Edwards and Mythen Extreme Physiology & Medicine 2014, 3:16

http://www.extremephysiolmed.com/content/3/1/16

REVIEW Open Access

Fluid therapy in critical illness

Mark R Edwards1,2* and Michael G Mythen3,4,5

Abstract
Major surgery and critical illnesses such as sepsis and trauma all disturb normal physiological fluid handling.
Intravenous fluid therapy for resuscitation and fluid maintenance is a central part of medical care during these
conditions, yet the evidence base supporting practice in this area lacks answers to a number of important
questions. Recent research developments include a refinement of our knowledge of the endothelial barrier
structure and function and a focus on the potential harm that may be associated with intravenous fluid therapy.
Here, we briefly describe the contemporary view of fluid physiology and how this may be disrupted by pathological
processes. The important themes in critical illness fluid research are discussed, with a particular focus on two emerging
ideas: firstly, that individualising fluid treatment to the patient, their underlying disease state and the phase of that
illness may be key to improving clinical outcomes using fluid interventions and, secondly, that fluids should be
considered to be drugs, with specific indications and contraindications, dose ranges and potential toxicities.
Keywords: Intravenous fluid, Critical illness, Surgery, Goal-directed therapy, Glycocalyx

Review address the limitations of the preceding literature, which

Background is based predominantly on relatively small, single-centre
Disturbances of body fluid homeostasis are common in studies.
major surgery and a range of critical illnesses including Some key themes have emerged from this body of re-
sepsis and trauma. Administering appropriate intraven- search. Firstly, it is increasingly clear that clinical context
ous fluid is a core part of medical care during these epi- is key to the success or failure of a given fluid strategy.
sodes. Despite years of research, there is still widespread A fluid strategy which appears to reduce morbidity in
debate about the best dosing strategy for these fluids one clinical population (e.g. patients undergoing major
and the optimum fluid composition for a given clinical surgery) may actually be harmful to another (patients with
situation. These are still important research questions, established sepsis). Fluid therapies are being increasingly
as there are clear signs from the current literature that tailored to very tightly defined clinical subgroups and
fluid administration strategies have the power to affect indeed to individual patients and the stage of their clinical
clinical outcomes in a variety of areas. course. A further theme is the emergence of toxicities
Our study of fluid therapy is being refined by recent related to fluid composition which have only become
advances in basic physiology research. In particular, our apparent through large clinical trials. Fluids are there-
knowledge of the structure of the vascular endothelium fore increasingly being seen as drugs, with specific
and the way in which fluid is handled at a capillary level indications, contraindications and dose ranges [1].
has recently been redeveloped. There has also been pro- In this review, we will outline the contemporary model
gress in pharmacokinetic modelling of the movement of of normal physiological fluid handling. The effects of
exogenous fluids between the body's fluid compart- a range of extreme physiological situations will be
ments. At the other end of the research spectrum, clin- described, including acute haemorrhage and systemic
ical fluid trials are becoming ever larger in an effort to inflammation. We will then review the latest literature
in relation to clinical fluid strategies in a number of
* Correspondence: [email protected] clinical settings Table 1.
1
Department of Anaesthesia, Southampton General Hospital, University
Hospital Southampton NHS Foundation Trust, Tremona Road, Southampton Fluid physiology
SO16 6YD, UK
2
NIHR Southampton Respiratory Biomedical Research Unit, Southampton, UK Water makes up approximately 60% of total body weight
Full list of author information is available at the end of the article in the average adult. Due to the low water content of
© 2014 Edwards and Mythen; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the
Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public
Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this
article, unless otherwise stated.
Edwards and Mythen Extreme Physiology & Medicine 2014, 3:16 Page 2 of 9
http://www.extremephysiolmed.com/content/3/1/16

