Alan D. Kaye and James M.
Riopelle
Intravascular Fluid and
54 Electrolyte Physiology
Key Points
1. Water is the major component of all fluid compartments
within the body and represents approximately 60% of
body weight.
2. Sodium is the most abundant positive ion of the
extracellular fluid (ECF) compartment and is crucial in
determining the extracellular and intracellular osmolality.
3. Potassium is the most abundant positive ion in the
intracellular fluid and plays an important role in the
membrane potential of cells.
4. Calcium is the key component that mediates muscle
contraction; exocrine, endocrine, and neurocrine
secretion; cell growth; and the transport and secretion
of fluids and electrolytes.
5. Magnesium is essential for many biochemical reactions;
its pharmacologic properties have only more recently
been appreciated.
6. Phosphate stores and releases energy through high-
energy phosphate bonds and is integral to the structure
of proteins, lipids, and bone.
7. Chloride is the predominant anion in the ECF.
Water Physiology
Water is the major component of all fluid compartments within
the body (see Chapter 88). Total body water represents approxi-
mately 60% of the bodys total weight in an average adult. In a
70-kg man, total-body water is about 600mL/kg, or 40L. The
percentage of water varies significantly, however, with age, gender,
and adiposity because muscle contains 75% water, whereas
adipose tissue contains only 10% water. The percentage water
content of a fetus is higher initially, but decreases progressively
during late gestation and the first 3 to 5 years of life.
Total-body water may be divided into its two basic compo-
nents, intracellular and extracellular. The compartments are sepa-
rated by water-permeable cell membranes. In a 70-kg man, the
intracellular fluid volume averages 400 to 450mL/kg (about 30L),
and the extracellular fluid (ECF) volume averages 150 to 200mL/
8. Glucose is a crucial fuel source, and insulin facilitates
glucose movement into cells in a process that also
requires potassium and phosphate.
9. Diabetes affects multiple organ systems, and the
perioperative effect of diabetes can be profound.
10. The most common causes of metabolic alkalosis are
antacid therapy, incidental administration of citrate with
blood products, sodium bicarbonate administration,
gastric drainage, and renal bicarbonate retention.
11. Metabolic acidosis is commonly caused by low cardiac
output and end-stage liver disease.
12. Transfusion of blood products improves tissue
oxygenation and decreases bleeding, but it also
increases the risk of transmission of infectious diseases,
transfusion reactions, immunosuppression, and
alloimmunization.
13. Anesthetics may blunt the normal physiologic responses
to hypovolemia and the stress response.
14. Shock is dysfunction of intracellular processes caused by
the lack of energy.
kg (about 14L). ECF volume is greater in young individuals and
men than in elderly individuals and women (Table 54-1). The
blood volume is 60 to 65mL/kg, and is distributed as 15% in the
arteries and 85% in the venous system.
The major components of the extracellular compartment
are the plasma volume (30 to 35mL/kg) and the interstitial fluid
(120 to 165mL/kg). Other constituents of the ECF include pleural
fluid, peritoneal fluid, aqueous humor, sweat, urine, lymph, and
cerebrospinal fluid. Plasma warrants special consideration because
of its importance in clinical practice. Plasma is the noncellular
component of blood and is continuously seeking equilibrium
with the interstitial fluid. A major difference of plasma compared
with interstitial fluid is the much higher concentration of pro-
teins. This higher concentration of protein results in plasma
having an osmotic pressure 20mmHg greater than the interstitial
fluid and ECF. This gradient helps maintain the intravascular
volume. The ECF compartment contains higher concentrations
1705
IV 1706 Anesthesia Management
Table 54-1 Electrolyte Composition in Body Fluids (Normal)
Interstitial Fluid Intracellular
Electrolyte Plasma (mEq/L) (mEq/L) Fluid (mEq/L)
Na+ 142 145 10
K+
4 4 159
2+
Mg 2 2 40
Ca 2+
5 3 1 1mOsm of solute excreted by the kidney. Normal kidneys respond
Cl 103 117 10
HCO3
25 27 7 tion, the kidney responds with antinatriuresis and antidiuresis
Adapted from Campbell I: Physiology of fluid balance. Anaesth Intensive Care Med
7:462-465 2006.
of sodium, chloride, and bicarbonate. Permeability to ions and
proteins varies with each organ, with the brain having the lowest
and the liver having the highest permeability.
Control of body water and its composition is multifactorial
and involves many factors, including atrial natriuretic peptide
(ANP), vasopressin (i.e., antidiuretic hormone [ADH]), aldoster-
one (e.g., renin, angiotensin), parathyroid hormone, calcitonin,
prostaglandins, dopaminergic receptors, -adrenergic receptors,
the thirst mechanism, and intrinsic renal properties. Water balance
describes the difference between water intake and water loss. The
kidneys are the major regulators of water output.
Approximately 60% of the daily water loss is excreted in
the urine. At high ambient temperatures or with significant exer-
cise, the amount of water lost by sweating increases and may
reflect most of the total daily water loss (Table 54-2). Heavy exer-
cise can increase the loss of water through sweating 50 times the
normal rate. Increased ventilation amplifies the insensible loss of
water through the respiratory tract. Under these conditions, renal
water loss decreases to compensate for the increased sweating and
insensible water losses.1
Sodium Physiology
Sodium is the most abundant positive ion of the ECF and is
crucial in determining the extracellular and intracellular osmolal-
ity. Serum osmolality is tightly regulated between 275 and
290mOsm/kg, primarily through the influence of ADH, and is
estimated by the following equation:
Serum osmolality = (2 Na ) + ( glucose 18) + (urea 2.8)
Osmoreceptors in the hypothalamus can detect changes of
1%. Most patients with hyponatremia have urine osmolality of
more than 200mOsm/kg, reflecting impairment in water excre-
tion.2 With regard to sodium physiology, the ratio of sodium
between plasma and interstitial fluid is roughly 5:1 at equilib-
rium. The distribution equilibrium is usually complete within 15
to 20 minutes. Extracellular sodium balance is determined by
sodium intake relative to sodium excretion. Most humans
consume far more salt than they need. In normal, healthy indi-
viduals, the kidneys main function in sodium balance is excretion
of excess sodium.
Sodium requirements vary with age. For infants born
before 32 weeks gestation, requirements are 3mEq/kg/day, and
at term, requirements are 2 to 3 mEq/kg/day (see Chapters 82
and 84).3 Neonatal stool losses are about 1mEq/kg/day, and the
growth process uses 0.5mEq/kg/day. Adult requirements decrease
to about 1.5mEq/kg/day. Urinary sodium excretion represents
most of the sodium loss and approximately equals the daily intake
of sodium. Extrarenal sodium loss by profuse sweating, burns,
severe vomiting, or diarrhea is important.
Normally, 10mL of urine can initially be formed for each
to a free water challenge with diuresis and to a sodium load with
natriuresis; if there is decreased sodium intake or volume deple-
(e.g., a patient undergoing surgery may excrete only 1.2 to
1.6mOsm/mL of solute). In various pathophysiologic processes,
abnormally high or low urinary sodium excretion can occur.
Many of the factors that control tubular sodium reabsorp-
tion are affected during the perioperative period, including hemo-
dynamic and physical factors (e.g., increased intra-abdominal
pressure during laparoscopic procedures), hormonal factors, and
renal sympathetic nerve activity. The balance of Starling forces
(hydrostatic versus osmotic) is responsible for sodium and water
transport across peritubular capillary walls. The net pressure in
the peritubular capillaries is about 10mmHg in favor of uptake
of reabsorbed fluid. Volume expansion with isotonic saline
decreases plasma protein concentration and reduces colloidal
osmotic pressure (COP) in peritubular capillaries.
The renin-angiotensin system (RAS) is involved in control
of arterial blood pressure and blood volume along with the sym-
pathetic nervous system, the kinin-kallikrein system, and arginine
vasopressin. The RAS plays a role in sodium homeostasis and
renal function, particularly under stress conditions. The RAS may
be initiated by decreases in blood pressure in the renal artery, by
decreases in sodium delivery to the macula densa, or by sympa-
thetic nervous system activation. In response, renin is synthesized
from its precursor, prorenin, and secreted by the juxtaglomerular
cells of the kidney. Renin, an aspartyl-protease (similar to pepsin
and cathepsin), cleaves its substrate, angiotensinogen, which is an
2-globulin, to generate the decapeptide angiotensin I. Although
renin is mostly produced by the kidney, renin isoenzymes have
been found in many tissues, including brain, adrenals, vascular
beds, uterus, and placenta.4 The gene for human renin has been
cloned. Angiotensinogen levels are increased after nephrectomy,
by estrogens, by thyroid hormones, by glucocorticoids, and during
angiotensin Iconverting enzyme inhibition.5,6
Angiotensin I is rapidly converted to the octapeptide angi-
otensin II by angiotensin Iconverting enzyme or by an endopepti-
Table 54-2 Daily Loss of Water
Normal Activity Normal Activity Prolonged
Source of and Temperature High Temperature Exercise
Loss (mL) (mL) (mL)
Urine 1400 1200 500
Sweat 100 1400 5000
Feces 100 100 100
Insensible 700 600 1000
losses
Total 2300 3300 6600
From Rhoades RA, Tanner GA: Medical Physiology. Boston, Little, Brown, 1995.

dase. The pulmonary circulation seems to be the major site of
angiotensin-converting enzyme activity, although angiotensin-
converting enzyme also is found in the vascular endothelium of
the heart, kidney, adrenal cortex, testes, and brain.7
Angiotensin II is a potent vasopressor that stimulates
aldosterone secretion through the adrenal cortex. Angiotensin II
has a vagal inhibitory effect and causes ganglionic stimulation.8-10
Angiotensin II partially suppresses renin secretion by a direct
effect on the juxtaglomerular cells. Angiotensin II may stimulate
a local increase of adenosine, a known inhibitor of renin release,
and may participate in a negative feedback loop whereby angi-
otensin II limits its own biosynthesis.11 Angiotensin II is degraded
in the plasma to the carboxyl-terminal heptapeptide angiotensin
III or the amino-terminal heptapeptide angiotensin (1-7), both
of which seem to be biologically active.12 A decline in arterial
blood pressure, a decrease in sodium delivery to the macula
densa, or sympathetic stimulation may activate the RAS, generat-
ing angiotensin II. This results in an increase in blood pressure
and sodium retention caused by enhanced aldosterone secretion.
The RAS does not have an active role in maintaining blood pres-
sure in a normal, sodium-replete, intact individual. Hydrostatic
forces act to maintain a stable glomerular filling pressure. This
activity affects venous return, cardiac output, and blood pressure.
Superimposed on this mechanism under stress conditions are
various neurohumoral control systems, including the sympathetic
nervous system, ADH, atrial natriuretic hormone, and prostag-
landins (Table 54-3).
Nitroprusside-induced hypotension is associated with
increases in renin activity and marked elevations in the plasma
concentrations of ADH, which are not seen with the older agent
trimethaphan used to induce blood pressure reductions.13 Con-
versely, administration of the -adrenergic blocker propranolol
during nitroprusside-induced hypotension prevents increases in
plasma renin activity.
Hyponatremia
Hyponatremia, typically defined as a plasma sodium concentra-
tion less than 135mEq/L, can result from excessive loss of sodium
from excessive sweating, vomiting, diarrhea, burns, and the
administration of diuretics. The most common cause of
Table 54-3 Major Causes of Natriuresis and Antinatriuresis
Mechanism Clinical Syndromes
Natriuresis
Volume-expanded state High sodium intake, inappropriate antidiuretic
hormone secretion
Volume-depleted state Addisons disease, renal salt wasting, diuretic
excess/abuse
Antinatriuresis
Edematous states Heart failure, chronic liver disease, nephrotic
syndrome, acute glomerulonephritis,
idiopathic edema
Nonedematous states Hemorrhage, low sodium intake, diuretic
withdrawal, acute mineralocorticoid
administration, nonrenal sodium loss by
sweating or vomiting or both
Intravascular Fluid and Electrolyte Physiology 1707 54
Table 54-4 Causes of Hyponatremia as Related to Intravascular
Volume Status
Hypovolemic Hypervolemic Euvolemic
Hemorrhage Congestive heart failure SIADH
Burn wound edema Nephrotic syndrome Pseudohyponatremia
Peritonitis Cirrhosis syndrome
Cerebral salt-wasting TURP syndrome
Section IV Anesthesia Management
syndrome
SIADH, syndrome of inappropriate secretion of antidiuretic hormone; TURP,
transurethral resection of prostate.
hyponatremia is an excess of total-body water, not a deficiency of
total-body sodium. Under normal conditions, water intake rarely
exceeds the kidneys ability to excrete free water. Hyponatremia
is usually associated with nonphysiologic vasopressin (i.e., ADH)
release and impairment of the renal diluting capacity.
Normal ADH secretion is mediated by an increase in
plasma osmolality greater than 280mOsm or a decrease in effec-
tive circulating volume. Other factors that result in release of
ADH include pain, sympathetic stimulation, and nausea. After
release, ADH binds to the V2 receptors in the medullary collect-
ing ducts. ADH increases the water permeability of these seg-
ments by facilitating the fusion of water channels to the apical cell
membrane.14,15 When ascertaining the cause of hyponatremia,
serum osmolality must be obtained to determine whether the
hyponatremia reflects true hypotonicity; this gives insight into
the underlying cause of low sodium levels. Hyponatremia can
occur in patients who are hypotonic, normotonic, or hypertonic.
Patients intravascular volume status must be evaluated to under-
stand further the underlying problem leading to abnormalities in
sodium physiology (Table 54-4).
Factitious hyponatremia (normotonic hyponatremia) can
be seen in hyperlipidemia (i.e., chylomicronemia) or hyperpro-
teinemia. Hyperosmolality resulting from nonsodium molecules
(e.g., hyperglycemia, mannitol overdose) draws water from the
intracellular space to dilute the intravascular sodium concentra-
tion. Significant decreases in total-body sodium most commonly
occur from diuretic administration.
Transurethral resection of prostate (TURP) syndrome is a
recognized cause of hyponatremia (see Chapter 65). TURP syn-
drome is caused by intravascular absorption of irrigation solu-
tion, which typically contains glycine, but no other electrolytes.16
The absorption of free water causes hyponatremia because of the
dilution of serum sodium. TURP syndrome is a case of true
hypotonic hyponatremia and is defined as a serum sodium level
of less than or equal to 125mmol/L with two or more associated
clinical symptoms or signs, which can be masked under general
anesthesia. A similar syndrome has been reported in 6% of
women undergoing transcervical endometrial ablation and in
patients undergoing neuroendoscopic ventriculostomy.17
The syndrome of inappropriate secretion of ADH (SIADH)
is associated with numerous processes, including pulmonary and
cranial disorders and several neoplasms, particularly oat cell car-
cinoma of the lung. Sympathetic activation (e.g., from postopera-
tive pain) also can lead to significantly increased levels of ADH
in the absence of volume constriction. SIADH may occur with
administration of various drugs, including oral hypoglycemics,
tricyclic antidepressants, and diuretics.18 High levels of ADH are
IV 1708 Anesthesia Management

secreted intermittently at an abnormally low threshold or con- Table 54-5 Major Causes of Hyponatremia
tinuously despite low osmolality. The presence of hyponatremia
Mechanism Clinical Syndromes
plus a urine osmolality higher than maximal dilution confirms
the diagnosis. In patients with SIADH, the urinary sodium con- Pseudohyponatremia (normal Hyperlipidemia, hyperproteinemia
centration usually exceeds 30mEq/L, the fractional excretion of osmolality)
sodium is greater than 1%, and the serum uric acid is reduced. Dilutional hyponatremia due to Hyperglycemia, mannitol administration
Patients with SIADH exhibit a characteristic response to water serum hyperosmolality
restriction; a 2- to 3-kg weight loss is accompanied by correction
of hyponatremia and cessation of salt wasting over 2 to 3 days. True hyponatremia with normal -globulins, lithium, THAM
effective osmolality
This is another example of true hypotonic hyponatremia.
Hyponatremia also occurs in mixed disorders, in which True hyponatremia with edema Congestive heart failure, nephrotic
nonosmotic ADH release and reductions in the rate of urinary (sodium excess)low effective syndrome, cirrhosis of the liver,
sodium excretion blunt urinary diluting capacity. This can occur osmolality idiopathic edema, hypoalbuminemia
in advanced volume contraction, intractable heart failure, and (e.g., due to malnutrition)
advanced hepatic cirrhosis with ascites. True hyponatremia (sodium Renal wasting (diuretics, renal tubular
It is crucial for the anesthesiologist to recognize the signs depletion)low effective acidosis, cerebral salt-wasting
and symptoms associated with hyponatremia, particularly for osmolality syndrome); extrarenal wasting
TURP syndrome managed with regional anesthesia. An awake (vomiting, diarrhea, ostomy losses)
patient permits detection of the signs of hyponatremia, including Hyponatremia with normal or SIADH, water intoxication due to
nausea, vomiting, visual disturbances, depressed level of con- expanded effective arterial primary polydipsia, water overload
sciousness, agitation, confusion, coma, seizures, muscle cramps, volume due to advanced renal failure or
weakness, myoclonus, coma, and death. Cerebral edema occurs at vascular or inflammatory renal
or below a serum level of 123mEq/L, and cardiac symptoms disease
occur at 100 mEq/L. Hyponatremia associated with increased
Reduced sodium delivery to Starvation, myxedema (?)
intravascular volume can result in pulmonary edema, hyperten- diluting segment
sion, and heart failure.19
Initially, stopping the infusion of hypotonic glycine solu- SIADH, syndrome of inappropriate secretion of antidiuretic hormone;
THAM, tris(hydroxymethyl)aminomethane.
tion can resolve or at least improve the progression of Adapted from Oh M, Carrol H: Disorders of sodium metabolism: Hypernatremia and
hyponatremia. Because TURP syndrome is a case of water over- hyponatremia. Crit Care Med 20:94, 1992; and from Andreoli TE: Disorders of fluid
load, the treatment should include water restriction. A loop diu- volume, electrolyte, and acid-base balance. In Wyngaarden JB, Smith LH Jr (eds):
retic may be added to facilitate free water excretion. Hypertonic Cecil Textbook of Medicine, 17th ed. Philadelphia, WB Saunders, 1985, p 525.
saline should be used only in cases of severe hyponatremia with
neurologic symptoms.20
In any patient whose hyponatremia warrants correction,
the dose of sodium required to correct a deficit may be calculated antagonists and the administration of urea as an osmotic diuretic
using the following formula: are under investigation.
Rapid treatment of hyponatremia can lead to cerebral
( ( ))
Dose [mEq ] = weight [ kg ] 140 Na + mEq L 0.6 edema or central pontine myelinolysis. Management of
hyponatremia involves elimination of the underlying condition
The optimal rate of correction seems to be 0.6 to 1mmol/L/hr when possible (e.g., stop TURP syndrome as soon as possible).
until the sodium concentration is 125mEq/L; correction then The use of normal saline (308mOsm/L) alone may worsen the
proceeds at a slower rate. One half the deficit can be administered hyponatremia, depending on the patients serum and urine osmo-
over the first 8 hours, and the next half can be administered over lality.23 Severe coma or seizures can be managed with one or more
1 to 3 days if symptoms remit. The appropriate treatment of approaches, including 3% hypertonic saline (513mEq/L), fluid
hyponatremia, particularly in patients with neurologic restriction, or furosemide.24 The most likely cause of postopera-
symptoms, is, however, an area of controversy. Sodium concentra- tive hyponatremia is SIADH (Table 54-5). SIADH and cerebral
tion should be monitored every 1 to 2 hours during rapid salt-wasting syndrome can best be differentiated by determina-
correction. tion of total-body water by clinical measurements, including skin
Menstruant female patients are at greater risk for develop- turgor, body weight, central venous pressure (CVP), and overall
ing significant neurologic sequelae after hyponatremia. Ayus and vital signs.
colleagues21 found that despite equal incidence of postoperative
hyponatremia among men and women, 97% of patients with per-
manent brain damage were women, and 75% of those were of Hypernatremia
reproductive age. The mechanism of this gender difference is
unknown; however, a potential explanation involves alterations in Hypernatremia is defined as an increase in extracellular sodium
the brains adaptive processes to hyponatremia. Estrogens seem concentration, and may be accompanied by the presence of low,
to alter the function of Na+/K+-ATPase in the rat brain, which normal, or high total-body sodium content. The major causes of
could alter the brains compensatory mechanisms for hypernatremia are excessive loss of water, inadequate intake of
hyponatremia.22 Several agents that interfere with urine concen- water, a lack of ADH, or excessive intake of sodium (e.g., with
tration at the collecting duct, including lithium and demeclocy- solutions containing a high sodium concentration, such as sodium
cline, have been used to manage chronic hyponatremia. ADH bicarbonate or hypertonic saline).

