items were extracted from the eligible studies by two
investigators (LBJ AND MYF) independently: characteristics of studies,
characteristics of patients, interventions and outcomes. Two reviewers (LBJ and
MZ) independently and in duplicate assessed the methodological quality of each
study by applying the following items: randomization-sequence generation,
allocation concealment, blinding, intention-to-treat analysis, selective outcome
reporting and the number of patients lost to follow-up. Randomization-sequence
generation was considered adequate when the study described the method to
generate the randomization sequence (such as computer-generated random
numbers or random number table). Allocation concealment was considered
adequate if researchers screening patients could not predict the next treatment for
a patient. Blinding was considered adequate if both patients and investigators did
not know which treatment the patients received. There was no evidence of
selective outcome reporting if all stated endpoints were reported on and
presented. Completeness of outcome data for each outcome was considered
adequate if intention-to-treat analysis was performed and the lost follow-up rate
should be within 10% [18].

Statistical Analysis
Pooled risk ratios (RRs) with 95% confidence intervals (CIs) for all-cause
mortality were calculated with RevMan 5.2.10 (http://tech.cochrane.org/revman/
download) and STATA 12.0 (SERIAL NO. 40120519635). Heterogeneity between
studies was measured by chi2 statistic (p,0.1) and quantified with I2 statistic [19].
If the I2 value was less than 50%, the fixed effects model was used to pool studies;
otherwise, the random effects model was used. Several predefined subgroup
analyses were performed according to patients age (adult or pediatrics), type of
resuscitation fluid in the control group (crystalloid or gelofusine or starch),
concentration of albumin (45% solution or 2025% solution), follow-up interval
(ICU mortality, hospital mortality, 28/30 days mortality and 90 days mortality),
disease severity (sepsis, severe sepsis and septic shock), and definition of sepsis
(American College of Chest Physicians/Society of Critical Care Medicine, ACCP/
SCCM, criteria or other criteria). Given the ALBIOS study was not limited to the
resuscitation phase but included albumin supplementation for 28 days after
enrollment, another subgroup analysis was conducted by the time interval
between patients enrollment and randomization in the ALBIOS study (,6 h or 6
24 h) [7]. Meanwhile, we conducted sensitivity analyses and verified the
robustness of our results by excluding either or both of the following studies: the

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

3 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

EARSS study, which has not yet been published [20]; and ALBIOS study, in which
albumin was used mainly for maintaining the serum albumin concentration of
.30 g/L (not merely volume expansion) [21]. In addition, studies [2230] with
small sample sizes (,100 patients), studies with large sample sizes (.100
patients) [13, 20, 21, 31, 36, 37] and studies [26, 30, 31] on malaria which has a
pathophysiology with many features in common with sepsis [32] were excluded to
confirm the robustness of our results. Both random and fixed effects models were
used. The sample size of a meaningful meta-analysis should be at least as large as a
powered RCT. And updated meta-analyses of studies are vulnerable to random
errors due to sparse data and repetitive testing of accumulated data [33].
Therefore, we conducted trial sequential analysis (TSA) to calculate the optimal
required information size [34] (meta-analysis sample size) for our meta-analysis
based on a baseline mortality rate of 31.7% in the control group which was
calculated according to the 3 largest trials [13, 20, 21], a relative risk reduction of
10% [20, 21], 80% of power, and a type I error of 5%. We constructed monitoring
boundaries to determine whether clinical trials could be terminated early when a p
value is small enough to detect the expected effect. TSA was performed in TSA
V.0.9 b (http://www.ctu.dk/tsa/). Publication bias was assessed by funnel plots
and Eggers test [35].

Results
Search Results and Study Characteristics
A total of 1460 articles were identified through the literature search. According to
our predefined inclusion criteria, 15 studies were included finally (Figure 1). Data
on mortality were available in the published papers of all [13, 21, 23
28, 30, 31, 36, 37] but three of trials [20, 22, 29]; data for these were extracted from
Delaney et als analysis [16] and Zheng yam et als analysis [38]. A total of 6998
septic patients were analyzed. Of these, 3225 patients received the albumincontaining fluids for resuscitation. The characteristics of all included studies are
showed in Table 1.

Quality Assessment
The methodical quality of all included studies was summarized in Table 2.

Mortality
Data on all-cause mortality were available from 15 RCTs [13, 20
29, 30, 31, 36, 37]. Although 90-day mortalities were reported in two studies,
mortalities at 28 days which were the primary endpoints in these two studies were
used to calculate the overall pooled RR for mortality [21, 37]. The results
indicated that there was no effect of albumin on all-cause mortality in the fixedeffects model (RR: 0.94, 95% CI: 0.87, 1.02; p50.15) (Figure 2) or random-effects
model (RR: 0.95, 95% CI: 0.88, 1.03; p50.20), with no heterogeneity between

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

4 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Figure 1. Flow chart for study selection.


doi:10.1371/journal.pone.0114666.g001

studies (I250%, p50.56). Trial sequential adjusted 95% CI of RR was 0.85 to 1.04
in the fixed effects model, and 0.86 to 1.04 in the random effects model. TSA
showed that the diversity adjusted information size was 6576 which was less than
that in our study (n56998) and the cumulative Z-curve surpassed the futility
boundary, but it did not cross the trial sequential monitoring boundary for benefit
or harm, indicating further studies are not required as they can unlikely change
the current conclusion (whether benefit or harm) (Figure 3). The shape of the
funnel plot and results of Eggers test (p50.264) suggested no publication bias
(Figure 4).

Subgroup Analysis
In order to further evaluate the effect of albumin-containing fluids on all-cause
mortality in patients with sepsis, several subgroup analyses were performed
according to patients age (adult or pediatrics), type of resuscitation fluid in the
control group (crystalloid or gelofusine or starch), concentration of albumin (4
5% solution or 2025% solution), follow-up interval (ICU mortality, hospital
mortality, 28/30 days mortality and 90 days mortality), disease severity (sepsis,
severe sepsis and septic shock), and definition of sepsis (American College of

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

5 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Table 1. Characteristics of included studies.

