NDT Advance Access published January 23, 2014

Nephrol Dial Transplant (2014) 0: 110

doi: 10.1093/ndt/gfu005

Full Review
A critical appraisal of intravenous uids: from the
physiological basis to clinical evidence
David Severs1, Ewout J. Hoorn1 and Maarten B. Rookmaaker2
Department of Internal Medicine Nephrology & Transplantation, Erasmus Medical Center, Rotterdam, The Netherlands and 2Department of

Correspondence and offprint requests to: David Severs; E-mail: d.severs@erasmusmc.nl

A B S T R AC T

INTRODUCTION

Fluid management has been a vital part of routine clinical care

for more than 180 years. The increasing number of available
uids has generated controversy about the optimal choice of resuscitation uid. In this review, we provide a critical overview of
the different uids available, their composition, the relevant
physiology as well as the published evidence on clinical outcomes to guide their use. Commonly used infusion uids
include semisynthetic colloids and crystalloids; the latter comprises both normal saline (NaCl 0.9%) and the more chloride-restricted balanced crystalloids. Despite their signicantly greater
intravascular persistence, semisynthetic colloids have an importantly adverse safety prole and are associated with greater incidence of renal failure and increased mortality; their use should
be restricted. To date, evidence for clinical benets associated
with albumin solutions is generally lacking; its merits in specic
clinical situations are the subject of further investigation. Infusion of normal saline, with its supraphysiological chloride
content, is associated with higher serum chloride concentrations
and metabolic acidosis, as well as renal vasoconstriction in
animal and human models. Infusion of balanced crystalloids is
not linked to such changes. Although data on clinical outcomes
associated with crystalloid infusion are heterogeneous, advantages of balanced salt solutions might include a lower need of
blood products, and lower incidence of renal replacement
therapy, hyperkalaemia and postoperative infections. Taken together, a critical appraisal of the data suggests that balanced salt
solutions deserve consideration as infusates of rst choice.

In 1832, Dr Thomas Latta rst reported on an attempt to intravenously replace uids and salts lost in cholera sufferers. He
made up a solution containing two to three drachms of
muriate of soda and two scruples of the subcarbonate of soda
in six pints of water. Faced with failure of earlier attempts at
oral rehydration, and with no precedent to direct him, he
proceeded to throw the uid immediately into the circulation
[1]. This highly successful intervention led to a number of
important experimentsin 1834, Dr John Mackintosh rst
used intravenous albumin [2], Ringers solution was introduced in 1876 by Dr Sydney Ringer and a modication was
made by Dr Alexis Hartmann in 1932 to include lactate [3].
Given impetus by a better understanding of the principles
of membrane permeability, uid tonicity and osmosis at the
time, it appears that the in vitro studies of the Dutch physiologist Hartog Jakob Hamburger in the 1890s led to the acceptance of NaCl 0.9% as isotonic to human blood. Noting that
erythrocytes were least likely to lyse in this solution, he called
it indifferent saline [4, 5]. While the use of intravenous
gelatin in animals was reported as early as 1894 [6], and a rst
study in humans was performed in 1915 [7], interest in gelatins
as well as other larger molecules to expand the intravascular
volume was renewed only in the Second World War [8]. Asanguineous infusion uids are most commonly classied as crystalloids (small electrolytes in water) or colloids (larger
molecules that are meant to remain in circulation for a longer
time). Crystalloids include 0.9% NaCl (normal saline) as well
as buffered and more balanced solutions such as lactated
Ringers solution.

Keywords: chloride, colloids, crystalloids, saline

The Author 2014. Published by Oxford University Press

on behalf of ERA-EDTA. All rights reserved.

Downloaded from http://ndt.oxfordjournals.org/ by guest on January 27, 2015

Nephrology & Hypertension, University Medical Center Utrecht, Utrecht, The Netherlands

REVIEW CRITERIA
We performed a search for original full-text English language
papers in Medline, EMBASE and the Cochrane library, using
Table 1. Summary of haemodynamic effects,
disadvantages of commonly used infusion uids

advantages

and

Colloids

Normal saline

Balanced
crystalloids

Intravascular
persistence
Advantages

Long

Short

Short

Larger
haemodynamic
effects

Extensive
experience

Disadvantages

Nephrotoxicity,
greater mortality
High

Hyperchloraemic
acidosis
Low

Most closely
resemble ionic
composition of
blood
Slightly
hypotonic
Low

Costs

D. Severs et al.

FULL REVIEW

the search terms crystalloid, saline, Ringer, Hartmann,

Plasma-Lyte, Sterofundin, colloid and their synonyms,
alone and in combination (last accessed 21 November 2013).
We also selected further relevant papers by following reference
lists of identied articles.

I N F U S I O N F L U I D C H A R AC T E R I S T I C S
A N D P H Y S I O LO G Y
The ionic composition of infusion uids is important because
it directly affects biological processes like signal transduction
and coagulation. The choice of infusate can be guided by the
intention to maintain or deliberately change the composition
of body uids. It is of note that clinical laboratories generally
report the electrolyte concentrations in serum. Concentrations
in the aqueous fraction of plasma are 7% higher due to the
presence of proteins and lipids. It follows that if the electrolyte
contents of an infused uid are to be proportionate to the corresponding aqueous fraction of blood plasma and effects on
acidity are to be minimal, it should contain 153 mmol/L Na+,
111 mmol/L Cl, 4.8 mmol/L K+, 2.7 mmol/L Ca2+ and 1.35
mmol/L Mg2+, and be able to maintain a plasma pH of 7.35
7.45 [12]. Surprisingly, however, no such uid is available
(Table 2). A change in plasma Ca2+ concentration, for
example, will change cellular (de)polarization, whereas
changes in acidity directly inuence processes like coagulation
[1317]. This is in line with the profound impairment in
factor VIIa (FVIIa), FVIIa/tissue factor and FXa/FVa activity
with decreases in pH [16]. In vitro and animal models also
suggest increased extracellular acidity may alter the immune
response in multiple ways, ranging from a change in production of proinammatory mediators to impaired leucocyte
chemotaxis and lymphocyte cytotoxicity [18, 19].
The distribution of infused uids between the intracellular
and extracellular compartments is dictated by the osmotic
pressure in the different compartments. The osmotic pressure
is determined by the amount of solutes and their osmotic coefcients, which describe the deviation of the osmotic behaviour
of solutes from their theoretical osmotic behaviour. When
considering the tonicity of an infused uid, it is important to
appreciate the differences between theoretical osmolarity (obtained by addition of all osmotically active molecules to 1 L of
solution) and actual osmolality (mOsmol/kg solvent),
especially in vivo. As sodium and chloride, when dissolved,
only partially dissociate, and as such are only partially osmotically active (osmotic coefcient 0.926), the actual osmolality of
a uid in which they are dissolved is lower than their added
molarities. The osmolality of a uid is derived from the ideal
osmolarity, the osmotic coefcient and the solvent content
(93% water for plasma) and may be directly measured via
freezing point depression. Surprisingly, the measured osmolality of plasma (288 mOsmol/kg H2O) is practically identical to
the calculated osmolarity (291 mOsmol/L) [12, 20]. This is in
stark contrast to the difference between the theoretical osmolarity of NaCl 0.9% (308 mOsmol/L) and its actual osmolality
of 286 mOsmol/kg H2O. It is of note that changes in the osmolality and volume of the blood plasma lead to more limited,

