Fluids, pH, ions and electrolytes

Lewis J. Kaplana and John A. Kellumb

a
Yale University School of Medicine; Department of
Surgery, Section of Trauma, Surgical Critical Care and
Surgical Emergencies; New Haven, Connecticut and
b
University of Pittsburgh School of Medicine;
Department of Critical Care Medicine; Pittsburgh,
Pennsylvania, USA
Correspondence to John A. Kellum, MD, FCCM, FCCP,
Department of Critical Care Medicine, University of
Pittsburgh, 608 Scaife Hall, 3550 Terrace Street,
Pittsburgh, PA 15261, USA
Tel: +1 412 647 0781; fax: +1 412 647 3791;
e-mail: kellumja@upmc.edu
Current Opinion in Critical Care 2010,
16:323–331
Purpose of review
Significant attention has been recently focused on both maintenance fluid and
resuscitation fluid use in critical care. Accordingly, a focused review of the properties of
crystalloid and colloid fluids, their expected benefits, and potential deleterious side
effects is appropriate and timely.
Recent findings
Despite their ubiquitous use, well described side effects, and ability to be titrated to a
physiologic endpoint, fluids are rarely considered in a fashion similar to other
pharmacologic agents. Understanding their physical and chemical properties allows
the clinician to understand, anticipate and deliberately harness their expected impact on
acid–base balance. Expanded insights into the pathogenesis of common acid–base
disorders may be gleaned from utilizing a physicochemical approach that allows the
precise quantification of the ionic species that impact pH.
Summary
This focused review further enables the clinician to appropriately investigate, modify,
and optimize bedside clinical care related to fluid and acid–base management.

Keywords
acid–base balance, colloids, crystalloids, strong ions, unmeasured ions

Curr Opin Crit Care 16:323–331

ß 2010 Wolters Kluwer Health | Lippincott Williams & Wilkins
1070-5295

Introduction Chemical properties of crystalloid fluids

Intravenous fluids are often treated by clinicians in a
fashion similar to that of oxygen. Both are common
commensals of admission to a healthcare facility, gener-
ally viewed as required elements of inpatient care, and
typically regarded as benign except in special circum-
stances. Fluid administration for critical illness manage-
ment, however, may be excessive, proinflammatory, and
potentially injurious [1,2,3]. These issues have come to
the fore in the settings of hemorrhagic and septic shock as
both conditions commonly entail large volume plasma
volume expansion (PVE). Indeed, it is more appropriate
to consider intravenous fluids as drugs with specific
indications, therapeutic windows and adverse effects.
Therefore, it is appropriate to understand the specific
chemical properties of the fluids one administers, includ-
ing the benefits as well as complications that follow
infusion. To this end, we will discuss general properties
of fluids, including how the specific ionic composition of
fluids influences their clinical effects, how fluids alter
plasma electrolytes, and how fluid infusion induced
electrolyte alterations impact plasma pH. Fluids for
human infusion may be either crystalloid or colloid,
and clinical care often uses a combination of each to
address plasma volume deficits, whereas maintenance
fluid is only provided as crystalloid infusion.
1070-5295 ß 2010 Wolters Kluwer Health | Lippincott Williams & Wilkins
Crystalloid fluids for human administration have three
general components: water, electrolytes, and sugar. As
sugars are generally provided as dextrose, utilized to
decrease gluconeogenic stress and reduce lean body mass
catabolism [4], and appropriate only for maintenance, but
not resuscitation, fluid administration (when provided in
low concentration, that is, 5% dextrose) our discussion of
these solutions will be narrow in scope. Dextrose infusion
in large quantity may induce an osmotic diuresis as the Km
for glucose transport is overwhelmed. This may lead to
confusion with regard to the significance of an increased
urine output during resuscitation. Thus, dextrose-contain-
ing fluids should only be provided as part of maintenance
fluid that is prescribed according to body weight and in
conjunction with overall nutritional support plans.
By contrast, resuscitation fluids are infused in large
volume at a rapid rate and designed to achieve physio-
logic endpoints very different from caloric support.
Unlike maintenance fluids which are provided in only
low volume and generally minimally perturb acid–base
and electrolyte balance except in the setting of organ
failure, resuscitation fluids administered in large volume
may have rather significant and even dangerous effects on
diverse systems, organ function and outcome.
DOI:10.1097/MCC.0b013e32833c0957