Diabetes insipidus may result from a deficiency of ADH
(vasopressin) or inability of the kidney to produce a hypertonic
medullary interstitium. Diabetes insipidus is characterized by
production of a large volume of dilute urine. Deficiency of vaso-
pressin is known as central diabetes insipidus and is an endocrine
disorder. Vasopressin deficiency is seen after pituitary surgery,
basal skull fracture, and severe head injury. Conversely, nephro-
genic diabetes insipidus results if the kidney cannot produce a
hypertonic medullary interstitium and is unable to concentrate
urine. Nephrogenic diabetes insipidus is defined as renal tubule
cell insensitivity to the effects of vasopressin. Nephrogenic diabe-
tes insipidus can result from any systemic or kidney disease that
impairs tubular function. In central or nephrogenic diabetes
insipidus, the patient loses a significant amount of body water in
a short period, which can cause profound hypovolemia if the
patient does not have adequate access to water.
Approximately 50% of patients with central diabetes insip-
idus are classified as having idiopathic disease. This condition is
thought to be caused by an autoimmune process affecting the
response of the hypothalamus to hypertonicity. Patients with
nephrogenic diabetes insipidus have marked increases in the
plasma levels of vasopressin because of significant hyperosmolal-
ity of the plasma. Renal tubular cells can become poorly respon-
sive to the effects of vasopressin for a variety of reasons. Various
insults to the collecting system, such as after ureteral obstruction
or medullary cystic disease, lead to decreased sensitivity to vaso-
pressin. Many pharmacologic agents impair the ability of vaso-
pressin to affect tubular water transport and result in nephrogenic
diabetes insipidus (Table 54-6).
Patients with continued urine output of more than 100
mL/hr who develop hypernatremia should be evaluated for dia-
betes insipidus by determining the osmolalities of urine and
serum. If the urine osmolality is less than 300mOsm/L, and
serum sodium exceeds 150mEq/L, the diagnosis of diabetes
insipidus is likely. In patients with central diabetes insipidus,
1-deamino-8-d-arginine vasopressin (DDAVP), a vasopressin
analog also known as desmopressin acetate, may be administered
to correct the deficiency in vasopressin. The underlying cause of
nephrogenic diabetes insipidus should be sought and treated if
possible.
In the presence of an intact thirst mechanism, a slight
increase in serum sodium concentration (e.g., 3 to 4mmol/L)
above baseline values elicits intense thirst. The lack of thirst in the
presence of hypernatremia in a mentally alert patient indicates a
defect in the osmoreceptor or in the cortical thirst center. The
most common objective sign of hypernatremia is lethargy or
mental status changes, which can proceed to coma and convul-
sions. Additional signs and symptoms of hypernatremia include
thirst, shock, peripheral edema, myoclonus, ascites, muscular
tremor, muscular rigidity, hyperactive reflexes, pleural effusion,
and expanded intravascular fluid volume. With acute and severe
hypernatremia, the osmotic shift of water from the cells can lead
Table 54-6 Drugs Causing Nephrogenic Diabetes Insipidus
Lithium
Foscarnet
Glyburide
Demeclocycline
Methoxyflurane
Amphotericin B
Intravascular Fluid and Electrolyte Physiology 1709 54
Table 54-7 Major Causes of Hypernatremia
Mechanism Clinical Syndrome(s)
Impaired thirst Stupor/coma, essential hypernatremia
Solute (osmotic) diuresis Diabetic ketoacidosis, hyperglycemic
nonketotic coma, mannitol administration
Excessive water losses
Section IV Anesthesia Management
Renal Pituitary diabetes insipidus, nephrogenic
diabetes insipidus
Extrarenal Sweating
Combined Stupor/coma plus hypertonic enteric feeding
Note: Hyperglycemia or mannitol administration also can result in hyponatremia.
Sweating also can result in hyponatremia.
Adapted from Andreoli TE: Disorders of fluid volume, electrolyte, and acid-base
balance. In Wyngaarden JB, Smith LH Jr (eds): Cecil Textbook of Medicine, 17th ed.
Philadelphia, WB Saunders, 1985, p 528.
to shrinkage of the brain with tearing of the meningeal vessels
and intracranial hemorrhage. Slowly developing hypernatremia is
usually well tolerated because of the brains ability to regulate its
volume. Treatment involves restoring normal osmolality and
volume and includes removal of excess sodium by the administra-
tion of diuretics and hypotonic crystalloid solutions. The speed
of correction depends on the rate of development of hyper-
natremia and associated symptoms. Because chronic hyper-
natremia is well tolerated, rapid correction offers no advantage
and may be extremely harmful or lethal because it may result in
brain edema.25,26 Typically, a maximum of 10% of the serum
sodium concentration, or about 0.7mmol/L/hr, should be the
goal rate of correction. Hypernatremia increases the minimum
alveolar concentration of inhaled anesthetic agents, possibly
because of enhanced sodium conductance during depolarization
of excitatory membranes (Table 54-7).
Potassium Physiology
Potassium is the most abundant positive ion in the intracellular
fluid, and nearly 98% of total-body potassium is intracellular. In
the short-term (minutes), potassium balance is influenced by
insulin, pH, -adrenergic agonists, and bicarbonate concentra-
tion. Long-term regulation of potassium excretion and balance
primarily involves the kidney and aldosterone. Potassium concen-
tration to average 0.4 to 0.5 mEq/L lower when measured on
heparinized arterial samples compared with clotted venous
samples is due to several factors, including extrusion of potassium
from erythrocytes during the clotting process.
Increases in potassium intake increase renal excretion of
potassium through various cellular mechanisms. In response to
increases in extracellular potassium levels, aldosterone is secreted
from the zona glomerulosa of the adrenal gland and acts on corti-
cal collecting ducts to increase potassium secretion into the
tubular fluid and increase potassium excretion.
Many of the effects seen during alterations in normal
potassium levels result from potassiums importance in the mem-
brane potential of cells. Potassium is the principal intracellular
cation, with more than 98% of the bodys potassium found within
the intracellular water. In the resting state, cell membrane con-
ductance is higher for potassium than for sodium. This increased
IV 1710 Anesthesia Management
conductance leads to transmembrane potential being closer to the
transmembrane value of potassium (90mV). Alterations in
extracellular potassium concentration alter the resting membrane
potential, which may cause the cell to be unresponsive or over-
responsive to sodium shifts into the cell. High or low potassium
levels can result in potentially lethal problems in excitatory tissue,
particularly cardiac tissue.
Multiple factors regulate the normal maintenance of the
transmembrane potassium gradient. The most important trans-
cellular enzyme involved in potassium regulation is Na+/K+-
ATPase, which maintains the transcellular gradient. 2-Adrenergic
drugs increase the activity of Na+/K+-ATPase by binding to cell
surface receptors, linking potassium transport to the sympathetic
nervous system. Insulin causes more sodium to enter the cell
through an Na+/H+-antiporter, which decreases intracellular
proton concentration. To maintain electric neutrality, any increase
in sodium must occur via exchange for potassium, decreasing
extracellular potassium. Shock can have an adverse impact on the
activity of Na+/K+-ATPase by limiting the amount of adenosine
triphosphate (ATP) available for ion transport because of the shift
to anaerobic metabolism.27 Potassium transport is affected by pH.
The body uses potassium to decrease excess extracellular hydro-
gen ions by moving potassium out of cells and hydrogen ions into
cells. Acidemia potentiates hyperkalemia by moving potassium
out of cells.
Potassium requirements vary with age and growth. A
typical term infant requires 2 to 3 mEq/kg/day,28 whereas adults
use 1 to 1.5mEq/kg/day (see Chapters 82 and 84). Potassium
demands are related to metabolic rate (2mEq/100kcal). In this
regard, the requirement increases dramatically during cell growth
after establishment of nutrition in previously starved individuals.
An extremely high or low level of potassium is life-threatening.
Hypokalemia
Hypokalemia (<3.5mEq/L) may occur because of an absolute
deficiency or redistribution into the intracellular space. A reduc-
tion in serum potassium of 1mEq/L indicates a net loss of 100
to 200mEq of potassium in a normal adult. Hypokalemia in the
range of 2 to 2.5mEq/L is likely to cause muscular weakness,
arrhythmias, and electrocardiographic abnormalities, including
sagging of the ST segment, depression of the T wave, and U wave
elevation. These electrocardiographic changes do not correlate
with the severity of potassium depletion. Cardiac dysrhythmias
are more predictable, however, and most frequently involve atrial
fibrillation and premature ventricular systoles.
The four most common causes of hypokalemia are reduced
intake, gastrointestinal losses, excessive renal losses of potassium
(e.g., with excess of mineralocorticoids or diuretics), and potas-
sium shifts from the ECF to the intracellular fluid (e.g., owing to
insulin administration). Such shifts can occur with acute alkalosis,
insulin therapy, stress-related catecholamine activity, and hypoka-
lemic periodic paralysis. Surgical stress may reduce the serum
potassium concentration by 0.5mEq/L. The administration of
exogenous catecholamines, such as isoproterenol, terbutaline,
epinephrine, and ritodrine, also decreases potassium levels29,30;
clinically, this includes pregnant patients receiving tocolytic
therapy or respiratory treatment with 2-agonists and critically ill
patients requiring pharmacologic cardiovascular support (see
Chapters 69 and 92).
No studies show increased morbidity or mortality for
patients undergoing anesthesia with a potassium level of at least
2.6mEq/L. Anesthesiologists commonly administer supplemen-
tal potassium chloride, but studies show that in many instances
therapy is ineffective, with most of the supplemental potassium
being excreted in the urine despite continued hypokalemia.31,32 If
perioperative therapy is indicated, intravenous potassium chlo-
ride should be used. Because the rate of administration of potas-
sium must be adjusted for the rate of distribution through the
extracellular space before entry into the intracellular space, the
rate of potassium administration is typically limited to approxi-
mately 0.5mEq/kg/hr. The typical replacement rate is 10 to
20mEq/hr for a normal adult with constant monitoring of the
electrocardiogram.33,34 Clinical anesthesiologists should be aware
that safe administration of potassium is enhanced by the use of
continuous electrocardiogram monitoring, bedside potassium
measurement, and delivery with an infusion pump (Table 54-8).
Hyperkalemia
Hyperkalemia (>5.5mEq/L) can occur in various disease states,
in response to drugs that diminish renal potassium excretion
(Table 54-9), or after sudden transcellular shifts of potassium
from the intracellular fluid to the ECF. Potentially lethal hyperka-
lemia during anesthesia may occur with reperfusion of a large
vascular bed after a period of ischemia (usually >4 hours).
Ischemia results in significant acidosis in the affected area, which
causes an outflow of intracellular potassium. When the area is
reperfused, the body receives a large bolus of potassium that
cannot be redistributed rapidly enough, resulting in potentially
fatal hyperkalemia. Any condition or drug resulting in adrenal
inhibition or decreasing aldosterone levels can cause potassium
retention. Factitious (e.g., spurious) hyperkalemia also should be
considered in the differential diagnosis and occurs in response to
lysis of the cellular components of blood.
Table 54-8 Major Causes of Hypokalemia
Mechanism Clinical Syndromes
Inadequate intake Anorexia nervosa, starvation, alcoholism,
mineralocorticoid excess (primary and
secondary hyperaldosteronism)
Excess renal loss Bartters syndrome; diuresis (diuretics with a
pre-late distal locus osmotic diuresis, chronic
metabolic alkalosis); impermeant anion
antibiotics (carbenicillin, penicillin, nafcillin,
ticarcillin); renal tubular acidosis;
hypomagnesemia; myelomonocytic leukemia
Gastrointestinal losses Vomiting; diarrhea, particularly secretory
diarrheas; villous adenoma
Shifts of extracellular 2-agonists, acute alkalosis, hypokalemic
fluid to intracellular periodic paralysis, insulin therapy, vitamin B12
fluid with altered therapy, lithium overdose
internal potassium
balance
Adapted from Andreoli TE: Disorders of fluid volume, electrolyte, and acid-base
balance. In Wyngaarden JB, Smith LH Jr (eds): Cecil Textbook of Medicine, 17th ed.
Philadelphia, WB Saunders, 1985, p 532.

Table 54-9 Drugs Causing Hyperkalemia
Amiloride
Angiotensin II antagonists
Angiotensin-converting enzyme inhibitors
Mannitol
Pentamidine
Spironolactone
Succinylcholine
Triamterene
Trimethoprim
Hyperkalemia can be separated into acute and chronic
processes. Acute hyperkalemia can occur in various circum-
stances and usually is more poorly tolerated than chronic hyper-
kalemia. The most common cause of chronic hyperkalemia
occurring with anesthesia is renal failure. A patient undergoing
anesthesia with even moderately elevated potassium concentra-
tion (>5.5mEq/L) should have continuous electrocardiographic
monitoring to help determine the severity of hyperkalemia.
Clinically, hyperkalemia can cause muscle weakness and
paralysis. Alterations in cardiac conduction increase automaticity
and enhance repolarization. Mild elevations in potassium levels
(6 to 7mEq/L) may manifest with peaked T waves; as levels
approach 10 to 12mEq/L, a prolonged PR interval, widening of
the QRS complex, ventricular fibrillation, or asystole can occur.
Conceptually, treatments are divided into (1) physiologic
antagonists (calcium); (2) agents to shift potassium into cells
(glucose, insulin, hyperventilation, bicarbonate, and -adrenergic
agonists); and (3) drugs and treatments to eliminate potassium
from the body (sodium polystyrene sulfonate [Kayexalate],
diuretics, aldosterone agonists, dialysis). Clinical treatment of
hyperkalemia is determined by the setting and presence of elec-
trocardiographic changes; it involves first stabilization of the
heart from effects of the potassium with intravenous calcium
(500mg of calcium chloride or 1g of calcium gluconate in an
adult) and redistribution of potassium from the plasma into cells.
As stated earlier, in addition to intravenous calcium, intravenous
glucose, insulin, and bicarbonate (1mEq/kg) and hyperventila-
tion are the major therapies used in the operating room. Intrave-
nous regular insulin, 5 to 10U, administered with dextrose at a
dose of 25 to 50g (i.e., one to two 50-mL ampules of 50% dextrose
solution), reduces serum potassium levels within 10 to 20 minutes,
and the effects last 4 to 6 hours.35 Resin exchange, dialysis, diuret-
ics (e.g., furosemide, 20-40mg intravenously), aldosterone ago-
nists (e.g., fludrocortisone), and inhaled or intravenous
-adrenergic agonists are established additional therapies (Table
54-10).
Calcium Physiology
Calcium is the key component that mediates muscle contraction;
exocrine, endocrine, and neurocrine secretion; cell growth; and
the transport and secretion of fluids and electrolytes. A 70-kg
adult contains approximately 1300g of calcium, 99% of which is
in the bones and teeth. The kidneys are the major organ respon-
sible for regulating calcium between 4.5 and 5mEq/L (8.5 to
10.5mg/dL). Calcium is abundant in milk and milk products, is
poorly absorbed by the intestine, and is excreted primarily in the
Intravascular Fluid and Electrolyte Physiology 1711 54
feces and urine. Circulating calcium exists in three forms: bound
to plasma proteins (primarily albumin) and not filtered by
glomerular capillaries (40% to 50%); ionized, physiologically
active, filtered at the glomerular membrane, and maintained at a
concentration of 2 to 2.5mEq/L (50%); and nonionized and
chelated with phosphate, sulfate, and citrate (10%). Because
changes in pH alter the fraction of calcium that is bound to
albumin, the level of ionized calcium can change without altera-
Section IV Anesthesia Management
tion of total calcium. Total calcium varies with protein-binding
capacity and results, under hyperaluminemia, in pseudohypercal-
cemia (most commonly associated with dehydration and multiple
myeloma). Most filtered calcium is reabsorbed in the proximal
tubule, the thick ascending limb of the loop of Henle, and the
distal tubule.
Because of the importance of calcium in virtually all cel-
lular functions, its intracellular and extracellular concentrations
are tightly controlled. Energy is expended to pump intracellular
calcium out of the cytosol into the sarcoplasmic reticulum or ECF.
As with potassium, during shock and depletion of intracellular
energy, calcium accumulates within cells and can facilitate cell
death.
The concentration of serum proteins is an important deter-
minant of calcium ion concentration. Ionized calcium can be
measured directly with the use of calcium-specific electrodes.
When ionized calcium cannot be measured, the approximate
amount of calcium bound to protein is given by the following
equation:
Protein-bound calcium (% ) = 0.8 albumin (g L ) + 0.2
globulin (g L ) + 3
Alternatively, a correction of 0.8mg/dL is added to the
serum calcium for every 1g/dL that serum albumin is less than
4g/dL.36 If the serum calcium is 7.8mg/dL (a subnormal value),
and the serum albumin is only 3mg/dL, the stated serum calcium
is corrected by adding 0.8mg/dL; the corrected value of 8.6mg/
dL is within the normal range. One advance in the management
of perioperative calcium includes the ease of bedside serum
ionized calcium measurement.
Within minutes of a slight decrease in extracellular
calcium concentration, the parathyroid glands release parathy-
Table 54-10 Major Causes of Hyperkalemia
Mechanism Clinical Syndromes
Pseudohyperkalemia Sample lysis, technical problems
Altered internal Acidosis; insulin deficiency; hypoaldosteronism;
potassium balance malignant hyperthermia; periodic paralysis; cell
necrosis; drugs (succinylcholine, digitalis,
nonselective -blockers)
Altered external Increased uptake by replacement therapy;
potassium balance transfusions; antibiotics containing potassium
salts; decreased excretion by renal disease;
hypoaldosteronism; drugs (heparin, amiloride,
triamterene, spironolactone, nonsteroidal drugs,
angiotensin-converting enzyme inhibitors,
angiotensin-receptor antagonists)
Adapted from Solomon RJ, Katz JD: Disorders of potassium homeostasis. In
Stoelting RK (ed): Advances in Anesthesia. Chicago, Year Book Medical Publishers,
1986.
IV 1712 Anesthesia Management
roid hormone, which increases calcium reabsorption in the thick
ascending limb and distal tubule, decreasing calcium excretion.
Excision of the parathyroid glands eliminates parathyroid
hormone secretion, which can significantly disrupt calcium
homeostasis.
Calcitonin, produced in the thyroid gland, decreases renal
reabsorption of calcium acutely, but has little effect on long-term
calcium homeostasis. Surgical removal of the thyroid gland elimi-
nates calcitonin without changing extracellular calcium ion
concentration.
Bone acts as the bodys major reservoir of calcium. When
the parathyroid gland releases parathyroid hormone in response
to decreased calcium levels, bone reabsorption is favored, and
calcium is released. Vitamin D increases absorption of calcium
from the gastrointestinal tract, and its action is potentiated by
parathyroid hormone.
Hypercalcemia
Hypercalcemia is associated with many disease processes and has
many signs and symptoms. Mild to moderate hypercalcemia (11
to 14mg/100mL) often has no symptoms, but when levels reach
15mg/100mL, clinical changes become more common. Hyper-
calcemia produces changes primarily in the central nervous
system (e.g., mental status changes), the gastrointestinal tract
(e.g., vomiting), the kidneys (e.g., polyuria, renal calculi, oliguric
renal failure), and the heart (e.g., cardiac conduction distur-
bances). Today, hypercalcemia is most commonly diagnosed in
asymptomatic patients, whereas clinical features previously were
the earliest manifestations. Potential causes of hypercalcemia
include thiazide diuretic therapy, malignancy (now known as
humoral hypercalcemia of malignancy), granulomatous disease
(related to increased intestinal absorption of calcium induced by
elevated calcitriol levels), vitamin D intoxication, or parathyroid
hormone adenoma. Additional causes of hypercalcemia include
lithium therapy, thyrotoxicosis, pheochromocytoma, immobiliza-
tion, vitamin A intoxication, renal failure, and theophylline
therapy.
Treatment of hypercalcemia essentially involves diuresis
and administration of normal saline to dilute plasma calcium.
These primary treatments also are useful because sodium inhibits
the renal reabsorption of calcium. Additional therapies include
bisphosphonates (pamidronate is the most commonly used),
calcitonin, ambulation, hemodialysis, and treatment of the under-
lying condition. Certain conditions, including numerous cancer-
related hypercalcemias, can be treated with calcium-lowering
agents, such as mithramycin and glucocorticoids. The anesthetic
management of a patient with hypercalcemia should involve
avoidance of thiazide diuretics and maintenance of hydration and
urine output with sodium-containing fluids. Monitoring the
patient by means of electrocardiograms is useful to detect cardiac
conduction abnormalities with shortened PR or QT interval,
with or without widening of the QRS complex. Patients who have
muscle weakness should receive decreased doses of nondepolar-
izing muscle relaxants (Table 54-11).
Hypocalcemia
In the operating room, ionized hypocalcemia is most commonly
caused by acute hyperventilation or the infusion of citrated blood
Table 54-11 Major Causes of Hypercalcemia
Mechanism Clinical Syndromes
Increased Primary hyperparathyroidism (solitary or multiple
parathyroid adenoma, multiple endocrine neoplasia syndrome),
hormone lithium therapy, familial hypocalciuric hypercalcemia
Vitamin D related Vitamin D intoxication; idiopathic hypercalcemia of
infancy; increase in 1,25-dihydroxyvitamin D;
sarcoidosis and other granulomatous diseases
Associated with Hyperthyroidism, immobilization, thiazides, vitamin A
high bone intoxication
turnover
Malignancy Solid tumors with metastases, solid tumors with
related, humoral mediators of hypercalcemia, hematologic
including malignancies, severe secondary
associated with hyperparathyroidism (from chronic renal failure),
renal failure aluminum intoxication, milk-alkali syndrome
Adapted from Potts JT: Diseases of the parathyroid gland and other hyper- and
hypocalcemic disorders. In Isselbacher KJ, Braunwald E, Wilson JD, et al (eds):
Harrisons Principles of Internal Medicine, 13th ed. New York, McGraw-Hill, 1995, p
2151.
greater than 1.5mL/kg/min (see Chapter 55). The most common
cause of hypocalcemia (plasma concentration <4.5mEq/L) in
hospitalized patients is a low albumin level, such as in critically
ill patients with severe sepsis, burns, or acute renal failure, and in
patients after extensive transfusions. Critically ill patients can
have low plasma albumin and low plasma calcium levels with
normal ionized calcium levels. There is no reason to correct the
calcium deficiency, but overall nutrition should be improved. The
major signs and symptoms of hypocalcemia include mental status
changes, tetany, positive Chvostek and Trousseau signs, laryngo
spasm, hypotension, and dysrhythmias. Electrocardiographic
evaluation may show prolongation of the QT interval or heart
block in severe cases. Treatment, often prompted by hypotension,
involves intravenous infusion of 10% calcium chloride (1.36mEq/
mL) or calcium gluconate (0.45mEq/mL). When equivalent
calcium doses are administered, both preparations are equally
efficacious in restoring the calcium level to normal (Table 54-12).
In this regard, calcium gluconate is especially advantageous in
peripheral venous administration because extravasated calcium
chloride solution can result in severe tissue destruction.
When hypocalcemia is caused by large-volume infusion of
isotonic saline (as seen during resuscitation in shock), it may be
accompanied by hypomagnesemia. Hypomagnesemia may impair
the actions of vitamin D and delay correction of postresuscitation
hypocalcemia. The magnesium level should be checked and cor-
rected if necessary. In this regard, fluid resuscitation involving
certain colloids can contribute to a hypocalcemic state.37
Magnesium Physiology
Magnesium sulfate has been used for many years on an empirical
basis to control convulsions in patients with preeclamptic toxemia.
Magnesium ions are essential for many biochemical reactions,
and a deficiency may produce clinically important consequences.
Many of the pharmacologic properties have only more recently
been appreciated. Magnesium is excreted through the gastroin-
testinal tract and kidneys.