Reference

Year

Adult/
Children

Rackow
et al [23]

1983

Adult

Septic
/hypovolaemic

Metildi
et al [25]

1984

Adult

Severe
pulmonary

Patient

No.
of
cases

Albumin
type (s)

Control
fluid (s)

Resuscitation Albumin
goal (s)
volume (ml)

Duration of
follow-up

26

5%

0.9% saline

PCWP$15

2833

Hospital

46

5%

Maintenance
of

9400

Hospital

shock

6% HES
Ringers lactate

insufficiency

normal BE, Ph,


SvO2

Rackow
et al [24]

1989

Adult

Severe
sepsis

20

5%

10%
Pentastarch

PCWP$15

975

Hospital

SAFE [13]

2004

Adult

Patients
in ICU

6997

4%

0.9% saline

The discretion
of the

2376, first 3 days

28 days,

requiring
fluid

treating
clinicians

ICU

resuscitation
Veneman et al 2004
[29]

Adult

Sepsis
and post

Hospital
63

20%

surgical
patients

0.9% saline

MAP.70 mmHg

10% HES

CVP 5
10 mmHg

900

30 days

with SIRS
Maitland
et al [30]

2005

Children

Severe malaria,

61

4.5%

0.9% saline

20 mL/kg

20 mL/kg

ICU

150

4.5%

0.9% saline

To avoid

Moderate acidosis:

ICU

hypotension,

45 mL/kg

sustained
oliguria,

Severe acidosis:

anaemia,
severe
acidosis,
respiratory
distress
Maitland
et al [31]

2005

Children

Severe malaria
and
metabolic
acidosis

worsening

Akech
et al [26]

2006

Children

Severe
falciparum

88

4.5%

Gelofusine

malaria,
metabolic

metabolic
acidosis

63 mL/kg

Resolution of

Moderate acidosis:

shock

46 mL/kg

acidosis,
shock

ICU

Severe acidosis:
50 mL/kg

Friedman
et al [22]

2008

Adult

sepsis and
suspected

42

4%

hypovolemia
Van der
et al [27]

2009

Adult

Septic and
non-septic

6% HES

Fixed volume
400 mL

400

Hospital

According to
fluid

1500

ICU

10% HES
48

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

5%

0.9% saline

6 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Table 1. Cont.

Reference

Year

Adult/
Children

Patient

No.
of
cases

Albumin
type (s)

patients with
or at risk
for ALI/ARDS
Dolecek
et al [28]

2009

Adult

severe
sepsis

Control
fluid (s)

Resuscitation Albumin
goal (s)
volume (ml)

4% gelatin

challenge
protocol

Duration of
follow-up

6% HES
56

20%

6% HES

Intrathoracic
blood volume

600

28 days

index 850 mL/


m2,
cardiac index
3.5 l/min/m2
FEAST [36]

2011

Children

Severe
febrile illness

3141

5%

0.9% saline

and impaired
perfusion
EARSS [20]

CRISTAL [37]

2011

2013

Adult

Adult

Septic shock

Patients
in ICU

Resolution of

40 mL/kg, first 8 h 48 h,

impaired
perfusion
792

2857

requiring
fluid

20%

4%/5%
or

0.9% NaCl

Isotonic saline

20%/25%

28 days

Fixed volume
100 mL

Fixed volume
100 mL,

Every 8 h for
3 days

every 8 h for 3
days

The discretion
of the

Not reported

investigators

28 days

28 days,
90 days

resuscitation
ALBIOS [21]

2013

Adult

Severe
sepsis or

1810

septic
Shock

20%

Crystalloid

300 mL/day
until day

300 mL/day until


day

28 days,

solution

28 or ICU
discharge

28 or ICU discharge

90 days

PCWP: Pulmonary capillary wedge pressure; MAP: Mean arterial pressure; CVP: Central venous pressure.
ALI: Acute lung injury; ARDS: Acute respiratory distress syndrome.
ICU: Intensive care unit.
doi:10.1371/journal.pone.0114666.t001

Chest Physicians/Society of Critical Care Medicine, ACCP/SCCM, criteria or


other criteria). Given the ALBIOS study was not limited to the resuscitation phase
but included albumin supplementation for 28 days after enrollment, another
subgroup analysis was conducted by the time interval between patients enrollment
and randomization in the ALBIOS study (,6 h or 624 h) [21]. As illustrated in
Table 3, our results suggested that there was no significant effect of albumin on
all-cause mortality in both adult and pediatric patients with sepsis. Eleven studies
compared albumin with crystalloid [13, 20, 21, 23, 25, 27, 29, 30, 31, 36, 37].
Albumin was not associated with a significant reduction in all-cause mortality
when compared to crystalloid (RR: 0.95, 95% CI: 0.87, 1.04; p50.25). Trial
sequential adjusted 95% CI of RR was 0.87 to 1.04 in the fixed effects model, and
0.86 to 1.05 in the random effects model. The required information size was 7635
and the cumulative Z-curve crossed the boundary of futility, but it did not cross
the trial sequential monitoring boundary for benefit or harm, indicating further

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

7 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Table 2. Qualitative assessment of included studies.

Blinding

Intention
to Treat
Analysis

Loss to
follow-up

Reference

Randomisation

Allocation
concealment

Rackow et al [23]

Low risk

Unclear riskd

High risk

Low risk

Low risk

Metildi et al [25]

Low risk

Unclear risk

High risk

Low risk

Low risk

Rackow et al [24]

Unclear riska

Unclear riskd

High risk

Low risk

Low risk

SAFE [13]

Low risk

Low risk

Low risk

Low risk

Low risk

Veneman et al [29]

Unclear riska

Low risk

High risk

Low risk

Low risk

Maitland et al [30]

Unclear riska

Low risk

High risk

Low risk

Low risk

Maitland et al [31]

Unclear risk

Low risk

High risk

Low risk

Low risk

Akech et al [26]

High riskb

Unclear riskd

High risk

Low risk

Low risk

Friedman et al [22]

Unclear riska

Low risk

High risk

Low risk

High riskc

van der et al [27]

Unclear riska

Low risk

High risk

Low risk

Low risk

Dolecek et al [28]

Low risk

Unclear riskd

High risk

Low risk

Low risk

FEAST [36]

Low risk

Low risk

High risk

Low risk

Low risk

EARSS [20]

Unclear riske

Unclear riske

High risk

Unclear riske

Unclear riske

CRISTAL [37]

Low risk

Low risk

High risk

Low risk

Low risk

ALBIOS [21]

Low risk

Low risk

High risk

Low risk

Low risk

Just mention the word of random.