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 324 Intravenous fluids
Recall that all crystalloid fluids are prepared in a water
base. Thus, administration of 1000 ml of any crystalloid
provides 1000 ml of water regardless of the additional
electrolyte content. Although seemingly simple, this
cornerstone concept is often misunderstood when addres-
sing disorders of sodium balance, in particular dilutional
hyponatremia. Crystalloids for human infusion that are
used for resuscitation purposes are constructed with
sufficient electrolyte additives to approximate human
plasma osmolality and tonicity. These requirements help
support retention of the administered fluid and salt
within the vascular space, as opposed to the extravascular
extracellular space. The reader should note that in health,
approximately 75% of an administered normotonic crys-
talloid infusion extravasates into the extracellular extra-
vascular space; fluids of lesser tonicity demonstrate
greater extravasation and less resuscitative efficacy.
Assuming a typical 5000 ml circulating blood volume, a
10% PVE bolus (500 ml of retained fluid) requires the
administration of 2000 ml of a normotonic fluid, and is
consistent with advanced trauma life support directed
resuscitation [5].
Osmolality and tonicity
Proper fluid management is aided by understanding the
properties of osmolality and tonicity. Osmolality refers to
the concentraiton of osmotically active particles per
1000 ml of fluid and is generally expressed as millios-
moles (mOsm)/l of H2O; normal blood osmolality is 285–
295 mOsm/l H2O. Tonicity is a relative term reflecting a
comparison of the osmolality of a particular fluid to that of
plasma. Thus, a normotonic solution has an osmolality
similar to that of plasma – a general property of resuscita-
tion fluids such as lactated Ringer’s solution and 0.9%
normal saline solution (NSS). However, tonicity also
refers to the effects in a given environment with respect
to the permeability of membranes in that environment.
Thus, a solution that is isotonic to plasma may not be
isoosmolar if the particles in question are freely diffusible
and create no osmotic effect.
For example, if red blood cells are placed in a solution
containing 400 mOsm/l of urea they will behave comple-
tely differently if they are placed in a solution of
400 mOsm/l of NaCl. Placed in the hypertonic saline
solution the cells will shrink as water diffuses out of
Table 1 Common crystalloid resuscitation fluids
Fluid Na K Cl
Plasma 140 4 100
Lactated Ringer’s 130 4 109
0.9% NSS 154 0 154
Normosol-R 140 5 98
1/2 NSS þ 75 mEq NaHCO 152 0 77
3
0.9% NSS, normal saline solution; Lact, lactate; Acet, acetate; Glucon, gluconate; mOsm, milliosmoles/l; var, variable as this solution is locally prepared
without pH balancing; pH may be further influenced over time with CO2 equilibration; Plasma, normal human plasma.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
the cells—NaCl is not freely diffusible and creates an
osmotic gradient between the internal environment of
the cell and solution. By contrast, cells placed in a
solution of urea will swell, as urea is freely diffusible
and has no osmotic effect. This is why a solution contain-
ing dextrose may be isotonic (and hence not cause injury
to vessles when infused) but hypoosmotic when infused
into a patient. Commonly available resuscitation fluids
and their compositions are provided in Table 1.
Maintenance fluids are hypotonic by design, providing
electrolytes that meet the minimum daily requirements
for salt administration (see Fluids and Their Clinical
Effects). Some fluids are constructed to be hypertonic
and have specific uses such as hypertonic saline (HTS)
whose use is designed to provide concentrated salt with-
out an accompanying large volume of water for the acute
therapy of severe and symptomatic hponatremia. How-
ever, HTS may also be used as an initial resuscitation
fluid for those with traumatic brain injury when elevated
plasma osmolality is desired as part of a management
strategy for cerebral edema. As crystalloid fluids are
protein-free and comprised of only small particles, their
ability to remain in the vascular space is limited.
Water and electrolytes
Although crystalloids are water based, they contain a
variety of electrolytes. Each of the electrolyte additives
to the fluid base carries a charge in an aqueous milieu.
Using physicochemical principles articulated by Peter
Stewart in 1983 [6], and reviewed elsewhere [7–9], ions
in an aqueous milieu are termed strong ions if at physio-
logic pH, they separate from their partners and contribute
a charge to the aqueous milieu. Such strong ions are
commonly used to construct resuscitation and mainten-
ance fluids and include Naþ, Kþ, Cl, Ca2þ, Mg2þ and
PO42þ. As such, these electrolytes establish a net charge
that will influence plasma pH as they are infused intra-
venously (see Physicochemistry of Aqueous Solutions).
On the basis of the particular composition of electrolytes
in the fluid preparation, each fluid should impact plasma
pH in a predictable fashion that may be utilized in a
therapeutic fashion. Moreover, a particular therapeutic
goal may help guide the clinician to select a parti-
cular fluid for resuscitation or maintenance use, most
Ca Mg Lact Acet Glucon pH mOsm
9 2 2 0 0 7.4 285–295
2.7 0 28 0 0 6.5 273
0 0 0 0 0 5.5 308
0 3 0 27 23 7.4 280
0 0 0 0 0 var 302

commonly to impact plasma electrolyte composition and
pH (see how fluids influence pH below).
Chemical properties of colloids
Colloids are defined as preparations of homogeneous
noncrystalline substances that are dispersed throughout
another substance that is usually water based (for medical
use) [10]. The colloid may be large macromolecules or
smaller particles, but do not precipitate and are not
separable from their suspending solution by filtration
or centrifugation. Colloids are generally polydispersed,
representing a span of molecular sizes that characterize a
single preparation. Molecular weight may be described in
two different fashions:
(1) weight-averaged MW: (# molecules at each weight 
particle weight)/total weight of all molecules
(2) number-averaged MW: mean of all particle weights
Furthermore, the weight distribution pattern may be
assessed by the colloid oncotic pressure ratio, a ratio that
reflects the osmotic activity of a colloid solution across
membranes with different pore sizes.
In general practice, the size, persistence, efficacy at
plasma volume expansion, side effect profile, and of
course, product approval by regulatory agencies, tend
to govern product selection by clinicians. The clinician
should remain acutely aware that the colloid preparations
contribute very little free water to the patient’s system
and therefore, should always be utilized with mainten-
ance solutions to avoid inadvertently creating a hyper-
oncotic state leading to acute kidney injury or acute renal
failure [11]. Several different colloids are available and
have different chemical properties.
Starches
Starches are synthetic colloid preparations derived from
amylopectin extracted from either maize or sorghum.
Amylopectin is a D-glucose polymer that is synthetically
modified with hydroxyethyl substitutions at the second
carbon (C2) as well as the sixth carbon (C6) with rather
few substitutions occurring at the third carbon (C3);
Table 2 Common starch-based colloids used for resuscitation
Colloid MW/DS