Table 1 Fluid therapy terminology within blood vessels, and transcellular fluid including
Term Summary gastrointestinal secretions, joint fluid, cerebrospinal fluid
Osmotic and oncotic — osmotic pressure is the hydrostatic pressure and pleural, peritoneal and pericardial fluid.
pressure that would be required to resist the diffusion The vascular endothelium is the key barrier between
of water across a semipermeable membrane the intravascular and interstitial spaces. It is particularly
from a higher solute concentration to a lower
solute concentration. Oncotic pressure is the relevant clinically as it can be damaged during patho-
portion of osmotic pressure which is due to physiological states such as inflammation, allowing po-
large molecular weight particles, particularly tentially harmful fluid accumulation in the interstitial
proteins.
space. The vascular endothelium in a typical capillary is
Fluid tonicity — the effective osmolality of a solution in
relation to a specific semipermeable
composed of a single layer of endothelial cells with inter-
membrane and therefore a useful way of cellular clefts closed by tight junctions. The cells sit on a
describing a given fluid's in vivo behaviour. For continuous basement membrane. On the vascular aspect
example, although 5% dextrose has a similar ex
vivo osmolality to 0.9% sodium chloride, after
of the endothelial cells lies a continuous layer of glycos-
infusion the dextrose is taken up into cells, aminoglycan chains, membrane-bound proteoglycans
rendering the solution effectively hypo- and glycoproteins. These form the endothelial glycocalyx
osmolar with respect to the cell membrane, i.e.
hypotonic; 0.9% sodium chloride remains
layer (EGL), which is approximately 1 μm thick.
isotonic due to the retention of sodium and Water and electrolytes can pass freely across the
chloride ions in the extracellular space. glycocalyx and then beyond the endothelial cells via the
Crystalloid — solutions of glucose and/or electrolytes in intercellular clefts. Proteins and other large molecules
water. however are unable to pass through an intact glycocalyx
Colloid — a dispersion of large molecules or and are transported in relatively small quantities across
ultramicroscopic non-crystalline particles in a the endothelial cells by active processes. The EGL is
carrier crystalloid. It includes gelatins, starches
and dextrans. therefore still part of the intravascular space but con-
Balanced solutions — those with a composition more similar to
tains up to 700–1,000 ml of almost protein-free fluid
plasma than to 0.9% sodium chloride. It is with the same electrolyte composition as plasma. This
achieved by replacing a proportion of the creates a large protein concentration gradient—and
chloride with stable organic anionic buffers
such as lactate, gluconate or acetate.
oncotic pressure gradient—between the plasma and the
space immediately below the EGL. This opposes the
Goal-directed — the use of cardiac output monitoring to
haemodynamic therapy guide fluid and inotrope therapy. Key hydrostatic pressure gradient at the arteriolar end of
physiological goals are targeted in specific capillaries, reducing the volume of water and electrolytes
treatment algorithms. This may be a filtered out of the capillary. Net filtration across a ca-
predefined increase in global oxygen delivery
or stepwise increases to wards a maximal pillary is expressed by the modified Starling equation
cardiac stroke volume. This treatment—as (Figure 1).
compared with fluid dosing based on clinical
assessment or ‘per weight’ basis—has been
used in various forms for over 40 years. Early
variants were guided by the pulmonary artery Pathophysiology of fluid homeostasis in critical illness
catheter, but several minimally invasive devices A wide variety of illnesses cause disturbances in fluid
are now in use.
balance. This is usually due to inadequate or excessive
volume in one or more functional fluid compartments,
adipose tissue, this proportion varies widely with obesity which may be associated with a disturbance in the
(as low as 45%), age (higher in childhood, reducing to barrier between those compartments.
approximately 50% in the elderly) and sex (50% in an Dehydration is a reduction in the water—and possibly
average adult female). Body water occupies functional sodium—content in all fluid compartments. It is a
and anatomical compartments. Fluid within cells—the feature of excessive fluid loss through the skin (pyrexia,
intracellular compartment—is separated from the extra- sweating), gastrointestinal tract (vomiting, diarrhoea,
cellular compartment by the cell membrane. This is im- bowel obstruction) or kidneys (osmotic diuresis or poly-
permeable to large hydrophilic molecules and charged uric renal failure) without adequate replacement.
particles, although these may cross it by specific trans- Acute blood loss causes a sudden reduction in intra-
port mechanisms. vascular volume, cardiac filling and cardiac output. The
A proportion of extracellular water is contained within transient reduction in arterial pressure is rapidly sensed
the bone and dense connective tissue, although due to by the high-pressure baroreceptors in the aortic arch
slow kinetics this is viewed as non-functional. The extra- and carotid sinus. The resultant sympathetic activation