Table 54-12 Major Causes of Hypocalcemia Excluding Neonatal Conditions
Mechanism Clinical Syndromes
Parathyroid hormone absent Hereditary hypoparathyroidism, acquired
hypoparathyroidism, hypomagnesemia
Parathyroid hormone Lack of active vitamin D (decrease in intake
ineffective or lack of sunlight), defective metabolism
from anticonvulsant therapy, vitamin D
dependent rickets type I
Pseudohyperparathyroidism Ineffective vitamin D (intestinal
malabsorption), vitamin Ddependent
rickets type II
Parathyroid hormone Severe, acute hyperphosphatemia; tumor
overwhelmed lysis; acute renal failure; rhabdomyolysis;
osteitis fibrosa after parathyroidectomy
Adapted from Potts JT: Diseases of the parathyroid gland and other hyper- and
hypocalcemic disorders. In Isselbacher KJ, Braunwald E, Wilson JD, et al (eds):
Harrisons Principles of Internal Medicine, 13th ed. New York, McGraw-Hill, 1995, p
2165.
Total-body magnesium is approximately 2000mEq. Mag-
nesium is the fourth most prevalent cation in the body and acti-
vates approximately 300 enzyme systems, including many involved
in energy metabolism.40 It is essential for the production and
functioning of ATP, which is fully functional only when chelated
to magnesium. Other processes dependent on magnesium include
the production of DNA, RNA, and protein synthesis.40
Magnesium is an essential regulator of calcium access
into the cell and of the actions of calcium within the cell. Mag-
nesium plays an essential role in the regulation of most cellular
functions and may be regarded as a natural physiologic calcium
antagonist.40
Hypomagnesemia
With an absence of magnesium in the diet, the kidneys are able
to decrease excretion significantly; however, hypomagnesemia is
common in hospitalized patients, especially patients in critical
care areas, and it manifests with signs and symptoms similar to
hypocalcemia. Slight hypomagnesemia occurs in athletes, in
hypermetabolic states such as pregnancy, and during cold accli-
matization. Magnesium stores can become depleted in patients
undergoing prolonged diuretic therapy or patients with chronic
diarrhea. Chronic alchohol ingestion leads to significant loss of
magnesium, and most hospitalized alcoholics have low magne-
sium levels. Magnesium deficiency alone or in combination with
diuretic-induced hypokalemia and digitalis-induced arrhythmia
responds to magnesium therapy.38
Anesthetizing patients with magnesium deficiency is likely
to increase the risk of perioperative arrhythmias. Respiratory
muscle power is impaired by hypomagnesemia, which may have
important clinical consequences for anesthesia and critical care.
Additional manifestations include central nervous system irrita-
bility with seizures and hyperreflexia (e.g., Chvostek sign) and
skeletal muscle spasm (e.g., Trousseau sign). Treatment in a 70-kg
adult with normal renal function in whom intravenous therapy
is warranted includes magnesium sulfate 1 to 2g (8 to 16mEq)
over 15 minutes followed by 1g/hr until serial serum magnesium
levels indicate the deficiency has been corrected.39 For acute
arrhythmias, magnesium sulfate can be administrated in a dose
Intravascular Fluid and Electrolyte Physiology 1713 54
of 1 to 2g over approximately 5 minutes intravenously with close
monitoring of blood pressure and heart rate. Arterial pressure,
deep tendon reflexes, and magnesium concentration may aid the
clinical anesthesiologist when overtreatment is suspected. In
asymptomatic patients with mild hypomagnesemia, oral replace-
ment is preferred. Beyond the theoretical risks of magnesium
deficiency and neuromuscular blockade, the clinical importance
of hypomagnesemia is related to the conditions and pathophysi-
Section IV Anesthesia Management
ologic processes associated with management of these conditions
(Table 54-13).
Hypermagnesemia
Hypermagnesemia (>2.5mEq/L) occurs most commonly from
iatrogenic causes and excessive use of magnesium-containing
antacids or laxatives. It is rare in clinical medicine because mag-
nesium is relatively poorly absorbed from the gastrointestinal
tract, and renal elimination of excess magnesium is extremely
rapid (within 4 to 8 hours of a magnesium load). Because elimina-
tion is directly related to the glomerular filtration rate, patients
with kidney failure are at increased risk of developing hypermag-
nesemia. Signs and symptoms are directly related to the blood
level and include alterations in the nervous, cardiovascular, res-
piratory, and genitourinary systems.
Magnesium depresses the central nervous system. In the
early 1900s, it was used effectively as a general anesthetic. Mag-
nesium penetrates the blood-brain barrier poorly, however, and
its level in the cerebrospinal fluid is well controlled by an active
transport mechanism.40
In the peripheral nervous system, magnesium interferes
with the release of neurotransmitters at all synaptic junctions and
potentiates the action of local anesthetics.40 At the neuromuscular
junction, magnesium concentrations of 5mmol/L cause signifi-
Table 54-13 Major Causes of Hypomagnesemia
Mechanism Clinical Syndromes
Primary nutritional Inadequate intake, total parenteral nutrition,
disturbances refeeding syndrome
Gastrointestinal disorders Specific absorptive defects, malabsorption
syndromes, prolonged diarrhea, prolonged
nasogastric suction, pancreatitis
Endocrine disorders Hyperparathyroidism, hypoparathyroidism,
hyperthyroidism, primary
hyperaldosteronism, Bartters syndrome,
diabetic or alcoholic ketoacidosis,
administration of epinephrine, SIADH,
hungry bone syndrome after
parathyroidectomy
Chronic alcoholism, Ethanol ingestion; idiopathic; after renal
alcoholic withdrawal, transplantation; drugs (cisplatin,
increased renal excretion aminoglycoside, amphotericin B, diuretics,
pentamidine, theophylline); recovery phase
of acute tubular necrosis; colony-
stimulating factor therapy
SIADH, syndrome of inappropriate secretion of antidiuretic hormone.
Adapted from Potts JT: Diseases of the parathyroid gland and other hyper- and
hypocalcemic disorders. In Isselbacher KJ, Braunwald E, Wilson JD, et al (eds):
Harrisons Principles of Internal Medicine, 13th ed. New York, McGraw-Hill, 1995,
p 2188.
IV 1714 Anesthesia Management
cant presynaptic neuromuscular blockade and enhance the action
of the nondepolarizing muscle relaxants.40 It can precipitate
severe muscle weakness in patients with Eaton-Lambert syn-
drome, patients with myasthenia gravis, or patients pretreated
with a small dose of a defasciculating agent.40 Magnesium pro-
longs the action of depolarizing neuromuscular blocker drugs
(e.g., succinylcholine); administration before the use of succinyl-
choline prevents the release of potassium provoked by the neu-
romuscular blocking drug (see Chapter 29).
In the cardiovascular system, magnesium produces vasodi-
lation by direct action on blood vessels and by interfering with a
wide range of vasoconstrictor substances. It also reduces periph-
eral vascular tone by sympathetic blockade and inhibition of cat-
echolamine release.40 In the isolated heart, increased concentrations
of extracellular magnesium ion markedly depress contractile
force. Decreased myocardial performance has been shown after a
bolus of 2.5g of magnesium sulfate. Severe myocardial depres-
sion can occur with combinations of magnesium and diltiazem.
In the isolated heart, magnesium produces bradycardia, but in an
intact subject, the inhibition of vagal acetylcholine release pro-
duced by magnesium overrides the intrinsic slowing, and a mild
tachycardia occurs. Magnesium is effective in treating various
arrhythmias, including ventricular arrhythmias, torsades de
pointes, arrhythmias associated with epinephrine administration,
and digitalis-associated arrhythmias. It also is efficacious in
certain arrhythmias induced by hypokalemia, alcoholism, and
myocardial infarction, and may offer partial protection against
bupivacaine-induced arrhythmias.
In the respiratory system, magnesium has no effect on
central respiratory drive, and its only respiratory depressant effect
is caused by the neuromuscular blockade that it produces. It is an
effective bronchodilator and has been used successfully in severe
acute asthma, although no study to date has shown benefit in
chronic asthma.
In obstetric practice, magnesium, at supraphysiologic con-
centrations, is a powerful tocolytic and has been used for many
years in the management of premature labor. Magnesium also is
used in obstetrics to prevent patients with preeclampsia from
developing seizures. Animal studies have shown that magnesium
suppresses electroencephalographic spike activity (see Chapter
69). Therapeutic magnesium levels are 5 to 7mg/dL when admin-
istered for preeclampsia. When levels exceed 15 to 20mg/dL,
respiratory depression can become profound. In the kidney, mag-
nesium is a renal vasodilator and a diuretic.
Symptoms and electrocardiographic changes of hypermag-
nesemia correspond to serum levels. Depressed cardiac conduc-
tion, widened QRS complexes, prolonged PR intervals, and
nausea appear with levels of 5 to 10mg/dL. Sedation, hypoventila-
tion, decreased deep tendon reflexes, and muscle weakness appear
with levels of 20 to 34mg/dL, with hypotension, bradycardia, and
diffuse vasodilation occurring at levels of 24 to 48mg/dL. Are-
flexia, coma, and respiratory paralysis occur at 48 to 72mg/dL.
For these reasons, all patients being treated with magnesium
therapy are clinically observed for magnesium intoxication.
Elimination of magnesium involves fluid loading followed
by or with concomitant diuresis. Definitive therapy involves dialy-
sis. Temporary reversal of the effects of magnesium can be
managed with calcium therapy. Because hypermagnesemia poten-
tiates the effects of depolarizing and nondepolarizing muscle
relaxants, these agents must be carefully titrated in conjunction
with appropriate assessment of neuromuscular blockade.
The deleterious effects of magnesium deficiency are well
recognized, and most critical care units monitor magnesium
levels. In coronary care units, several studies have shown that the
infusion of magnesium sulfate can reduce the incidence and
severity of cardiac arrhythmias associated with myocardial infarc
tion. Numerous reports in the literature describe the bronchodi-
lator effect of magnesium and its successful use in the management
of asthma. Magnesium may decrease the incidence of adrener
gically mediated arrhythmias without interfering with the bron-
chodilating action of -stimulants and contribute to smooth
muscle relaxation of bronchioles.40
Magnesium has several important pharmacologic actions.
Its route of elimination is renal, and any patient who is oliguric
or in a reduced urine output state requires downward dosing
adjustment of magnesium therapy. Magnesium should be regarded
as a cardiovascular drug, first and foremost, with calcium antago-
nistic and antiadrenergic properties that may be accompanied by
minimal myocardial depression.40
Phosphate Physiology
Phosphate is the most abundant intracellular anion, and is essen-
tial for membrane structure, cellular energy, and cell transport.
Specifically, phosphate is required to produce ATP, DNA, RNA,
and 2,3-diphosphoglycerate, which facilitates oxygen release from
hemoglobin. About 1g of phosphorus is ingested daily. Intake
generally exceeds metabolic requirements. Approximately 70%
(700mg) is absorbed primarily from the small intestine, with the
rest (300mg) eliminated in the feces. In addition to 1,25-dihy-
droxyvitamin D3 (vitamin D), parathyroid hormone facilitates
absorption of phosphate from the gut lumen. The gut secretes
phosphate into the lumen and then reabsorbs it, unless it is bound
there by calcium or antacids or is lost by diarrhea or drainage
through ostomies or fistulas. Under normal dietary conditions,
absorption of phosphate occurs through paracellular diffusive
pathways with little regulation. When luminal phosphate concen-
trations are low, an active sodium-dependent transport mecha-
nism is activated, and additional phosphate is absorbed.41 When
dietary intake is normal, phosphate absorption is essentially an
unregulated process. Cations such as calcium, magnesium, and
aluminum decrease phosphate absorption, however, by directly
binding with phosphate in the gut lumen. This is useful clinically
in renal failure to limit the amount of phosphate reabsorbed by
the gut. Because of the unregulated nature of phosphate absorp-
tion in the gut, the kidneys become the primary organ of excre-
tion of excess phosphate. The kidneys normally excrete 700mg/day
by filtering 6g and reabsorbing 5.3g. Urinary phosphate elimina-
tion approximately equals intestinal absorption.
As mentioned, phosphate stores and releases energy
through high-energy phosphate bonds and is integral to the
structure of proteins, lipids, and bone (hydroxyapatite). Bone
functions as the primary reservoir of phosphate and calcium in
the body. When the body requires extra calcium, bone is broken
down to release calcium for the bodys use. In addition to calcium,
significant amounts of phosphate also are liberated. This balance
is controlled by the regulatory processes discussed in the previous
section.
Factors favoring cellular uptake include ingestion of glucose
or fructose, alkalosis, insulin level, -adrenergic stimulation, and
anabolism. Phosphate occurs in organic or inorganic forms. Most