Voluven HES 130/0.4
Volulyte HES130/0.4
Pentastarch HES 200/0.5

Hextend HES 670/0.7

Hespan HES 670/0.7
DS, degree of substitution; HES, hydroxyethyl starch; MW, molecular weight in kiloDaltons (kDa); NSS, 0.9% normal saline solution. Note: not all
colloids are available in the US. Food and Drug Administration approved colloids are indicated by ( ).
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Fluids, pH, ions and electrolytes Kaplan and Kellum 325
hydroxylation retards the rate of hydrolysis by plasma
nonspecific a-amylases. Starches are characterized by
their average molecular weight and average molecular
size, as they exist as a polydispersed preparation of
different molecular weight and sizes. Thus, starches
may be further classified by their average molecular
weight into high molecular weight (>450 kDa), medium
molecular weight (200kDa), and low molecular weight
(70–130 kDa). Furthermore, they are characterized by
the C2/C6 substitution ratio also known as the degree of
substitution or molar substitution ratio; ratios are expressed
as a number spanning 0–1. The greater the degree of
substitution or molar substitution ratio, the longer the
plasma persistence and plasma half-life (t / ). 1 2
By way of example, Hextend is a commercially available
starch used in the United States. It may be characterized
as a large molecular weight starch (670 kDa) with a high
degree of substitution (0.7). The last two characteristics
are the concentration of the preparation and the diluent
in which the colloid is prepared. Hextend is a 6% starch
preparation in a balanced salt solution. Changing the
diluent may change important consequences of admin-
istration rendering the product functionally different
[12]. For instance, Hextend’s predecessor, Hespan, is
the identical starch in every way but was prepared in a
0.9% saline base. Table 2 presents some commonly
utilized starch preparations and their key features.
Gelatins
Gelatins are preparations created from the hydrolysis of
bovine collagen and then further modified by either
succinylation (polygeline; Gelofusine) of urea-linkage
(Hemaccel) [13]. Succinylation results in no change in
molecular weight but a significant increase in molecular
size; no such changes occur with urea-linkage. The
diluents are different between the two products with
only Hemaccel being prepared with calcium and potass-
ium. It is important to note that the only cases of prion
related disease derived from cattle involve food-based
disease transmission, not pharmaceutical preparations.
Gelatins are principally used in the United Kingdom,
and have never been approved by the United States Food
and Drug Administration.
Concentration Diluent C2/C6
6% NSS 9:1
10%
6% Balanced solution 6:1
6% NSS 5:1
10%
6% Balanced solution 4.5 : 1
6% NSS 4.5 : 1

 326 Intravenous fluids
Dextrans
Dextrans are fairly homogeneous preparations of D-glu-
cose polymers principally joined by a-1,6 bonds creating
linear macromolecules that are characterized by their
concentration into two commercially available prep-
arations, Dextran 40 (molecular weight avg ¼ 40kDa)
and Dextran 70 (molecular weight avg ¼ 70kDa) [14].
The glucose moieties are derived from enzymatic clea-
vage of sucrose generated by Leuconostoc bacteria utilizing
dextran sucrase yielding high molecular weight detrains
that are modified into the final product using acid
hydrolysis and ethanol-based fractionation processes.
Clearance is proportional to molecular weight with 50–
55 kDa molecules being readily renally filtered and
excreted unchanged in the urine such that 70% of a
Dextran 40 dose is excreted unchanged over a 24-h period.
Molecules with a larger molecular weight undergo gastro-
intestinal clearance or cleavage within the reticuloen-
dothelial system via extant dextranases. Only Dextran
40 appears to have clinical use at present due to issues
with allergic reaction and bleeding with Dextran 70.
Combination preparations
Combinations of hypertonic saline and hyperoncotic
starch are available outside of the United States as well.
These preparations rely on starch PVE and the concen-
tration dependent movement of water from the extra-
vascular space to the intravascular domain on the basis of
creating a hypertonic plasma space. Their efficacy or
outcome advantage over other colloid solutions has yet
to be demonstrated.
Albumin
Albumin is a biologically active protein with a molecular
weight of 60 kDa that is normally found in human and
animal plasma. Used for PVE, it comes (in the United
States) as a 5% or 25% formulation with the sole difference
between the two formulations being that the dilute
solution is suspended most commonly in 0.9% saline with
or without other additives (e.g. acetate). Despite its rela-
tively small molecular size compared with the starches,
albumin continues to be commonly utilized for fluid
resuscitation. Its major evidence supported uses for
PVE include large volume paracentesis (>5 l), acute hepa-
tic failure in the pretransplant setting, and in combination
with antimicrobials for the management of spontaneous
bacterial peritonitis, as well as the management regimen in
a small number of studies for hepatorenal syndrome with or
without diuretics or vasoactive agents [15,16,17].
Ionic composition of fluids and their clinical
effects
Tables 1 and 2 list some common crystalloid and colloid
resuscitation fluids as well as their electrolyte compo-
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
sitions and key characteristics. It is important to note
some generalities regarding these fluids: many are pre-
pared in NSS, many are prepared in a solution similar to
lactated Ringer’s solution, and most have ionic compo-
sitions that are dissimilar in important ways from plasma.
Accordingly, a wide variety of plasma electrolyte abnorm-
alities arise when patients receive all of their fluids and
electrolytes via the intravenous route. In order to under-
stand how these abnormalities arise, one must understand
how the provided fluid and electrolytes differ from the
human minimum daily requirements for each.
Water
Humans require approximately 1.5–2.0 l of water per day
to maintain health and maintain a balance between
evaporative loss, external loss as water in urine and stool,
as well as water used to rebuild rapidly differentiating and
renewing cells such as those of the gastrointestinal
lumen. Excess intake as well as water generated as ‘water
of metabolism’ (approximately 1 l per day) is eliminated
as urine and is principally managed by altering the
activity of antidiuretic hormone at the level of the renal
tubules in response to changes in plasma Naþ con-
centration.
Sodium and potassium minimum requirements
An adult human requires 1–2 mEq Naþ, and 0.5 mEq Kþ
per kg body weight per day. Therefore, for a 70 kg
individual, between 70–140 mEq Naþ and 35 mEq Kþ
in a maintenance infusion would suffice to preserve
balance. Recall that 1000 ml of 0.9% NSS provides
9 gm of salt per day – the salt content of an inpatient
regular diet. For those NPO, these minimum require-
ments are conveniently provided as D5 1/2 NSS þ 20 mEq
KCl at a body weight calculated maintenance rate, which
for a 70 kg individual is 110 ml/h for a total of 2640 ml/
24 h. This fluid prescription provides 203.28 mEq Naþ
and 52.8 mEq Kþ, as well as 256.08 mEq Cl. Assuming
normal renal function, this approximation will not sig-
nificantly perturb normal homeostasis. However, under
the influence of increased ADH activity, tissue injury,
capillary leak, or septic physiology, homeostasis is per-
turbed and abnormalities of electrolyte balance result.
Sodium, potassium and chloride balance
The administration for intravenous solutions containing
electrolytes and water can obviously influence the con-
centrations of electrolytes in the body. Changes in
sodium, potassium and chloride are common and each
has specific consequences.
Dilutional hyponatremia
Dilutional hyponatremia is the most common inpatient
disorder of sodium balance. As most patients have
received maintenance fluid and many have also received