cellular compartment is further divided into interstitial results in vasoconstriction, increased cardiac inotropy
fluid between cells and in lymphatics, intravascular fluid and increased heart rate. A range of other neurohumoral
Edwards and Mythen Extreme Physiology & Medicine 2014, 3:16 Page 3 of 9
http://www.extremephysiolmed.com/content/3/1/16

‘autotransfusion’ of interstitial fluid into the intravascular

space, although this is limited to approximately 500 ml
[6]. Ongoing reduced capillary pressure reduces net
filtration of fluid out of the intravascular space, and this
state is maintained until capillary pressure returns to
normal [6].
Fluid homeostasis becomes even more disturbed if the
endothelium becomes damaged and loses its barrier func-
tion. A range of insults has been associated with degrad-
ation of the glycocalyx, including inflammatory mediators
[7] or even natriuretic peptides associated with acute
iatrogenic hypervolaemia [8]. Systemic inflammation also
leads to an altered endothelial cell phenotype and in-
creased endothelial pore size. This endothelial dysfunction
allows the loss of proteins into the interstitial space, ac-
companied by excess water. This overloads the lymphatic
system and results in oedema within the compliant tissues
such as connective tissue, lungs and gut. The loss of
plasma proteins and oncotic pressure also aggravates
excessive capillary filtration, causing hypovolaemia and
Figure 1 The revised Starling principle and equation. Hydrostatic
organ dysfunction.
pressures are higher within the vascular lumen (Pv) than within the
interstitium (Pi), favouring outward fluid filtration. The endothelial The clinical management of intravenous fluid adminis-
surface layer (ESL) is formed by the endothelial glycocalyx (EG), which tration therefore has the potential to influence organ
binds plasma proteins and excludes them from the subglycocalyx layer function, morbidity and mortality in pathophysiological
(S). This forms an oncotic gradient from the low protein concentration settings. If fluid replacement is inadequate during critical
of the subglycocalyx (IIs) to the intravascular space (IIv). This gradient
illness, inadequate organ blood flow and therefore
opposes outward fluid filtration. Net transcapillary flow (Jv)—dashed
arrows—can be expressed using the revised Starling equation: Jv = inadequate delivery of oxygen and energy substrates may
Kf([Pv − Pi] − σ[πv − πs]), where Kf is the filtration coefficient and σ is occur. This results in inefficient anaerobic cellular me-
the reflection coefficient (the degree to which the tendency of a tabolism and in severe cases cell death and organ failure.
macromolecule to cross the endothelial barrier is resisted). EC, Conversely, excessive salt and water administration,
endothelial cell. NB: Reproduced with permission from reference [2].
particularly in the setting of endothelial dysfunction,
may worsen organ oxygenation and function through
oedema formation.
mechanisms are activated, including the renin- Again, the capillary handling of administered intraven-
angiotensin-aldosterone axis and antidiuretic hormone ous fluids is critical and has recently been re-examined.
release. These further contribute to vasoconstriction, Free water contained in hypotonic solutions distributes
reduction in glomerular filtration rate and renal retention evenly throughout all fluid compartments, leaving only a
of salt and water to compensate for the acute volume loss. small proportion within the intravascular space. Trad-
These early mechanisms may initially result in ap- itionally, isotonic electrolyte solutions were thought to
parently normal haemodynamic observations. However, distribute evenly through the extracellular (intravascular
these responses divert blood away from ‘non-essential’ and interstitial) space only, leaving approximately 20%–
organs (kidney, gut, liver, skin) and towards essential or- 25% of their volume in the intravascular space. Colloids,
gans (brain, heart). Healthy volunteer experiments have due to their superior oncotic pressure, were thought to
demonstrated signs of critical organ hypoperfusion, mea- initially remain almost completely within the intravascu-
sured by a reduction in gut pH, despite normal-range lar space and were seen as the best fluid for acute plas-
blood pressure and heart rate after up to 30% blood vol- ma volume expansion.
ume loss [3]. This subclinical gut hypoperfusion may be It is now proposed that the ‘context’ of capillary pres-
a potent inflammatory stimulus, as reduced gut barrier sure is important [2]. At low capillary pressures, infusion
function allows translocation of bacterial endotoxin into of colloid will expand the plasma volume, but the large
the systemic circulation [4,5]. molecules should not pass beyond an intact glycocalyx.
Acutely reduced capillary pressure, seen in blood loss Crystalloid infusions will expand the entire intravascular
or due to widespread vasodilatation due to sepsis or volume, i.e. both plasma and EGL. In both cases, net ca-
anaesthesia, alters fluid movement at the capillary pillary filtration will remain close to zero until capillary
level also. In the initial phase, there may be an hydrostatic pressure becomes normal or supra-normal.
Edwards and Mythen Extreme Physiology & Medicine 2014, 3:16 Page 4 of 9
http://www.extremephysiolmed.com/content/3/1/16