Downloaded from http://ndt.oxfordjournals.org/ by guest on January 27, 2015

Today, uid management is a vital part of routine care for a

wide range of settings. In volume replacement, the goal is to
replete intravascular uid volume in trauma or sepsis, while in
dehydration (contraction of the extracellular uid volume),
the total extracellular space is the target of uid repletion.
A third potential goal of infusion is to restore the composition
of the extracellular uid by altering its electrolyte contents or
acidity, which will not be discussed in this review. Finally,
another aim of uid infusion may be the maintenance of ow
in the distal renal tubules and urinary tract to prevent toxicity
or precipitation of drugs or radiocontrast. Indeed, the question
of the optimal choice of infusion uid for this indication has
led to a number of trials, mostly making a comparison
between colloid and crystalloid infusions or more specically
between infusion of normal saline and sodium bicarbonate
solutions. A more in-depth discussion of this indication is
beyond the scope of this article (for review, see [911]). Of
note, infusion therapy often serves more than one purpose, e.
g. uid repletion and electrolyte correction.
The choice of a particular infusion uid has generated considerable controversy. Arguments are predominantly based on
theoretical and physiological considerations as well as preclinical studies (Table 1). The number of clinical studies is
limited and restricted to the eld of general surgery, kidney
transplantation and intensive care medicine. The aim of this
review is to provide a nephrological readership with an overview of the different uids available for volume resuscitation
and uid repletion. This review will not consider blood products other than albumin. First, we give an introduction on infusion uid characteristics and physiology. We then
summarize the relevant chemical properties of commonly
used infusates and discuss studies that evaluate their physiological effects and potential side effects. We proceed by highlighting the available evidence on outcomes in clinical studies
in general surgery, kidney transplantation and intensive care
units (ICUs). We conclude that the available evidence merits
consideration as infusates of rst choice for balanced salt solutions, although denitive large-scale studies are still needed.

Table 2. Characteristics of commonly available infusion uids

Plasma

Na (mmol/L)
Cl (mmol/L)
K+ (mmol/L)
Ca2+ (mmol/L)
Mg2+ (mmol/L)
Bufferb (mmol/L)
Colloid (g/L)
Oncotic pressure (mmHg)
Osmolarity (mOsmol/L)
Osmolality (mOsmol/kg)

142
103
4.5
2.5
1.25
24
3545
25
291
288

Crystalloids

Colloids

0.9% NaCl

Lactated Ringers

Plasma-Lyte

Sterofundin

Voluven (HES 6%)

Gelofusinea

Dextran 40

154
154
0
0
0
0
0
0
308
286

130
109
4
2.7
0
28
0
0
273
254

140
98
5
0
1.5
50
0
0
295
N/K

140
127
4
2.5
1
29
0
0
304
290

154
154
0
0
0
0
60
36
308
N/A

154
120
0
0
0
0
40
2629
274
N/A

154
154
0
0
0
0
100
168191
308
N/A

HES, hydroxyethyl starch; N/A, not applicable; N/K, not known.

a
Contains 4% gelatin.
b
The buffer in plasma is bicarbonate, in lactated Ringers lactate, in Plasma-Lyte acetate (27 mmol/L) and gluconate (23 mmol/L), in Sterofundin acetate (24 mmol/L) and maleate
(5 mmol/L).

Colloid is the collective noun for all infusion uids that

contain large molecules which are meant to keep the infusate
in the intravascular space for a longer time by exerting an
oncotic pressure [35]. Colloid infusates comprise blood products such as human albumin solution and the more commonly used semisynthetic products: gelatins, dextrans and
hydroxyethyl starch (HES) solutions. Their effect on plasma
volume expansion depends on their oncotic pressure, inuenced directly and indirectly by molecular weight (MW) and
plasma half-life [36].
Various colloids are currently available. Human albumin is
available in iso-oncotic 45% solutions and hyperoncotic 20
25% solutions. The high cost of human albumin has limited
its use. Gelatins are proteins formed by hydrolysis of animalderived connective tissue. Because high-MW gelatin solutions
tend to gel, the average MW of present solutions is limited to
3035 kDa, which reduces their oncotic pressure and intravascular persistence. Dextran 40 (MW 40 kDa) and 70 (MW
70 kDa) are the products of acid hydrolysis and ethanol fractionation of highly branched polysaccharide molecules synthesized by the B512 strain of Leuconostoc mesenteroides
[37]. Finally, HESs are derivatives of maize or potato starch,
amylopectin, in which intravascular amylase-driven degradation of the molecules is slowed by substitution of the
hydroxyl radicals in positions C2, C3 and C6 with hydroxyethyl radicals [38]. This molecular modication renders
HES more prone to accumulation in reticular connective
tissues, such as the spleen, skin, liver and kidneys [3942], and
it may explain the nding of lesions with a histological appearance of osmotic nephrosis in animals subjected to HES infusion [43]. Semisynthetic colloids are considerably cheaper
than human albumin, but still usually at least an order of magnitude more expensive than crystalloids.
The use of albumin solutions in volume replacement in resuscitation of critically ill patients has a number of theoretical
advantages over crystalloids, but its actual clinical benets and
safety prole are uncertain. The addition of human albumin to

Intravenous uids: a critical appraisal

FULL REVIEW

C O L LO I D S

Downloaded from http://ndt.oxfordjournals.org/ by guest on January 27, 2015

yet potentially important changes in erythrocyte volume, and

may have signicant rheological effects [21].
In contrast to small solutes, plasma proteins move across
the capillary wall only to a limited degree, subsequently acting
as effective osmoles, impeding ltration of water into the interstitium. The osmotic pressure generated by plasma proteins is
commonly referred to as the colloid osmotic or oncotic
pressure [22]. The oncotic pressure depends on both the
difference in plasma protein concentrations and the barrier
between the intravascular compartment and the interstitium.
This barrier is predominantly formed by the endothelial glycocalyx layer (EGL) but also by the endothelial basement membrane and extracellular matrix in the interstitial space [23, 24].
The EGL consists of a dense network of proteoglycans associated with various glycosaminoglycans and coats the luminal
side of healthy vascular endothelium [25]. The polyanionic
nature of the glycosaminoglycans makes the EGL semipermeable to anionic macromolecules, such as albumin, depending
on their size and structure [26]. Under physiological conditions, while continuously subject to degradation and reconstitution, its thickness is estimated to be as little as 0.2 m in
the microcirculation and up to 8 m in the large vessels [27].
Several factors that contribute to EGL degradation, leading to
shedding of its components and compromise of its function,
have now been identied. Among them are inammatory
mediators [2832], but also hypervolaemia, such as when
induced by rapid crystalloid infusion in healthy volunteers
[28, 33]. It is of note that the oncotic pressure gradient
between the intravascular and interstitial spaces is insufcient
to counter the hydraulic pressure found in the intravascular
space [24, 34]. Since the subglycocalyx space is nearly protein
free, it is likely the oncotic pressure gradient across the EGL,
and not, as previously assumed, across the endothelium, which
restricts water loss across the vascular wall [24]. Hence, in
summary, the effect of a uid infused in the intravascular compartment on its extravascular counterpart depends on hydrostatic and (colloid) osmotic pressure gradients as well as EGL
function and factors in the endothelial basement membrane
and the extracellular matrix.