resuscitation fluid in the emergency department (ED),
operating room or ICU, it is the rare patient who has not
received salt far in excess of their minimum require-
ments. This excess is complicated by a commonly high-
ADH level that supports water retention, and may be
especially marked on those with heart failure. Thus, a
falling Naþ generally indicates free water excess, rather
than a true total body sodium deficit. Frequently, those
with dilutional hyponatremia have a normal or nearly
normal plasma Cl and a high urinary Naþ as well. The
therapy for this disorder is fluid restriction to decrease
free water, not the delivery of a higher Naþ content fluid.
Judicious diuretic use may also help correct dilutional
hyponatremia, and particular efficacy may be realized
with the use of the new class of diuretics – the aqua-
porins – based on their pure aquaretic effect; though
experience is limited with these agents in patients with
critical illness and the authors recommend limiting use to
refractory cases and only with careful monitoring even
then [18].
A special note is made regarding the correction of hypo-
natremia with regard to timing. The Naþ concentration
may be corrected at the same rate as that with which it
was acquired. When the condition has developed over
several days (chronic hyponatremia), one should avoid
raising the Naþ more rapidly than 0.5–1 mEq per hour to
avoid central pontine myelinolysis (CPM) [19]. This is
especially important in those with a Naþ less than
120 mEq/l for more than 48 h. CPM may result in per-
manent injury with devastating neurological effects.
Extrapontine demyelination may also occur. These time
rules also apply to true total body salt depletion. In
general, the rapidity of correction is also driven by the
presence or absence of neurologic symptoms. Three
percent saline is commonly used to raise the Naþ con-
centration above 120 mEq/l in symptomatic patients but
should always be used with careful monitoring of the
serum Naþ concentration.
Salt depletion hyponatremia
True total body salt depletion leading to hyponatremia is
less common than dilutional hyponatremia in the inpa-
tient setting. However, this disorder is more common in
those on chronic diuretic therapy coupled with a salt
restricted diet (less than 9 gm NaCl per day) but may also
accompany cerebral salt wasting as well as a variety of
medical therapies that result in urinary salt loss; more rare
causes include biliary drains and high output proximal
gastrointestinal fistula with loss of bile salts. In these
unique patient sets, both Naþ and Cl are reduced, and
are coupled with low urine Naþ concentration. Therapy
consists of salt administration by either the intravenous or
oral route, or both. The following formula may be used to
estimate the anticipated change in serum Naþ when
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Fluids, pH, ions and electrolytes Kaplan and Kellum 327
correcting hyponatremia using IVF.
D Na for 1 liter of IVF ¼ ½ðinfusate Na þ infusate KÞ
 serum Na=½Total body water
þ 1
Where total body water ¼ weight (kg)  correction factor
(1) Correction factors
(a) Children: 0.6
(b) Nonelderly men: 0.6
(c) Nonelderly women: 0.5
(d) Elderly men: 0.5
(e) Elderly women: 0.45
For example, a liter of NSS (Naþ 154 mEq) would result
in a modest 0.8 mEq increase in the serum Naþ concen-
tration when given to a 70 kg man with a serum Naþ of
120 mEq/l:
D Na ¼ ½ð154 þ 0Þ  120=½42 þ 1 ¼ 0:8
However, if 40 mEq/l of Kþ was added to the infusate, the
change would be 1.7 mEq:
D Na ¼ ½ð154 þ 40Þ  120=½42 þ 1 ¼ 1:7
This is because Kþ and Naþ are ‘exchangeable’ cations
and Kþ administration will increase the plasma Naþ
concentration by exchanging with intracellular Naþ.
Indeed, NSS with 40 mEq KCl is modestly hypertonic.
Failure to appreciate this effect can result in an overly
rapid correction of serum Naþ and can be dangerous.
Note that the earlier formula is used to determine the
effect of administration of a liter of fluid, in order to
determine the effect of administration of multiple liters
of fluid it is necessary to use the following formula:
New Na ¼ ½ðserum Na  total body waterÞ
þ infused Na=ðtotal body water
þ infused fluidÞ:
Hypernatremia
Except in cases of hypertonic salt administration (e.g.
injudicious use of 3% saline, ingestion of sea water),
hypernatremia is the result of a free water deficit. Like
hyponatremia, hypernatremia is divided into acute and
chronic with 48 h as the dividing time. Some general rules
apply to the safe correction of symptomatic hypernatremia:
(1) Correct no more rapidly than 1–2 mEq/l per hour
(2) Provide 50% of the water deficit in the first 12–24 h
and the rest over the next 24 h
(3) Measure electrolytes q 2 h during correction to adjust
the rate of correction to avoid cerebral edema
 328 Intravenous fluids
(4) Asymptomatic chronic hypernatremia should be cor-
rected at a rate not exceeding 0.5 mEq/l per hour, and
not more than 10mEq/l over 24 h.
Intravenous free water, commonly provided as D5W, is
most commonly used but may be supplemented by
gastrointestinal luminal free water using either pure