This may help to explain the observation in numerous 1990s, they had fallen out of widespread use. New goal-
clinical studies that similar resuscitation end points can directed haemodynamic therapy strategies using less inva-
be achieved using 50% more crystalloid than colloid, ra- sive technologies such as the transoesophageal Doppler
ther than the much larger crystalloid volumes tradition- monitor were therefore developed and tested. However,
ally predicted [9-11]. Pharmacokinetic studies have also the evidence of benefit was accrued slowly through a
demonstrated that in anaesthetised patients—with pre- number of single-centre trials, and it was not until 2011
sumably low capillary pressure—around 60% of an iso- that the National Institute for Clinical Excellence recom-
tonic crystalloid solution is retained in the circulation mended their use in UK peri-operative practice [42].
during the infusion, as redistribution to other compart- Despite this, widespread uptake into clinical practice was
ments takes 20–25 min [12]. limited. Financial incentives through ‘quality payments’
For both fluid types, in the setting of normal or supra- have been used in an attempt to promote the technology
normal capillary pressure, net filtration will occur, with in the UK. However, in the intervening decades, changes
loss of fluid from the intravascular space and the poten- in surgical practices and care pathways have developed,
tial for oedema formation. Although in health the main- reducing baseline morbidity and mortality. Despite nearly
tained oncotic pressure associated with colloids helps to 40 years of research, it therefore remains uncertain whe-
limit this filtration, this effect is reduced in the presence ther goal-directed haemodynamic therapy is beneficial for
of endothelial dysfunction. Here, there is the potential contemporary surgical populations. Increasingly large-
for large colloid molecules to pass freely into the inter- scale trials are currently in progress or development in
stitial space. However, as more crystalloid volume is order to provide a definitive answer to this question.
needed to achieve similar resuscitation endpoints to col- Even at a basic physiological, pathological and pharma-
loids, patients resuscitated with crystalloid do ultimately cological level, there is considerable complexity which
gain a more positive fluid balance [13] Table 1. may be important in fluid therapy. The mechanisms of
action of fluid strategies at an organ and cellular level have
Therapeutic fluid strategies and clinical outcome been relatively under-explored in relevant human clinical
As a core part of medical treatment during acute illness settings. This gap in the knowledge base may help to
and trauma, it is not surprising that a number of fluid explain the variable success of many fluid clinical trials.
administration strategies have been trialled in various The ultimate goals of fluid therapy are to provide enough
clinical settings over the years. This evidence base has circulating volume and blood flow to organs, so that the
shown that fluid strategies have the potential to influ- quantities of oxygen and nutrients delivered are sufficient
ence morbidity and possibly mortality outcomes after to meet their metabolic requirements, while avoiding iat-
critical illness and major surgery. However, clear-cut an- rogenic harm through excess dosing of water, electrolytes
swers which can be confidently translated to widespread or exogenous compounds. Yet this apparent simplicity be-
clinical practice are rare. A summary of some of the crit- lies the biological complexity of achieving these goals
ical illness fluid research areas is contained in Table 2. in diverse disease settings. The clinical benefit or harm
There are a number of challenges which have brought by administering fluids may result from fluid
hampered evidence-based practice in this area. Firstly, composition, dose, physiological target, timing during the
fluid therapy is a prime example of what the Medical Re- illness, or complex interactions between these factors, the
search Council terms a ‘complex intervention’ [39]. This patient's pre-morbid phenotype and the acute but ever-
complexity lies at multiple levels, from the biological changing pathophysiological processes.
interaction between administered fluid and the patho- Developing the evidence base behind critical illness
physiological setting, through the interaction of the clin- fluid practices further therefore needs a combination of
ician with choice of fluid, monitoring and therapeutic basic science, translational trials and clinical trials.
strategy, to clinician beliefs, resource limitations, health- Controlled trials with large sample sizes are necessary to
care systems, guidelines and financial incentives which definitively detect differences in outcome which may be
may promote or inhibit the adoption of certain fluid small but important when applied to the huge number
interventions. of patients undergoing major surgery or critical illness.
One example of this complexity leading to ongoing But achieving this size often requires pragmatic trial
uncertainty is in fluid strategies around the time of major design and heterogeneous groups (e.g. ‘all major surgical
surgery. An approach to haemodynamic therapy, based on patients’ or ‘all critical care patients’) which potentially
targeting supra-normal global oxygen delivery—moni- obscure the nuances of which individual patients might
tored using the pulmonary artery catheter—was initially get benefit or harm and when. More mechanistic studies
investigated in the 1970s [20,40]. Despite early promise, are needed to explore what aspect(s) of the interven-
controversy developed over the safety of pulmonary artery tion strategy brought the benefit or harm and to address
catheters though a number of clinical trials [41]. By the apparently basic but as-yet unanswered questions, such as
Edwards and Mythen Extreme Physiology & Medicine 2014, 3:16 Page 5 of 9
http://www.extremephysiolmed.com/content/3/1/16