D. Severs et al.

FULL REVIEW

gelatin in 0.7% saline), HES (6% HES 130/0.4 in 0.9% saline)

or 0.9% saline alone. HES and gelatin suppressed plasma renin
activity (PRA) more than saline, but this difference did not
reach statistical signicance [57]. Finally, a study reported on a
nurse-delivered algorithm, in which, after cardiac surgery, 237
patients were randomly allocated to receive 250-mL boluses of
either normal saline or a HES solution based on haemodynamic parameters. Although the study did not report total
volumes of infused uids, signicantly more catecholamines
were used in the group assigned to receive normal saline [58].
However, serious concerns have risen over deleterious
effects of semisynthetic colloids on specic systems, such as
haemostasis [59] and renal function [6062]. Signicant
reductions in FVIII and von Willebrand factor are observed
after infusion of dextrans, gelatins and HES, with high-molecular HES and dextran having the greatest effect on haemostasis [59]. Moreover, concerns about nephrotoxicity have
assumed prominence since the introduction of semisynthetic
colloids, owing to a host of evidence published in the last few
decades to suggest greater incidence of renal dysfunction and
renal replacement therapy associated with semisynthetic
colloid infusion. Importantly, evidence of a survival benet of
semisynthetic colloids over crystalloids has been lacking.
Much of the attention has focused on HES. Four well-designed, large randomized controlled trials, published in 2008,
2012 and 2013, compared the use of HES with Ringers acetate
or lactate [63, 64] with normal saline [65, 66], and demonstrated increased risk of death [63] and the use of renal replacement therapy [6366] in the groups assigned to receive HES.
A recent Cochrane review, which also included two of these
studies, found no evidence from randomized controlled trials
that colloids (dextran 70, gelatins, HES, albumin and plasma
protein fraction) are superior to isotonic or hypertonic crystalloids as a treatment for intravascular volume resuscitation in
critically ill patients [67]. Notably, the pooled RR of mortality
with HES versus crystalloids was 1.10 (95% CI, 1.021.19), implying a potential mortality hazard associated with HES [67].
A 2011 Cochrane review found no evidence of superiority of
one particular colloid solution in critically ill and surgical
patients requiring volume replacement, noting, however, a
relative lack of direct comparisons and wide CIs [68]. Furthermore, two recent meta-analyses, including one Cochrane
review, exclusively compared HES with other resuscitation
uids and found a signicantly increased risk of mortality and
acute kidney injury associated with the use of HES [69, 70]. Of
particular interest is the nding that these detrimental effects
only became visible after exclusion of seven pre-1999 trials
initiated by Dr Joachim Boldt, whose subsequent publications
have been retracted because of scientic misconduct [71, 72].
In addition, observational data suggest that gelatins may be nephrotoxic similarly to HES [73]. A very recently published,
large, multi-centre, randomized controlled trial in patients admitted to the ICU and presenting specically with hypovolaemic shock (CRISTAL) allocated 2857 patients to receive either
exclusively colloids (various semisynthetic colloids and
albumin solutions) or crystalloids in an unblinded fashion.
Patients were randomized at the onset of resuscitation. Contrasting with previous studies, investigators found no

Downloaded from http://ndt.oxfordjournals.org/ by guest on January 27, 2015

normal saline, bringing the solution to iso-oncoticity, in

theory expands this uids capacity to provide intravascular
volume expansion. Indeed, in one large randomized controlled
trial comparing resuscitation with albumin or normal saline,
40% greater average volumes of saline were infused [44]. Similarly, greater increases in plasma volume and cardiac index
were obtained with infusion of 5% albumin compared with
normal saline in cardiac surgery and sepsis [45, 46]. Moreover,
pre-clinical studies suggest albumin may reduce vascular
endothelial leucocyte adhesion [4749].
A rst meta-analysis by the Cochrane collaboration was
published in 1998, comparing the use of albumin or plasma
protein fraction with crystalloids in patients with hypovolaemia,
burns or hypoalbuminaemia [50]. Its most important result, a
pooled relative risk (RR) of death of 1.68 [95% condence interval (CI), 1.262.23; P < 0.01] associated with albumin,
caused a large uproar and a signicant drop in the clinical use
of albumin, but the investigators also attracted staunch criticism pertaining to limitations of included studies, review methodology and interpretation. The last decade saw publications of
the Saline versus Albumin Fluid Evaluation (SAFE) study, in
which 6997 critically ill, ICU-admitted adults were randomized
between resuscitation with 4% albumin or normal saline [44],
and the Fluid Expansion as Supportive Therapy (FEAST)
study, in which 3141 febrile, critically ill children were randomly assigned to receive boluses of 5% albumin, normal
saline or no bolus of resuscitation uid [51]. While neither the
FEAST nor the SAFE study detected differences in clinical outcomes between groups receiving normal saline or albumin, a
post hoc subgroup analyses from the SAFE study showed
higher mortality associated with albumin in 460 patients with
traumatic brain injury (RR, 1.63; 95% CI, 1.172.26; P = 0.003)
[52]. In contrast, another post hoc subgroup analysis of the
SAFE study indicated a potential decrease in the adjusted risk
of death with albumin in severely septic patients [odds ratio
(OR), 0.71; 95% CI, 0.520.97; P = 0.03] [53]. A 2011 update of
the above Cochrane review, which relied heavily on the SAFE
study, no longer detected mortality differences between
albumin and crystalloid solutions (RR, 1.02; 95% CI, 0.921.13,
P = 0.35) [54]. Three trials addressing the use of albumin solutions specically in early septic shock are under way
(NCT00707122, NCT00327704 and NCT00819416).
Several clinical studies have conrmed that, in analogy to
albumin solutions, semisynthetic colloids provide more intravascular uid repletion than crystalloids. In a randomized
clinical trial in 67 patients who had undergone cardiac or
major vascular surgery, patients receiving colloid solutions (an
average of 1600 mL of gelatin 4%, HES 6% or albumin 5%)
had signicantly greater increases, during the 90 min of infusion, in plasma volume (19 versus 3%) and cardiac index (22
versus 13%) compared with patients receiving normal saline
[55]. Similarly, in a randomized trial with 25 septic patients
and 24 non-septic hypovolaemic patients, uid loading with
semisynthetic colloids resulted in greater increases in cardiac
lling, output and stroke work than loading with normal
saline did [56]. The increased intravascular volume repletion
was also demonstrated in a study with 10 healthy volunteers
who received infusion with 1 L of gelatin (4% succinylated

difference in 28-day mortality (RR, 0.96, 95% CI, 0.881.04; P

= 0.26), whereas the secondary outcome, 90-day mortality, was
lower in the group receiving colloids (RR, 0.92, 95% CI, 0.86
0.99; P = 0.03) [74]. The latter nding should be interpreted
with caution, since the CI approaches 1. No difference in the
risk of receiving renal replacement therapy was found. Unfortunately, the pragmatic design of this study leaves questions on
the contribution of individual uid types to these outcomes,
and especially the distinction between albumin and semisynthetic colloids.
Taken together, growing concerns about safety, especially
concerning semisynthetic colloids, and the lack of consistent
clinical superiority of colloids, suggest that their application
should be minimized and the less expensive crystalloids
should be the infusates of choice, while we await clarication
from ongoing trials on the potential benet of albumin infusion specically in septic shock.

Normal saline
Normal saline (NS, 0.9% or 154 mM NaCl) is the most
frequently prescribed crystalloid in clinical practice [75], yet
remarkably few studies address the time scale and extent of its
effects in normal subjects on haemodynamic parameters,
blood dilution and excretion. After infusion, normal saline is
rapidly distributed between the compartments of the extracellular space but remains in the body for a long time.
In normovolaemic subjects, expansion of the intravascular
volume persists for as long as 6 h after infusion of normal
saline, even though most of the infused volume is found in the
interstitial compartment. A study in 28 healthy volunteers
found that 30 min after completion of a 20-min infusion of
1030 mL/kg, 64% of the infused volume had diffused into the
interstitial compartment [76]. Studies in pre-surgical patients
and healthy volunteers receiving a 2-L normal saline infusion
over the course of 1 or 2 h reported intravascular retention of
the infused uid of 1824% (calculations based on haematocrit values) [7779]. After 6 h, 60% of the uid volume still remained in the body, an estimated 13% in the intravascular
space [78, 79]. Finally, it is of note that during anaesthesia and
surgery, distribution and elimination of infused crystalloids
are delayed when compared with healthy volunteers [80].
Owing to the non-physiological ion content and the lack of
buffering capacity, normal saline infusion can be complicated
by metabolic acidosis. The determination of saline-induced
acidosis is subject to vivid debate [81]. The most commonly
used method for interpreting acid-base disturbances is the
HendersonHasselbalch equation, in which the dependent
variable pH is calculated from the variables PaCO2 and plasma

Intravenous uids: a critical appraisal

FULL REVIEW

The term crystalloid is typically used to refer to solutions in

water of small inorganic ions and small organic molecules.
Purely NaCl-based solutions are most commonly used.
Alternatives include various concentrations of potassium,
calcium, magnesium and organic anions such as lactate to
more closely resemble the ionic pattern of blood plasma.