intracellular phosphate is organic. Plasma contains lipid phos-
phates, organic ester phosphates, and inorganic phosphates,
including divalent (HPO42) and monovalent (H2PO4) phos-
phate. At physiologic pH, 80% of the inorganic phosphate is diva-
lent. Normally, plasma inorganic phosphate is maintained at 3 to
4.5mg/100mL in adults and 4 to 5mg/100mL in children. Par-
athyroid hormone inhibits proximal tubular inorganic phosphate
reabsorption and increases inorganic phosphate excretion.
In animals that are thyroparathyroidectomized, parathyroid
hormone is absent, and reabsorption of inorganic phosphate
increases significantly and ultimately increases plasma inorganic
phosphate levels. In patients with primary hyperparathyroidism,
parathyroid hormone secretion is elevated, and plasma levels of
inorganic phosphate are low; however, steady-state urinary inor-
ganic phosphate excretion is not markedly increased because it
depends largely on intestinal inorganic phosphate absorption.
Dietary restriction of inorganic phosphate leads to almost 100%
reabsorption of filtered inorganic phosphate and to reduction of
urinary phosphate to zero.42
Hyperphosphatemia
Severe hyperphosphatemia occurs after tissue damage or cell
death. Moderate to severe hyperphosphatemia may be caused by
an impaired ability to excrete phosphorus because of renal failure.
As renal failure worsens, and the glomerular filtration rate declines
to less than 25mL/min, hyperphosphatemia may develop. Other
causes include iatrogenic, hypothermia, massive liver failure,
and certain hematologic malignancies associated with high cell
turnover. The increased cell turnover can be part of the malig-
nancy or may result from cell destruction when chemotherapy is
instituted.
Hypoparathyroidism can cause hyperphosphatemia in the
presence of normal renal function. Rapid increases in serum
phosphate can lead to development of severe hypocalcemia.
Hypocalcemia results from decreased calcitriol production, which
causes significant decline in gastrointestinal tract absorption of
calcium. There may be frank precipitation of calcium and phos-
phate, decreasing serum calcium levels further. When the calcium-
phosphorus product exceeds 70mg2/dL2, the risk of abnormal
calcification is increased. An elevated calcium phosphorus level
represents poor phosphorus control and often is associated with
metastatic tissue calcifications. Treatment involves administra-
tion of phosphate-binding antacids, such as aluminum antacids
and sucralfate, calcium citrate, or calcium carbonate, and may
include dialysis, especially in patients with renal failure (Table
54-14).
Hypophosphatemia
The typical normal range of phosphorus is 2.4 to 4.1mg/dL.
Hypophosphatemia has many causes and is severe when serum
phosphorus levels decline to less than 1mg/dL. Conditions
causing such low phosphate levels include prolonged respiratory
alkalosis and rapid cellular uptake. Severe hypophosphatemia
with total-body deficiency usually reflects poor dietary intake or
consumption of phosphate-binding antacids, or both.
Hypophosphatemia occurs in alcoholism (50% of hospital-
ized alcoholics), refeeding syndrome (commonly in the intensive
Intravascular Fluid and Electrolyte Physiology 1715 54
Table 54-14 Major Causes of Hyperphosphatemia
Mechanism Clinical Syndromes
Binding to serum Including plasma cell dyscrasias
proteins
Decreased renal Renal insufficiency, hypoparathyroidism,
excretion pseudohypoparathyroidism types I and II, tumoral
calcinosis, pseudoxanthoma elasticum, infantile
Section IV Anesthesia Management
hypophosphatasia, hyperostosis, hyperthyroidism,
adrenal insufficiency, bisphosphonate therapy
Increased Phosphorus-containing cathartics; medication with
intestinal vitamin D compounds; granulomatous disease
absorption producing vitamin D, including sarcoidosis and
tuberculosis
Internal Acute metabolic or respiratory acidosis, reduced
redistribution insulin level, clonidine administration
Cellular release Rhabdomyolysis; organ infarction; tumor lysis, as in
Burkitts or lymphoblastic lymphomas or metastatic
small cell carcinoma
Parenteral Intravenous phosphate salts, lipid (phospholipid)
administration infusion
Adapted from Potts JT: Diseases of the parathyroid gland and other hyper- and
hypocalcemic disorders. In Isselbacher KJ, Braunwald E, Wilson JD, et al (eds):
Harrisons Principles of Internal Medicine, 13th ed. New York, McGraw-Hill, 1995, p
2186.
care unit), malnutrition, sepsis, trauma, diuretic or steroid therapy,
ketoacidosis, osmotic diuresis, acidosis, and catabolic states.
Depressed intake or absorption and increased urinary losses are
common causes. Starvation for 48 hours and poor nutritional
status predispose to hypophosphatemia. In individuals with
chronic alcoholism, reduction of the phosphorus content of
skeletal muscle occurs as a result of renal phosphate loss. The
hypophosphatemic syndrome includes phosphate trapping, rhab-
domyolysis, cardiomyopathy, respiratory insufficiency from pro-
found muscle weakness, erythrocyte and leukocyte dysfunction,
skeletal demineralization, metabolic acidosis, and nervous system
dysfunction (Table 54-15).
Before initiating treatment, the cause of hypophosphatemia
should be clearly identified with measurement of arterial blood
gases and the concentrations of ionized calcium, magnesium,
potassium, and serum and urinary phosphorus. Phosphate salts,
such as sodium and potassium phosphate, are available for oral
or intravenous administration. Multiplying the volume of distri-
bution (400mL/kg) by the desired change in inorganic phosphate
provides the total amount to be administered. The rate of intra-
venous administration should not exceed 0.25mmol/kg over 4 to
6 hours to avoid hypocalcemia and tissue damage. Oral supple-
mentation is often limited to 30mmol/day (1g/day) because of
the induction of diarrhea. Hyperphosphatemia should be avoided
because it can cause hypocalcemia and crystal deposition in the
eyes, heart, lung, blood vessels, and kidneys. Most hypophos-
phatemic patients, such as patients with diabetic ketoacidosis or
recovering from exercise, are not severely phosphorus-depleted,
unless they have been sick for an extended time. They typically
may be treated with a glass of milk (100mg/dL or 33mmol/L of
phosphorus). After achieving normal serum phosphate levels, the
concentrations of serum inorganic phosphate and ionized calcium
and a 24-hour urine sample should be monitored to ensure that
balance has been achieved.
IV 1716 Anesthesia Management
Table 54-15 Major Causes of Hypophosphatemia
Chronic alcoholism and alcohol withdrawal
Dietary deficiency and phosphate-binding antacids
Severe thermal burns
Recovery from diabetic ketoacidosis
Hyperalimentation
Nutritional recovery syndrome
Respiratory alkalosis
Therapeutic hyperthermia
Neuroleptic malignant syndrome
Recovery from exhaustive exercise
Renal transplantation
Acute renal failure
Adapted from Potts JT: Diseases of the parathyroid gland and other hyper- and
hypocalcemic disorders. In Isselbacher KJ, Braunwald E, Wilson JD, et al (eds):
Harrisons Principles of Internal Medicine, 13th ed. New York, McGraw-Hill, 1995, p
2185.
Significant hypophosphatemia is common after surgery
and can contribute to respiratory and cardiac failure. In one study
of elective cardiac surgery, 34.3% of patients had significant hypo-
phosphatemia. These patients were characterized by prolonged
ventilation, had a larger cardioactive drug requirement, and had
prolonged hospital stays.43
Chloride Physiology
Chloride is the predominant anion in the ECF volume. Hyper-
chloremic, metabolic acidosis results from excess intake or inad-
equate excretion because of renal dysfunction. When administering
infusions to such patients, bicarbonate, acetate, citrate, or phos-
phate salts should be substituted for chloride salts.
Excess loss of chloride in gastric secretions or urine causes
hypochloremic alkalosis. Chloride depletion tends to limit bicar-
bonate excretion, and this may be caused by reduced delivery of
chloride to the collecting tubules, where chloride is needed for
bicarbonate secretion by means of bicarbonate-chloride exchange.
Sodium reabsorption is enhanced in chloride-depleted states
because it is generally associated with ECF volume depletion.
When less chloride is available for reabsorption, a greater fraction
of the sodium must be reabsorbed with bicarbonate through
increased proton secretion.44 Sodium or potassium chloride
should be administered if intravascular volume depletion or
hypokalemia is present. If these are not a problem, 0.1N hydro-
chloric acid should be administered through a central catheter:
Chloride dose = (Cl desired Cl measured ) 0.2 weight ( kg )
Glucose Physiology and
Fluid Implications
Close Monitoring
Glucose is a crucial fuel source, and insulin facilitates glucose
movement into cells in a process that also requires potassium and
phosphate. Red blood cells, healing wounds, brain, and adrenal
medulla require glucose for fuel, totaling approximately 2mg/kg/
min. When controlling blood glucose, close monitoring is crucial.
Reagent strips containing glucose oxidase, used in conjunction
with glucometers, provide rapid and reliable results. It is impera-
tive to know the insulin regimen or oral hypoglycemic agent of
each patient and then measure the blood glucose preoperatively,
intraoperatively, and postoperatively. The degree of end-organ
damage, coronary artery disease, and autonomic neuropathy can
contribute to the risk of aspiration, myocardial infarction, and
peripheral neuropathy.
Hyperglycemia
Hyperglycemia (>180 to 200mg/dL) is most often caused by
insulin deficiency, insulin receptor resistance, or glucose overad-
ministration. Hyperglycemia produces osmotic diuresis; exacer-
bation of brain, spinal cord,45 and renal damage46 by ischemia;
delayed gastric emptying47; hypophosphatemia47; delayed wound
healing47,48; and impaired white blood cell function.48 Maternal
hyperglycemia increases the risk of neonatal jaundice, the risk of
neonatal brain damage, and fetal acidosis if the fetus becomes
hypoxic.
Even with supramaximal levels of insulin, adults can use
glucose only at a rate of 3 to 5mg/kg/min at rest (approximately
240mL/hr of 5% solution). The maximal rate of metabolism is
less in stress states and more with increased metabolic rates. Gen-
erally, the rate of administration should be limited to 2 to 3mg/
kg/min (120 to 180mg/kg/hr), which is 100g/hr for a 70-kg
person (200mL/hr of a 5% dextrose solution). Healthy infants
and children become hyperglycemic if 5% dextrose is included in
maintenance fluids.49 The maximal rate of glucose disposition in
young children is 4 to 8mg/kg/min, and the optimal rate is less
than 5mg/kg/min.50 It is unclear that glucose administration is
necessary for intraoperative management of most patients.51
Diabetes Mellitus
Diabetes mellitus (see Chapter 35) is the most common endo-
crine disease and is characterized by long-term complications
involving the eyes, kidneys, nerves, and blood vessels. Diabetes is
a major risk factor for heart disease, stroke, kidney disease, blind-
ness, and nontraumatic amputations. The cause of diabetic com-
plications is multifactorial, including glycosylation of proteins
and glucose reduction to sorbitol, which functions as a tissue
toxin. This pathophysiologic process is associated with a decrease
in myo-inositol content and metabolism and with a decrease in
Na+-K+/ATPase activity. Hyperglycemia is recognized as a major
factor in the development of complications associated with dia-
betes. The patient population is not homogeneous, and several
diabetic syndromes have been delineated. There are almost 8
million diagnosed diabetics in the United States and another 8
million who are unaware of their diabetes. These numbers
approach 10% of the overall U.S. population (Table 54-16).52
Pathology
Hyperglycemia of diabetes is the consequence of relative or abso-
lute deficiency of insulin and a relative or absolute excess of glu-
 Intravascular Fluid and Electrolyte Physiology 1717 54
Table 54-16 Classification of Diabetes pregnancies in the United States, resulting in about 135,000 cases
annually.55 Clinical recognition of gestational diabetes is impor-
Class Causes and Characteristics
tant because therapy and antepartum fetal monitoring can reduce
Primary IDDM, type 1without insulin, the patient develops perinatal morbidity and mortality. Maternal complications related
ketoacidosis and dies to gestational diabetes include an increased rate of cesarean
NIDDM, type 2the patient does not always develop delivery.
ketoacidosis without insulin During the past decade, a new disorder known as syn-
Nonobese NIDDM (type 1 IDDM in evolution?) drome X has been described. As the name implies, it is a syn-

Section IV Anesthesia Management

Obese NIDDM drome rather than a specific disease state. The hallmark of
Maturity-onset diabetes of the young syndrome X is insulin resistance with hyperinsulinemia. The
underlying pathology is similar to type 2 diabetes; however,
Secondary Pancreatic disease, hormonal abnormalities, drug or
chemical induced, insulin receptor abnormalities, genetic
patients with syndrome X do not exhibit hyperglycemia. Patients
syndrome, other with syndrome X may never develop type 2 diabetes. The clinical
significance of this condition stems from its association with mul-
IDDM, insulin-dependent diabetes mellitus; NIDDM, noninsulin-dependent
tiple metabolic abnormalities, including low levels of high-density
diabetes mellitus.
Adapted from Foster DW: Diabetes mellitus. In Isselbacher KJ, Braunwald E, Wilson
lipoprotein, increased blood pressure, and increased plasminogen
JD, et al (eds): Harrisons Principles of Internal Medicine, 13th ed. New York, activator inhibitor-1 levels. All these abnormalities have a definite
McGraw-Hill, 1995, p 1980. or possible association with coronary artery disease. Whether
syndrome X and type 2 diabetes are on a spectrum of disease with
insulin resistance as a common denominator or are totally sepa-
rate entities has yet to be clarified.

cagon. In type 1 diabetes, there is an absolute deficiency in insulin

production, and without insulin, patients die. These patients
eventually become dependent on exogenous insulin to prevent
lipolysis and eventually ketoacidosis. The onset of type 1 diabetes
usually occurs by adolescence, although it may occur at any age
and is thought to result from autoimmune destruction of islet
cells in the pancreas.
Type 2 diabetes is characterized by a relative deficiency in
insulin, typically caused by insulin resistance. The onset of this
type of diabetes is usually in adulthood, although evidence sug-
gests that the mean age of onset of type 2 diabetes is decreasing.53
Type 2 diabetes almost certainly has a heterogeneous group of
etiologic factors. Commonly associated findings in type 2 diabe-
tes are obesity, abnormal insulin levels, and a strong genetic com-
ponent (Table 54-17).54
A third type of diabetes mellitus is gestational diabetes (see
Chapter 69). Gestational diabetes is defined as any degree of
glucose intolerance with the onset first recognized during preg-
nancy. Gestational diabetes complicates approximately 4% of all
Management and Evaluation of
Diabetes Mellitus
Treatment and evaluation of diabetes (see Chapter 35) includes
diet, oral hypoglycemic drugs, exercise, exogenous insulin, and
weight reduction if warranted. Insulin preparations are produced
from pork, beef, and human or recombinant sources. Most dia-
betics are prescribed a combination of rapid-acting and interme-
diate-acting insulin before breakfast and before bedtime. Rebound
hyperglycemia may follow a hypoglycemic reaction (i.e., Somogyi
effect). In the past decade, newer therapies for new-onset insulin-
dependent diabetes mellitus (type 1 diabetes) have been intro-
duced, including transplantation of pancreatic tissue and
immunosuppression. Fifteen percent of patients with type 1 dia-
betes have other autoimmune processes, and the elevated glucose
levels are probably caused by destruction of pancreatic cells in
these instances (Table 54-18).

Table 54-17 Characteristics of Type 1 and Type 2 Diabetes Mellitus

Characteristic Type 1 Type 2

Table 54-18 Classification of Insulin Preparations
Genetic locus Chromosome 5 Chromosome 11 (?)
Preparation Type Onset Peak Duration
Age at onset 40 >40
Fast acting Insulin lispro 15min 30-90min 3.5-4hr
Body habitus Normal to wasted Obese injection
Plasma insulin Low to absent Normal to high (Humalog)
Regular 30-60min 2-4hr 6-8hr
Plasma glucagons High, suppressible High, resistant Semilente 1-3hr 5-10hr 16hr
Acute complications Ketoacidosis Hyperosmolar coma Intermediate acting Isophane (NPH) 2-4hr 6-12hr 18-26hr
Insulin therapy Responsive Responsive to resistant Lente (Humulin) 2-4hr 6-12hr 18-26hr

Sulfonylurea therapy Unresponsive Responsive Long acting Protamine zinc 4-8hr 14-24hr 28-36hr
Ultralente 4-8hr 14-24hr 28-36hr
Foster DW: Endocrinology and metabolism. In Wilson JD, Braunwald E, Fauci AS,
et al (eds): Harrisons Principles of Internal Medicine, 12th ed. New York, McGraw- Adapted from Stoelting RK, Dierdorf SF: Endocrine disease. In: Anesthesia and Co-
Hill, 1991, p 1743. Existing Disease, 3rd ed. New York, Churchill Livingstone, 1993, p 340.
IV 1718 Anesthesia Management
Table 54-19 Insulin Analog Comparison
Brand Onset Peak Duration
Analog Name Manufacturer (min) (hr) (hr)
Lispro Humalog Eli Lilly 5-15 1-2 2-3
Glargine Lantus Hoechst- 1-3hr No true 24
Marion peak
Roussel for
Aventis
Aspart NovoRapid NovoNordisk 5-15 1-2 2-3
Insulin therapy can result in anaphylactic and anaphylac-
toid reactions, especially when using protamine-containing
insulin preparations. Protamine-derived insulin is made from fish
sperm and can cause immunologic sensitization when protamine
reversal is administered after cardiopulmonary bypass or heparin
reversal.56 The protamine reaction can be devastating and includes
profound hypotension, pulmonary vasoconstriction, and noncar-
diogenic pulmonary edema. Insulin analogs have been developed
with improved time-action profiles.
The first insulin analog was introduced in the mid-1990s.
Several other insulin analogs subsequently have been formulated
(Table 54-19). Insulin lispro (Humalog) is a genetically engi-
neered protein identical to human insulin except for a reversal of
the amino acids, proline and lysine, at positions 28 and 29 of the
chain.57
This reversal in the amino acid sequence allows for a more
rapid absorption and faster onset of action compared with regular
insulin.58-61 Reversal of the proline-lysine sequence eliminates two
crucial hydrophobic interactions responsible for the -sheet con-
formational changes and dimerization. Decreased propensity for
dimerization is the key to understanding the absorptive charac-
teristics of insulin lispro.
Another insulin analog that has been released is insulin
glargine (Lantus), which resembles the human insulin protein
with addition of two arginine amino acids to the chain and
replaces asparagine with glycine at the -21 position. The addi-
tion of the two arginine units causes a shift in the isoelectric point
to 6.7 0.2, which allows for precipitation of the insulin glargine
in the neutral pH of human tissue. The substitution of glycine in
the chain and the addition of 30 g/mL of zinc increase the
hexamer stability of the precipitate.62,63 This precipitate slowly
dissolves and results in a constant time-action profile with no
pronounced peak. This time-action profile allows for once-daily
dosing and a duration of action of 18 to 24 hours. These new
preparations are being encountered more often in patients being
assessed for surgery, and understanding their time-action profiles
is essential in clinical management.
Anesthetic Considerations
Because diabetes affects multiple organ systems, the perioperative
impact can be profound. Several clinically relevant issues should
be considered during perioperative anesthetic management, as
follows:
1. Diabetes affects oxygen transport by causing glucose to
bind covalently to the hemoglobin molecule, and alters the
allosteric interactions between chains. This alteration in
normal hemoglobin has been shown to decrease oxygen
saturation and red blood cell oxygen transport in pregnant
diabetic patients.64
2. A common complication of diabetes is autonomic dysfunc-
tion. Patients in whom diabetes has been poorly controlled
for many years often have damage to the autonomic nervous
system. One study65 showed that diabetic patients with pre-
viously diagnosed autonomic dysfunction are at increased
risk for intraoperative hypothermia. The pathogenesis may
be related to inappropriate regulation of peripheral vaso-
constriction to conserve body heat.
3. Autonomic dysfunction also affects the bodys ability to
regulate arterial blood pressure, leading to significant
orthostatic hypotension. This underlying defect is caused
by a lack of appropriate vasoconstriction. This denervation
also may involve vagal control of the heart rate. The changes
in heart rate seen with atropine and -blockers are blunted
in patients with significant autonomic dysfunction.66
4. Damage to the autonomic nervous system can significantly
affect the choice of anesthetic technique. Patients are at
significantly increased risk of hypotension caused by induc-
tion agents, such as thiopental or propofol. For this reason,
etomidate may be a better induction agent because of
its considerably lower incidence of cardiovascular side
effects.
5. Diabetes has well-defined adverse effects on the cardiovas-
cular system. Men who have diabetes have twice the age-
adjusted risk for coronary artery disease. The risk for
women is tripled, indicating that they may be even more
sensitive to the cardiovascular effects of diabetes.67 Data
show that patients with diabetes may be at even greater risk
for coronary artery disease than was previously suspected.
One study revealed that patients with type 2 diabetes had
as great a risk for myocardial infarction as nondiabetic
patients who already had a previous myocardial infarc-
tion.68 This information reinforces the point that diabetic
patients must be carefully evaluated preoperatively for
coronary artery disease. Also, diabetic patients are more
likely to have silent ischemia. They may not experience the
classic chest pain and tightness associated with ischemic
heart disease. Questions regarding exercise tolerance and
shortness of breath with exertion may provide important
information regarding underlying heart disease or the
degree of compensation.
6. Diabetes affects the gastrointestinal tract in several ways.
First, it damages the ganglion cells of the gastrointestinal
tract, inhibiting motility, which delays gastric emptying and
overall transit time through the gut. This increased transit
time has a significant impact on the practice of anesthesia
in that all diabetic patients should be treated as if they have
a full stomach. Preoperative treatment with agents that
inhibit acid secretion and neutralize stomach acid is essen-
tial. Rapid-sequence induction is commonly employed to
try to minimize this risk of aspiration.
Perioperative and intraoperative glycemic-control regi-
mens depend on several factors. First, differentiating type 1 from
type 2 diabetes is crucial. Patients with type 1 diabetes are at risk
for ketonemia if they are without insulin. The risk of ketosis is
amplified when the patient undergoes the stress of surgery.