Table 2 Key themes in critical illness fluid research

Patient Research question Early trial evidence Subsequent developments and ongoing
group research questions
Sepsis Does early goal-directed bolus Key single-centre study by Rivers showing Further developments in sepsis care
fluid therapy improve survival reduced mortality when bolus fluid therapy through treatment ‘bundles’ and the need
from severe sepsis? targeting central venous pressure and mixed to test the intervention have led to a re-
venous saturations was used at presentation of appraisal of the benefits of this intervention.
septic patients to hospital [14]. The ProCESS trial recently published
comparing protocol-driven goal-directed
therapy with protocol (but not goal-
directed) therapy and with standard care in
1,341 patients. No differences in mortality at
60, 90 or 365 days [15]. Further multicentre
trials based on River's original protocol are
nearing completion in the UK (ProMISe) and
Australasia (ARiSE).
Mixed Is fluid composition used for Early starch-based colloids with high molecular A number of large trials have demonstrated an
critical care resuscitation associated with acute weight were associated with renal dysfunction increase in the need for renal replacement
populations kidney injury and increased [16]. Lower molecular weight starches were therapy with modern starches, when compared
mortality? brought to market based on efficacy data but to isotonic crystalloids. This has been shown in
without trials large enough to detect possible both septic [17,18] and mixed critical care [13]
harm. populations. Despite concerns over trial
methods, for example that some patients had
been partially resuscitated before trial entry, use
of starches has now been restricted (USA,
Europe) or stopped (UK). A recent network
meta-analysis has suggested that renal
replacement requirement is highest in
association with starches, followed by isotonic
saline then balanced crystalloids [19].
Surgery Does goal-directed Survivors of high-risk surgery were found to Reduced popularity of the pulmonary artery
haemodynamic therapy improve achieve higher global oxygen delivery levels catheter followed by minimally invasive cardiac
outcomes for patients undergoing than non-survivors. These ‘survivor values’ were output monitors and an evolution of fluid
major surgery? then used as therapeutic targets in subsequent administration protocols. Single-centre trials
interventional trials, with benefit shown [20,21]. from the 1990s onwards showed reduction in
hospital length of stay and postoperative
morbidity in goal-directed therapy intervention
groups. A recent Cochrane review points to no
harm and probably morbidity benefit from goal-
directed therapy, although studies span over
30 years of trials [22]. A large contemporary
multicentre trial also involving an inotrope in
the intervention group (OPTIMISE) again
suggested benefit but lacked statistical
significance. The accompanying update to the
meta-analysis strengthens the evidence for a
reduction in postoperative morbidity [23].
Studies in emergency surgery are lacking.
Surgery Can targeting a fluid ‘dose’ lead to Patients receiving ‘standard’ volumes of fluid in This research theme has been hampered by
improved outcomes in patients the peri-operative phase were shown to have varying definitions of what constitutes liberal
undergoing major surgery? more postoperative morbidity than those and restrictive volumes of fluid. Subsequent
receiving ‘restrictive’ volumes [24]. trials have therefore shown contrasting results. A
common theme though is the association
between a positive fluid balance in the
immediate peri-operative period (>3,500–
5,000 ml) and worse postoperative outcomes.
Giving larger volumes of fluid without robust
physiological monitoring appears to be
associated with worse outcomes, even when
overall quantities given may be similar to those
used in goal-directed therapy interventions [25].
A large and hopefully definitive multicentre trial
(RELIEF) is underway in Australasia and the UK.
Surgery Does isotonic saline cause harm to Concerns that isotonic saline is unphysiological Observational studies suggest harm with saline
patients undergoing surgery when in composition and may cause harm through as compared to balanced solutions [26,27]. A
compared with balanced reduction in renal blood flow and Cochrane review suggests an increase in
solutions? hyperchloraemic acidosis. These have never postoperative acidosis and need for
compensatory hyperventilation with saline used
Edwards and Mythen Extreme Physiology & Medicine 2014, 3:16 Page 6 of 9
http://www.extremephysiolmed.com/content/3/1/16

Table 2 Key themes in critical illness fluid research (Continued)

been backed up by trials with adequate in the peri-operative setting [28]. No increase in
statistical power. postoperative renal morbidity. However, there is
a lack of high-quality large trials, and this remains
an important topic for ongoing research.
Trauma Does limiting volumes of early Early suggestions from battlefield that fluid Permissive hypotension with limited volumes of
resuscitation fluid improve resuscitation may be harmful in ‘wound shock’ clear fluid given prior to achieving haemostasis
outcomes from trauma? [29]. Renewed interest following pre-hospital is proposed as a part of ‘damage control
randomised trial in penetrating trauma showing resuscitation’ [31]. This also includes early
reduced mortality in limited vs. standard surgical or radiological control of haemorrhage
resuscitation [30]. and steps to limit coagulopathy, including
tranexamic acid and high ratios of plasma and
platelets to red blood cell units transfused [32].
Controversy remains about the place of limited
volume early resuscitation, as follow-up trials
have not shown in-hospital or 30-day mortality
benefit when blunt trauma patients are included
[33,34]. Hypotension may also worsen co-
existing brain injury [35].
Trauma Does hypertonic saline improve Hypertonic solutions may have the ability to Limited human studies examining hypertonic
outcomes from trauma? draw water from the intracellular to the saline used early in trauma resuscitation. One
extracellular compartment, achieving plasma large trial stopped early as interim analysis
volume expansion with minimal volume of fluid demonstrated futility [36]. Hypertonic solutions
administered. not routinely recommended in trauma
resuscitation. Although hypertonic saline is
effective at reducing raised intracranial pressure
[37], there is no benefit in early traumatic brain
injury when intracranial pressure is not
monitored [38].