A quasi-randomised design was used, whereby fluid interventions were allocated sequentially in blocks of ten.
c
4 patients (11%) were excluded because of inadequate data collection.
d
not reported.
e
The research has not yet been published.
b

doi:10.1371/journal.pone.0114666.t002

Figure 2. Forest plot showing the effects of albumin-containing fluids on all-cause mortality in patients with sepsis.
doi:10.1371/journal.pone.0114666.g002

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

8 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Figure 3. Trial sequential analysis of all-cause mortality in patients with sepsis. Trial sequential analyses assessing the effect of albumin on all-cause
mortality in 15 studies. The diversity-adjusted required information size (6576 participants) was based on a relative risk reduction of 10%; an alpha of 5%; a
beta of 2% and an event proportion of 31.7% in the control arm. The blue cumulative z curve was constructed using a random effects model.
doi:10.1371/journal.pone.0114666.g003

studies are unlikely to change the current conclusion (Figure 5). Six studies [22
24, 2729] compared albumin with hydroxyethyl starch and two studies [26, 27]
compared albumin with gelofusine. The results indicated there is no evidence that
albumin reduces mortality when compared with hydroxyethyl starch or
gelofusine. And given the small sample sizes of these studies, we were unable to
perform TSA. Although the difference of mortality between albumin group (both
4%5% and 20%25%) and control group did not reach statistical significance,
we found 4%5% albumin may be relative safer than 20%25% albumin for fluid
resuscitation (Table 3). In addition, we found albumin did not reduce all-cause
mortality regardless of the follow-up time point. And these results were also not
affected by the sepsis definition (5 studies [13, 20, 21, 28, 37] fitted the ACCP/
SCCM criteria). Finally, a subgroup analysis was performed based on the disease
severity. Albumin was associated with a small reduction in all-cause mortality
when compared to an alternative resuscitation fluid in patients with septic shock
(RR: 0.89, 95% CI: 0.80, 0.99; p50.04) [20, 21, 23]; however, this mortality benefit
became insignificant when the comparison was limited to patients with sepsis
[22, 37] and severe sepsis [13, 21, 24, 25, 2729]. Moreover, this mortality benefit

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

9 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Figure 4. Funnel plot showing no significant publication bias.


doi:10.1371/journal.pone.0114666.g004

in patients with septic shock was not robust to TSA. Trial sequential adjusted 95%
CI of RR was 0.74 to 1.07 both in the fixed and random effects model. Although
the cumulative Z curve crossed conventional monitoring boundary for benefit, it
did not cross the trial sequential monitoring boundary (Figure 6), suggesting that
there was insufficient evidence to show a 10% reduction of all-cause mortality for
80% power, an a of 0.05, and a mortality rate of 40.3% in the control group.
Given the potential bias of our study, further studies are needed to confirm
whether albumin has an impact on mortality of patients with septic shock.

Sensitivity analysis
Sensitivity analyses were performed by excluding the following studies
successively: EARSS study [20], which was abstract from conference proceedings;
ALBIOS study [21], in which the primary aim of albumin-containing fluids
administration was not for initial resuscitation; small studies (,100 patients) [22
25, 2631]; large studies (.100 patients) [13, 20, 21, 31, 36, 37]; and studies
[26, 30, 31] on malaria which has a pathophysiology with many features in
common with sepsis [39]. The results indicated that the exclusion of these studies
did not change our primary outcomes (Table 4).

Discussion
A total of 15 studies enrolling 6998 patients were eligible for evaluating the effect
of albumin-containing fluids on all-cause mortality in patients with sepsis [13, 20
29, 30, 31, 36, 37]. The results of this meta-analysis indicated that the use of

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

10 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Table 3. Subgroup analyses.


Fixed
RR
(95%CI)

Random
RR (95%CI)

6983

0.94
(0.87 to 1.02)

0.95
(0.88 to 1.03)

15

5757

0.93
(0.85 to 1.02)

15

6400

Children

Adults

Fixed
RD (95%CI)

Random
RD (95%CI)

0%

20.01
(20.03 to 0.01)

20.02
(20.04 to 0.00)

9%

0.94
(0.86 to 1.03)

0%

20.02
(20.04 to 0.00)

20.02
(20.05 to 0.00)

7%

0.91
(0.84 to 0.99)

0.92
(0.85 to 1.00)

0%

20.02
(20.04 to 0.00)

20.03
(20.05 to 0.00)

8%

2345

0.92
(0.74 to 1.14)

0.55
(0.21 to 1.45)

63%

20.01
(20.04 to 0.02)

20.07
(20.16 to 0.03)

73%

11

4638

0.95
(0.87 to 1.03)

0.95
(0.87 to 1.03)

0%

0.02
(0.05 to 0.01)

0.02
(0.05 to 0.01)

0%

,6 h

11

3412

0.93
(0.84 to 1.03)

0.94
(0.85 to 1.03)

0%

0.02
(0.06 to 0.01)

0.02
(0.06 to 0.01)

0%

624 h

11

4055

0.91
(0.84 to 1.00)

0.92
(0.84 to 1.00)

0%

0.03
(0.06 to 0.00)

0.03
(0.06 to 0.00)

0%

Crystalloid

11

6741

0.95
(0.88 to 1.04)

0.95
(0.87 to 1.04)

5%

0.01
(0.03 to 0.01)

0.02
(0.05 to 0.02)

38%

,6 h

11

5515

0.94
(0.86 to 1.03)

0.94
(0.86 to 1.04)

3%

0.01
(0.04 to 0.01)

0.02
(0.06 to 0.02)

38%

624 h

11

6158

0.92
(0.85 to 1.00)

0.92
(0.85 to 1.00)

0%

0.02
(0.04 to 0.00)

0.03
(0.06 to 0.01)

40%

Gelofusine

100

0.33
(0.10 to 1.12)

0.42
(0.05 to 3.18)

59%

0.12
(0.24 to 0.00)

0.13
(0.24 to 0.02)

0%

Starch

169

0.91
(0.62 to 1.32)

0.93
(0.65 to 1.33)

0%

0.04
(0.17 to 0.10)

0.05
(0.18 to 0.09)

0%

Subgroup

No. of
Studies

Patients

Total

15

,6 h
624 h

I2

I2

Age of
patients

Types of
control fluids

Concentrations of albumin
4%5%

10

3691

0.90
(0.79 to 1.01)

0.91 (0.80 to 1.04)

7%

0.02
(0.05 to 0.00)