Second, the degree to which blood glucose levels are controlled
long-term affects management. Glycosylated hemoglobin (hemo-
globin A1c) is the most accurate way to assess glucose control over
the previous 2 to 3 months. As levels of hemoglobin A1c increase,
so does the complication rate of diabetes. The amount of exoge-
nous insulin a patient normally requires is important in deciding
how blood glucose should be treated intraoperatively. The mag-
nitude of the surgery plays an important role in determining
therapy. There are many different protocols for preoperative and
intraoperative insulin management, but there are few prospective
studies comparing regimens. Common protocols are discussed
subsequently.
Successful perioperative glucose management depends on
careful monitoring. Perioperative management of blood glucose
during a brief surgical procedure in a patient with diet-controlled
diabetes generally involves only monitoring of blood glucose
immediately perioperatively and every 3 hours until oral intake
is resumed.69,70 The preoperative physical examination and history
may reveal extensive diabetic neuropathy, which may be seen as
orthostatic hypotension, syncopal episodes, mononeuropathies or
polyneuropathies, erectile or bladder dysfunction, and an electro-
cardiogram showing a loss of R-to-R variability. Patients may
present with numerous additional findings, including nonfamilial
short stature, cerebrovascular disease, renal dysfunction, micro-
albuminemia, and tight, waxy skin. In an estimated 30% to 40%
of diabetic patients, glycosylation of the atlanto-occipital joint
may limit joint mobility and cause difficulty with airway manage-
ment (i.e., stiff-neck syndrome).71,72 Laboratory evaluation of
hemoglobin A1c is an accurate measure of the severity of hyper
glycemia and has been shown to correlate directly with increasing
rates of complications. Conversely, lower hemoglobin A1c values
are associated with decreased risk and can be considered a
measure of the quality of the diabetic care. Hemoglobin A1c pro-
vides the best evidence of overall blood glucose control over the
past 1 to 2 months and should replace the oral glucose tolerance
test as the gold standard for diagnosing diabetes.73 Basic electro-
lyte and renal function tests should be evaluated, especially if
the patient has frequent urinary tract infections or renal
impairment.74
The preoperative hemoglobin A1c level gives the anesthesi-
ologist a reasonable idea about the patients blood glucose level
over the past several months, and this information can be used
to evaluate the need for preoperative and intraoperative insulin
requirements. The regimen selected to manage patients with dia-
betes undergoing surgery depends on the severity of the diabetes
and the magnitude of the surgery. Frequent glucose monitoring
and preparation for insulin administration are essential for a
diabetic patient. Recommendations include discontinuation of
long-acting insulin or oral hypoglycemic agents 1 to 2 days preop-
eratively. Short-acting insulin should be administered every 4 to
6 hours subcutaneously, with the dose adjusted according to
glucose levels just before administration. Metformin, a biguanide
oral hypoglycemic, also possesses a low risk of lactic acidosis (0.03
cases per 1000 patient-years). Despite these low figures, met-
formin should not be used in patients with mild renal dysfunction
(serum creatinine >1.5mg/dL in male patients or >1.4mg/dL in
female patients), congestive heart failure, recent myocardial infarc
tion or any other condition producing a hypoxic state, current
alcohol abuse, or impaired hepatic function. Metformin should
be discontinued 24 hours before and for at least 48 hours after
any procedure using intravenous contrast dye. It should be rein-
Intravascular Fluid and Electrolyte Physiology 1719 54
stituted only after renal function has been reevaluated and found
to be normal.75,76
The typical sliding scale is destined to fail because it
involves the administration of a fixed dose after documentation
of hyperglycemia. A small modification improves control. The
selected dose should be administered every 4 to 6 hours, based
on response. If the glucose level is less than 60mg/dL, the dose
should be with held for at least 1 hour, and 50% dextrose should
Section IV Anesthesia Management
be given intravenously (0.01 to 0.02mL/kg/min), with blood
glucose monitored hourly. When the blood glucose is greater than
125mg/dL without supplemental dextrose infusion, the next
insulin dose should be 20% to 40% smaller. If the glucose is less
than 100mg/dL or is less than 125mg/dL and decreasing, the
scheduled dose should be maintained until the hourly measured
glucose is greater than 125mg/dL, followed by resumption with
a 10% to 20% smaller dose. If the glucose level is 100 to 200mg/
dL and stable, the current dose and interval are continued. If the
glucose level is 200 to 350mg/dL, the scheduled dose is increased
by 10% to 20%. If the glucose level is more than 350mg/dL, the
dose is increased by 20% to 40%.
On the day of surgery, a dextrose infusion (2mg/kg/min)
is started at the time a meal would have been ingested, and glucose
is measured preoperatively. For patients currently receiving
insulin, an insulin infusion is started at a rate of 0.5 to 1.25U/hr,
depending on the amount of insulin normally administered and
the current glucose level. Blood glucose is monitored hourly, and
the infusion rate is adjusted to maintain glucose concentration
based on which glucose homeostasis protocol has been selected
for use. Patients with obesity, liver disease, severe infections or on
steroid therapy may require increased hourly insulin doses as may
patients undergoing cardiopulmonary bypass.77 After the blood
glucose level is stable, urine glucose levels and ketone bodies are
checked to ensure that glycosuria resulting from a low threshold
does not confuse interpretation of urine output.
Oral intake may be stopped 12 hours before anesthesia
because some element of gastroparesis exists in these patients.
Typically, patients are administered a histamine blocker along
with a gastric-emptying drug, such as metoclopramide, the night
before and the morning of surgery. Because the gastrointestinal
tract is a prime target for autonomic neuropathy, there may be
esophageal dysfunction with difficulty swallowing, constipation,
or diarrhea.
Diabetic ketoacidosis, dehydration, impaired wound
healing, and electrolyte imbalance are minimized with the proper
use of exogenous insulin. There is no clear consensus about the
method of insulin therapy or the exact range of blood glucose that
can affect morbidity or mortality.78
If a general anesthetic is used, clinical consideration should
include a rapid-sequence induction because of the high rate of
gastroparesis. Cerebrovascular accidents, peripheral vascular
disease, and cardiovascular infarction are 2 to 10 times more
common in diabetics. Strategies designed to reduce the risk of
labile blood pressures and myocardial ischemia related to auto-
nomic or vascular disease may include -blockade to blunt the
stress of induction, a narcotic-based anesthetic to minimize car-
diopulmonary depression, and prophylactic nitroglycerin in these
patients with their significant risk of coronary artery disease.
Commonly associated conditions include obesity and stiff cervi-
cal joints, which may make airway management challenging.
Associated cardiovascular conditions often result in the need for
additional invasive monitoring.
IV 1720 Anesthesia Management
Acute Hyperglycemia
The cellular effects of acute hyperglycemia are poorly understood,
but reports have begun to shed some light on alterations in
normal cell functions at the molecular level. One study79 showed
that acutely elevated glucose levels depress endothelin-induced
calcium signaling in rat mesangial cells, depressing the contractile
state of glomerular cells. Such information supports the hypoth-
esis that acute hyperglycemia can affect cellular functions and
may lead to clinically evident pathology.
National trends show a movement toward tighter control
of glucose levels perioperatively. Acute consequences of elevated
serum glucose levels include impaired wound healing, dehydra-
tion, impaired immune system response, and proteolysis. The
dehydration seen during acute hyperglycemia results from the
osmotic diuretic effect of high serum glucose levels. Patients with
diabetes are well known for having poor wound healing, which is
commonly believed to be caused by impaired blood flow to the
wound. There is evidence, however, that acute hyperglycemia may
impair fibroblast activity and impair vitamin C uptake by cells,
inhibiting new collagen synthesis.80
In addition to delayed wound healing, elevated glucose
levels inhibit immune function and increase the risk of postop-
erative infection. One study81 suggested that continuous insulin
infusions, used to control intraoperative and postoperative glucose
levels, decrease the incidence of sternal wound infections after
cardiac surgery. A second study82 found that aggressive glucose
control intraoperatively increased neutrophil activity in vitro sig-
nificantly. Alveolar macrophages from normal hosts show
impaired respiratory burst when exposed in vitro to elevated
glucose concentrations.83 Taken together, these data indicate that
tight glucose control intraoperatively may significantly influence
the ability of patients to recover more quickly from surgery and
reduce morbidity.
Diabetic Ketoacidosis
Diabetic ketoacidosis is an emergent condition that often mani-
fests in a diabetic patient with leukocytosis and an acute surgical
abdominal emergency or with nausea, vomiting, lethargy, and
signs of hypovolemia. The priorities are to restore intravascular
volume (usually rapid intravenous administration of 1L of saline),
administer regular insulin (traditionally at a starting infusion rate
of 0.1U/kg/hr), eliminate the ketonemia, control blood glucose,
and correct the underlying problem (e.g., antibiotics for urosepsis
or pneumonia). Patients with ketoacidosis are dehydrated as a
result of glucosuria; because the dehydration is caused by water
and electrolyte loss, colloids are not indicated. If the patients
osmolality is elevated, 0.45% sodium chloride can be adminis-
tered, and the volume administered should be guided by the
hemodynamic response and urinary output. When the blood
glucose declines to less than 250mg/dL, dextrose should be
added to the fluids. Urine or blood ketones are monitored every
2 hours after the blood glucose is within 100 to 200mg/dL. If
ketones are still present, the rates of glucose and insulin infusions
should be increased.
Osmotic diuresis promotes loss of sodium, potassium,
magnesium, and phosphate. Despite total-body deficiencies,
however, their concentrations may be elevated at the time of
presentation because of severe water loss. The best treatment for
hyperkalemia in these patients is appropriate therapy for the dia-
betic ketoacidosis. Potassium levels decrease rapidly with appro-
priate volume replacement and insulin therapy. Most patients
require electrolyte replacement after volume expansion has begun.
Alternatively, severe hyperglycemia may extract water from the
intracellular space and may dilute the electrolyte concentrations.
If the initial potassium level is elevated, or the patient is anuric,
potassium should not be administered. As hydration improves,
and urinary output increases, potassium, magnesium, and phos-
phorus should be administered, and levels should be monitored
frequently. Generally, acidosis should not be treated with buffers.
Ketoacidosis is corrected as insulin and glucose levels improve,
and lactic acidosis resulting from poor perfusion responds to
intravascular fluid replacement. If the pH approaches 7.15, the
bicarbonate ion concentration is less than 10mEq/L, and hypo-
tension fails to respond to intravascular fluid administration,
sodium bicarbonate therapy may be required.
Hypoglycemia
Hypoglycemia (<50mg/dL) is dangerous because glucose is the
sole fuel source for much of the brain. Threshold levels depend
on age. Signs of hypoglycemia include irritability, seizures, brady-
cardia, hypotension, and respiratory failure. Symptoms com-
monly occur in adults with blood glucose concentrations less
than 57mg/dL or in infants with levels of 30 to 50mg/dL. Symp-
toms are seen at higher levels in diabetic patients than in nondia-
betic patients and are obscured by general anesthesia. At about
70mg/dL, a biochemical stress response occurs, including sym-
pathetic nervous system stimulation and elevated levels of growth
hormone or cortisol, or both. Neurologic and electroencephalo-
graphic depression appears at 50 to 55mg/dL in nondiabetic
patients and 70 to 85mg/dL in diabetic patients.84 Selective neu-
ronal necrosis, not infarction, occurs in the caudate, putamen,
and cortex. There is a compensatory increase in cerebral blood
flow.
During labor, maternal starvation-induced ketosis has
adverse fetal effects, including fetal ketonemia, hypoxia, and fetal
lactic acidosis.85 Neonates are at an increased risk for hypogly
cemia because of limited glycogen stores and from large amounts
of fetal insulin production in response to gestational hypergly
cemic states. Adults also are at risk for hypoglycemia from inad-
equate gluconeogenesis coupled with inadequate nutritional
intake or from excess insulin (e.g., from an insulinoma, pancre-
atic islet cell adenoma, or carcinoma; iatrogenic overadministra-
tion). Hypoglycemia also can follow too abrupt cessation of
dextrose infusion during total parenteral nutrition (i.e., reactive
hyperinsulinemic states). Inadequate gluconeogenesis occurs in
liver failure, cortisol deficiency (primary or secondary), inade-
quate glucagon response, and growth hormone deficiency, and
during -adrenergic blockade. Fasting in women is likely to
produce hypoglycemia in 24 hours, whereas men tolerate about
72 hours of fasting.47 The incidence of hypoglycemia in healthy
infants and children is low (2 of 446) with 4 to 8 hours of fasting.50
Fetal hypoglycemia occurs if maternal glucose is greater than
150mg/dL because glucose crosses the placenta, inducing fetal
insulin secretion. Treatment consists of an intravenous bolus of

5g of dextrose followed by increasing the rate of dextrose infu-
sion by 1 to 2mg/kg/min.
Acid-Base Factors
Metabolic Acidosis
The necessity for treating metabolic acidosis is judged largely on
clinical grounds, with the metabolic acid level base excess (BE)
used to determine the correct dose of bicarbonate (see Chapter
49). This prediction is based on the size of the treatable space and
the magnitude of the problem:
Dose (mEq ) = 0.3 weight ( kg ) BE (mEq L )
The dose calculated is sufficient to return the metabolic
disturbance almost to zero. Bicarbonate therapy is reserved,
however, for emergencies and situations in which indications for
therapy are compelling. It is usual to give a reduced dose (about
one half) for several reasons, as follows:
1. The bicarbonate is injected initially into the plasma volume
(about 3L) instead of into the calculated treatable space
(21L).
2. When bicarbonate is added to acid, it fizzes. This does not
occur literally in the blood. Nevertheless, most bicarbonate
is converted to carbon dioxide and has to be eliminated.
For each 100mEq that is converted, about 2.24L of carbon
dioxide has to be exhaled, equivalent to 10 minutes of
normal production.
3. The carbon dioxide that is produced enters the cells freely,
in contrast to the bicarbonate ions that have been admin-
istered. The inside of the cell initially may become even
more acid. Direct studies with nuclear magnetic resonance
have not confirmed this, however (Severinghaus, personal
communication, 1986).
4. The bicarbonate is accompanied by sodium ions that are
responsible for a residual increase in the osmolality of the
ECF. In combination with other therapy, such as intrave-
nous glucose, the hyperosmolality may be critical and cause
coma. In neonates, rapid infusion of bicarbonate may cause
intracranial hemorrhage.
5. If the body deals with its metabolic acidosis, there is a
residual metabolic alkalosis.
Anion Gap
Some causes of metabolic acidosis release anions into the ECF
that are not normally measured. When this occurs, there is
an unexpected discrepancy between the sums of the principal
cations and anions. The usual sum is shown in the following
equation:
Na + + K + = Cl + HCO3 + gap
140 + 5 = 105 + 25 + ( 15)
When there are any additional, unmeasured anions, they
become part of the gap, which is then correspondingly larger. A
Intravascular Fluid and Electrolyte Physiology 1721 54
gap greater than 30 suggests that there is an increase in the con-
centration of the unmeasured anions. This method relies on the
accuracy of the other measurements. Small errors in these large
numbers cause a proportionately greater error in the result. If
information is required about these anions, it is more appropriate
to measure their concentration. In practice, it suffices to analyze
lactate in cases of tissue hypoxia, 3-hydroxybutyrate in diabetic
ketosis, and phosphate or sulfate in renal failure.
Section IV Anesthesia Management
The common causes of perioperative metabolic alkalosis
include antacid therapy, incidental administration of citrate with
blood products, sodium bicarbonate administration, gastric
drainage, or renal bicarbonate retention caused by diuresis or in
compensation for respiratory acidosis. Sodium chloride or potas-
sium chloride can be administered orally or through a peripheral
intravenous catheter, or 0.1N hydrochloric acid may be adminis-
tered slowly through a central venous catheter. The required dose
is calculated as follows:
Dose = (Cl desired Cl measured ) 0.2 (L ) weight ( kg )
Clinical Acid-Base Balance
Disturbances
This section provides practical clinical examples to assist in
understanding acid-base balance.
Metabolic Acidosis from
Low Cardiac Output
A patient with coronary artery disease and an ejection fraction
of 15% experienced myocardial perforation while undergoing
balloon angioplasty in the cardiac catheterization laboratory.
After resuscitation, intubation, insertion of an arterial line, and
initiation of dopamine by infusion, the patient was transferred to
the operating room for emergency thoracotomy. Blood pressure
was 80/50mmHg, and the heart rate was 120 beats/min. During
preparation for central venous cannulation, the blood pressure
decreased, followed by cardiac arrest. The patient underwent
immediate thoracotomy. Blood gas determinations obtained at
the time were as follows: pHa 7.15, Paco2 35mmHg, BE 15mEq/
L, bicarbonate 12mEq/L, and Pao2 90mmHg. These values indi-
cated a severe metabolic acidosis with mild respiratory alkalosis.
After the administration of 4U of packed red blood cells (PRBCs)
and 88mEq of sodium bicarbonate, repeat blood gas determina-
tions were as follows: pHa 7.14, Paco2 39mmHg, BE 14mEq/L,
bicarbonate 13mEq/L, and Pao2 95mmHg. These values still
indicated a severe metabolic acidosis, but with no respiratory
alkalosis.
The first blood gas determination revealed a metabolic aci-
dosis attributable to the low cardiac output with tissue ischemia
and with slight hyperventilation. The second blood gas determi-
nation showed almost no benefit from the bicarbonate, probably
because of continuing low cardiac output. Bicarbonate is con-
verted to carbon dioxide, which may explain the slightly higher
level of Paco2.
IV 1722 Anesthesia Management
Metabolic Acidosis During
Anhepatic Phase
A patient with end-stage liver disease caused by viral hepatitis
underwent orthotopic liver transplantation. Blood gas analysis
performed late in the anhepatic phase (see Chapter 67), after
earlier administration of bicarbonate, revealed the following: pHa
7.10, Paco2 28mmHg, BE 18mEq/L, bicarbonate 8.6mEq/L,
Pao2 260mmHg, and Na+ 153mEq/L. These values indicated a
severe metabolic acidosis with a moderate respiratory alkalosis.
The patient was treated with 500mL (150mEq, 18g) of 0.3M
solution of tris(hydroxymethyl)aminomethane (THAM) by infu-
sion over 20 minutes. Repeat blood gas analysis showed the fol-
lowing: pHa 7.09, Paco2 30mmHg, BE 18mEq/L, bicarbonate
9mEq/L, Pao2 251mmHg, and Na+ 154mEq/L. A second dose
of THAM was administered by infusion. After transplantation,
blood gas analysis showed the following: pHa 7.30, Paco2
33mmHg, BE 9mEq/L, bicarbonate 16mEq/L, Pao2 283mmHg,
and Na+ 149mEq/L. These values indicated a marked metabolic
acidosis with a mild respiratory alkalosis typical of a partially
compensated metabolic disturbance.
The metabolic acidosis during the anhepatic phase showed
the importance of the liver in metabolizing lactate to bicarbonate.
The high sodium militated against repeat administration of
sodium bicarbonate. In this case, THAM was used to minimize
sodium administration while controlling the metabolic acidosis.
On reperfusion (phase 3), the new liver started to function, and
there was improvement in the metabolic acidosis.
Metabolic Alkalosis After Diuretics
A 3-month-old, 4-kg infant boy was presented for repair of an
atrioventricular canal (see Chapter 83). The patients congenital
heart failure had been controlled with digitalis and furosemide.
Before induction of anesthesia, blood gas analysis showed the
following: pHa 7.53, Paco2 44mmHg, BE 14mEq/L, bicarbonate
35.9mEq/L, and Pao2 110mmHg. These values showed a severe
metabolic alkalosis with a minimal respiratory acidosis typical of
a partially compensated metabolic disturbance. After induction
of anesthesia, blood gas analysis showed the following: pHa 7.61,
Paco2 35mmHg, BE 14mEq/L, bicarbonate 34mEq/L, and Pao2
60mmHg. After completion of the procedure and cessation of
cardiopulmonary bypass, the blood gas analysis showed the fol-
lowing: pHa 7.40, Paco2 33mmHg, BE 4mEq/L, and bicarbo-
nate 20mEq/L.
The hypochloremic alkalosis that accompanies prolonged
administration of furosemide explains the first set of results. After
induction of anesthesia, the hyperventilation decreased the Paco2,
but had no effect on the base excess. Two major factors explain
the change during the procedure. Hemodilution tends to return
the pHa toward neutral (6.8 at body temperature), and low per-
fusion during bypass frequently produces a metabolic acidosis
that, in this case, also helped to overcome the initial metabolic
alkalosis.
Chronic Respiratory Failure in a Neonate
A 2.5-month-old, 3.5-kg infant boy with bronchopulmonary dys-
plasia was tracheally intubated and received supplemental oxygen
(Fio2 0.3). He presented for cryoablation for proliferative retin-
opathy (see Chapters 82 and 83). After the patient was trans-
ported to the operating room, blood gas analysis showed the
following: pHa 7.27, Paco2 65mmHg, BE 3mEq/L, bicarbonate
29mEq/L, and Pao2 35mmHg. Anesthesia was induced, and ven-
tilation was controlled, with a tidal volume of 60mL, peak inspir-
atory pressure of 25cm H2O, positive end-expiratory pressure of
4cm H2O, respiratory rate of 40 breaths/min, no inspiratory
pause, and Fio2 of 0.3. Repeat analysis of blood gases showed the
following: pHa 7.36, Paco2 50mmHg, BE 3mEq/L, bicarbonate
28mEq/L, and Pao2 48mEq/L.
The mild metabolic alkalosis of the first analysis is compat-
ible with compensation for chronic respiratory failure. The high
Paco2 value represented hypoventilation during transport. Appro-
priate selection of the ventilator settings returned the Paco2 to a
value more typical for a neonate with bronchopulmonary dyspla-
sia with a pHa about halfway between no compensation and com-
plete compensation (see Chapters 82 and 83).
Preparation for Chronic Hyperventilation
A clinician was a member of a team providing plastic surgery at
high altitude. For 2 days preceding the trip, he was advised to take
a carbonic anhydrase inhibitor, acetazolamide, which would effec-
tively slow the normal reversible reaction between carbonic acid
and its dissociation products, carbonic acid and water. The drug
alkalinizes the urine and produces a moderate metabolic acidosis,
the typical response to hypoxic hyperventilation at altitude. A
typical blood gas determination after acclimatization to altitude
shows the following: pHa 7.42, Paco2 31mmHg, BE 4mEq/L,
bicarbonate 19mEq/L, and Pao2 80mmHg.
The benefit of taking acetazolamide is debated by climbers,
and the drug is not without side effects. During the initial experi-
ence of taking acetazolamide at sea level, a striking, if minor, side
effect is the metallic taste associated with drinking any carbon-
ated beverage, presumably attributable to the inability to convert
the high carbon dioxide content in the beverages to carbonic
acid.
Acute Hyperventilation and Hypotension
A 5-ft, 48-kg elderly woman had an arterial blood pressure of
115/75mmHg and heart rate of 70 beats/min (see Chapter 71).
She was receiving no medications and was in otherwise normal
health. She was scheduled for a laparotomy and hysterectomy.
After induction of anesthesia and orotracheal intubation, control-
led ventilation was initiated with a tidal volume of 900mL and a
respiratory rate of 12 breaths/min. Shortly before incision, the
arterial blood pressure was 58/35mmHg, and the end-tidal Pco2
was 22mmHg. (Blood gases were not tested, but typical values
for a patient experiencing such acute hyperventilation are as
follows: pHa 7.53, Paco2 26mmHg, bicarbonate 21mEq/L, and
BE 0mg/L.)
The ventilator was immediately reset to provide 4 breaths/
min at a tidal volume of 500mL. Arterial blood pressure about 2
minutes later was 85/55mmHg, and the next reading about 2
minutes later still was 95/65mmHg. The patients end-tidal Pco2
increased more slowly, and after about 8 minutes, it had returned
to 36mmHg when the ventilatory rate was increased to 8 breaths/