how the body handles the currently available range of Timing: tailoring the fluid strategy to the
exogenously administered anions. pathophysiological phase
There is growing evidence that the timing of plasma vol-
ume expansion is important in its potential success. This
Individualisation of fluid strategies
may be seen even within the relatively short time frame
It is implausible that one fluid administration strategy
of a surgical operation. One study showed that despite
will ever be beneficial to all patients in all clinical
only modest differences in the overall volume of fluid
settings at all time points in their critical illness or
given to patients undergoing surgery, the use of cardiac
physiological insult. There is an increasing realisation
output monitoring to guide fluid input led to fluid being
that trial findings in one group cannot be readily extrap-
given earlier in the operation, which led to sustained in-
olated to other groups. The theme of ‘individualisa-
creases in cardiac output not seen in the control group
tion’—tailoring fluid interventions to each patient and
[43]. This was associated with a reduction in postopera-
each setting—is gaining momentum and is seen in the
tive morbidity. More broadly, initial studies pointed to
following areas.
potential benefits of goal-directed fluid resuscitation early
in severe sepsis [14], whereas later aggressive fluid admin-
Patient: tailoring to patient physiology istration during established critical illness appears to be
Goal-directed haemodynamic therapy in patients under- harmful [44-46]. Differences in the time of entry into clin-
going major surgery is a longstanding example of adapt- ical trials of fluid resuscitation may also explain differing
ing fluid dosing to a patient's baseline physiology. This is messages about potential harm from colloids used for
also true of contemporary algorithms aimed at optimis- volume expansion [13,47].
ing cardiac stroke volume. These involve a physiological The differing physiological requirements from fluid
‘test’ in the form of a fluid bolus with ongoing moni- therapy during the developing stages of critical illness
toring of stroke volume. If the patient is fluid replete— and circulatory shock have been recognised in a new
or cardiac performance is beyond the optimum point of proposed treatment approach [48]:
the Starling curve—there will be no further increase in
stroke volume and fluid bolus administration can stop.  Salvage phase—the immediate phase where fluid
This is in contrast to other dosing strategies based on resuscitation is used as a lifesaving measure, guided
patient weight or physiological signs which are poor at by immediately available information such as clinical
predicting a response to fluid. assessment, blood pressure and heart rate
Edwards and Mythen Extreme Physiology & Medicine 2014, 3:16 Page 7 of 9
http://www.extremephysiolmed.com/content/3/1/16

 Optimisation phase—guided by more in-depth why they are giving a fluid in a particular situation and
monitoring, and aiming to optimise global oxygen match the fluid composition and volume to one of the
delivery, with resultant improvements in arterial following physiological needs:
lactate and mixed venous blood desaturation
 Stabilisation phase—a more cautious approach to  Resuscitation (bolus therapy recommended, using
fluid administration, aiming to minimise iatrogenic fluids with sodium in the range 130–154 mmol.L−1
harm caused by fluid overload and institute more but avoiding starches)
general organ support where required  Routine maintenance (25–30 ml.kg−1.day−1 with
 De-escalation phase—appropriate during the start of 1 mmol.kg−1.day−1 each of sodium, potassium and
resolution of critical illness, aiming to reduce chloride, plus 50–100 g.day−1 of glucose)
vasoactive treatments and achieve a negative fluid  Replacement due to ongoing losses and
balance redistribution (fluid composition tailored to match
the estimated water and electrolyte loss)
Disease: tailoring fluid strategy to the underlying
pathophysiology Health-care setting: tailoring fluid strategy to the wider
A number of clinical syndromes (‘critical illness’ or ‘cir- clinical context
culatory shock’) appear to have common features which
A further layer of complexity in fluid treatments may
may be addressed by similar fluid approaches. Yet the
come from the health-care setting in which they are
underlying disease process seems to be critical to the
used. Fascinating insights have come from the FEAST
benefit or harm brought by these fluid strategies. In
study, which took the principle of fluid bolus therapy
mixed critical care populations, including severe sepsis,
early in the presentation of patients with sepsis but
there is growing evidence that starch-based colloids are
applied it in the setting of children presenting to hospital
associated with an increased requirement for renal re-
in Africa [53]. In contrast to the potential benefits found
placement therapy [17,18]. This has led to restrictions
in previous studies, in this population there was a 3.3%
on the licences for these products. In contrast, many of
absolute increase in mortality at 48 h in the fluid inter-
the trials showing the benefits of goal-directed haemo-
vention group. Subsequent analysis has suggested that
dynamic therapy in major surgical patients have used
these excess deaths were probably not attributable dir-
colloids—including starches—for their intervention fluid
ectly to fluid overload (e.g. through respiratory or neuro-
boluses.
logical events) [54]. Instead, the bolus fluid group had
Another example of this is the Saline vs. Albumin
an initial improvement in haemodynamic parameters,
Fluid Evaluation (SAFE) study, a large multicentre trial
but then went on to have an excess of cardiogenic or
comparing albumin with isotonic saline for fluid resusci-
shock-related terminal events. One interpretation of this
tation in a mixed critical illness population [49]. Overall
is that early large-volume fluid resuscitation sets up a
no difference in outcome was found, but importantly
series of adverse processes, including vascular dysfunc-
predefined subgroup analyses highlighted clinical po-
tion and impaired myocardial performance. Thus, the
pulations worthy of further investigation. Patients with
rapid reversal of shock may then lead to a requirement
brain injury appeared to have worse outcomes in the
for inotropic and other systems support, which are hard
albumin arm. It has been suggested that this was due to
to deliver in health-care systems without intensive care
the slight hypotonicity of the albumin preparation used
facilities.
(260 mOsm.L−1), which may have contributed to intra-
cranial hypertension [50]. Conversely, patients with sep-
sis seemed to have better outcomes with albumin. This Conclusions
led to the recently published ALBIOS trial of albumin Abnormal fluid handling is a central part of many
supplementation in patients with sepsis, powered spe- critical illnesses, and exogenous fluid administration has
cifically to detect mortality difference in this more ho- become a core part of medical care in these settings.
mogeneous group [51]. No differences in the primary Despite apparently rational physiological goals, the rea-
outcome were found, and despite differences in protocol lity of successful fluid intervention is extremely complex
from the original SAFE study, this trial demonstrates the at a basic biological, clinical and health-care system level.
utility of focussing on more tightly defined disease As a result, many of the key research questions in this
phenotypes even in large randomised trials. field have not yet been definitively answered despite years
Lastly, the most basic example of matching fluid input of research. Large trials and meta-analyses have raised the
to the pathophysiological situation has been summarised possibility of fluid-related toxicities long after these fluids
in the UK national guidance on intravenous fluid use have been brought to market. Intravenous fluids must be
[52]. Here, clinicians are encouraged to consider exactly re-framed as drugs, with indications, contraindications,
Edwards and Mythen Extreme Physiology & Medicine 2014, 3:16 Page 8 of 9
http://www.extremephysiolmed.com/content/3/1/16