Downloaded from http://ndt.oxfordjournals.org/ by guest on January 27, 2015

C R Y S TA L LO I D S

bicarbonate. The acidosis resulting from normal saline infusion is often interpreted as dilutional, i.e. the result of decreased bicarbonate concentrations relative to stable PaCO2
[81]. The model assumes changes in the volume of distribution of bicarbonate with changes in pH [82, 83]. Advocates
of the alternative Stewart approach, introduced in 1983 [84],
argue that the essential change is not dilution of bicarbonate

(HCO
3 ) but infusion of the strong ion Cl . This explanation
is based on the concept that independent variablesPaCO2,
the strong ion difference and the total concentration of nonvolatile weak acidsinuence the dependent variables pH and
bicarbonate concentration [85]. Indirect effects of chloride
loading might also be responsible for the acidosis. Intraluminal chloride is needed for pendrin-mediated bicarbonate
excretion [8688]. We therefore speculate that excessive chloride loading might be involved in disproportionate urinary
bicarbonate loss. However, it should be noted that in animal
models, pendrin protein expression is rapidly downregulated
in response to induction of acidosis [89]. In addition, the presence of chloride ions in an infusion uid is instrumental in
this uids capacity to suppress renin release, and subsequently
its capacity to suppress aldosterone availability and bicarbonate retention [90]. The latter was demonstrated in experiments
in healthy, sodium-restricted volunteers, in whom no suppression of PRA was found after infusion of 750 mL 1.4%
NaHCO3, whereas infusion of 500 mL 0.9% NaCl led to a 50%
decrease in PRA [91]. Regardless of the explanatory model,
reductions in blood pH on infusion of saline, not in the more
balanced Ringers solution, were demonstrated in healthy
human volunteers who received 50 mL/kg over 1 h (average
pH reduction 0.04 0.04) [92], and, as described later in this
review, in patients undergoing gynaecological surgery (average
pH reduction 0.13) [93] and pancreatico-hepatobiliary surgery
[94].
As normal saline contains a supraphysiological concentration of chloride [12], its infusion increases plasma chloride
concentrations [93], which is associated with renal vasoconstriction and decreased glomerular ltration rate (GFR). In an
animal model, intrarenal infusion of chloride-rich solutions to
denervated kidneys resulted in decreases in renal blood ow
(RBF) and GFR compared with low-chloride solutions of
equal tonicity [95, 96]. Two studies with healthy human volunteers consistently showed a longer time to rst micturition
after infusion of normal saline compared with a low-chloride
solution [79, 92]. Changes in plasma osmolality must be taken
into account in these studies, where the low-chloride solution
had a lower osmolality than normal saline, with a potential
effect on the release of antidiuretic hormone. Strengthening
the hypothesis of renal vasoconstriction by chloride infusion,
however, a study in healthy volunteers showed signicant
reductions in renal artery ow velocity and renal cortical tissue
perfusion during and after infusion of normal saline, but not
with a low-chloride solution (Plasma-Lyte 148) [97]. Such increases in renal cortical tissue perfusion were also seen when
healthy volunteers received 1-L infusions of 6% HES suspended in a low-chloride solution compared with 6% HES
suspended in normal saline [98]. The effect may be due to
luminal chloride concentrations in the cortical thick ascending

Balanced crystalloids
Some of the above effects can be prevented by using a
solution that more closely mimics the ionic make-up of
the aqueous fraction of plasma. Acidosis can be avoided by
inclusion of bicarbonate or metabolizable anions, such as
acetate, lactate, malate and citrate. Examples of more balanced
crystalloid solutions are lactated Ringers solution, Hartmanns
solution, featuring slightly different ionic concentrations,
Plasma-Lyte and Sterofundin. Their characteristics are summarized in Table 2. As described above, the use of infusates
with metabolizable anions instead of chloride has been shown
not to increase plasma acidity [92, 93]. In addition, balanced
salt solutions did not reduce renal artery ow and renal cortical
perfusion seen after infusion of normal saline (see above) [97].
Unfortunately, some of the most commonly used balanced
solutions (like lactated Ringers solution) are neither isotonic
nor precisely balanced. With an osmolarity of 273 mOsmol/L
and a measured osmolality of 254 mOsmol/kg, infused lactated Ringers solution leads to a small decrease in plasma osmolality [79, 92]. Theoretical concerns with slightly hypotonic
infusion uids include a potential increase in brain water
[102] and effects on diuresis. Interestingly, Hartmanns solution (a slightly modied form of Ringers solution) when
compared with normal saline, doubled urine output in the 6 h
after infusion, and subsequently had a shorter persistence in
the body [79]. Such effects may be relevant considering the
goals of infusion therapy. Commercial initiatives, such as
Plasma-Lyte 148 and Sterofundin, have attempted to address
concerns over differences in tonicity and ionic composition
between plasma and the infused uid. Unfortunately, there are
no published studies that address potential differences in kinetics and the effect on plasma osmolality between the more
balanced solutions.
CLINICAL STUDIES
In light of concerns raised around colloids, their use should be
restricted. For (intravascular) volume repletion, crystalloids
deserve consideration as infusates of rst choice. Below, we
compare the use of normal saline with balanced salt solutions
in clinical outcome studies.
Surgery
The risk of hyperchloraemic acidosis in patients receiving
normal saline was demonstrated in surgical patients. Subjects
undergoing intra-abdominal gynaecological surgery were randomly assigned to receive normal saline or lactated Ringers,
with an average volume of 6 L over 2 h. Predictably, hyperchloraemia and metabolic acidosis were more prevalent in the
normal saline group (average pH 7.28 versus 7.41, chloride
115 versus 107 mmol/L) [93]. Such metabolic derangements