min. (Typically, her blood gas values would be near-normal, such
as pHa 7.44, Paco2 38mmHg, bicarbonate 23.3mEq/L, and BE
0mg/L). Hyperventilation after induction is common and fre-
quently contributes to hypotension by various methods, includ-
ing increased intrathoracic pressure, diminished venous return,
and depressed cardiac output.
Acute Hypoventilation
An obese, middle-aged woman was undergoing laparoscopic
cholecystectomy. After a routine induction and orotracheal intu-
bation, low-flow anesthesia was initiated, and the ventilator was
set to provide a tidal volume of 700mL at a rate of 10 breaths/min
(measured exhaled tidal volume was 560mL). Ten minutes after
induction of anesthesia, the Pco2 was 35mmHg, and the peak
airway pressure was 28mmHg. Arterial blood gas determina-
tions showed the following: pHa 7.43, Paco2 35mmHg, bicarbo-
nate 22.4mEq/L, and BE 1mEq/L. Thirty minutes into the
procedure, the peak airway pressure was 48mmHg, the end-
expired Pco2 was 49mmHg, and the measured exhaled tidal
volume was 380mL. Repeat blood gas analysis showed the fol-
lowing: pHa 7.27, Paco2 55mmHg, bicarbonate 24.9mEq/L, and
BE 1mEq/L. The principal abnormality was a marked respira-
tory acidosis with no metabolic acidosis. It is characteristic of an
acute respiratory disturbance.
The increase in Paco2 can be attributed to two factors:
uptake of the inflating gas (CO2) in the abdomen and reduction
of effective ventilation owing to increased airway pressure
required to overcome abdominal distention. The discrepancy
between end-tidal Pco2 and arterial blood gas Paco2 is typical
and results from the patients temperature being slightly less than
that of the analyzer, lack of time between breaths to reach true
end-tidal equilibration, normal pulmonary maldistribution, and
the slight loss occurring on the length of the tubing.
Fluid Balance and Fluid
Replacement Solutions
Transfusions of red blood cells, platelets, fresh frozen plasma,
factor solutions, synthetic crystalloids and colloids, and cryopre-
cipitate have the potential for improving clinical outcomes in
perioperative and peripartum settings. Potential benefits include
improved tissue oxygenation and decreased bleeding; however,
transfusions are not without risks or costs. Blood product trans-
fusion can cause many complications. Although red blood cells
and coagulation products have excellent volume-expanding
capacity, they are not considered in this chapter (see Chapter
55).
The American Society of Anesthesiologists and the Ameri-
can Red Cross Blood Services have convened task forces and
medical advisory boards on blood component therapy over the
past decade to develop evidence-based guidelines on the proper
indications for perioperative and peripartum administration of
red blood cells, platelets, fresh frozen plasma, and cryoprecipitate.
This section details only basic crystalloid, colloid, and hypertonic
Intravascular Fluid and Electrolyte Physiology 1723 54
solution recommendations applicable to typical surgical and
obstetric patients (Table 54-20). Information about infants, chil-
dren, and special clinical settings can found elsewhere in this
book (see Chapter 82).
Surgical Fluid Balance and Shock
Section IV Anesthesia Management
Physiologic changes during surgery and anesthesia lead to shifts
in fluid balance. Epidural, spinal, and caudal anesthesia all may
cause variable amounts of sympathetic blockade (see Chapter 51).
Although young, healthy, and well-hydrated patients may tolerate
sympathectomy, patients who are older, severely dehydrated, or
on antihypertensive drugs or diuretics may not be able to respond
to the effects of sympathectomy. It is common to administer 1L
of fluid intravenously before placement of a spinal anesthetic, or
to administer fluid concurrently when epidural anesthesia is
being induced. Vasopressors, typically ephedrine or phenyl
ephrine, are needed to overcome the hemodynamic effects of
sympathetic block.
Although inhaled anesthetics do not directly alter fluid
losses, all anesthetics may blunt the normal physiologic responses
to hypovolemia and the stress response. The stress response to
surgery involves an increase in ADH production, which can be
blocked with anesthetics. Superimposed are the variable effects of
intravenous and inhaled agents on the myocardium, venous
return, blood pressure, and the vasculature. Mechanical ventila-
tion can decrease the release of atrial natriuretic hormone and
increase the release of ADH, resulting in retention of sodium and
fluids.
Simple restoration of blood volume may not ensure sur-
vival. In addition to blood loss, significant third space loss may
occur, which essentially involves fluid that is still in the body, but
not contributing to intravascular volume, oxygen delivery, or
waste removal; this is often difficult to measure. Patients undergo-
ing major surgical procedures require fluid replacement beyond
simple blood loss, and the anesthesiologist plays a vital role in
assessing and ultimately administering appropriate fluid therapy
in intraoperative and postoperative clinical settings. Even after
normalization of cardiac output, cells in the liver and gut may
remain ischemic after flow is reestablished in the macrocircula-
tion because of occlusion of capillary networks caused by
edema.86
The controversial issue of timing and volume of fluid resus-
citation in uncontrolled hemorrhage has been reviewed. In 52
animal trials, fluid resuscitation seemed to decrease the risk of
death in models of severe hemorrhage, but increased the risk of
death in models with less severe hemorrhage. Hypotensive resus-
citation, whenever tested, reduced the risk of death.87
Shock
Shock can be defined as dysfunction of intracellular processes as
a result of lack of energy (see Chapter 72). This lack of energy is
caused by deficient delivery of oxygen to tissue, resulting in an
inability to maintain normal cellular respiration. Although shock
is a cellular problem, therapy is directed at the systemic delivery
of oxygen to the tissue.
IV 1724 Anesthesia Management

Table 54-20 Composition of Replacement Fluids

Fluid Sodium (mEq/L) Potassium (mEq/L) Glucose (g/L) Osmolarity pH Other

Whole blood in CPD* 168-156 3.9-21 7.20-6.84 Hct = 35-40

PRBC, AS-1 117 ?-49 552 6.6 Hct = 59

PRBC, CPD ?-95 6.6 Hct = 77

PRBC, CPDA-1 169-111 5.1-78.5 7.55-6.71 Hct = 65-80

FFP 15.4

5% Albumin 145 15 <2.5 0 330 7.4 COP = 32-35mmHg

2.5% Albumin 145 15 <2 0 330

Plasmanate 145 15 <2 7.4 COP = 20mmHg

10% Dextran 40 0 0 50 255 4.0

Hetastarch 154 0 0 310 5.9

0.9% Sodium chloride 154 0 0 308 6.0

Lactated Ringers solution 130 4 0 273 6.5 Lactate = 28

5% Dextrose 0 0 50 252 4.5

D5LR 130 4 50 525 5.0

D50.45% NaCl 77 0 50 406 4.0

Normosol 140 5 100 555 7.4 Mg = 3, acetate = 27,

gluconate = 23

Normosol-M 40 13 50 363 5.0 Mg = 3, acetate = 16,

gluconate = 27

Normosol-R 140 5 0 294 6.6 Mg = 3, acetate = 27,

gluconate = 23

D5 Normosol-R 140 5 0 547 5.2 Mg = 3, acetate = 27,

gluconate = 23

Normosol-R pH 7.4 140 5 0 295 7.4 Mg = 3, acetate = 27,

gluconate = 23

*The range of values refers to concentrations on day 1 and day 21 (CPD), 35 (CPDA-1), or 42 (AS-1) of storage.

Fructose substituted for dextrose.
COP, colloid oncotic pressure; D5LR, 5% dextrose in lactated Ringer solution; D50.45% NaCl, 5% dextrose in 0.45% sodium chloride; FFP, fresh frozen plasma; Hct, hematocrit;
Mg, magnesium; PRBC, packed red blood cells.

In normal patients, the body maintains a balance between
delivery of oxygen (Do2) and global oxygen consumption (Vo2).
Global oxygen consumption is a measure of the total amount of
oxygen consumed by all tissues per minute. The amount of oxygen
consumed as a fraction of the amount delivered defines the
oxygen extraction ratio (OER):
OER = VO2 DO2
Vo2 in a normal adult undertaking routine activities is
approximately 250mL/min with an extraction ratio of 25%. The
OER can be increased to 75% during extreme exercise in a healthy
adult. Oxygen not extracted from the blood returns to the lungs
and can be measured as the mixed venous saturation (Svo2),
which is a measure of global tissue oxygenation. Svo2 less than
60% to 70% usually indicates insufficient Do2. A mixed venous
blood sample is needed because the saturation of venous blood
varies depending on the organ being measured. Venous blood
returning from the liver has a saturation of 40% to 50%, and blood
from the kidneys may have a saturation greater than 80%.
Oxygen consumption is directly related to cellular meta-
bolic rate. Sympathetic activity, shivering, and physical activity all
increase the cellular metabolic rate. As the delivery of oxygen
decreases or the consumption increases, the OER increases to
maintain aerobic metabolism. The OER can increase to approxi-
mately 60% to 70%, at which point further increase in consump-
tion or decrease in delivery results in tissue hypoxia and anaerobic
metabolism.
Tissue hypoxia and shock can result from inappropriate
delivery of cellular oxygen or inappropriate cellular use of oxygen.
There are two distinct types of shock. Type I involves patients who
experience decreased delivery of oxygen, such as in hypovolemic
or cardiogenic shock. Type II involves problems with inappropri-

ate distribution of oxygen delivery, such as in septic or neurogenic
shock. Regardless of the underlying cause, appropriate fluid man-
agement is an absolute necessity.
Proper fluid management of shock can have a significant
impact on the delivery of oxygen to tissues and prevention of
anaerobic metabolism. Until more recently, vigorous fluid loading
and inotropic agents were commonly used in what was called
goal-directed therapy during the care of critically ill patients.
Goal-directed therapy was based on the principle that critically
ill patients had supranormal Vo2 values. It was thought that by
increasing Do2 to supraphysiologic levels, cellular metabolism
could be maintained in an aerobic state, and multisystem organ
dysfunction could be avoided.88 In the 1990s, two large, random
ized trials showed this therapy to be ineffective and potentially
dangerous, however.89,90 Continuing research has shown that
identifying and treating volume depletion and poor cardiac func-
tion early in shock is beneficial.91-93
The key to the different results between the early study by
Shoemaker and colleagues88 and later trials is recognizing the
difference between early and late shock. Early shock involves
acute tissue hypoxemia with end organs not yet sustaining irre-
versible damage. In type I shock (hypovolemic or cardiogenic),
prompt treatment of the underlying cause of tissue hypoperfusion
may prevent patients from progressing to late shock and multi-
system organ failure. If tissue oxygenation is not restored soon
enough, end-organ failure and endothelial cell dysfunction occur,
signaling the onset of late shock. With end-organ failure and
endothelial cell dysfunction, the bodys normal control of regional
blood flow is impaired, and further oxygen delivery to tissue is
compromised. At this point, increasing the blood flow further
with use of overaggressive volume loading and inotropic agents
does not improve tissue oxygenation because blood is not being
delivered to the tissue that needs it. During early shock, aggres-
sive fluid resuscitation is appropriate to minimize end-organ
hypoxemia as quickly as possible. After the patient has progressed
to late shock, it is still important to maintain normal intravascular
volumes, but overaggressive therapy with fluids and pressors has
been shown to be detrimental.
Crystalloids
Crystalloids are fluids that contain water and electrolytes. They
are grouped as isotonic, hypertonic, and hypotonic salt solutions.
Crystalloid solutions are used to provide maintenance water and
electrolytes and to expand intravascular fluid. The replacement
requirement is threefold or fourfold the volume of blood lost
because administered crystalloid is distributed in a ratio 1:4
similar to ECF, which is composed of about 3L intravascularly
(plasma) and about 12L extravascularly (i.e., about 20% should
remain in the intravascular space).
Balanced Salt Solutions
Balanced salt solutions have an electrolyte composition similar to
ECF (e.g., lactated Ringer solution, Plasma-Lyte, Normosol).
With respect to sodium, they are hypotonic. A buffer is included
in place of bicarbonate, which hydrates to carbonic acid with the
production of carbon dioxide, which diffuses from the solution.
Intravascular Fluid and Electrolyte Physiology 1725 54
Compared with 0.9% NaCl, these solutions provide small quanti-
ties of other electrolytes, which are inadequate to meet daily
maintenance requirements.
Normal Saline
Normal saline (0.9% NaCl) is isotonic and iso-osmotic, but con-
Section IV Anesthesia Management
tains more chloride than ECF. When used in large volumes, mild
hyperchloremia (i.e., nonanion gap metabolic acidosis) results.
Normal saline contains no buffer or other electrolytes. It is pre-
ferred to lactated Ringers solution (which contains a hypotonic
concentration of sodium) when brain injury, hypochloremic
metabolic alkalosis, or hyponatremia occurs. Many patients with
hyperkalemia, including patients with renal failure who are fre-
quently in the operating room for vascular access bypass proce-
dures, routinely receive normal saline because it contains no
potassium.
Hypertonic Salt Solutions
Hypertonic salt solutions are less commonly used, and their
sodium concentrations range from 250 to 1200mEq/L. The
greater the sodium concentration, the less the total volume is
required for satisfactory resuscitation. This difference reflects the
movement owing to osmotic forces of water from the intracellular
space into the extracellular space. The reduced volume of water
injected may reduce edema formation; this could be crucial in
patients predisposed to tissue edema (e.g., prolonged bowel
surgery, burns, brain injuries). Clinical studies have confirmed
that a moderately hypertonic solution (250mEq/L of sodium)
can produce lower muscle interstitial pressure than lactated
Ringers solution. Bowel function returned earlier, although the
pulmonary shunt fraction was no different.94 Experimental studies
have shown decreased intracranial pressures in animals receiving
hypertonic solutions. The intravascular half-life of hypertonic
solutions is no longer, however, than isotonic solutions of an
equivalent sodium load. In most studies, sustained plasma volume
expansion was achieved only when colloid was present in the
resuscitation solution. The osmolality of these solutions can cause
hemolysis at the point of injection.95
Hypertonic saline has been used for almost a century in
clinical practice for a variety of reasons. Interest in its use
in hemorrhagic shock was rekindled in 1980 in an experiment
in dogs with severe hemorrhagic shock. These dogs were resusci-
tated solely with 7.5% NaCl or equal volumes of normal saline.
All dogs in the hypertonic saline group survived; all dogs in the
normal saline group died.96 Many randomized trials have been
conducted in the past 2 decades. Hypertonic saline increases
mean arterial pressure, decreases systemic vascular resistance,
decreases pulmonary vascular resistance, and reduces subsequent
blood requirements.97-99 Studies have shown no increase in
complications, and there has been a trend toward decreased
mortality.
There continues to be controversy regarding the use of
hypertonic saline in hemorrhagic shock. During mass casualties
as seen in natural or man-made disasters, hypertonic saline can
be used during triage. It is particularly useful in this setting
because of its ease of storage, low cost, and ability to expand
IV 1726 Anesthesia Management
plasma volume rapidly. Results can be achieved with 250mL of
7.5% saline that are comparable to resuscitation with 2 to 3L of
0.9% saline.87 Larger, randomized controlled studies are needed
to evaluate thoroughly its use in standard practice. In this regard,
to date, improved rates of survival with hypertonic saline were
reported in seven of eight clinical trials, although statistically
significant improvement in overall survival was seen in only one
trial.87
Five Percent Dextrose
Five percent dextrose functions as free water because the dextrose
is metabolized. It is iso-osmotic and does not cause the hemolysis
that would occur if pure water were injected intravenously.
Although it may be used to correct hypernatremia, 5% dextrose
is most often used in the prevention of hypoglycemia in diabetic
patients who have had insulin administered.
Crystalloids Versus Colloids
Much controversy exists about the role of crystalloids and colloids
in fluid therapy (see Chapters 55 and 88). Proponents of colloid
fluid point out that resuscitation with crystalloid solution dilutes
the plasma proteins, with a subsequent reduction of plasma
oncotic pressure resulting in fluid filtration from the intravascular
to the interstitial compartment and the development of interstitial
pulmonary edema. Proponents of crystalloid solutions have
argued that albumin molecules normally enter the pulmonary
interstitial compartment freely and then are cleared through the
lymphatic system returning to the systemic circulation. Addi-
tional albumin should merely increase the albumin pool cleared
by the lymphatics. A review of the literature by Moss and Gould100
confirmed that all unflawed clinical and experimental studies
showed that isotonic solutions are effective plasma volume
expanders for resuscitation without the addition of a variety of
colloid fluids. The additional cost and potential risks of colloids
compared with crystalloids is another argument against colloid
administration.
Colloids used in the United States include albumin, hydrox-
yethyl starch (hetastarch), and dextran. Because the molecules are
large, colloids usually do not cross capillary membranes and
remain intravascular. Distribution of fluid throughout the body
is represented by the Starling-Landis equation:
J v = K h A ([ PM PT ] [COPM COPT ]) for volume resuscitation may play a role in the coagulopathy often
In this equation, Jv represents the net volume of fluid
moving across the capillary wall per unit of time, expressed as
m3/min; Kh is the hydraulic conductivity for water, which is the
fluid permeability of the capillary wall, expressed as m3/min per
square micrometer of capillary surface area per 1mmHg pres-
sure difference. The value of Kh increases up to fourfold from the
arterial to the venous end of a typical capillary. PM is the capillary
hydrostatic pressure; PT is the tissue hydrostatic pressure; A is the
capillary surface area; and is the reflection coefficient for plasma
proteins. This coefficient is necessary because the plasma proteins
are slightly permeable to the microvascular wall, preventing full
expression of the two COPs. When is 0, molecules cross the
membrane freely; when is 1, molecules cannot cross the mem-
brane. Typical values for plasma proteins in the microvascula-
ture exceed 0.9 in most organs, and these values remain constant,
but can be decreased significantly by pathophysiologic processes,
including hypoxia, inflammation, and tissue injury. COPM is the
colloid oncotic pressure, and COPT is the colloid oncotic pressure
in the tissue.
The hydrostatic and colloid pressure differences across
capillary walls (i.e., Starling forces) cause movement of water
and dissolved solutes into the interstitial spaces. These move-
ments play a minor role with regard to tissue nutrition relative to
simple diffusion. The reflection coefficient that expresses the
ability of the semipermeable membrane to prevent movement of
a solute varies greatly among tissues. The lungs are moderately
permeable relative to other organs, and during pathophysiologic
processes such as surgical trauma, the reflection coefficient
may change further to alter capillary permeability, resulting in
increased capillary permeability or leak. In this setting, colloids
move more easily into the interstitium and increase interstitial
edema.
With leakage of colloid molecules into the interstitial space,
further swelling of tissues occurs because of the unfavorable
oncotic pressure gradient, and these molecules are removed by
the lymphatic system. Removal of colloids requires longer periods
than for crystalloids and can be a significant problem in burn
patients and patients undergoing major surgery. A well-known
meta-analysis by Velanovich101 of mortality from eight studies
published in 1989 concluded that trauma patients should be
resuscitated with crystalloid solutions, whereas colloids were
more effective in nonseptic, nontraumatic, elective surgical
patients. Current data seem to support the concept that blood
composition should be maintained as close to normal as possible
in terms of directed transfusion of red blood cells and coagulation
factors.86
Another factor that must be considered is the overall effect
of resuscitation fluid on the coagulation cascade. It is a well-
known phenomenon that after trauma, patients may have low
levels of circulating coagulation factors. This deficiency may lead
to transfusion of fresh frozen plasma to restore deficient factors.
The clotting cascade is an extremely complex string of events,
with many factors affecting the progression to clot. Studies using
a thromboelastogram to measure trauma patients clotting status
have indicated that these patients may be hypercoagulable.102
Another study using thromboelastographic profiles showed that
patients undergoing major surgery who were given intraoperative
lactated Ringers solution also were hypercoagulable compared
with baseline.103 These studies suggest that the type of fluid used
seen in trauma patients (see Chapter 56). Definitive studies are
still needed, however, to elucidate the true impact that resuscita-
tion fluids play in the coagulation cascade.
Colloid Solutions and Blood
Substitutes
Colloid solutions generally are administered in a volume equiva-
lent to the volume of blood lost (see Chapter 55). The initial
volume of distribution is equivalent to the plasma volume. The