side effects and dose ranges. Lastly, the emergence of the 8. Bruegger D, Jacob M, Rehm M, Loetsch M, Welsch U, Conzen P, Becker BF:
theme of individualisation—to timing, patient, disease Atrial natriuretic peptide induces shedding of endothelial glycocalyx in
coronary vascular bed of guinea pig hearts. Am J Physiol Heart Circ Physiol
state and health-care setting—is an important step for- 2005, 289:H1993–H1999.
ward, but a combination of robust basic science, transla- 9. Finfer S, Bellomo R, Boyce N, French J, Myburgh J, Norton R: A comparison
tional trials and large clinical trials will be required to of albumin and saline for fluid resuscitation in the intensive care unit.
N Engl J Med 2004, 350:2247–2256.
progress this area further. 10. Brunkhorst FM, Engel C, Bloos F, Meier-Hellmann A, Ragaller M, Weiler N,
Moerer O, Gruendling M, Oppert M, Grond S, Olthoff D, Jaschinski U, John S,
Abbreviations Rossaint R, Welte T, Schaefer M, Kern P, Kuhnt E, Kiehntopf M, Hartog C,
EG: endothelial glycocalyx; EGL: endothelial glycocalyx layer; ESL: endothelial Natanson C, Loeffler M, Reinhart K: Intensive insulin therapy and
surface layer; IIs: subglycocalyx oncotic pressure; IIv: intravascular oncotic pentastarch resuscitation in severe sepsis. N Engl J Med 2008,
pressure; Jv: net transcapillary flow; Pi: interstitial hydrostatic pressure; 358:125–139.
Pv: vascular lumen hydrostatic pressure; S: subglycocalyx layer.
11. James MFM, Michell WL, Joubert IA, Nicol AJ, Navsaria PH, Gillespie RS:
Resuscitation with hydroxyethyl starch improves renal function and
Competing interests
lactate clearance in penetrating trauma in a randomized controlled
MRE declares that he has no competing interests. MGM has received
study: the FIRST trial (Fluids in Resuscitation of Severe Trauma). Br J Anaesth
honoraria for speaking, or consultation and/or travel expenses from Baxter,
2011, 107:693–702.
B. Braun, Covidien, Fresenius Kabi, Hospira and LiDCO. He is the National
Clinical Advisor for the Department of Health Enhanced Recovery 12. Hahn RG: Volume kinetics for infusion fluids. Anesthesiology 2010, 113:470–481.
Partnership, is a Smiths Medical Professor of Anaesthesia and Critical Care 13. Myburgh JA, Finfer S, Bellomo R, Billot L, Cass A, Gattas D, Glass P, Lipman J,
at UCL, has consulted AQIX (start-up company with a novel crystalloid Liu B, McArthur C, McGuinness S, Rajbhandari D, Taylor CB, Webb SAR:
solution—pre-clinical), is the Director of Medical Defence Technologies LLC Hydroxyethyl starch or saline for fluid resuscitation in intensive care.
(‘Gastrostim’ patented) and is a co-inventor of ‘QUENCH’ (pump) IP being N Engl J Med 2012, 367:1901–1911.
exploited by UCL Business. Professor Mythen's institution has also received 14. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E,
charitable donations and grants from Smiths Medical Endowment and Deltex Tomlanovich M: Early goal-directed therapy in the treatment of severe
Medical. Professor Mythen is also a co-author of the GIFTASUP guidelines on sepsis and septic shock. N Engl J Med 2001, 345:1368–1377.
peri-operative fluid management, the Chair of the National Institute of 15. The ProCESS Investigators: A randomized trial of protocol-based care for
Academic Anaesthesia, a Board member of the Faculty of Intensive Care early septic shock. N Engl J Med 2014, 370:1683–1693.
Medicine, a Council member of the Royal College of Anaesthetists of Great 16. Dart AB, Mutter TC, Ruth CA, Taback SP: Hydroxyethyl starch (HES) versus
Britain and Ireland, the Editor in Chief of Peri-operative Medicine, on the other fluid therapies: effects on kidney function. Cochrane Database Syst
Editorial Board of the BJA and Critical Care, a member of the Improving Surgical Rev, 2010, (1):1–104.
Outcomes Group, a member of the NICE IV Fluids Guideline Development 17. Haase N, Perner A, Hennings LI, Siegemund M, Lauridsen B, Wetterslev M,
Group (now an expert advisor) and the Co-Director of Xtreme Everest. Wetterslev J: Hydroxyethyl starch 130/0.38-0.45 versus crystalloid or
albumin in patients with sepsis: systematic review with meta-analysis
Authors’ contributions and trial sequential analysis. BMJ 2013, 346:f839.
MRE and MGM both conceived and planned this review. MRE drafted the 18. Perner A, Haase N, Guttormsen AB, Tenhunen J, Klemenzson G, Åneman A,
manuscript. MGM helped to draft the manuscript and revised it critically. Madsen KR, Møller MH, Elkjær JM, Poulsen LM, Bendtsen A, Winding R,
Both authors read and approved the final manuscript. Steensen M, Berezowicz P, Søe-Jensen P, Bestle M, Strand K, Wiis J, White
JO, Thornberg KJ, Quist L, Nielsen J, Andersen LH, Holst LB, Thormar K,
Author details Kjældgaard A-L, Fabritius ML, Mondrup F, Pott FC, Møller TP, et al:
1 Hydroxyethyl starch 130/0.42 versus Ringer's acetate in severe sepsis.
Department of Anaesthesia, Southampton General Hospital, University
Hospital Southampton NHS Foundation Trust, Tremona Road, Southampton N Engl J Med 2012, 367:124–134.
SO16 6YD, UK. 2NIHR Southampton Respiratory Biomedical Research Unit, 19. Raghunathan K, Murray P, Beattie WS, Lobo D, Myburgh JA, Sladen R,
Southampton, UK. 3Department of Anaesthesia and Critical Care, University Kellum JA, Mythen M, Shaw AD: Choice of fluid in acute illness: what
College London Hospitals, London, UK. 4University College London, London, should be given? Management strategies from the twelfth consensus
UK. 5UCLH/UCL NIHR Comprehensive Biomedical Research Centre, London, UK. conference of the acute dialysis quality initiative (ADQI) London 2013.
Br J Anaesth 2014. in press.
Received: 6 June 2014 Accepted: 13 August 2014 20. Shoemaker WC, Appel PL, Kram HB, Waxman K, Lee TS: Prospective trial of
Published: 29 Sep 2014 supranormal values of survivors as therapeutic goals in high-risk surgical
patients. Chest 1988, 94:1176–1186.
References 21. Boyd O, Grounds RM, Bennett ED: A randomized clinical trial of the effect
1. Myburgh JA, Mythen MG: Resuscitation fluids. N Engl J Med 2013, of deliberate perioperative increase of oxygen delivery on mortality in
369:1243–1251. high-risk surgical patients. JAMA 1993, 270:2699–2707.
2. Chappell D, Jacob M, Hofmann-Kiefer K: A rational approach to 22. Grocott MPW, Dushianthan A, Hamilton MA, Mythen MG, Harrison D, Rowan
perioperative fluid management. Anesthesiology 2008, 109:723–740. K: Perioperative increase in global blood flow to explicit defined goals
3. Hamilton-Davies C, Mythen MG, Salmon JB, Jacobson D, Shukla A, and outcomes following surgery. Cochrane Database Syst Rev Online 2012,
Webb AR: Comparison of commonly used clinical indicators of 11, CD004082.
hypovolaemia with gastrointestinal tonometry. Intensive Care Med 23. Pearse RM, Harrison DA, MacDonald N, Gillies MA, Blunt M, Ackland G,
1997, 23:276–281. Grocott MP, Ahern A, Griggs K, Scott R, Hinds C, Rowan K, OPTIMISE Study
4. Mythen MG, Webb AR: Intra-operative gut mucosal hypoperfusion is Group: Effect of a perioperative, cardiac output-guided hemodynamic
associated with increased post-operative complications and cost. therapy algorithm on outcomes following major gastrointestinal surgery: a
Intensive Care Med 1994, 20:99–104. randomized clinical trial and systematic review. JAMA, 2014,
5. Deitch EA: The role of intestinal barrier failure and bacterial translocation 311(21):2181–2190.
in the development of systemic infection and multiple organ failure. 24. Brandstrup B, Tønnesen H, Beier-Holgersen R, Hjortsø E, Ørding H,
Arch Surg Chic Ill 1960 1990, 125:403–404. Lindorff-Larsen K, Rasmussen MS, Lanng C, Wallin L, Iversen LH, Gramkow
6. Levick JR, Michel CC: Microvascular fluid exchange and the revised CS, Okholm M, Blemmer T, Svendsen P-E, Rottensten HH, Thage B, Riis J,
Starling principle. Cardiovasc Res 2010, 87:198–210. Jeppesen IS, Teilum D, Christensen AM, Graungaard B, Pott F: Effects
7. Henry CB, Duling BR: TNF-alpha increases entry of macromolecules into of intravenous fluid restriction on postoperative complications:
luminal endothelial cell glycocalyx. Am J Physiol Heart Circ Physiol 2000, comparison of two perioperative fluid regimens. Ann Surg 2003,
279:H2815–H2823. 238:641–648.
Edwards and Mythen Extreme Physiology & Medicine 2014, 3:16 Page 9 of 9
http://www.extremephysiolmed.com/content/3/1/16