D. Severs et al.

associated with the use of normal saline in comparison with

Plasma-Lyte were also seen in a study including 30 patients
undergoing major hepatobiliary or pancreatic surgery [94]. In
a randomized controlled trial in 46 trauma patients, resuscitation with Plasma-Lyte resulted in a quicker resolution of acidbase disturbances compared with normal saline [103].
Although the data on clinical outcomes in patients undergoing surgery are heterogeneous, advantages of balanced salt
solutions might include lower need of blood products, lower
incidence of renal replacement therapy and postoperative infections. A 2012 Cochrane review, using pooled data, found no
inuence of the choice of buffered or non-buffered infusion
uids in the perioperative period on mortality, renal function
and blood loss [104]. In the non-buffered group, however, the
average volume of platelet concentrates transfused was 242%
greater (95% CI, 24.61848.77; P = 0.02) [104]. Unfortunately,
the trials included in this review were highly heterogeneous
with regard to study groups, outcomes reported and administered infusion uids, which included hypertonic and colloid
solutions. In a recent large observational study, outcomes of
30 994 adult patients undergoing major abdominal surgery
who received normal saline were compared with 926 patients
who received only a non-calcium-containing balanced crystalloid solution (Plasma-Lyte) [105]. Unadjusted in-hospital
mortality was far greater in the saline group than in the balanced crystalloid group (5.6 versus 2.9%). After propensity
matching, mortality in the saline group remained higher, but
the difference lost its statistical signicance. However, treatment with balanced crystalloids was associated with fewer
complications (OR, 0.79, 95% CI, 0.660.97; P < 0.05). Recipients of normal saline infusions were 4.8 times more likely to
require haemodialysis (P < 0.001), and had greater blood
transfusion requirements and more frequent postoperative infections [105]. Another double-blind randomized trial in 66
patients undergoing abdominal aortic reconstructive surgery
found no statistically signicant difference in blood loss, but
an increased need for fresh frozen plasma and platelet transfusions in the group that received normal saline compared with
the lactated Ringers group [106].
Kidney transplantation
Balanced salt solutions appear to reduce the incidence of
metabolic acidosis and hyperkalaemia after kidney transplantation, but no differences in transplant outcome have been reported. In one Turkish double-blind trial, 90 recipients of
kidney transplants of living related donors were randomized
to receive either normal saline, lactated Ringers or PlasmaLyte. Patients receiving normal saline had a signicant decrease in pH and increase in serum Cl, which was not seen in
patients receiving lactated Ringers or Plasma-Lyte. No signicant differences in renal function were seen, although the 24-h
urine output in the rst three postoperative days was signicantly larger in the group receiving normal saline, which is in
contrast with the expected effect of the increased chloride load
in the normal saline group [107]. In an Iranian double-blind
randomized trial, 52 adults undergoing kidney transplantation
were either prescribed normal saline or lactated Ringers.
Mean serum potassium values were lower in the group

Downloaded from http://ndt.oxfordjournals.org/ by guest on January 27, 2015

FULL REVIEW

limb of the loop of Henle being a rate-limiting step in tubuloglomerular feedback [95, 99101], a mechanism that induces
alterations in the GFR by inuencing afferent arteriolar vasoconstriction. Finally, chloride has been shown to be actively involved in the suppression of renin release, as described above.

Miscellaneous
A small number of studies compared the use of balanced
crystalloids with saline in resuscitation in diabetic ketoacidosis
and choleriform diarrhoea, and found faster resolution of

CONCLUSIONS
We have reviewed the available experimental and clinical data
on colloid and crystalloid infusates. In view of an overall
lack of survival benet with albumin, recent evidence on
adverse safety outcomes associated with semisynthetic colloids, as well as the effects of the retraction of a large body of
work by Dr Joachim Boldt due to integrity concerns, the use
of colloid infusates is to be limited. Therefore, we have focused
on balanced salt solutions and normal saline.
Infusion of normal saline is clearly associated with hyperchloraemia and metabolic acidosis [93, 94], and such derangements can be prevented by using balanced crystalloids [93,
94]. Studies in surgical and ICU patients reported a signicantly lower incidence of renal replacement therapy in groups
receiving balanced crystalloids compared with those receiving
normal saline [105, 114]. Interestingly, one study involving
kidney transplant recipients found that correction of acidosis
was associated with signicantly lower creatinine levels and increased diuresis [110]. One explanation for this phenomenon
is that by inducing renal vasoconstriction, hyperchloraemia
appears to reduce RBF and GFR [9597]. Acidosis leads to a
dysfunction of various cellular and extracellular mechanisms,
most notably coagulation [1317]. This may explain the increased need for blood products in surgical patients receiving
normal saline [104, 105].
A shortcoming in the literature is the absence of a direct
comparison between the various balanced crystalloids and the
relatively small study participant numbers. Furthermore, in
the largest meta-analysis on this subject to date, relatively heterogeneous data on various salt solutions were pooled to
compare balanced with unbalanced crystalloids. Even though
the documented negative safety outcomes pertaining to the
use of normal saline seem relatively persistent across studies,
the heterogeneity of studies leaves questions on the size of the
mentioned effects in clinical scenarios, especially mortality,
unanswered.
In conclusion, there is a growing body of evidence that balanced crystalloids are preferred over normal saline in clinical
practice and that the positive effects of these solutions are not
restricted to one specic application only. In our opinion,

Intravenous uids: a critical appraisal

FULL REVIEW

Critical care medicine

Recently, a large prospective, non-randomized, before-andafter study in a tertiary ICU was published [113]. Investigators
allowed intravenous uid administration as usual in patients
admitted to the ICU during a 6-month control period. During
the 6-month intervention period, which was timed to start at
the same season a year later, the use of chloride-rich uids was
restricted to specic conditions. Instead, patients received
Hartmanns, Plasma-Lyte or, less frequently, human albumin.
Biochemical effects of the policy change included fewer episodes of severe acidosis, more episodes of metabolic alkalosis,
slightly higher lactate levels and no differences in sodium or
potassium concentrations [113]. Importantly, in a separate
publication the investigators reported a signicantly better
kidney function during ICU stay in the chloride-restricted
group as well as signicantly less renal injury and failure according to the RIFLE criteria (OR, 0.52, 95% CI, 0.370.75;
P < 0.001), and subsequently, fewer episodes of renal replacement therapy (OR, 0.52, 95% CI, 0.330.81; P = 0.004) [114].
Interestingly, no signicant difference in mortality was seen
between the chloride-liberal and chloride-restricted groups
[114]. The latter nding is difcult to reconcile with the 50%
reduction in renal replacement therapy in the chloriderestricted group.

acidosis with the use of balanced crystalloids. One study in

diabetics presenting with ketoacidosis randomly allocated 45
patients to groups resuscitated either with Plasma-Lyte or
normal saline, and found that postresuscitation serum chloride was signicantly higher (111 [110112] versus 105 [103
108]) and bicarbonate signicantly lower (17 [1518] versus
20 [1821]) in the normal saline group compared with the
Plasma-Lyte group, respectively [115]. A retrospective analysis
of 23 patients treated for diabetic ketoacidosis with either
normal saline or Plasma-Lyte by another group yielded comparable results [116]. Finally, in a Peruvian study, 40 severely
dehydrated patients presenting with choleriform diarrhea received infusions of either lactated Ringers or normal saline,
and faster resolution of acidosis was attained in the group receiving lactated Ringers [117].

Downloaded from http://ndt.oxfordjournals.org/ by guest on January 27, 2015

receiving lactated Ringers, despite the presence of potassium

in this infusion uid. No statistically signicant differences in
urine output or renal function were seen [108]. In a Korean
study in 60 kidney transplant recipients, the use of PlasmaLyte was associated with lower plasma Cl concentrations and
lower incidence of metabolic acidosis, and no differences in
postoperative creatinine values or urine output [109]. Finally,
another double-blinded randomized trial in 51 renal transplant recipients was suspended prematurely after a planned
interim analysis showed a greater incidence of hyperkalaemia
and metabolic acidosis in patients administered normal saline
compared with the group receiving lactated Ringers. No
difference in the primary outcome, the serum creatinine concentration on the third postoperative day, was seen [110].
Eight patients (31%) in the group receiving normal saline were
prescribed NaHCO3 in an attempt to correct acidosis. Interestingly, those patients treated for acidosis had signicantly
higher 4- and 48-h postoperative urine output, and signicantly lower 24-h and 1-week postoperative creatinine values
compared with the patients who received normal saline but no
correction for acidosis [110].
Published evidence on the optimal uid choice in preconditioning kidney donors revolves around a very limited number
of trials, in which comparisons usually involve a colloid. These
studies have been reviewed elsewhere [111, 112]. There is no
clinical evidence to guide the choice of a particular crystalloid
in healthy donors in the preoperative and perioperative
period.

balanced crystalloids therefore deserve consideration as rst

choice infusates. Obviously, individual situations might
require different infusion uids. Knowledge of the physiological characteristics of the different infusion uids is important
in guiding this choice. However, the quality of the evidence
available leaves room for a large, well-designed randomized
controlled trial.