water, or diluted tube feeds. The same formula used
for correcting hyponatremia may be used to calculate free
water deficit:
D Na ¼ ½ðinfusate Na þ infusate KÞ
 serum Na=½Total body water þ 1
For example, administration of 1 l of D5W to a 70 kg man
with a serum Naþ of 160 would result in a decrease of
3.7 mEq/l assuming no loss of water.
Hypokalemia
Hypokalemia is much more common in hospitalized
patients than hyperkalemia. In those with normal renal
function, hypokalemia is often related to diminished
intake, the infusion of Kþ free fluids, and/or the use of
kaliuretic diuretics (i.e. loop diuretics such as furose-
mide). It is important to recall that serum Kþ deficit does
not demonstrate linearity with the amount needed to
restore a normal concentration. As Kþ is principally an
intracellular cation, extracellular deficits draw on intra-
cellular stores to maintain homeostasis. Thus, a total body
deficit exists when serum Kþ is less than 3.0 mEq/l, and
patients generally require 200 mEq Kþ (and often more)
to replace intracellular and extracellular Kþ to normal,
especially in the setting of ongoing renal or gastrointes-
tinal losses. As acute and potentially life-threatening
dysrhythmias are common with Kþ less than 3.0 mEq/l,
and the concentration of the replacement solutions is
typically higher than may be administered on the general
ward, continuous ECG monitoring is warranted. Addi-
tionally, restoration of normokalemia relies on the esta-
blishment of normomagnesemia as both Kþ and Mg2þ
cotransport in the kidney.
Hyperchloremia and hyperchloremic metabolic acidosis
As NSS is the most commonly prescribed fluid in the
United States, and 1L provides the minimum daily
requirements for maintenance electrolyte balance, let us
consider the impact of infusing a large volume of NSS into
a healthy individual with no acid–base disturbance, a
normal pco2 and normal pH before examining the results
from the same infusion into a patient with critical illness.
Using a prototypical 70 kg male, one may calculate total
body water as 42 l (70  0.6). Assuming standard electro-
lytes including a Naþ of 140 mEq/l and a Cl of 100 mEq/l,
we may calculate the total body Na (140  42 ¼ 5880 mEq)
and total body Cl (100  42 ¼ 4200 mEq). Infusing 10 l of
0.9% NSS (154 mEq/l of each Naþ and Cl) contributes an
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
additional 1540 mEq Naþ and 1540 mEq Cl into a new
total body water of 52 kg (42 kg þ 10 kg H2O from the NSS
infusion). Thus, one may derive the new Naþ concen-
tration by totaling the existing Naþ 5880 plus the infused
Na 1540 and dividing by the new water volume of 52 kg to
yield 142.7 mEq/l. However, the same calculation for Cl
yields 4200 þ 1540/52 ¼ 110.4 mEq/l. The net impact of
this infusion, in addition to an increase in plasma volume is
the induction of a hyperchloremic metabolic acidosis with
an important, but tolerable decrease in pH in this healthy
individual. However, large volumes such as those used in
this example are not provided to the healthy, but are
instead infused in those with critical illness or injury. In
this patient population, the pCO2 and pH prior to infusion
are typically abnormal, and the patients often suffer from
lactic acidosis, or other metabolic acidoses from preexisting
organ failure or ketoacidosis. These acidoses may be
aggravated by a rising pCO2 from impending respiratory
failure as well. In these circumstances, the induction of a
HCMA may be disastrous and potentially lethal.
Physical chemistry of aqueous solutions
Biological solutions are both water based and primarily
alkaline (as opposed to being acidic) in nature. Using
Stewart’s physicochemical approach to acid–base bal-
ance, there are only three independent variables that
determine pH in human plasma: pCO2; the strong ion
difference (SID); and the sum of the total weak acid
concentration (ATOT) [6,7]. The last two components
represent nonvolatile acids, whereas CO2 serves as a
volatile component. Note that HCO3 does not serve as
an independent determinant of pH.
Strong ions
Plasma contains a variety of ions that are classified as
positive (cations) or negative (anions) depending on their
charge. Some ions, as noted earlier, exist completely
dissociated from their ionic partners at physiologic pH
and are termed strong ions to separate them from weak
ions, species that may exist in dissociated or associated
forms; only the dissociated species manifest a charge.
Strong ions include Naþ, Kþ, Caþþ, Mgþþ, Cl, and
lactate (LA), whereas albumin and PO42 are weak ions.
By way of example, a solution of only water and NaCl, has
equal amount of strong cations (Naþ) and strong anions
(Cl), establishing a SID of zero. Normal plasma does not
have a SID of zero as in the earlier example. Instead
plasma has more strong cations (mainly Naþ) than strong
anions (mainly Cl). The difference between the sum of
all strong cations minus the sum of all strong anions is
known as the strong ion difference (SID).
Plasma SID normally ranges between 40–42 mEq/l in
health but is typically much lower in those with critical
illness. As SID is positive, it must be balanced by an equal