half-life in circulation of albumin is normally 16 hours, but it can
be only 2 to 3 hours in pathophysiologic conditions.104 The syn-
thetic colloids, processed albumin, and protein fractions have
minimal or no risks of infection. Blood substitutes are useful to
restore intravascular fluid volume temporarily until definitive
treatment can be established. They are inexpensive, have a long
storage life, and lack the risk of transmitting viral diseases.
Five Percent Albumin
Five percent albumin or plasma protein fractions (e.g., Plas-
manate) have a COP of about 20mmHg (i.e., near-normal COP).
Preparation methods eliminate infectious agents. These solutions
are chosen when crystalloids fail to sustain plasma volume for
more than a few minutes because of low COP. These solutions are
most appropriate when there is an abnormal loss of protein from
the vascular space, such as in cases of peritonitis or extensive
burns.
Twenty-Five Percent Albumin
A colloid solution of 25% (salt-poor) albumin contains purified
albumin at five times the normal concentration. When adminis-
tered, it has the potential to expand the plasma volume by up to
five times the volume provided. It is selected when the current
plasma volume is diminished, but blood pressure is acceptable,
and the total ECF volume is expanded.
Six Percent Dextran 70
Dextran is available as dextran 40 or dextran 70. The numbers 40
and 70 refer to the average molecular mass of the molecules in
the solution. The mean molecular weight of dextran 40 is about
40,000 daltons (40kD), and the mean molecular mass of dextran
70 is about 70,000 daltons (70kD). The dextran solutions are
water-soluble glucose polymers synthesized by certain bacteria
from sucrose. Both dextran solutions are ultimately degraded
enzymatically to glucose. A colloid solution of 6% dextran 70 is
administered for the same indications as 5% albumin. Dextran 40
is used in vascular surgery to prevent thrombosis and is rarely
used as a volume expander.
Side effects include anaphylactic or anaphylactoid reactions
in about 1 in every 3300 administrations, increased bleeding time
caused by decreased platelet adhesiveness (at doses of 20mL/
kg/24hr), rouleaux formation (i.e., interfering with crossmatch-
ing of blood), and rare cases of noncardiogenic pulmonary edema
thought to be a direct toxic effect on pulmonary capillaries after
intravascular absorption.105
Hydroxyethyl Starch
Hydroxyethyl starch (hetastarch) is a synthetic colloid solution in
which the molecular mass of at least 80% of the polymers ranges
from 10,000 to 2 million daltons. It is available in the United States
as a 6% solution in 0.9% NaCl (Hespan). The pH of hetastarch is
about 5.5, and the osmolarity is about 310mOsm/L. The larger
molecules are degraded enzymatically by amylase. It is stored in
Intravascular Fluid and Electrolyte Physiology 1727 54
the reticuloendothelial system for several hours and is believed
to be renally excreted. It produces dilutional effects similar to
other volume expanders and reduces factor VIII:C and vWF
levels by 50% to 80% in a dose of 1 to 1.5L, with prolongation of
the partial thromboplastin time.106 Hetastarch also can interfere
with platelet adhesion and clot formation by reduction in
glycoprotein IIb/IIIa availability and direct movement of the
hetastarch molecules into the fibrin clot.106 In clinically recom-
Section IV Anesthesia Management
mended volumes (i.e., 20mL/kg), there is minimal interference
with subsequent crossmatching. Repeated doses can result in
accumulation and side effects, which include allergic reactions
and bleeding as described with higher doses (20 to 25mL/kg).
The combination of dextran with Hespan seems to blunt Hespan-
induced neutrophil activation.87
A new form of hydroxyethyl cellulose has been approved
for use in the United States. Hextend (Biotime, Berkeley, CA)
contains 6% hetastarch in a solution that approaches physiologic
concentrations of the major electrolytes. It also contains physio-
logic levels of glucose and lactate as a buffer. In contrast to Hespan,
Hextend polymer units have an average molecular mass of 670kd,
with 80% within the range of 20 to 2500kd, making these units
significantly smaller than Hespan. Early studies have suggested
that Hextend might not affect the coagulation process to the same
degree as Hespan. In a study of 90 patients undergoing major
surgery, the patients who were given Hextend had no significant
change in their thromboelastogram profiles. Patients receiving
lactated Ringer solution were found to be hypercoagulable, and
patients receiving Hespan were found to be hypocoagulable com-
pared with baseline.103 In clinical practice, there is movement
away from Hespan toward lower and medium molecular weight
hetastarchs, which have less inhibitory effect on platelet function
and coagulation factors, but a shorter intravascular half-life.
Pentastarch
Pentastarch is a lower molecular weight hetastarch with fewer
hydroxyethyl groups per molecule. Pentastarch is undergoing
clinical trials in the United States and has anticoagulant effects
similar to those of hetastarch.
Perfluorochemical Emulsions and
Hemoglobin-Based Oxygen Carriers
Perfluorochemical emulsions have linear oxygen-carrying capac-
ity, but are insufficient to sustain human cellular function, and the
first-generation emulsion, Fluosol-DA 20, was discontinued and
is no longer available. The second-generation emulsion, Oxygent,
discontinued clinical trials. Third-generation emulsions are cur-
rently in preclinical stages.
There are numerous classes of hemoglobin-based oxygen
carrier solutions; these include cross-linked, polymerized, and
conjugated hemoglobin solutions. Hemopure and PolyHeme have
completed Food and Drug Administration phase II trials, and
phase III trials are under way. Several newer generation hemo-
globin-based oxygen carriers are in early preclinical evaluation
(see Chapter 55).
IV 1728 Anesthesia Management

Table 54-22 Volume and Composition of Gastrointestinal Fluids

Fluid Management of Specific 24-Hour
Clinical Conditions Fluid Volume Na+ K+ Cl HCO3
Source (mL) (mEq/L) (mEq/L) (mEq/L) (mEq/L)

The following guidelines are intended to facilitate initiating Saliva 500-2000 2-10 20-30 8-18 30
therapy, but the choice of fluid and rate of administration must Stomach 1000-2000 60-100 10-20 100-130 0
be adjusted to achieve physiologic goals. These guidelines are only
a starting point for patients without other major comorbid dis- Pancreas 300-800 135-145 5-10 70-90 95-120
eases of vital organs. Careful observation of the patients response Bile 300-600 135-145 5-10 90-130 30-40
forms the basis for ongoing modification in a continuous feed-
Jejunum 2000-4000 120-140 5-10 90-140 30-40
back loop.
Ileum 1000-2000 80-150 2-8 45-140 30

Routine Maintenance Fluids
Routine maintenance fluids are described for a 70-kg postopera-
tive patient. The patient requires 110mL H2O and 110kcal/hr, or
2640mL and 2640kcal/day. This example is based on the 4-2-1
rule (Table 54-21), which provides a close approximation of water
requirements. The sodium requirement (1.5mEq/kg/day) is dis-
solved in the daily fluid requirement of 2.64L; the 100mEq/kg/
day requirement for potassium is placed in the 2.64L/day water
requirement: 100mEq K/2.64L = 42mEq/L. The potassium con-
centration needs to be limited, however, if the fluid is to be infused
into a peripheral vein because of the chemical irritation induced
by high concentrations of potassium.
The long-standing concept first presented in 1957 by Holli-
day and Seger of the 4-2-1 rule for crystalloid delivery has been
challenged more recently. The focus of this challenge is on reduc-
tion of free water administration because of increased postopera-
tive hyponatremia related to increased secretion of ADH. An
alternative concept is for much lower glucose administration and
isotonic sodiumcontaining solutions to avoid the rare problem
of hyponatremic seizure and brain damage (seen mainly in
children).107
The obligatory glucose needs of the brain and red blood
cells are roughly 2 mg/kg/min. Because dextrose contains
3.41kcal/g instead of 4kcal/g of glucose, about 17% more dex-
trose than glucose is required. If carbohydrate is not provided,
glycogenolysis and gluconeogenesis from amino acid pools
provide the necessary glucose, but accelerate protein
catabolism.
Carbohydrate is said to prevent catabolism (i.e., protein
sparing), but the benefit of this dose of dextrose is unclear. Total
starvation may be preferable because insulin concentrations
decrease to very low levels, facilitating lipolysis as a caloric source.
Table 54-21 Calculations of Fluid Requirements by 4-2-1 Rule*
Body Weight Fluid Rate Weight Category
(kg) (mL/kg) (kg)
0-10 4 10
11-20 2 10
>21 1 5
Total 25
*Assumes a patient weighing 25kg, resulting in an estimated fluid requirement of
65mL/hr.
Colon 60 30 40
The osmolarity of 7.5% dextrose is 417mOsm/L, to which is
added 156mOsm/kg H2O, resulting in a highly hyperosmolar
solution. A compromise between the need for glucose and hyper-
osmolality has been 5% dextrose.
If there are other losses (e.g., gastric drainage), additional
sodium and water are required. Gastric drainage of 0.5L/day
loses 30 to 50mEq of sodium and 50 to 65mEq of chloride (Table
54-22). When these are added to the maintenance fluid, the con-
centration approximates 0.45% NaCl. This solution is commonly
used as a maintenance intravenous fluid postoperatively in
patients with nasogastric drainage.
Routine Intraoperative Fluid Administration
The goals of intraoperative fluid administration are to maintain
adequate oxygen delivery, normal electrolyte concentrations, and
normoglycemia. The total fluid requirement is composed of com-
pensatory intravascular volume expansion (CVE), deficit replace-
ment, maintenance fluids, restoration of losses, and substitution
for fluid redistribution (i.e., third space fluids):
Rate of fluid = CVE + deficit + maintenance administration +
loss + third space
Compensatory Intravascular
Volume Expansion
Intravascular volume usually must be supplemented to compen-
sate for the venodilation and cardiac depression caused by
anesthesia. Sustaining adequate oxygen delivery in relation to
oxygen consumption is an important goal of fluid therapy. Tissue
Fluid oxygen delivery depends on hemoglobin concentration, oxygen
(mL/hr) tension, organ perfusion pressures, and organ vascular resistance.
40
Organ perfusion pressures depend on systemic arterial pressure
and the higher of organ venous pressure or tissue pressure. Arte-
20 rial pressure depends on cardiac output and systemic vascular
5 resistance. Cardiac output is related to stroke volume and heart
rate, and stroke volume depends on preload, contractility, and
65 afterload. Most general and regional anesthetics cause arteriolar
and venous dilation, expanding the vascular capacity. The latter
reduces peripheral venous pressure and venous return and cardiac

output. Fluid must be administered to expand the blood volume
to compensate for venodilation.
General anesthetics produce myocardial depression (see
Chapter 23). Increasing cardiac preload by infusing fluid intra-
vascularly to take advantage of the Starling mechanism often
returns stroke volume to an acceptable range. Postoperatively,
venodilation and myocardial depression rapidly subside when
administration of the anesthetic is stopped. Patients with impaired
cardiac or renal responses may then become acutely hypervo-
lemic. Compensatory intravascular volume expansion with 5 to
7mL/kg of balanced salt solution must occur before or simulta-
neous with the onset of anesthesia.
Deficits
Characteristics of Fluid Deficits
The fluid deficit equals the maintenance fluid requirement mul-
tiplied by the hours since last intake plus unreplaced preoperative
external and third space losses. When hypovolemia is present,
sufficient fluid should be infused to restore mean arterial pressure,
heart rate, and filling pressures to near-normal values before
induction. If sufficient time is available, restoration of normal
urine flow rate also is desirable. The fluid infusion rate for normal
patients should be set to deliver three to four times the mainte-
nance rate until the calculated deficit has been corrected.
Electrolyte abnormalities are common among hospitalized
patients. Surgical urgency often forces us to evaluate and correct
preexisting abnormalities that may be directly related to the surgi-
cal problem or to comorbid diseases or their therapy. The princi-
ples of management are outlined in other sections of this chapter.
The induction of anesthesia and the onset of mechanical ventila-
tion, fluid shifts, and stress responses induced by surgical trauma
all lead to redistribution of water, protein, and electrolytes. The
most common and most studied of these are abnormalities of
potassium, calcium, and magnesium.
Maintenance Fluid
The maintenance fluid meets ongoing basal needs for water and
electrolytes as described earlier. The onset of surgical stimulation
and, to a smaller extent, onset of anesthesia elicit changes in cat-
echolamines, cortisol, and growth hormone. These tend to reduce
insulin secretion or impede its glucose-lowering effect, leading to
hyperglycemia. If dextrose-containing solutions are infused at the
usual 5% concentration at the rates often required during surgery,
severe hyperglycemia results. The fluid used for volume mainte-
nance should not contain dextrose.
Losses
External losses (e.g., blood, ascites) should be replaced to main-
tain normal blood volume and normal composition of the ECF
volume. Blood loss is replaced initially with 3mL of balanced salt
solution or 0.9% NaCl for each 1mL of blood loss. For each 1mL
Intravascular Fluid and Electrolyte Physiology 1729 54
of blood lost, 1mL of colloid solution should be administered to
provide improvement of filling pressures, arterial blood pressure,
and heart rate. PRBC infusions are used at roughly 1mL for each
2mL of blood lost plus crystalloid or colloid, as previously
described. Although the effect on volume is 1:1, the hematocrit
of PRBCs (60% to 70%) is about twice the necessary hematocrit
of the patient. In patients with reasonable cardiac reserve and
without compromised regional circulations (e.g., coronary, cere-
Section IV Anesthesia Management
bral, renal, intestinal), hemoglobin levels of 7.5g/dL or greater
are usually well tolerated.108 If blood volume is normal, and
cardiac failure is not a problem, signs of sympathetic activation,
signs of mixed venous oxygen desaturation, or electrocardiogram
signs of myocardial ischemia suggest the need for red blood cell
administration. The following equation is used to calculate the
necessary volume of red blood cells to be infused, and weight is
measured in kilograms; because PRBCs have a hematocrit of 60%,
the volume to be infused can be estimated:
PRBC infused = ( Hct desired EBV weight )
(Hct observed EBV weight ) 0.60
where EBV is estimated blood volume per kg: 55mL/kg for
women and 70mL/kg for men. For a man with a hematocrit of
21% for whom the target hematocrit is 30%, the calculation would
be 0.30 (70 70) 0.21 (70 70) = PRBC infused, estimated
at 735mL, or about 3U of PRBCs.
Ascites and pleural effusions drained during surgery re-
form at extremely variable rates. The electrolyte composition is
the same as that at the ECF volume, but it also contains protein
at concentrations 30% to 100% of the plasma value. Balanced salt
solutions are most appropriate for replacement, but colloid should
be added when dilution of the patients COP becomes severe (<15
to 17mmHg), and the apparent volume of redistribution of crys-
talloid begins to increase.
The electrolyte composition of gastrointestinal tract losses
depends on the site (see Table 54-22). Most gastrointestinal losses
removed at the time of surgery entered the bowel lumen preop-
eratively and should be considered to be part of the deficit. Evapo-
ration from exposed viscera is entirely water, but the electrolyte
is left behind, leading to a need for free water. The amount evapo-
rated is directly proportional to temperature and exposed surface
area, and is inversely proportional to relative humidity. Excess
urine caused by diuretics, glycosuria, or diabetes insipidus should
be replaced with a solution based on urinary electrolyte measure-
ments. The sodium concentration generally ranges from 50 to
100mEq/L, and the potassium concentration ranges from 20 to
60mEq/L.
Redistribution
Redistribution (so-called third space losses) results primarily
from tissue edema and transcellular fluid displacement. Func-
tionally, this fluid is not available to the vascular space.109,110 The
volume of edema formed is governed by the principles discussed
earlier. Colloid enters the injured tissue at a more rapid rate than
normal, but at a slower rate than electrolyte. Bowel wall edema
may be lessened by using colloid-containing fluids compared
with crystalloid.111 The composition of third space losses is equiv-
alent to the ECF volume electrolyte concentration plus a smaller
amount of protein. Balanced salt solution is the most appropriate
IV 1730 Anesthesia Management
Table 54-23 Fluid Calculations for a Simulated Case
Compensatory Deficit Maintenance
Time (mL) (mL) (mL)
P-I 350 220 110
I-S 220 110
First hour 220 110
Second hour 220 110
Third hour 220 110
Fourth hour 0 110
*Reflects fluid replacement for blood loss. Preinduction phase lasts 15 to 20 minutes.
Total fluid administered during the hour.
Grand total since beginning of the case.
Preinduction phase lasts 15 to 20 minutes.
Induction until intra-abdominal surgical entry (assumed to be 1 hour).
Operative time.
replacement fluid. The volume redistributed correlates roughly
with the degree of tissue manipulation. Intra-abdominal proce-
dures with small incisions (e.g., hysterectomy) may require an
additional 2mL/kg/hr, whereas a major bowel resection requires
an additional 4 to 6mL/kg/hr.
Table 54-23 describes the fluid management, starting with
a hemoglobin level of 15g/dL, for a 70-kg patient undergoing
gastrectomy who has been fasting for 10 hours. The maintenance
rate is 110mL/hr, producing a deficit of 1100mL.
During the first and second hours of intra-abdominal
activity, 100mL of blood was lost and replaced at a 3:1 rate with
a balanced salt solution. By the fourth hour, the deficit had been
replaced, and because the abdomen was being closed during part
of that hour, the estimate of third space losses was likewise
reduced, and no further blood loss occurred. The assumption is
made that urine flow was 50 to 80mL/hr, and that heart rate and
arterial blood pressure were in an acceptable range, and the CVP
remained 6 to 9mmHg. Had the CVP or urine output begun to
increase, the rate of fluid administration would have been slowed.
If oliguria and tachycardia had occurred, a fluid bolus would have
been administered.
Pediatric Patient
There are few special considerations relevant to pediatric fluid
management in addition to those for adults (see Chapter 82).
Neonates have limited ability to dilute or concentrate urine and
have a high fluid requirement. Neonates should not be without
fluid for more than 3 to 4 hours; otherwise, significant dehydra-
tion may result. Food should be offered until 6 to 8 hours before
anesthesia induction, and glucose-containing clear liquids should
be given about 4 hours before induction. Clear liquids are defined
as transparent liquids that do not contain particulate material or
protein, which coagulate in the acid medium of the stomach.
Using the same principles as outlined for adults, a neonate requires
maintenance fluids of 0.3% NaCl with potassium. Dextrose
administration should not exceed 5mg/kg/min. This requirement
generally can be met by using 2.5% dextrosecontaining fluids.
During extensive, prolonged procedures, blood glucose should be
monitored, and the rate of dextrose administration should be
Blood Loss Third Space This Hour
(mL)* (mL) (mL) Cumulative
0 0 680 680
0 0 330 1010
300 350 980 1990
300 350 980 2970
150 350 830 3800
0 200 330 4130
modified accordingly. Otherwise, the choice and volumes of fluid
are as described for an adult.
Postoperative Patient with
Bowel Obstruction
Patients with bowel obstruction are often older and have limited
reserves in several organ systems. Estimating the degree of loss
of fluids is difficult because fluid is retained in the bowel lumen,
where it cannot be measured. The slowly developing volume
depletion allows time for full expression of compensatory mecha-
nisms that mask the degree of the deficit. Patients have often
ingested no fluids for many hours before entering the hospital and
are often vomiting. The patient may have ischemic bowel injury
with severe bowel wall edema and continued sequestration of
luminal fluid and formation of ascites. Perioperative fluid intake
is much larger than measured output, leading consultants to
suggest diuresis for fluid overload. Nutrition is impaired preop-
eratively, and protein losses into the bowel are increased, leading
to hypoalbuminemia and exacerbating the loss of fluid from the
vascular space. The clinical picture is one of ongoing fluid require-
ment in the absence of external fluid loss, commonly referred to
as third spacing.
The goals of fluid management in this group are similar to
the goals for other patients and include restoration of the vascular
volume and interstitial volume, correction of electrolyte deple-
tion, correction of acidosis, reduction of systemic vascular resist-
ance into the normal range, and optimization of oxygen delivery
and use. Initial fluid infusion can restore intravascular volume
sufficiently for blood pressure and heart rate to improve. Intra-
vascular volume still may be depleted, however, with a low cardiac
output and severe arterial and venous vasoconstriction. The ECF
volume (except for the bowel) remains dehydrated and continues
to accept fluids from the vascular space. A patient with vasocon-
striction fails to perfuse all tissues, limiting the rate at which the
ECF volume can be replenished. If fluids are infused more rapidly
than the rate at which they can enter the ECF volume, increased
filling pressures may result in pulmonary edema. It is crucial to
estimate filling pressures, cardiac output, and systemic vascular