C O N F L I C T O F I N T E R E S T S TAT E M E N T
All authors conrm they have no conicts of interest with
regard to this submission.

1. Latta TA. Malignant cholera. Documents communicated by the Central

Board of Health, London, relative to the treatment of cholera by the
copious injection of aqueous and saline uids into the veins. Lancet
1832; 18: 274280
2. Barsoum N, Kleeman C. Now and then, the history of parenteral uid
administration. Am J Nephrol 2002; 22: 284289
3. Hartmann AF, Senn MJ. Studies in the metabolism of sodium
r-lactate. I. Response of normal human subjects to the intravenous injection of sodium r-lactate. J Clin Invest 1932; 11: 327335
4. Awad S, Allison SP, Lobo DN. The history of 0.9% saline. Clin Nutr
2008; 27: 179188
5. Lazarus-Barlow WS. On the initial rate of osmosis of blood-serum with
reference to the composition of physiological saline solution in
mammals. J Physiol. 1896; 20: 145157
6. Czerny A. Versuche ber Bluteindickung und ihre Folgen. Archiv fr Experimentelle Pathologie und Pharmakologie 1894; 34: 268280
7. Hogan JJ. The intravenous use of colloidal (gelatin) solutions in shock.
J Am Med Assoc 1915; LXIV: 721726
8. Saddler JM, Horsey PJ. The new generation gelatins. A review of their
history, manufacture and properties. Anaesthesia 1987; 42: 9981004
9. Barrett BJ, Parfrey PS. Clinical practice. Preventing nephropathy induced
by contrast medium. N Engl J Med 2006; 354: 379386
10. Hoste EA, De Waele JJ, Gevaert SA et al. Sodium bicarbonate for prevention of contrast-induced acute kidney injury: a systematic review and
meta-analysis. Nephrol Dial Transplant. 2010; 25: 747758
11. Pannu N, Wiebe N, Tonelli M, Alberta Kidney Disease Network. Prophylaxis strategies for contrast-induced nephropathy. J Am Med Assoc
2006; 295: 27652779
12. Anonymous. Case records of the Massachusetts General Hospital.
Weekly clinicopathological exercises. Normal reference laboratory
values. N Engl J Med 1992; 327: 718724
13. Crowell JW, Houston B. Effect of acidity on blood coagulation. Am J
Physiol 1961; 201: 379382
14. Dunn EL, Moore EE, Breslich DJ et al. Acidosis-induced coagulopathy.
Surg Forum 1979; 30: 471473
15. Cosgriff N, Moore EE, Sauaia A et al. Predicting life-threatening coagulopathy in the massively transfused trauma patient: hypothermia and
acidoses revisited. J Trauma 1997; 42: 857861; discussion 861862.
16. Meng ZH, Wolberg AS, Monroe DM, 3rd et al. The effect of temperature
and pH on the activity of factor VIIa: implications for the efcacy of
high-dose factor VIIa in hypothermic and acidotic patients. J Trauma
2003; 55: 886891
17. Engstrom M, Schott U, Romner B et al. Acidosis impairs the coagulation:
a thromboelastographic study. J Trauma 2006; 61: 624628
18. Kellum JA, Song M, Li J. Science review: extracellular acidosis and the
immune response: clinical and physiologic implications. Crit Care 2004;
8: 331336

D. Severs et al.

Downloaded from http://ndt.oxfordjournals.org/ by guest on January 27, 2015

FULL REVIEW

REFERENCES

19. Lardner A. The effects of extracellular pH on immune function. J Leukoc

Biol 2001; 69: 522530
20. Greger R. Units used in physiology and their denition. In: Greger R,
Winhorst U (eds). Comprehensive Human Physiology From Cellular
Mechanisms to Integration. Berlin: Springer, 1996, pp. 2421
21. Stuart J. Erythrocyte rheology. J Clin Pathol 1985; 38: 965977
22. Weisberg HF. Osmotic pressure of the serum proteins. Ann Clin Lab Sci
1978; 8: 155164
23. Woodcock TE, Woodcock TM. Revised Starling equation and the glycocalyx model of transvascular uid exchange: an improved paradigm for
prescribing intravenous uid therapy. Br J Anaesth 2012; 108: 384394
24. Adamson RH, Lenz JF, Zhang X et al. Oncotic pressures opposing ltration across non-fenestrated rat microvessels. J Physiol 2004; 557:
889907
25. Pries AR, Secomb TW, Gaehtgens P. The endothelial surface layer. Pugers Arch 2000; 440: 653666
26. Vink H, Duling BR. Capillary endothelial surface layer selectively
reduces plasma solute distribution volume. Am J Physiol Heart Circ
Physiol 2000; 278: H285H289
27. Reitsma S, Slaaf DW, Vink H et al. The endothelial glycocalyx: composition, functions, and visualization. Pugers Arch 2007; 454: 345359
28. Berg S, Golster M, Lisander B. Albumin extravasation and tissue washout of hyaluronan after plasma volume expansion with crystalloid or
hypooncotic colloid solutions. Acta Anaesthesiol Scand 2002; 46:
166172
29. Devaraj S, Yun JM, Adamson G et al. C-reactive protein impairs the
endothelial glycocalyx resulting in endothelial dysfunction. Cardiovasc
Res 2009; 84: 479484
30. Henry CB, Duling BR. TNF-alpha increases entry of macromolecules
into luminal endothelial cell glycocalyx. Am J Physiol Heart Circ Physiol
2000; 279: H2815H2823
31. Rehm M, Bruegger D, Christ F et al. Shedding of the endothelial glycocalyx in patients undergoing major vascular surgery with global and
regional ischemia. Circulation 2007; 116: 18961906
32. Steppan J, Hofer S, Funke B et al. Sepsis and major abdominal surgery
lead to aking of the endothelial glycocalyx. J Surg Res 2011; 165:
136141
33. Berg S, Engman A, Hesselvik JF et al. Crystalloid infusion increases
plasma hyaluronan. Crit Care Med 1994; 22: 15631567
34. Levick JR. Capillary ltrationabsorption balance reconsidered in light of
dynamic extravascular factors. Exp Physiol 1991; 76: 825857
35. Kaye AD, Kucera IJ. Intravascular uid and electrolyte physiology. In:
Miller RD (ed). Millers Anesthesia. Philadelphia: Churchill Livingstone,
2005, pp. 1763
36. Dubois MJ, Vincent JL. Colloid uids. In: Hahn RG, Prough DS, Svensen
CH (eds). Perioperative Fluid Therapy. New York: Wiley, 2007, pp. 153
37. Atik M. Dextran 40 and dextran 70: a review. Arch Surg 1967; 94:
664672
38. Westphal M, James MF, Kozek-Langenecker S et al. Hydroxyethyl
starches: different productsdifferent effects. Anesthesiology 2009; 111:
187202
39. Christidis C, Mal F, Ramos J et al. Worsening of hepatic dysfunction as a
consequence of repeated hydroxyethylstarch infusions. J Hepatol 2001;
35: 726732
40. Dickenmann M, Oettl T, Mihatsch MJ. Osmotic nephrosis: acute kidney
injury with accumulation of proximal tubular lysosomes due to administration of exogenous solutes. Am J Kidney Dis 2008; 51: 491503
41. Jungheinrich C, Neff TA. Pharmacokinetics of hydroxyethyl starch. Clin
Pharmacokinet 2005; 44: 681699
42. Kamann S, Flaig MJ, Korting HC. Hydroxyethyl starch-induced itch: relevance of light microscopic analysis of semi-thin sections and electron
microscopy. J Dtsch Dermatol Ges 2007; 5: 204208
43. Huter L, Simon TP, Weinmann L et al. Hydroxyethylstarch impairs renal
function and induces interstitial proliferation, macrophage inltration
and tubular damage in an isolated renal perfusion model. Crit Care 2009;
13: R23
44. Finfer S, Bellomo R, Boyce N et al. A comparison of albumin and saline
for uid resuscitation in the intensive care unit. N Engl J Med 2004; 350:
22472256