0.05
(0.09 to 0.00)

32%

20%25%

2676

0.97
(0.86 to 1.09)

0.98 (0.87 to 1.09)

0%

0.01
(0.04 to 0.03)

0.01
(0.04 to 0.02)

0%

,6 h

1450

0.96
(0.82 to 1.11)

0.97
(0.83 to 1.12)

0%

0.01
(0.06 to 0.03)

0.02
(0.06 to 0.03)

0%

624 h

2093

0.92
(0.82 to 1.03)

0.92 (0.82 to 1.03)

0%

0.03
(0.07 to 0.01)

0.03
(0.07 to 0.01)

0%

ICU
mortality

3587

0.88
(0.76 to 1.02)

0.86 (0.64 to 1.14)

40%

0.02
(0.04 to 0.00)

0.05
(0.10 to 0.00)

58%

Hospital
mortality

1322

0.88
(0.76 to 1.02)

0.89 (0.78 to 1.02)

0%

0.04 (0.09 to
0.01)

0.04
(0.09 to 0.01)

0%

28/30-day
mortality*

6607

0.96
(0.88 to 1.04)

0.96 (0.88 to 1.04)

0%

0.01
(0.03 to 0.01)

0.01
(0.03 to 0.01)

0%

90-day
mortality

2397

0.95
(0.86 to 1.06)

0.95
(0.86 to 1.06)

0%

0.02
(0.06 to 0.02)

0.02
(0.06 to 0.02)

0%

,6 h

1185

0.01
(0.85 to 1.21)

0.01 (0.85 to 1.21)

0%

0.01
(0.06 to 0.07)

0.01
(0.06 to 0.07)

0%

Endpoints

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

11 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Table 3. Cont.
Fixed
RR
(95%CI)

Fixed
RD (95%CI)

Random
RD (95%CI)

0%

0.03
(0.08 to 0.02)

0.03
(0.08 to 0.02)

0%

0.94
(0.86 to 1.04)

0%

0.02
(0.05 to 0.01)

0.02
(0.05 to 0.01)

0%

0.93
(0.83 to 1.03)

0.93 (0.84 to 1.03)

0%

0.02
(0.06 to 0.01)

0.03
(0.06 to 0.01)

0%

3894

0.91
(0.83 to 1.00)

0.91 (0.83 to 1.00)

0%

0.03
(0.06 to 0.00)

0.03
(0.06 to 0.00)

0%

10

2506

0.93
(0.78 to 1.12)

0.96
(0.81 to 1.14)

4%

0.01
(0.04 to 0.02)

0.05
(0.11 to 0.01)

27%

Sepsis

658

1.10
(0.77 to 1.57)

1.10 (0.77 to 1.57)

0%

0.03
(0.09 to 0.14)

0.03
(0.09 to 0.14)

0%

Severe
sepsis

2035

0.95
(0.84 to 1.07)

0.95 (0.85 to 1.07)

0%

0.02
(0.06 to 0.02)

0.02
(0.06 to 0.02)

0%

Septic
shock

1931

0.89
(0.80 to 0.99)

0.89 (0.80 to 0.99)

0%

0.04
(0.09 to 0.00)

0.04
(0.09 to 0.00)

0%

Subgroup

No. of
Studies

Patients

Random
RR (95%CI)

I2

624 h

1828

0.94
(0.83 to 1.06)

0.93 (0.83 to 1.05)

ACCP/SCCM

4477

0.94
(0.86 to 1.03)

,6 h

3251

624 h

Non ACCP/SCCM

I2

The definition of
sepsis

Severity of
disease

RR, Relative Risk; 95% CI, 95% Confidence Intervals; RD, Risk difference; Fixed, fixed - effects model; Random, random - effects model.
*, 28/30 day mortality were not stratified according to the time interval between patient enrollment and randomization in the ALBIOS study.
ACCP/SCCM, American College of Chest Physicians/Society of Critical Care Medicine.
doi:10.1371/journal.pone.0114666.t003

albumin-containing fluids for the resuscitation of patients with sepsis of any


severity was not associated with lower death rates compared with other fluid
resuscitation regimens.
Cardiovascular system can be impaired by sepsis which may be mediated by
multiple mechanisms, with the result of tissue hypo-perfusion. Meanwhile, the
increased intravascular space and capillary permeability which result from direct
cell damage and the release of inflammatory mediators can further increase the
amount of fluid required and thus complicate the resuscitative process [40].
Therefore, maintenance of adequate intravascular volume and tissue perfusion is
critical with regard to patients outcome, and early adequate fluid resuscitation
has been shown to improve the prognosis of septic patients [41]. Until now, the
choice of fluid for resuscitation in patients with sepsis remains controversial. In
recent years, several large RCTs and systematic reviews have reported that the use
of hydroxyethyl starch, the commonly used colloid solution, is associated with a
significant increased risk of acute kidney injury and death in critically ill patients
[810]. Meanwhile, the results from the SAFE study and a subsequent Metaanalysis have shown that albumin as a resuscitation fluid for patients with sepsis
may significantly reduce the risk of death [13, 16]. These conflicting results have

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

12 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Figure 5. Trial sequential analysis of all-cause mortality in patients with sepsis comparing albumin with crystalloid solutions. Trial sequential
analyses assessing the effect of albumin on all-cause mortality in 11 studies. The diversity-adjusted required information size (7635 participants) was based
on a relative risk reduction of 10%; an alpha of 5%; a beta of 2% and an event proportion of 31.7% in the control arm. The blue cumulative z curve was
constructed using a random effects model.
doi:10.1371/journal.pone.0114666.g005

raised a re-emerging debate regarding which fluid on earth should be used for
fluid resuscitation in patients with sepsis.
Crystalloid solutions are widely used in fluid resuscitation of critically ill
patients. In addition to their efficiency, crystalloids are popular also because they
are readily available and cheap. However, the use of crystalloids is not without
drawbacks. Because crystalloids are composed of only small particles such as
sodium ions and chlorine ions, large infusion of crystalloids especially normal
saline may result in hypernatronemia and hyperchloremic acidosis which have
proven to be associated with coagulation derangements and renal, cerebral,
gastrointestinal and respiratory dysfunction [40]. In addition, due to their lower
molecular weight, crystalloid solutions can easily across the damaged semipermeable membrane of capillaries which often results in a shorter intravascular
persistence of fluids and may aggravate lung edema [42]. Human albumin is a
natural protein which accounts for 50%60% of all plasma proteins and nearly
80% of plasma colloid osmotic pressure [11]. Compared with crystalloids,
albumin can efficiently hold intravenous fluids due to their larger molecular
weight [35, 37]. Although there is the possibility that the increased membrane