negative charge to conform to the principle of electrical
neutrality. These balancing charges stem from both CO2
and the weak acids (A). In the absence of renal failure or
ketoacidosis, the total weak acid concentration derives
principally from albumin’s exposed histidine residues
and PO42. As weak acids, ATOT exists in both associated
and dissociated forms represented by the following
equation: ATOT ¼ AH þ A. In order to preserve electo-
neutrality, it then follows that plasma SID may be
derived from the remaining negative charges: SID –
(CO2þA) ¼ 0. This estimated value of SID is termed
the effective SID (SIDe) and is also known as ‘buffer
base’ [20]. These relationships also enable one to use the
base excess to understand an additional aspect of the
SID. As the base excess is the change in ‘buffer base’
required to reestablish a pH of 7.4 when pCO2 ¼ 40 torr
base excess identifies the change require in SID to
normalize pH [21].
A more clinically appealing and clinically malleable
method of estimating SID is (Naþ þ Kþ þ Caþþ þ
Mgþþ) – (Cl þ lactate). This value is termed the
‘apparent’ SID (SIDa) reflecting the fact that some
‘unmeasured’ ions might also be present. Although both
SIDe and SIDa are convenient and fairly readily measur-
able, neither value is a completely accurate representa-
tion of the actual SID. The discrepancy stems from the
presence of clinical entities that generate unmeasured
ions such as the sulfates and ketones mentioned earlier.
Regulation of the strong ion difference
It is important to recall that regulation of the SID utilizes
three interrelated organ systems: the kidney, the liver and
the gastrointestinal tract. The kidney is the dominant
organ and discharges its role principally via Cl excretion.
With each Cl that is not reclaimed after filtration, the
SID will increase, and so will pH (hypochloremic alka-
losis). As normal dietary intake is rich in strong ions,
gastrointestinal uptake is sufficient to maintain a normal
balance. As regulation of intravascular volume, especially
during critical illness and after injury establishes a strong
imperative for Naþ uptake and reclamation, and Kþ
handling is driven by plasma concentration as well as
glucose and insulin levels, renal Cl handling is especi-
ally important in pH maintenance [22].
Albumin and the weak acids
The second nonvolatile determinant of blood pH is the
total weak acid concentration (ATOT). The weak acids,
are mostly proteins (predominantly albumin) and phos-
phates, and they contribute the remaining charges to
satisfy electroneutrality such that SID – (CO2 þ
A) ¼ 0. As A varies with changes SID and pCO2, A
cannot be an independent variable. Instead, the inde-
pendent variable is the sum of the associated and dis-
sociated weak acids represented by ATOT. It is important
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Fluids, pH, ions and electrolytes Kaplan and Kellum 329
to note that the loss of weak acid (ATOT) is an alkaliniz-
ing process [23]. However, the critically ill are only
infrequently alkalemic despite often profound hypo-
albuminemia. This seemingly counterintuitive obser-
vation is understood by recalling that the SID and
ATOT are best evaluated in relation to one another
rather than as absolute values. Hypoalbuminemic
patients often also manifest a reduced SID, perhaps
as compensation for their reduced ATOT from hypo-
albuminemia [24,25].
How do fluids influence plasma pH?
Metabolic acidoses and alkaloses are categorized accord-
ing to the ions that are responsible. Thus, there is lactic
acidosis and chloride responsive alkalosis, etc. It is
important to recognize that metabolic acidosis is pro-
duced by a decrease in the SID [26]. A decrease in
SID may be brought about by the generation of organic
anions (e.g. lactate, ketones) the loss of cations (e.g.
diarrhea), the mishandling of ions (e.g. renal tubular
acidosis) or the addition of exogenous anions (e.g. iatro-
genic acidosis, poisonings). By contrast metabolic alka-
loses occur as a result of an inappropriately large SID,
although the SID need not be greater than the ‘normal’
40–42 mEq/l. This may be brought about by the loss of
anions in excess of cations (e.g. vomiting, diuretics), or
rarely by administration of strong cations in excess of
strong anions (e.g. transfusion of large volumes of banked
blood).
In the acute setting, acidosis is usually more of a problem
than alkalosis and in the critically ill, the most common
sources of metabolic acidosis are disorders of chloride
homeostasis, lactate, and other anions. Hyperchloremic
metabolic acidosis occurs either as a result of chloride
administration or secondary to abnormalities in chloride
handling or related to movements of chloride from one
compartment to another. The effect of chloride admin-
istration on the development of metabolic acidosis has
been known for many years [27,28]. Recently, new
attention has been paid to this area in light of better
understanding of the mechanisms responsible for this
effect [29–32]. It has now been shown in animal models
of sepsis [29] and in patients undergoing surgery [23,
30–32] that 0.9% saline causes metabolic acidosis not by
‘diluting’ HCO3 but rather by its chloride content
[33,34]. From a physical chemical prospective this is
completely expected. HCO3 is a dependent variable
and cannot be the cause of the acidosis. Instead, Cl
administration decreases the SID (an independent vari-
able) and produces a decrease in pH. The reason this
occurs with 0.9% saline administration is that although
0.9% saline contains equal amounts of both Naþ and Cl,
the plasma does not. When large amounts of salt are
added, the Cl concentration increases much more than
 330 Intravenous fluids
the sodium concentration as per the example provided
earlier.
There are other important causes of hyperchloremia
(renal tubular acidosis, diarrhea, etc.) and in addition,
this form of metabolic acidosis is common in critical
illness, especially sepsis. Although 0.9% saline resuscita-
tion undoubtedly plays a role, there appear to be unex-
plained sources of Cl, at least in animal models of sepsis
[29]. One possible explanation is that this Cl is coming
from intracellular and interstitial compartments as a