Intravenous uids: a critical appraisal

FULL REVIEW

69. Mutter TC, Ruth CA, Dart AB. Hydroxyethyl starch (HES) versus other
uid therapies: effects on kidney function. Cochrane Database Syst Rev
2013; 7: CD007594
70. Zarychanski R, Abou-Setta AM, Turgeon AF et al. Association of hydroxyethyl starch administration with mortality and acute kidney injury in
critically ill patients requiring volume resuscitation: a systematic review
and meta-analysis. J Am Med Assoc 2013; 309: 678688
71. Anonymous. Editors-in-Chief statement regarding published clinical
trials conducted without IRB approval by Joachim Boldt. Minerva Anestesiol 2011; 77: 562563
72. Miller DR. Update to readers and authors on ethical and scientic misconduct: retraction of the Boldt articles. Can J Anaesth 2011; 58:
777779, 779781
73. Bayer O, Reinhart K, Sakr Y et al. Renal effects of synthetic colloids and
crystalloids in patients with severe sepsis: a prospective sequential comparison. Crit Care Med 2011; 39: 13351342
74. Annane D, Siami S, Jaber S et al. Effects of uid resuscitation with colloids vs crystalloids on mortality in critically ill patients presenting with
hypovolemic shock: the CRISTAL randomized trial. J Am Med Assoc
2013; 310: 18091817
75. Stoneham MD, Hill EL. Variability in post-operative uid and electrolyte
prescription. Br J Clin Pract 1997; 51: 8284
76. Greeneld RH, Bessen HA, Henneman PL. Effect of crystalloid infusion
on hematocrit and intravascular volume in healthy, nonbleeding subjects.
Ann Emerg Med 1989; 18: 5155
77. Mullins RJ, Garrison RN. Fractional change in blood volume following
normal saline infusion in high-risk patients before noncardiac surgery.
Ann Surg 1989; 209: 651659; discussion 659661
78. Lobo DN, Stanga Z, Simpson JA et al. Dilution and redistribution effects
of rapid 2-litre infusions of 0.9% (w/v) saline and 5% (w/v) dextrose on
haematological parameters and serum biochemistry in normal subjects: a
double-blind crossover study. Clin Sci(Lond) 2001; 101: 173179
79. Reid F, Lobo DN, Williams RN et al. (Ab)normal saline and physiological Hartmanns solution: a randomized double-blind crossover study.
Clin Sci(Lond) 2003; 104: 1724
80. Hahn RG. Volume kinetics for infusion uids. Anesthesiology 2010; 113:
470481
81. Prough DS, White RT. Acidosis associated with perioperative
saline administration: dilution or delusion? Anesthesiology 2000; 93:
11671169
82. Garella S, Dana CL, Chazan JA. Severity of metabolic acidosis as a determinant of bicarbonate requirements. N Engl J Med 1973; 289: 121126
83. Fernandez PC, Cohen RM, Feldman GM. The concept of bicarbonate
distribution space: the crucial role of body buffers. Kidney Int 1989; 36:
747752
84. Stewart PA. Modern quantitative acid-base chemistry. Can J Physiol
Pharmacol 1983; 61: 14441461
85. Kellum JA. Saline-induced hyperchloremic metabolic acidosis. Crit Care
Med 2002; 30: 259261
86. Vallet M, Picard N, Lofng-Cueni D et al. Pendrin regulation in mouse
kidney primarily is chloride-dependent. J Am Soc Nephrol 2006; 17:
21532163
87. Frische S, Kwon TH, Frokiaer J et al. Regulated expression of pendrin in
rat kidney in response to chronic NH4Cl or NaHCO3 loading. Am J
Physiol Renal Physiol 2003; 284: F584F593
88. Galla JH, Bonduris DN, Dumbauld SL et al. Segmental chloride and uid
handling during correction of chloride-depletion alkalosis without
volume expansion in the rat. J Clin Invest 1984; 73: 96106
89. Purkerson JM, Tsuruoka S, Suter DZ et al. Adaptation to metabolic
acidosis and its recovery are associated with changes in anion exchanger
distribution and expression in the cortical collecting duct. Kidney Int
2010; 78: 9931005
90. Karet FE. Mechanisms in hyperkalemic renal tubular acidosis. J Am Soc
Nephrol 2009; 20: 251254
91. Julian BA, Galla JH, Guthrie GP, Jr et al. Renin and aldosterone
responses to short-term NaCl or NaHCO3 loading in man. J Lab Clin
Med 1982; 100: 261268
92. Williams EL, Hildebrand KL, McCormick SA et al. The effect of intravenous lactated Ringers solution versus 0.9% sodium chloride solution on

Downloaded from http://ndt.oxfordjournals.org/ by guest on January 27, 2015

45. Ernest D, Belzberg AS, Dodek PM. Distribution of normal saline and 5%
albumin infusions in cardiac surgical patients. Crit Care Med 2001; 29:
22992302
46. Ernest D, Belzberg AS, Dodek PM. Distribution of normal saline and 5%
albumin infusions in septic patients. Crit Care Med 1999; 27: 4650
47. Nathan C, Xie QW, Halbwachs-Mecarelli L et al. Albumin inhibits neutrophil spreading and hydrogen peroxide release by blocking the shedding of CD43 (sialophorin, leukosialin). J Cell Biol 1993; 122: 243256
48. Bignold LP, Rogers SD, Harkin DG. Effects of plasma proteins on the
adhesion, spreading, polarization in suspension, random motility and
chemotaxis of neutrophil leukocytes on polycarbonate (Nuclepore) ltration membrane. Eur J Cell Biol 1990; 53: 2734
49. Horstick G, Lauterbach M, Kempf T et al. Early albumin infusion improves global and local hemodynamics and reduces inammatory
response in hemorrhagic shock. Crit Care Med 2002; 30: 851855
50. Cochrane Injuries Group Albumin Reviewers. Human albumin administration in critically ill patients: systematic review of randomised controlled trials. Br Med J 1998; 317: 235240
51. Maitland K, Kiguli S, Opoka RO et al. Mortality after uid bolus in
African children with severe infection. N Engl J Med 2011; 364:
24832495
52. SAFE Study Investigators, Australian and New Zealand Intensive Care
Society Clinical Trials Group, Australian Red Cross Blood Service et al.
Saline or albumin for uid resuscitation in patients with traumatic brain
injury. N Engl J Med 2007; 357: 874884
53. SAFE Study InvestigatorsFinfer S, McEvoy S et al. Impact of albumin
compared to saline on organ function and mortality of patients with
severe sepsis. Intensive Care Med 2011; 37: 8696
54. Roberts I, Blackhall K, Alderson P et al. Human albumin solution for resuscitation and volume expansion in critically ill patients. Cochrane Database Syst Rev 2011; 11: CD001208
55. Verheij J, van Lingen A, Beishuizen A et al. Cardiac response is greater
for colloid than saline uid loading after cardiac or vascular surgery. Intensive Care Med 2006; 32: 10301038
56. Trof RJ, Sukul SP, Twisk JW et al. Greater cardiac response of colloid
than saline uid loading in septic and non-septic critically ill patients
with clinical hypovolaemia. Intensive Care Med 2010; 36: 697701
57. Lobo DN, Stanga Z, Aloysius MM et al. Effect of volume loading with 1
liter intravenous infusions of 0.9% saline, 4% succinylated gelatine (Gelofusine) and 6% hydroxyethyl starch (Voluven) on blood volume and
endocrine responses: a randomized, three-way crossover study in healthy
volunteers. Crit Care Med 2010; 38: 464470
58. Magder S, Potter BJ, Varennes BD et al. Canadian Critical Care Trials
Group. Fluids after cardiac surgery: a pilot study of the use of colloids
versus crystalloids. Crit Care Med 2010; 38: 21172124
59. de Jonge E, Levi M. Effects of different plasma substitutes on blood
coagulation: a comparative review. Crit Care Med 2001; 29: 12611267
60. Schortgen F, Lacherade JC, Bruneel F et al. Effects of hydroxyethylstarch
and gelatin on renal function in severe sepsis: a multicentre randomised
study. Lancet 2001; 357: 911916
61. Schortgen F, Girou E, Deye N et al. CRYCO Study Group. The risk
associated with hyperoncotic colloids in patients with shock. Intensive
Care Med 2008; 34: 21572168
62. Rioux JP, Lessard M, De Bortoli B et al. Pentastarch 10% (250 kDa/0.45)
is an independent risk factor of acute kidney injury following cardiac
surgery. Crit Care Med 2009; 37: 12931298
63. Perner A, Haase N, Guttormsen AB et al. Hydroxyethyl starch 130/0.42
versus Ringers acetate in severe sepsis. N Engl J Med 2012; 367: 124134
64. Brunkhorst FM, Engel C, Bloos F et al. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med 2008; 358: 125139
65. Myburgh JA, Finfer S, Bellomo R et al. Hydroxyethyl starch or saline for
uid resuscitation in intensive care. N Engl J Med 2012; 367: 19011911
66. Phillips DP, Kaynar AM, Kellum JA et al. Crystalloids vs colloids: KO at
the twelfth round? Crit Care 2013; 17: 319
67. Perel P, Roberts I, Ker K. Colloids versus crystalloids for uid resuscitation in critically ill patients. Cochrane Database Syst Rev 2013; 2:
CD000567
68. Kruer RM, Ensor CR. Colloids in the intensive care unit. Am J Health
Syst Pharm 2012; 69: 16351642