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

13 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Figure 6. Trial sequential analysis of all-cause mortality in patients with septic shock. Trial sequential analyses assessing the effect of albumin on allcause mortality in 3 studies. The diversity-adjusted required information size (4586 participants) was based on a relative risk reduction of 10%; an alpha of
5%; a beta of 2% and an event proportion of 40.3% in the control arm. The blue cumulative z curve was constructed using a random effects model.
doi:10.1371/journal.pone.0114666.g006

permeability can augment the extravasation of fluid into the interstitium due to
leakage of albumin [31], it has been reported that the required amount of fluid to
achieve the same resuscitation endpoint can be two to three times higher in the
crystalloids group than in the colloids group [13, 14, 21]. As a natural colloid,
human albumin is supported by the findings that septic patients receiving
albumin-containing fluids usually have higher colloid osmotic pressure, central
venous pressure, and slower heart rate than those who received crystalloids [21
23, 27, 28]. Its worth noting that whether patients are in septic or non-septic
status, resuscitation with albumin showed greater cardiac responses than normal
saline [43]. Unfortunately, until now, data regarding differences in the above
mentioned indicators between albumin and artificial colloids were controversial
[2325, 27, 29, 43].
Human albumin also has multiple roles other than its oncotic properties: 1)
transporting other biologically active molecules; 2) antioxidant; 3) antiinflammatory action; 4) inhibition of platelet aggregation; 5) capacity for reducing
capillary permeability and maintaining endothelial cell integrity; and 6) buffering
the acid-base equilibrium [4, 16, 21]. It has been reported that hypoalbuminemia,

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

14 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Table 4. Sensitivity analysis.


RD(95%CI)
Fixed

Random

I2

0%

0.01
(0.04 to 0.01)

0.02
(0.05 to 0.00)

16%

0.94
(0.86 to 1.04)

0%

0.02
(0.04 to 0.01)

0.03
(0.06 to 0.00)

14%

0.91
(0.84 to 1.00)

0.92
(0.85 to 1.00)

0%

0.02
(0.05 to 0.00)

0.03
(0.06 to 0.00)

15%

5188

0.91
(0.82 to 1.01)

0.93
(0.84 to 1.02)

0%

0.02
(0.04 to 0.00)

0.03
(0.06 to 0.00)

13%

Both
[22, 33]

4396

0.91
(0.81 to 1.02)

0.93
(0.83 to 1.03)

0%

0.02
(0.04 to 0.00)

0.04
(0.07 to 0.00)

19%

Small
studies [2230]

6650

0.95
(0.87 to 1.03)

0.95
(0.85 to 1.07)

31%

0.01
(0.03 to 0.01)

0.02
(0.05 to 0.01)

46%

,6 h

5409

0.94
(0.85 to 1.03)

0.95
(0.84 to 1.07)

29%

0.02
(0.04 to 0.01)

0.02
(0.06 to 0.01)

45%

624 h

6052

0.92
(0.84 to 1.00)

0.93
(0.83 to 1.03)

24%

0.02
(0.04 to 0.00)

0.03
(0.06 to 0.01)

48%

Large studies

348

0.85
(0.64 to 1.13)

0.94
(0.75 to 1.17)

0%

0.05
(0.13 to 0.04)

0.08
(0.15 to 0.00)

0%

Trials on
malaria [26, 30, 31]

6750

0.96
(0.88 to 1.04)

0.96
(0.88 to 1.04)

0%

0.01
(0.03 to 0.01)

0.01
(0.03 to 0.01)

0%

,6 h

5508

0.95
(0.86 to 1.04)

0.95
(0.87 to 1.04)

0%

0.01
(0.04 to 0.01)

0.01
(0.03 to 0.01)

0%

624 h

6152

0.93
(0.86 to 1.01)

0.93
(0.86 to 1.01)

0%

0.02
(0.04 to 0.00)

0.01
(0.03 to 0.01)

0%

No. of
Patients

RR(95%CI)
Fixed

Random

I2

EARSS
[22]

6206

0.94
(0.86 to 1.03)

0.95
(0.88 to 1.04)

,6 h

4965

0.93
(0.84 to 1.03)

624 h

5638

ALBIOS [33]

Excluding studies

[13, 20,
21, 31, 36, 37]

RR, Relative Risk; 95% CI, 95% Confidence Intervals; RD, Risk difference; Fixed, fixed - effects model; Random, random - effects model.
Both, EARSS+ALBIOS were excluded.
doi:10.1371/journal.pone.0114666.t004

which is common in critically ill patients (including septic patients), is associated


with poor clinical outcomes [4446]. Chou et al reported that for patients with
severe sepsis due to secondary peritonitis, albumin administration may reduce 28day mortality, however this mortality benefit was limited to patients whose
baseline serum albumin is 20 g/L or lower [45].
The abnormal accumulation of fluid in the extravascular space of the lung along
with severe inflammation may cause impairment of oxygenation and are strongly
associated with a high risk of death [40, 47]. Therefore, whether albumin can be
used as a resuscitation fluid has been comprehensively appraised by oxygenation,
pulmonary edema, organ performance, and resource utilization. As mentioned
above, due to its oncotic properties and non-oncotic properties, albumin may
decrease the extravasation of fluid from vessels into interstitial spaces and thus
reduce the degrees of pulmonary edema and improve oxygenation [40, 4851]. A
recent meta-analysis has demonstrated that the use of albumin is associated with
improved oxygenation when compared to crystalloid solutions [47]. However,
this effect was not observed in the study by Van der et al [27]. This inconsistent