result of the partial loss of Donnan equilibrium due to
albumin exiting the intravascular space [29]. However,
this hypothesis is yet unproven.
Consequences of hyperchloremia: immune
activation and inflammation
A growing body of literature documents that crystalloid
fluids are proinflammatory and serve as potent immune
activation triggers [35,36]. Importantly, evidence sug-
gests, at least in a cell-culture model, that the impact
of an induced acidosis hinges not only on pH, but also
how the acidosis is induced with differential effects noted
when comparing lactic acidosis to HCMA [2]. Intact
animals with sepsis exhibit increased cytokine pro-
duction when HCMA is produced via infusion of
chloride-rich fluids [1]. Buttressing this putative differ-
ence are clinical data identifying a higher in-hospital
mortality rate for patients admitted to an ICU with lactic
acidosis versus those with HCMA [37]. Other lines of
evidence identify that crystalloids serve also as potent
triggers of macrophage stimulation, and activation of
MAP-kinase, P38, and NF-KB pathways when compared
with starch-based colloids or albumin [35].
Although the significance of these findings in the clinical
circumstance remains to be clarified, the impact of not
understanding the genesis of an induced acidosis has
been quanitfied. In one study comparing two different
methods of assessing acid–base balance in critically ill
and injured ICU patients, misinterpretation of a falling
plasma bicarbonate (HCO3) or a more negative base
deficit (i.e. increasing base excess) as evidence of hypo-
perfusion instead of a HCMA led to 110 l of excess and
unnecessary fluids prescribed for PVE in 34 patients. The
untoward effects of unnecessary PVE are measured in
edema, anasarca, ventilator length of stay, as well as a
rising incidence of secondary abdominal compartment
syndrome due to visceral edema and acute large volume
ascites [38,39].
Conclusion
It is important to remember that intravenous fluids are
drugs. They should be prescribed with care and with
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
attention to the potential for harm as well as benefit.
Intravenous fluids can correct or induce electrolyte and
acid–base abnormalities. In experienced hands, intra-
venous fluids can be life saving—if used improperly, like
most drugs, they can lead to adverse effects and contribute
to morbidity and even mortality in critically ill patients.
References and recommended reading
Papers of particular interest, published within the annual period of review, have
been highlighted as:
 of special interest
 of outstanding interest
Additional references related to this topic can also be found in the Current
World Literature section in this issue (pp. 394–395).
1 Kellum JA, Song M, Venkataraman R. Effects of hyperchloremic acidosis on
arterial pressure and circulating inflammatory molecules in experimental
sepsis. Chest 2004; 125:243–248.
2 Kellum JA, Song M, Li J. Lactic and hydrochloric acids induce different
patterns of inflammatory response in LPS-stimulated RAW 264.7 cells. Am
J Physiol Regul Integr Comp Physiol 2004; 286:R686–R692.
3 Kaplan LJ, Maerz LL, Schuster K, et al. Uncovering system errors using a rapid
 response team: cross-coverage caught in the crossfire. J Trauma 2009;
67:173–178; discussion 178–179.
This manuscript details the relevant impact of ACGME work hours restrictions as it
applies to postoperative care. One major deleterious impact was inadequate fluid
management.
4 Wasserman D. Four grams of glucose. Am J Physiol Endocrinol Metab 2009;
E11–E21.
5 Shock. Advanced Trauma Life Support, 8th ed. In: J Fildes and J Wayne
Meredith, editors. American College of surgeons Committee on Trauma
2008. pp. 55–71.
6 Stewart P. Modern quantitative acid-base chemistry. Can J Physiol Pharmacol
1983; 61:1444–1461.
7 Kellum JA, Kramer DJ, Pinsky MR. Strong ion gap: a methodology for exploring
unexplained anions. J Crit Care 1995; 10:51–55.
8 Kellum JA. Clinical review: reunification of acid-base physiology. Crit Care
2005; 9:500–507.
9 Kaplan LJ, Kellum JA. Comparison of acid base models for prediction of early
mortality following trauma. Shock 2008; 29:662–666.
10 Roche AM, James MF. Colloids and crystalloids: does it matter to the kidney?
 Curr Op Crit Care 2009; 15:520–524.
Given the recent focus on acute kidney injury and acute renal failure, this review
addresses fluid selection as it applies to renal injury.
11 Lipfert B, Standl T, Dullmann J, et al. Histology and ultrastructure of liver and
kidney following blood exchange with ultrapurified, polymerised bovine hemo-
globin in comparison with hydroxyethyl starch. Lab Invest 1999; 79:573–582.
12 Wilkes NJ, Woolf R, Mutch M, et al. The effects of balanced versus saline-
based hetastarch and crystalloid solutions on acid-base and electrolyte status
and gastric mucosal perfusion in elderly surgical patients. Anes Analg 2001;
93:811–816.
13 Ertmer C, Rehberg S, Van Aken H, Westphal M. Relevance of nonalbumin
 colloids in intensive care medicine. Best Pract Res Clin Anaesthesiol 2009;
23:193–212.
This is a balanced review of available nonalbumin colloids for use in the intensive
care or high-dependency unit.
14 Bunn F, Trivedi D, Ashraf S. Colloid solutions for fluid resuscitation. Cochrane
Database Syst Rev 2008. doi: 001319.
15 Luca A, Garcia-Pagan JC, Bosch J, et al. Beneficial effects of intravenous
albumin infusion on the hemodynamic and humoral changes after total
paracentesis. Hepatology 1995; 22:753–758.
16 Fernandez J, Monteagudo J, Bargallo X, et al. A randomized unblinded pilot
study comparing albumin versus hydroxyethyl starch in spontaneous bacterial
peritonitis. Hepatology 2005; 42:627–634.
17 Charlton MR, Wall WJ, Akinlolu OO, et al. Report of the first international liver
 transplantation society expert panel consensus conference on renal insuffi-
ciency in liver transplantation. Liver Transpl 2009; 15:S1–S34.
As the first such publication from this multispecialty group, this manuscript is
comprehensive and provides an up-to-date review of the available evidence
addressing renal-hepatic dynamics. A well done and detailed review of fluid
management as it applies to hepatorenal syndrome, tense ascites, and hepatic
failure is provided.