93.

94.
95.
96.

97.

98.

99.

101.

FULL REVIEW

102.
103.

104.

105.

D. Severs et al.

106. Waters JH, Gottlieb A, Schoenwald P et al. Normal saline versus lactated
Ringers solution for intraoperative uid management in patients undergoing abdominal aortic aneurysm repair: an outcome study. Anesth
Analg 2001; 93: 817822
107. Hadimioglu N, Saadawy I, Saglam T et al. The effect of different crystalloid solutions on acid-base balance and early kidney function after
kidney transplantation. Anesth Analg 2008; 107: 264269
108. Khajavi MR, Etezadi F, Moharari RS et al. Effects of normal saline vs lactated ringers during renal transplantation. Ren Fail 2008; 30: 535539
109. Kim SY, Huh KH, Lee JR et al. Comparison of the effects of normal
saline versus plasmalyte on acid-base balance during living donor kidney
transplantation using the Stewart and base excess methods. Transplant
Proc 2013; 45: 21912196
110. OMalley CM, Frumento RJ, Hardy MA et al. A randomized, doubleblind comparison of lactated Ringers solution and 0.9% NaCl during
renal transplantation. Anesth Analg 2005; 100: 15181524, table of
contents
111. Dikdan GS, Mora-Esteves C, Koneru B. Review of randomized clinical
trials of donor management and organ preservation in deceased donors:
opportunities and issues. Transplantation 2012; 94: 425441
112. Rech TH, Moraes RB, Crispim D et al. Management of the brain-dead
organ donor: a systematic review and meta-analysis. Transplantation
2013; 95: 966974
113. Yunos NM, Kim IB, Bellomo R et al. The biochemical effects of restricting chloride-rich uids in intensive care. Crit Care Med 2011; 39:
24192424
114. Yunos NM, Bellomo R, Hegarty C et al. Association between a chlorideliberal vs chloride-restrictive intravenous uid administration strategy
and kidney injury in critically ill adults. J Am Med Assoc 2012; 308:
15661572
115. Mahler SA, Conrad SA, Wang H et al. Resuscitation with balanced electrolyte solution prevents hyperchloremic metabolic acidosis in patients
with diabetic ketoacidosis. Am J Emerg Med 2011; 29: 670674
116. Chua HR, Venkatesh B, Stachowski E et al. Plasma-Lyte 148 vs 0.9%
saline for uid resuscitation in diabetic ketoacidosis. J Crit Care 2012; 27:
138145
117. Cieza JA, Hinostroza J, Huapaya JA et al. Sodium chloride 0.9% versus
Lactated Ringer in the management of severely dehydrated patients with
choleriform diarrhoea. J Infect Dev Ctries 2013; 7: 528532
Received for publication: 14.10.2013; Accepted in revised form: 3.1.2014

10

Downloaded from http://ndt.oxfordjournals.org/ by guest on January 27, 2015

100.

serum osmolality in human volunteers. Anesth Analg 1999; 88:

9991003
Scheingraber S, Rehm M, Sehmisch C et al. Rapid saline infusion produces hyperchloremic acidosis in patients undergoing gynecologic
surgery. Anesthesiology 1999; 90: 12651270
McFarlane C, Lee A. A comparison of Plasmalyte 148 and 0.9% saline
for intra-operative uid replacement. Anaesthesia 1994; 49: 779781
Wilcox CS. Regulation of renal blood ow by plasma chloride. J Clin
Invest 1983; 71: 726735
Quilley CP, Lin YS, McGiff JC. Chloride anion concentration as a determinant of renal vascular responsiveness to vasoconstrictor agents. Br J
Pharmacol 1993; 108: 106110
Chowdhury AH, Cox EF, Francis ST et al. A randomized, controlled,
double-blind crossover study on the effects of 2-L infusions of 0.9%
saline and plasma-lyte(R) 148 on renal blood ow velocity and renal cortical tissue perfusion in healthy volunteers. Ann Surg 2012; 256: 1824
Chowdhury AH, Cox EF, Francis ST et al. A randomized, controlled,
double-blind crossover study on the effects of 1-L infusions of 6% hydroxyethyl starch suspended in 0.9% saline (Voluven) and a balanced solution (Plasma Volume Redibag) on blood volume, renal blood ow
velocity, and renal cortical tissue perfusion in healthy volunteers. Ann
Surg 2013; doi:10.1097/SLA.0000000000000324.
Greger R, Velazquez H. The cortical thick ascending limb and early
distal convoluted tubule in the urinary concentrating mechanism.
Kidney Int 1987; 31: 590596
Imig JD, Passmore JC, Anderson GL et al. Chloride alters renal blood
ow autoregulation in deoxycorticosterone-treated rats. J Lab Clin Med
1993; 121: 608613
Bullivant EM, Wilcox CS, Welch WJ. Intrarenal vasoconstriction during
hyperchloremia: role of thromboxane. Am J Physiol 1989; 256: F152F157
Zornow MH, Todd MM, Moore SS. The acute cerebral effects of changes in
plasma osmolality and oncotic pressure. Anesthesiology 1987; 67: 936941
Young JB, Utter GH, Schermer CR et al. Saline versus Plasma-Lyte a in
initial resuscitation of trauma patients: a randomized trial. Ann Surg
2014; 259: 255262.
Burdett E, Dushianthan A, Bennett-Guerrero E et al. Perioperative buffered versus non-buffered uid administration for surgery in adults. Cochrane Database Syst Rev 2012; 12: CD004089
Shaw AD, Bagshaw SM, Goldstein SL et al. Major complications, mortality, and resource utilization after open abdominal surgery: 0.9% saline
compared to Plasma-Lyte. Ann Surg 2012; 255: 821829