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

15 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

result may be partially explained by that 5% albumin was used in the study by Van
der et al, whereas 25% albumin was used in the meta-analysis. Dolecek et al
reported that 20% albumin could significantly reduce the amount of extravascular
lung water when compared to 6% HES [28]. Nevertheless, oxygenation was not
shown to be better in patients treated with 20% albumin [22, 28, 52]. In addition,
there was no significant difference in pulmonary edema and oxygenation function
between the 5% albumin group and the 6% HES group [24, 27]. As for organ
function, patients in the albumin group, as compared with those in the crystalloid
group, had a higher SOFA sub-scores for liver [13, 21]. It may be explained by the
presence of bilirubin which was associated with the methods used to prepare
albumin solutions [13, 21]. In addition, in the ALBIOS study, the authors also
found a slightly higher SOFA sub-scores for coagulation in the albumin group,
which was attributed to the dilution of the hemoglobin content due to early and
large intravascular volume expansion [21]. Finally, most studies showed that there
was no effect of albumin on the length of stay in ICU/hospital, duration of
mechanical ventilation, requirement of renal replacement therapy [14, 20, 21].
Hitherto, there are still many unsolved issues about albumin administration in
patients with sepsis. Firstly, timing of albumin administration. The optimal time
to administer albumin to patients with sepsis has not yet been explored. However,
it has been reported that fluid resuscitation improve microvascular perfusion in
the early but not in the late phase of sepsis, and this effect is independent of the
type of fluid [53]. Thus, it seemed that the timing of fluid resuscitation is more
important than the type of fluid [54]. Secondly, concentration of albumin. In
general, 4%5% albumin is usually used for resuscitation and 20%25% albumin
is usually used for maintaining normal serum albumin levels. However, in a large
meta-analysis, the authors reported that hyperoncotic albumin decreased the odds
of acute kidney injury and death by 76% and 48%, respectively [55]. And
hyperoncotic albumin seems to improve oxygenation better than hypooncotic
albumin [47]. Moreover, the results of the SAFE study indicated that resuscitation
with 4% albumin might increase mortality in patients with traumatic brain injury
[13]. It is worth noting that the choice of albumin concentration may also depend
on the type of fluid which is administrated simultaneously. In the present metaanalysis, we found 4%5% albumin may be relative safer than 20%25% albumin
for fluid resuscitation. Thirdly, dose of albumin. Until now, no researches have yet
been designed to assess the dose-response relationship between albumin exposure
and mortality rate in patients with sepsis. Recent evidence have suggested that
whether in the early or late phase of resuscitation, net positive fluid balance is
associated with worse outcome [56, 57, 58]. In a large meta-analysis, the authors
reported that albumin reduces morbidity in acutely ill hospitalized patients,
however this effect was significantly influenced by the albumin dose in the control
group [59]. Finally, the high cost of albumin may limit its wide applicability.
Albumin can be anywhere between 20 and 100 times more expensive than
crystalloids, therefore, the cost effectiveness of albumin should be incorporated
into the stands of care. It has also been reported that the number of patients
needed to treat (NNT) to avoid one additional death is 45, namely the cost per

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

16 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

case avoided was $31,220, based on the results of the EARSS study [60, 61].
Therefore, if there is no significant advantage of albumin in reducing mortality
rate, it is difficult to justify unrestricted use of albumin for resuscitation of
patients with sepsis. Though our results indicated that further studies are unlikely
to change the current conclusion but considering the above mentioned issues and
potential bias, further studies are needed to confirm whether albumin has an
impact on mortality of patients with sepsis.

Strengths and Limitations of This Meta-Analysis


There are several strengths of this meta-analysis. First, the present meta-analysis
was performed according to the Preferred Reporting Items for Systematic Reviews
and Meta-Analyses (PRISMA) statement protocol [17]. Three electronic databases
which are recommended by the Cochrane Collaboration were searched for
relevant studies. The screening of eligible studies, assessment of methodological
quality and data extraction were conducted independently and in duplicate.
Second, we only included RCTs in this review to minimize potential bias and
there were enough number of patients to address this question. Meanwhile, there
was no significant heterogeneity among included studies. Third, several predefined subgroup analyses and sensitivity analyses were performed to verify the
robustness of our results and trial sequential analysis was performed to eliminate
random errors.
This study has several limitations. Firstly, although there was insignificant
heterogeneity between studies in this meta-analysis, the methodological quality of
all included studies was variable and all studies were open label except for the
SAFE study [13]. Secondly, patients with sepsis in six studies were only a
subgroup of the total populations studied. Thirdly, there is evidence that a longer
observation period for mortality, such as 90 days, is appropriate to assess the real
effects of treatments in critically ill patients [62, 63]. Unfortunately, in our metaanalysis, 90 days mortality was only reported in two studies [21, 37]. Another
limitation of our meta-analysis is that there were relatively few studies comparing
albumin and artificial colloids were included. As mentioned above, albumin was
compared with hydroxyethyl starch in six studies [2224, 2729], and compared
with gelofusine in two studies [26, 27]. As we all know, hydroxyethyl starch is
associated with an increased risk of acute kidney injury and death [810], thus,
indirectness is a major limitation for the comparison of albumin with
hydroxyethyl starch and there is insufficient evidence to make any firm
conclusions on comparisons of albumin with artificial colloids based on these
sparse data. In addition, because the first research included in our study can be
retrospect to 1983, the influence of existing standards of care on outcome may
have affected the results of this study and different albumin manufacturers may
also have an impact on the results.

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

17 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

Conclusion
Although albumin has many theoretical advantages, these have not been
supported by clinical trials. The present meta-analysis did not demonstrate
significant advantages of albumin over other fluids for resuscitation in patients
with sepsis of any severity. Given the tremendous economic burden of albumin,
crystalloids should be the first choice for fluid resuscitation in septic patients.

Supporting Information
Text S1. Search strategy.
doi:10.1371/journal.pone.0114666.s001 (PDF)
Checklist S1. The Preferred Reporting Items for Systematic Reviews and MetaAnalyses (PRISMA) statement.
doi:10.1371/journal.pone.0114666.s002 (DOC)

Author Contributions
Conceived and designed the experiments: LBJ MZ YFM. Performed the
experiments: LBJ SYJ ZJZ. Analyzed the data: LBJ SYJ ZJZ YFM MZ. Contributed
reagents/materials/analysis tools: LBJ ZJZ MZ. Contributed to the writing of the
manuscript: LBJ YFM ZJZ MZ.