18 Hasler U, Leroy V, Martin PY, Feraille E. Aquaporin-2 abundance in the renal
collecting duct: new insights from cultured cell models. Am J Physiol Renal
Physiol 2009; 297:F10–F18.
19 Sterns RH, Silver S, Kleinschmidt-DeMasters BK, Rojiani AM. Current per-
spectives in the management of hyponatremia: prevention of CPM. Expert
Review of Neurotherapeutics 2007; 7:1791–1797.
20 Singer RB, Hastings AB. An improved clinical method for the estimation of
disturbances of the acid-base balance of human blood. Medicine (Baltimore)
1948; 27:223–242.
21 Kellum JA, Bellomo R, Kramer DJ, Pinsky MR. Fixed acid uptake by visceral
organs during early endotoxemia. Adv Exp Med Biol 1997; 411:275–279.
22 Mackenzie W. Roles of urea production, ammonium excretion, and amino acid
oxidation in acid-base balance. Am J Physiol 1986; 250 (2 Pt 2):F181–F188.
23 Morgan TJ, Venkatesh B, Hall J. Crystalloid strong ion difference determines
metabolic acid-base change during in vitro hemodilution. Crit Care Med
2002; 30:157–160.
24 Kellum JA. Determinants of blood pH in health and disease. Critical Care
2000; 4:6–14.
25 Wilkes P. Hypoproteinemia, SID, and acid-base status in critically ill patients.
J Appl Physiol 1998; 84:1740–1748.
26 Kaplan L, Kellum JA. Initial pH, base deficit, lactate, anion gap, strong ion
difference, and strong ion gap predict outcome from major vascular injury. Crit
Care Med 2004; 32:1120–1124.
27 Wilcox CS. Regulation of renal blood flow by plasma chloride. J Clin Invest
1983; 7:726–735.
28 Williams EL, Hildebrand KL, McCormick SA, Bedel MJ. The effect of intra-
venous lactated ringer’s solution versus 0.9% sodium chloride solution on
serum osmolality in human volunteers. Anesth Analg 1999; 88:999–1003.
29 Kellum JA, Bellomo R, Kramer DJ, Pinsky MR. Etiology of metabolic acidosis
during saline resuscitation in endotoxemia. Shock 1998; 9:364–368.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Fluids, pH, ions and electrolytes Kaplan and Kellum 331
30 Scheingraber S, Rehm M, Sehmisch C, Finsterer U. Rapid saline infusion
produces hyperchloremic acidosis in patients undergoing gynecologic sur-
gery. Anesth 1999; 90:1265–1270.
31 Waters JH, Bernstein CA. Dilutional acidosis following hetastarch or albumin
in healthy volunteers. Anesth 2000; 93:1184–1187.
32 Waters JH, Miller LR, Clack S, Kim JV. Cause of metabolic acidosis in
prolonged surgery. Crit Care Med 1999; 27:2142–2146.
33 Ring T, Kellum JA. Comment on ‘Changes in acid-base balance following
bolus infusion of 20% albumin solution in humans’ by Bruegger et al. Int Care
Med 2006; 32:480; author reply 481–482.
34 Bruegger D, Jacob M, Scheingraber S, et al. Changes in acid-base balance
following bolus infusion of 20% albumin solution in humans. Int Care Med
2005; 31:1123–1127.
35 Rhee P, Wang D, Ruff P, et al. Human neutrophil activation and increased
adhesion by various resuscitation fluids. Crit Care Med 2000; 28:74 –
78.
36 Bulger EM, Cuschieri J, Warner K, Maier RV. Hypertonic resuscitation
modulates the inflammatory response in patients with traumatic hemorrhagic
shock. Ann Surg 2007; 245:635–641.
37 Gunnerson KJ, Saul M, He S, et al. Lactate versus nonlactate metabolic
acidosis: a retrospective outcome evaluation of critically ill patients. Crit Care
2006; 10:R22–R26.
38 Kaplan LJ, Cheung NH-T, Maerz LL, et al. A physicochemical approach to
 acid-base balance in critically ill trauma patients minimizes errors and reduces
inappropriate plasma volume expansion. J Trauma 2009; 66:1045–1051.
One of the few manuscripts evaluating the benefits of fluid management using a
physico-chemical approach in the critically ill injured patient. Specific benefits
include reductions in unnecessary plasma volume expansion, and more rapid
correction of acid-base abnormalities.
39 Maerz L, Kaplan LJ. Abdominal compartment syndrome. Crit Care 2008; 36
(Suppl):S212–S215.