25. Corcoran T, Rhodes JEJ, Clarke S, Myles PS, Ho KM: Perioperative fluid 46. Boyd JH, Forbes J, Nakada T, Walley KR, Russell JA: Fluid resuscitation in
management strategies in major surgery: a stratified meta-analysis. septic shock: a positive fluid balance and elevated central venous
Anesth Analg 2012, 114:640–651. pressure are associated with increased mortality. Crit Care Med 2011,
26. Shaw AD, Bagshaw SM, Goldstein SL, Scherer LA, Duan M, Schermer CR, 39:259–265.
Kellum JA: Major complications, mortality, and resource utilization 47. Annane D, Siami S, Jaber S, Martin C, Elatrous S, Declère AD, Preiser JC,
after open abdominal surgery: 0.9% saline compared to Plasma-Lyte. Outin H, Troché G, Charpentier C, Trouillet JL, Kimmoun A, Forceville X,
Ann Surg 2012, 255(5):821–829 Darmon M, Lesur O, Reignier J, Abroug F, Berger P, Clec'h C, Cousson J,
27. McCluskey SA, Karkouti K, Wijeysundera D, Minkovich L, Tait G, Beattie WS: Thibault L, Chevret S, CRISTAL Investigation: Effects of fluid resuscitation
Hyperchloremia after noncardiac surgery is independently associated with colloids vs crystalloids on mortality in critically ill patients
with increased morbidity and mortality: a propensity-matched cohort presenting with hypovolemic shock: The CRISTAL randomized trial. JAMA
study. Anesth Analg 2013, 117:412–421. 2013, 310:(17)1809–1817.
28. Burdett E, Dushianthan A, Bennett-Guerrero E, Cro S, Gan TJ, Grocott MP, 48. Vincent J-L, De Backer D: Circulatory shock. N Engl J Med 2013,
James MF, Mythen MG, O'Malley CM, Roche AM, Rowan K: Perioperative 369:1726–1734.
buffered versus non-buffered fluid administration for surgery in adults. 49. The SAFE Study Investigators: A comparison of albumin and saline for
Cochrane Database Syst Rev Online 2012, 12, CD004089. fluid resuscitation in the intensive care unit. N Engl J Med 2004,
29. Cannon W, Fraser J, Cowell E: The preventative treatment of wound 350:2247–2256.
shock. Jama J Am Med Assoc 1918, 618:621. 50. Cooper DJ, Myburgh J, Heritier S, Finfer S, Bellomo R, Billot L, Murray L,
30. Bickell WH, Wall MJ, Pepe PE, Martin RR, Ginger VF, Allen MK, Mattox KL: Vallance S: Albumin resuscitation for traumatic brain injury: is intracranial
Immediate versus delayed fluid resuscitation for hypotensive patients hypertension the cause of increased mortality? J Neurotrauma 2013,
with penetrating torso injuries. N Engl J Med 1994, 331:1105–1109. 30:512–518.
31. Harris T, Thomas GOR, Brohi K: Early fluid resuscitation in severe trauma. 51. Caironi P, Tognoni G, Masson S, Fumagalli R, Pesenti A, Romero M, Fanizza
BMJ 2012, 345:e5752. C, Caspani L, Faenza S, Grasselli G, Iapichino G, Antonelli M, Parrini V, Fiore
32. Johansson PI, Oliveri RS, Ostrowski SR: Hemostatic resuscitation with G, Latini R, Gattinoni L: Albumin replacement in patients with severe
plasma and platelets in trauma. J Emerg Trauma Shock 2012, 5:120–125. sepsis or septic shock. N Engl J Med 2014, 370:1412–1421.
33. Morrison CA, Carrick MM, Norman MA, Scott BG, Welsh FJ, Tsai P, Liscum KR, 52. NICE: Intravenous Fluid Therapy in Adults in Hospital (Clinical Guidance 174).
Wall MJJ, Mattox KL: Hypotensive resuscitation strategy reduces London: NICE; 2014.
transfusion requirements and severe postoperative coagulopathy in 53. Maitland K, Kiguli S, Opoka RO, Engoru C, Olupot-Olupot P, Akech SO, Nyeko
trauma patients with hemorrhagic shock: preliminary results of a R, Mtove G, Reyburn H, Lang T, Brent B, Evans JA, Tibenderana JK, Crawley J,
randomized controlled trial. J Trauma-Inj Infect 2011, 70:652–663. Russell EC, Levin M, Babiker AG, Gibb DM, FEAST Trial Group: Mortality after
34. Dutton RP, Mackenzie CF, Scalea TM: Hypotensive resuscitation during fluid bolus in African children with severe infection. N Engl J Med 2011,
active hemorrhage: impact on in-hospital mortality. J Trauma 2002, 364:2483–2495.
52:1141–1146. 54. Maitland K, George EC, Evans JA, Kiguli S, Olupot-Olupot P, Akech SO, Opoka
35. Moppett IK: Traumatic brain injury: assessment, resuscitation and early RO, Engoru C, Nyeko R, Mtove G, Reyburn H, Brent B, Nteziyaremye J, Mpoya
management. Br J Anaesth 2007, 99:18–31. A, Prevatt N, Dambisya CM, Semakula D, Ddungu A, Okuuny V, Wokulira R,
36. Bulger EM, May S, Kerby JD, Emerson S, Stiell IG, Schreiber MA, Brasel KJ, Timbwa M, Otii B, Levin M, Crawley J, Babiker AG, Gibb DM: Exploring
Tisherman SA, Coimbra R, Rizoli S, Minei JP, Hata JS, Sopko G, Evans DC, mechanisms of excess mortality with early fluid resuscitation: insights
Hoyt DB: Out-of-hospital hypertonic resuscitation after traumatic from the FEAST trial. BMC Med 2013, 11:68.
hypovolemic shock. Ann Surg 2011, 253:431–441.
37. Kamel H, Navi BB, Nakagawa K, Hemphill JC 3rd, Ko NU: Hypertonic 10.1186/2046-7648-3-16
saline versus mannitol for the treatment of elevated intracranial Cite this article as: Edwards and Mythen: Fluid therapy in critical illness.
pressure: a meta-analysis of randomized clinical trials. Crit Care Med Extreme Physiology & Medicine 2014, 3:16
2011, 39:554–559.
38. Bulger EM, May S, Brasel KJ, Schreiber M, Kerby JD, Tisherman SA, Newgard
C, Slutsky A, Coimbra R, Emerson S, Minei JP, Bardarson B, Kudenchuk P,
Baker A, Christenson J, Idris A, Davis D, Fabian TC, Aufderheide TP, Callaway
C, Williams C, Banek J, Vaillancourt C, Van Heest R, Sopko G, Hata JS, Hoyt
DB: Out-of-hospital hypertonic resuscitation following severe traumatic
brain injury: a randomized controlled trial. Jama J Am Med Assoc 2010,
304:1455–1464.
39. Medical Research Council: Developing and Evaluating Complex Interventions:
New Guidance. London: Medical Research Council; 2008.
40. Shoemaker WC, Chang P, Czer L, Bland R, Shabot MM, State D:
Cardiorespiratory monitoring in postoperative patients: I. Prediction of
outcome and severity of illness. Crit Care Med 1979, 7:237–242.
41. Connors AF, Speroff T, Dawson NV, Thomas C, Harrell FE, Wagner D,
Desbiens N, Goldman L, Wu AW, Califf RM, Fulkerson WJ, Vidaillet H,
Broste S, Bellamy P, Lynn J, Knaus WA: The effectiveness of right heart
catheterization in the initial care of critically ill patients. SUPPORT Submit your next manuscript to BioMed Central
Investigators. Jama J Am Med Assoc 1996, 276:889–897. and take full advantage of:
42. National Institute for Health and Care Excellence: CardioQ-ODM Oesophageal
Doppler Monitor [MTG3]. London: National Institute for Health and Care
• Convenient online submission
Excellence; 2011.
43. Noblett SE, Snowden CP, Shenton BK, Horgan AF: Randomized clinical trial • Thorough peer review
assessing the effect of Doppler-optimized fluid management on • No space constraints or color figure charges
outcome after elective colorectal resection. Br J Surg 2006, 93:1069–1076.
44. Hayes MA, Timmins AC, Yau EH, Palazzo M, Hinds CJ, Watson D: Elevation • Immediate publication on acceptance
of systemic oxygen delivery in the treatment of critically ill patients. • Inclusion in PubMed, CAS, Scopus and Google Scholar
N Engl J Med 1994, 330:1717–1722. • Research which is freely available for redistribution
45. Gattinoni L, Brazzi L, Pelosi P, Latini R, Tognoni G, Pesenti A, Fumagalli R:
A trial of goal-oriented hemodynamic therapy in critically ill patients.
SvO2 Collaborative Group. N Engl J Med 1995, 333:1025–1032. Submit your manuscript at
www.biomedcentral.com/submit