References
1. Martin GS, Mannino DM, Eaton S, Moss M (2003) The epidemiology of sepsis in the United States
from 1979 through 2000. N Engl J Med 348: 15461554.
2. Dombrovskiy VY, Martin AA, Sunderram J, Paz HL (2007) Rapid increase in hospitalization and
mortality rates for severe sepsis in the United States: a trend analysis from 1993 to 2003. Crit Care Med
35: 12441250.
3. Angus DC, van der Poll T (2013) Severe sepsis and septic shock. N Engl J Med 369: 2063.
4. Karakala N, Raghunathan K, Shaw AD (2013) Intravenous fluids in sepsis: what to use and what to
avoid. Curr Opin Crit Care 19: 537543.
5. Martin GS, Eaton S, Mealer M, Moss M (2005) Extravascular lung water in patients with severe sepsis:
a prospective cohort study. Crit Care 9: R7482.
6. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, et al. (2001) Epidemiology of
severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit
Care Med 29: 13031310.
7. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, et al. (2013) Surviving sepsis campaign:
international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 41:
580637.
8. Haase N, Perner A, Hennings LI, Siegemund M, Lauridsen B, et al. (2013) Hydroxyethyl starch 130/
0.380.45 versus crystalloid or albumin in patients with sepsis: systematic review with meta-analysis and
trial sequential analysis. BMJ 346: f839.
9. Zarychanski R, Abou-Setta AM, Turgeon AF, Houston BL, McIntyre L, et al. (2013) Association of
hydroxyethyl starch administration with mortality and acute kidney injury in critically ill patients requiring
volume resuscitation: a systematic review and meta-analysis. JAMA 309: 678688.

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

18 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

10. Perner A, Haase N, Guttormsen AB, Tenhunen J, Klemenzson G, et al. (2012) Hydroxyethyl starch
130/0.42 versus Ringers acetate in severe sepsis. N Engl J Med 367: 124134.
11. Finfer S (2013) Reappraising the role of albumin for resuscitation. Curr Opin Crit Care 19: 315320.
12. (1998) Human albumin administration in critically ill patients: systematic review of randomised controlled
trials. BMJ 317: 235240.
13. Finfer S, Bellomo R, Boyce N, French J, Myburgh J, et al. (2004) A comparison of albumin and saline
for fluid resuscitation in the intensive care unit. N Engl J Med 350: 22472256.
14. Finfer S, McEvoy S, Bellomo R, McArthur C, Myburgh J, et al. (2011) Impact of albumin compared to
saline on organ function and mortality of patients with severe sepsis. Intensive Care Med 37: 8696.
15. (2011) Human albumin solution for resuscitation and volume expansion in critically ill patients. Cochrane
Database Syst Rev: CD001208.
16. Delaney AP, Dan A, McCaffrey J, Finfer S (2011) The role of albumin as a resuscitation fluid for
patients with sepsis: a systematic review and meta-analysis. Crit Care Med 39: 386391.
17. Moher D, Liberati A, Tetzlaff J, Altman DG (2009) Preferred reporting items for systematic reviews and
meta-analyses: the PRISMA statement. BMJ 339: b2535.
18. Schulz KF, Altman DG, Moher D (2010) CONSORT 2010 statement: updated guidelines for reporting
parallel group randomised trials. BMJ 340: c332.
19. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Measuring inconsistency in meta-analyses.
BMJ 327: 557560.
20. (2011) Abstracts of ESICM LIVES 2011, the 24th Annual Congress of the European Society of Intensive
Care Medicine. October 15, 2011. Berlin Germany. Intensive Care Med 37 Suppl 1: S6314.
21. Caironi P, Tognoni G, Masson S, Fumagalli R, Pesenti A, et al. (2014) Albumin replacement in
patients with severe sepsis or septic shock. N Engl J Med 370: 14121421.
22. Friedman G, Jankowski S, Shahla M, Gomez J, Vincent JL (2008) Hemodynamic effects of 6% and
10% hydroxyethyl starch solutions versus 4% albumin solution in septic patients. J Clin Anesth 20: 528
533.
23. Rackow EC, Falk JL, Fein IA, Siegel JS, Packman MI, et al. (1983) Fluid resuscitation in circulatory
shock: a comparison of the cardiorespiratory effects of albumin, hetastarch, and saline solutions in
patients with hypovolemic and septic shock. Crit Care Med 11: 839850.
24. Rackow EC, Mecher C, Astiz ME, Griffel M, Falk JL, et al. (1989) Effects of pentastarch and albumin
infusion on cardiorespiratory function and coagulation in patients with severe sepsis and systemic
hypoperfusion. Crit Care Med 17: 394398.
25. Metildi LA, Shackford SR, Virgilio RW, Peters RM (1984) Crystalloid versus colloid in fluid
resuscitation of patients with severe pulmonary insufficiency. Surg Gynecol Obstet 158: 207212.
26. Akech S, Gwer S, Idro R, Fegan G, Eziefula AC, et al. (2006) Volume expansion with albumin
compared to gelofusine in children with severe malaria: results of a controlled trial. PLoS Clin Trials 1:
e21.
27. van der Heijden M, Verheij J, van Nieuw Amerongen GP, Groeneveld AB (2009) Crystalloid or colloid
fluid loading and pulmonary permeability, edema, and injury in septic and nonseptic critically ill patients
with hypovolemia. Crit Care Med 37: 12751281.
28. Dolecek M, Svoboda P, Kantorova I, Scheer P, Sas I, et al. (2009) Therapeutic influence of 20%
albumin versus 6% hydroxyethylstarch on extravascular lung water in septic patients: a randomized
controlled trial. Hepatogastroenterology 56: 16221628.
29. Veneman TF, Oude Nijhuis J, Woittiez AJ (2004) Human albumin and starch administration in critically
ill patients: a prospective randomized clinical trial. Wien Klin Wochenschr 116: 305309.
30. Maitland K, Pamba A, English M, Peshu N, Levin M, et al. (2005) Pre-transfusion management of
children with severe malarial anaemia: a randomised controlled trial of intravascular volume expansion.
Br J Haematol 128: 393400.
31. Maitland K, Pamba A, English M, Peshu N, Marsh K, et al. (2005) Randomized trial of volume
expansion with albumin or saline in children with severe malaria: preliminary evidence of albumin benefit.
Clin Infect Dis 40: 538545.

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

19 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

32. Maitland K, Marsh K (2004) Pathophysiology of severe malaria in children. Acta Trop 90: 131140.
33. Brok J, Thorlund K, Gluud C, Wetterslev J (2008) Trial sequential analysis reveals insufficient
information size and potentially false positive results in many meta-analyses. J Clin Epidemiol 61: 763
769.
34. Wetterslev J, Thorlund K, Brok J, Gluud C (2009) Estimating required information size by quantifying
diversity in random-effects model meta-analyses. BMC Med Res Methodol 9: 86.
35. Hayashino Y, Noguchi Y, Fukui T (2005) Systematic evaluation and comparison of statistical tests for
publication bias. J Epidemiol 15: 235243.
36. Maitland K, Kiguli S, Opoka RO, Engoru C, Olupot-Olupot P, et al. (2011) Mortality after fluid bolus in
African children with severe infection. N Engl J Med 364: 24832495.
37. Annane D, Siami S, Jaber S, Martin C, Elatrous S, et al. (2013) Effects of fluid resuscitation with
colloids vs crystalloids on mortality in critically ill patients presenting with hypovolemic shock: the
CRISTAL randomized trial. JAMA 310: 18091817.
38. Yan Z, Peng-lin MA (2013) Albumin for fluid resuscitation sepsis:is just harmless or beneficial.
Chin J Crit Care med 33: 101105.
39. Maitland K, Marsh K (2004) Pathophysiology of severe malaria in children. Acta Trop 90: 131140.
40. Vincent JL, Gottin L (2011) Type of fluid in severe sepsis and septic shock. Minerva Anestesiol 77:
11901196.
41. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, et al. (2001) Early goal-directed therapy in the
treatment of severe sepsis and septic shock. N Engl J Med 345: 13681377.
42. Zampieri FG, Park M, Azevedo LC (2013) Colloids in sepsis: evenly distributed molecules surrounded
by uneven questions. Shock 39 Suppl 1: 4249.
43. Trof RJ, Sukul SP, Twisk JW, Girbes AR, Groeneveld AB (2010) Greater cardiac response of colloid
than saline fluid loading in septic and non-septic critically ill patients with clinical hypovolaemia. Intensive
Care Med 36: 697701.
44. Vincent JL, Dubois MJ, Navickis RJ, Wilkes MM (2003) Hypoalbuminemia in acute illness: is there a
rationale for intervention? A meta-analysis of cohort studies and controlled trials. Ann Surg 237: 319
334.
45. Chou CD, Yien HW, Wu DM, Kuo CD (2009) Albumin administration in patients with severe sepsis due
to secondary peritonitis. J Chin Med Assoc 72: 243250.
46. Dubois MJ, Orellana-Jimenez C, Melot C, De Backer D, Berre J, et al. (2006) Albumin administration
improves organ function in critically ill hypoalbuminemic patients: A prospective, randomized, controlled,
pilot study. Crit Care Med 34: 25362540.
47. Uhlig C, Silva PL, Deckert S, Schmitt J, de Abreu MG (2014) Albumin versus crystalloid solutions in
patients with the acute respiratory distress syndrome: a systematic review and meta-analysis. Crit Care
18: R10.
48. Margarido CB, Margarido NF, Otsuki DA, Fantoni DT, Marumo CK, et al. (2007) Pulmonary function
is better preserved in pigs when acute normovolemic hemodilution is achieved with hydroxyethyl starch
versus lactated Ringers solution. Shock 27: 390396.
49. Verheij J, van Lingen A, Raijmakers PG, Rijnsburger ER, Veerman DP, et al. (2006) Effect of fluid
loading with saline or colloids on pulmonary permeability, oedema and lung injury score after cardiac and
major vascular surgery. Br J Anaesth 96: 2130.
50. Marx G, Pedder S, Smith L, Swaraj S, Grime S, et al. (2004) Resuscitation from septic shock with
capillary leakage: hydroxyethyl starch (130 kd), but not Ringers solution maintains plasma volume and
systemic oxygenation. Shock 21: 336341.
51. Marx G, Pedder S, Smith L, Swaraj S, Grime S, et al. (2006) Attenuation of capillary leakage by
hydroxyethyl starch (130/0.42) in a porcine model of septic shock. Crit Care Med 34: 30053010.
52. Palumbo D, Servillo G, DAmato L, Volpe ML, Capogrosso G, et al. (2006) The effects of
hydroxyethyl starch solution in critically ill patients. Minerva Anestesiol 72: 655664.
53. Ospina-Tascon G, Neves AP, Occhipinti G, Donadello K, Buchele G, et al. (2010) Effects of fluids on
microvascular perfusion in patients with severe sepsis. Intensive Care Med 36: 949955.

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

20 / 21

Albumin vs Other Fluids in Sepsis Patients: A Meta-Analysis

54. Vincent JL, Gerlach H (2004) Fluid resuscitation in severe sepsis and septic shock: an evidence-based
review. Crit Care Med 32: S451454.
55. Wiedermann CJ, Dunzendorfer S, Gaioni LU, Zaraca F, Joannidis M (2010) Hyperoncotic colloids
and acute kidney injury: a meta-analysis of randomized trials. Crit Care 14: R191.
56. Boyd JH, Forbes J, Nakada TA, Walley KR, Russell JA (2011) Fluid resuscitation in septic shock: a
positive fluid balance and elevated central venous pressure are associated with increased mortality. Crit
Care Med 39: 259265.
57. Murphy CV, Schramm GE, Doherty JA, Reichley RM, Gajic O, et al. (2009) The importance of fluid
management in acute lung injury secondary to septic shock. Chest 136: 102109.
58. Pandey NR, Bian YY, Shou ST (2014) Significance of blood pressure variability in patients with sepsis.
World J Emerg Med 5: 4247.
59. Vincent JL, Navickis RJ, Wilkes MM (2004) Morbidity in hospitalized patients receiving human
albumin: a meta-analysis of randomized, controlled trials. Crit Care Med 32: 20292038.
60. Latour-Perez J (2013) New recommendations for the use of serum albumin in patients with severe
sepsis and septic shock. Crit Care Med 41: e289.
61. Lyu PF, Murphy DJ (2014) Economics of fluid therapy in critically ill patients. Curr Opin Crit Care 20:
402407.
62. Finfer S, Chittock DR, Su SY, Blair D, Foster D, et al. (2009) Intensive versus conventional glucose
control in critically ill patients. N Engl J Med 360: 12831297.
63. Brunkhorst FM, Engel C, Bloos F, Meier-Hellmann A, Ragaller M, et al. (2008) Intensive insulin
therapy and pentastarch resuscitation in severe sepsis. N Engl J Med 358: 125139.

PLOS ONE | DOI:10.1371/journal.pone.0114666 December 4, 2014

21 / 21

Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.

You might also like