resistance while aggressive intravascular fluid with balanced
salt solution and colloid is pursued. It may be necessary to admin-
ister arterial vasodilators, such as nitroprusside, to facilitate
correction.
Management of a typical patient includes frequent moni-
toring of arterial blood pressure, heart rate, CVP, pulse pressure,
respiratory variation, urine output, and electrolytes. If these
parameters are stable, the hemoglobin level and COP should be
monitored every 2 to 4 hours until stable. An increasing hemo-
globin level indicates ongoing loss of plasma water, with or
without protein loss. An increasing COP indicates continued loss
of plasma water greater than protein loss. Because maintenance
fluid requirements persist, dextrose 5% is infused in 0.45% NaCl
with 20 to 40mEq/L of potassium chloride at maintenance rate.
The fluid lost to the bowel and ECF volume is similar to
plasma water in electrolyte composition, and a balanced salt solu-
tion is a reasonable first choice for the fluid boluses required to
sustain plasma volume. Because the fluid lost also contains
protein, albumin or COP should be monitored, and colloid should
be infused. If a low COP (approximately 15 to 18mmHg) coex-
ists with hemodynamic instability, replacement should be started
at 3mL/kg/hr if the CVP (and pulmonary artery occlusion pres-
sure [PAOP] if a pulmonary artery catheter has been placed) is
within the patients usual range. If COP is high, replacement
should start at 1 to 2mL/kg/hr. If the urine output increases to
greater than 1.5mL/kg/hr, the physician should check for gluco-
suria. If glucosuria is absent, the fluid infusion rate is reduced by
0.5mL/kg/hr. If the CVP or PAOP value increases to greater than
the patients usual values, and urine output is 0.5 to 1.5mL/kg/hr,
the fluid infusion rate is reduced by 0.5mL/kg/hr. If the values
for CVP or PAOP or both decrease below the patients baseline,
and the urine flow rate is low, balanced salt or colloid solution is
administered rapidly (0.5 to 2mL/kg/min) with close monitoring
of filling pressures.
Patient with Liver Failure
Fluid management in a patient with liver failure (see Chapters 66
and 67) is complicated by several interacting problems. These
patients seem to be simultaneously hypervolemic and hypovo-
lemic. Most infused fluid is retained, but renal function deterio-
rates, along with avid sodium retention and arterial underfilling.
Neither sodium retention nor arterial underfilling explains all the
clinical findings or leads to consistently successful therapy.
According to the arterial underfilling (primary vasodila-
tion) hypothesis,112 some factor produced by or not catabolized
by the failing liver causes inappropriate arterial dilation. The rela-
tive hypotension leads to activation of the systematic nervous
system and the RAS and vasopressin release. Subsequent sodium
and water retention results in ascites and tissue edema.
Cirrhotic patients have a low systemic vascular resistance,
high cardiac output, and relative hypotension. Persistent endotox-
emia (shunting through portosystemic anastomoses and enhanced
endotoxin absorption from the intestine as a result of bile salt
deficiency) may contribute to the vasodilation by activation of a
cascade of secondary mediators, beginning with tumor necrosis
factor and interleukins. Other vasodilator neurotransmitters may
be produced or may not be cleared by the damaged liver.
The primary sodium retention hypothesis explains the avid
sodium retention on the basis of hormonal (aldosterone) hyper-
Intravascular Fluid and Electrolyte Physiology 1731 54
activity because of failure of the liver to metabolize aldosterone.
Given the ECF volume expansion, the distribution into ascites
and tissue edema is explained by the abnormally high portal
venous pressures and hypoalbuminemia. Abnormalities of ANP
have been investigated. ANP levels are normal or low in cirrhotic
patients; this contrasts with the increase usually found in volume-
expanded patients.113 ANP increased with water immersion or
fluid administration, but natriuresis was not closely correlated
Section IV Anesthesia Management
with ANP levels. Water immersion of the lower body compresses
the venous capacitance system, resulting in centralization of
blood volume. This centralization of blood volume simulates a
volume infusion without adding any sodium or water to the body
and should cause ANP release. Water immersion did not increase
ANP levels in patients with ascites, who have blunted responses.
Impaired ANP release failed to explain sodium retention.113
Patients with tense ascites had increased ANP, renin, and aldos-
terone concentrations. After paracentesis, ANP increased, but
renin and aldosterone levels decreased.114 The reasons for these
findings were not clarified, but a reduction in intra-abdominal
pressure could have decreased inferior caval pressure, facilitated
venous return, and increased ANP levels. Decreased intra-abdom-
inal pressures also would improve renal perfusion pressure,
causing a reduction in renin release and subsequent aldosterone
generation.
The role of increased intra-abdominal pressure may be
important in sustaining sodium retention after ascites develops.
Increased intra-abdominal pressure increases caval pressure; this
decreases renal blood flow and glomerular filtration rate because
the renal perfusion pressure gradient (i.e., mean arterial pressure
minus renal venous pressure) is decreased by systemic hypoten-
sion and increased caval pressure. The hypotension and reduction
in renal blood flow can lead to renin activation, with aldosterone
production. Hypotension, increased aldosterone, and a decreased
glomerular filtration rate lead to enhanced sodium reabsorption
and low fractional excretion of sodium.
Hypoalbuminemia results from impaired synthesis by the
liver, transudation into ascitic fluid owing to portal hypertension,
and malnutrition. Low COP favors loss of fluid from the vascular
space into the interstitial space, producing intravascular hypovo-
lemia. Ascites results from high portal venous pressure and
hypoalbuminemia, markedly increasing the volume of transcel-
lular fluid, which is functionally excluded from rapid exchange in
the ECF volume.
Splanchnic blood pooling resulting from increased portal
venous resistance plus lower body pooling resulting from elevated
caval pressures from ascites tend to decrease net systemic venous
return. Patients are functionally hypovolemic despite normal or
elevated total blood volume. Heart failure from alcoholic cardio-
myopathy may complicate the clinical picture further.
The goals in these patients are to avoid increasing intersti-
tial fluid overload, maintain normal potassium concentration,
and maintain intravascular volume. If cardiac failure is present,
treatment must include administration of inotropic drugs and
diuretics when filling pressures are increased. Intravascular COP
should be restored by infusion of 25% albumin when possible. If
the patient is acutely hypovolemic, 5% albumin solutions should
be preferred to crystalloid, which tend to expand further the
already overexpanded ECF volume (i.e., produce more edema and
ascites). Intra-abdominal pressure should be estimated from
urinary bladder pressure, and paracentesis should be performed
whenever it increases to greater than 20 to 25mmHg. Trials of
IV 1732 Anesthesia Management
dopamine, norepinephrine, phenylephrine, or vasopressin may be
performed in hypotensive patients with low vascular resistance
and hypotension to increase renal perfusion pressure and renal
blood flow.
Patient with Heart Failure
Fluid management of heart failure is directed to optimize cardiac
preload, avoid overadministration of sodium, diminish edema,
and correct common electrolyte abnormalities (see Chapter 60).
Maintaining ideal cardiac preload during rapid fluid shifts that
occur perioperatively is facilitated with direct or indirect meas-
ures of preload and cardiac contractile function. Measures of
preload include CVP, thermodilution cardiac output, end-diasto-
lic volume, echocardiography, PAOP, and left atrial pressure.
Measures of cardiac contractile function include stroke volume,
ejection fraction, and stroke work.
Patients with a history of cardiac failure scheduled for
major or prolonged surgery should have monitoring instituted
preoperatively, and an intravascular fluid challenge (i.e., 500 to
1000mL/70kg of crystalloid) should be performed to identify the
optimal preload level. Tissue edema is avoided by frequent moni-
toring of preload and arterial blood pressure coupled with support
of contractility and control of vascular resistance. These patients
have impaired ability to excrete fluids during the fluid mobiliza-
tion, which occurs postoperatively. Because the ECF volume is
usually already expanded in these patients, the initial rates of fluid
infusion intraoperatively should be at the lower ranges of esti-
mates. Similarly, maintenance of intravascular volume without
expansion of the interstitial space favors the use of colloid during
the immediate perioperative period.
Postoperatively, hemodynamic monitoring is continued
until fluid mobilization is complete. The goal postoperatively is to
give as little crystalloid as required to maintain adequate overall
cardiovascular performance. Perioperative patients commonly
receive more than 200mEq of sodium per day, including main-
tenance fluids, saline used to measure cardiac output, sodium
from antibiotics administration, and sodium infused with vasoac-
tive medications and inotropic agents. Fluid should be main-
tained at a low maintenance level, and flush fluids and sodium
dosage should be measured. As soon as urine output begins to
increase, or filling pressures or diastolic volume begins to increase,
maintenance fluids should be stopped completely. If preload
becomes excessive, diuretics should be administered.
Patients with heart failure have primary electrolyte prob-
lems because of compensatory physiologic mechanisms activated
by the impaired cardiac performance. These are complicated by
therapy with diuretics, digitalis, vasodilators, and angiotensin-
converting enzyme inhibitors. Hyponatremia is common because
of excess activation of the vasopressin system despite sodium
retention. Treatment is directed at excreting the excess water load
with diuretics, which should increase free water excretion more
than sodium excretion. Sodium administration is not indicated,
unless volume depletion is documented. Aldosterone activation
and diuretics lead to loss of potassium and magnesium. These
ions are crucial for maintaining the stability of cardiac electro-
physiology and the effectiveness of digitalis and catecholamine.
Ionized calcium is crucial for cardiac contractility, and hypocal-
cemia is extremely common during the perioperative period.
Ionized calcium must be measured and corrected routinely. Severe
hypophosphatemia often coexists with abnormalities of calcium,
potassium, and magnesium, and leads to depressed contractility.
Patient with Cerebral Edema
Fluid management in patients with cerebral edema (see Chapters
13 and 63) is directed at maintaining cerebral perfusion pressure,
avoiding elevations of cerebral venous pressure and hypertension,
preventing large changes in plasma osmolality (particularly
depression of plasma osmolality), and avoiding hyperglycemia.
Cerebral perfusion pressure should be maintained in the high-
normal range (see Chapter 63) because autoregulation is impaired
in the damaged brain, with a baseline increase in cerebrovascular
resistance. Avoidance of intracranial hypertension, the other
determinant of cerebral perfusion pressure, is crucial to preserve
blood flow to compromised but viable brain tissue. Cerebral
edema formation is related to capillary pressure, COP, and per-
meability. The capillary pressure is between the arterial and
venous pressures. Systemic arterial or venous hypertension should
be prevented. The normal brain capillary bed is essentially imper-
meable to sodium, mannitol, and protein, although water crosses
freely. The damaged capillary bed becomes excessively permeable,
with conductivity being greatest for the smallest molecules, but
less for colloids.
Free water restriction without hypovolemia is desired to
maintain the plasma sodium level between 142mEq/L and
148mEq/L. It is common to provide 75% to 90% of maintenance
fluids with 0.9% NaCl or lactated Ringers solution and to mini-
mize water volume required for other medications. Isotonic crys-
talloids or colloids do not cause edema in normal brain and can
be used to sustain intravascular volume.115 Hyponatremia is often
caused by hypovolemia with inappropriate sodium loss116 and
subsequent water retention. Hyponatremia should be treated with
intravascular volume expansion with isotonic or hypertonic
sodium chloride. Parenteral nutrition can be prepared with 15%
amino acid solutions, 20% lipid, and 70% dextrose to give full
protein and caloric support in a minimal volume. Similarly, tube-
feeding formulations with 2cal/mL should be prescribed. Hypo-
volemia must be carefully avoided. Colloid infusion to sustain
intravascular volume, guided by hemodynamic monitoring, tends
to provide better long-term control of the intravascular volume
than crystalloid. If diuretics or diabetes insipidus causes loss of
greater than 2mL/kg/hr for more than 2 to 4 hours, a pulmonary
artery catheter may be indicated for closer monitoring. Blood
glucose should be maintained at 80 to 175mg/dL by close moni-
toring, initially hourly, with decreasing frequency as stability is
exhibited. During acute resuscitation, dextrose administration is
limited to no more than 2mg/kg/min. Indications for steroid use
have become fewer in recent years, lessening the problems of
hyperglycemia.
Patient with Anuric Renal Failure
Undergoing Nontransplantation Surgery
Fluid management in a patient with anuric renal failure for non-
transplantation surgery should avoid excessive intravascular fluid

administration and ECF volume expansion, and should maintain
or correct electrolyte and acid-base status (see Chapter 96). The
anesthesiologist also should attempt to prevent precipitating con-
ditions that would require dialysis during the immediate postop-
erative period (e.g., hyperkalemia, pulmonary edema, metabolic
acidosis). Dialysis is difficult or impossible in a hemodynamically
unstable patient, and anticoagulation may be strongly contrain-
dicated for several hours postoperatively. Patients with chronic
renal failure often present with hypertension, diabetes, and vas-
cular disease. Avoidance of hypotension may be indicated to
maintain coronary or cerebral perfusion pressure. In patients
with acute renal failure, hypotension may worsen ischemic renal
injury. In contrast, there is a lack of concern for an adverse effect
on renal function in a patient with chronic renal failure. Recent
dialysis often induces acute electrolyte shifts, hypovolemia, and
acidosis. The best compromise seems to be dialysis 12 to 24 hours
preoperatively. If dialysis is performed close to the time of surgery,
volume removal should be minimal. Patients with renal failure
are often malnourished, have hypoproteinemia, are anemic, and
have poor glucose tolerance.
Hemodynamics should be closely monitored. If hypovo-
lemia develops, colloid should be used early, but the physician
should avoid producing vascular overexpansion. Likewise, inter-
stitial fluid overload should be avoided because it would require
acute postoperative dialysis. Crystalloid replacement of third
space losses should be limited to 1 to 2mL/kg/hr, whereas blood
loss should be replaced with colloid or RBCs. Correction of
sodium, potassium, and acidosis can be achieved by the use of
isotonic fluid without potassium (to correct sodium), with
reduced amounts of chloride (to correct potassium), and with
increased amount of buffer (to correct acidosis). Initially, an infu-
sion of 30% of calculated maintenance fluid rate is important
because approximately 70% of normal fluid requirements are
used in excreting solute through the kidney, a route no longer
available. The pH and levels of sodium, potassium, bicarbonate,
and glucose should be monitored at regular intervals.
Patient with Adult Respiratory
Distress Syndrome
Fluid management in a patient with adult respiratory distress
syndrome (ARDS) (see Chapters 91 and 92) involves maintaining
COP and requires a CVP and PAOP value as low as possible that
is consistent with good ventricular function. Oxygen delivery
should be sustained, ideally greater than 600mL/min/m2, by
increasing cardiac index and hemoglobin concentration. Com-
pensation for hemodynamic effects of increased airway pressure
is achieved by maintaining adequate right and left ventricular
filling volumes. When airway pressure increases, the lung expands
in proportion to its compliance. The intrapleural pressure
increases in direct proportion to lung expansion and inversely
with chest-diaphragm compliance. The increased intrapleural
pressure increases CVP, which reduces the gradient between the
periphery and the CVP. This reduces venous return, unless
peripheral venous tone increases, or fluids are administered. The
increased venous pressure increases fluid filtration out of the
capillary bed, tending to produce hypovolemia. The increased
Intravascular Fluid and Electrolyte Physiology 1733 54
pleural pressure is transmitted through the pericardium to reduce
the filling pressure gradient for all cardiac chambers. The conse-
quent reduction in end-diastolic volume diminishes preload to all
chambers. The elevation of intrapulmonary pressure increases
pulmonary artery pressure, which leads to augmented right
ventricular afterload. The reduced venous return, decreased
right ventricular preload, and increased afterload may severely
compromise right ventricular stroke volume, cause stiffening of
Section IV Anesthesia Management
the interventricular septum, and reduce preload for the left
ventricle.
All these consequences can be improved by intravascular
volume expansion except for the increased afterload on the right
ventricle. CVP and right ventricular end-diastolic volume are
monitored to aid in the decision to institute inotropic therapy. An
increase in CVP also compromises lymphatic return from the
lung and periphery. Sepsis, the most common coexisting condi-
tion with ARDS, causes increased vascular capacity and vasodila-
tion, decreases COP, and results in tissue edema. The net clinical
outcome of these factors is that fluid intake usually exceeds output
until ARDS and sepsis begin to subside. Studies comparing crys-
talloid and colloid management over the entire time course of
ARDS are being evaluated. After 48 hours, pulmonary function
may be better, however, in patients managed in part with colloid
than with crystalloids alone.
Acutely Burned Patient
Fluid management in an acutely burned patient focuses on res-
toration of plasma volume and a shift of the ECF volume into the
burned but viable tissue, accompanied by increased losses caused
by loss of the normal barrier function of the skin. The tissue
injury produced by the burn leads to an abrupt disruption of the
capillary bed, manifested by local vasodilation, increased perme-
ability, and, presumably, decreased reflection coefficient to pro-
teins. Vasodilation increases the surface area for filtration and
tends to increase capillary pressure. The lowered reflection coef-
ficient diminishes the ability of colloids to retain fluid in the
capillaries. Water, electrolytes, and protein enter the burned tissue
at the expense of the intravascular volume. Fluid is mobilized
from uninjured tissues by various processes, with a net result of
transfer of fluid from normal tissue into the injured tissue and
intravascular hypovolemia. Capillary permeability was formerly
thought to increase in all tissues after a serious burn injury, but
this is not the currently held view.117 In addition to the tissue
edema, there is a marked increase in loss of water by evaporation
from the wound surface, and the metabolic rate increases dra-
matically, leading to proportionate increases in fluid
requirements.
Several formulas have been developed to aid in writing the
initial fluid prescription. The key to success is close hemodynamic
monitoring with titration of therapy to the individual patients
physiology. If urine output increases progressively, or if filling
pressures increase, fluid administration must be decreased. The
Parkland formula118 prescribes fluids based on the percentage of
the body surface area (BSA) burned (% BSA-burned): 2mL/kg/%
BSA-burned during the first 8 hours and 2mL/kg/% BSA-burned
during the next 16 hours. The formula prescribes the following
regimen of fluids:
IV 1734 Anesthesia Management
(
Lactated Ringers solution at 0.25 mL kg [% BSA-burned ] ) and epinephrine solution. This is referred to as the tumescent
hr for 8 hours
(
Lactated Ringers solution at 0.125 mL kg [% BSA-burned ] ) the tissue firm and to facilitate removal of adipose tissue. Because
hr for 16 hours
(
5% dextrose in water at 0.8 mL kg [% BSA-burned ] ) shifts. Semitumescent liposuction tends to involve removal of
(
hr plus 5% albumin at 0.015 mL kg [% BSA-burned ] ) larger volumes of fat, larger volumes of lidocaine plus epineph
hr for 24 hours
A 50-kg person with 50% BSA burn should receive fluids
equal to (50kg) (0.25mL/kg) (50% BSA-burned), or 625mL/
hr for 8 hours. This is followed by fluids equal to (50kg)
(0.125mL/kg) (50% BSA-burned), or 310mL/hr for 16 hours.
During the second 24 hours, fluids administered equal (50kg)
(0.08mL/kg) (50% BSA-burned), or 200mL/hr of 5% dextrose
in water, plus (50kg) (0.015mL/kg) (50% BSA-burned), or
37.5mL/hr of 5% albumin. Colloids are not contraindicated, even
during the acute phase of fluid resuscitation. Although they pass
into the injured tissue at an accelerated rate, they have a more
sustained effect on plasma volume than crystalloid alone.
Pregnant Patient with Preeclampsia
Goals of fluid management in a pregnant patient with preeclamp-
sia (see Chapter 69) are to restore the contracted intravascular
volume,119 avoid excessive intravascular fluid administration
because of normal postpartum fluid mobilization, replace
increased sensible (sweat) and insensible (respiratory) losses
resulting from labor, be prepared to replace rapid blood loss,
avoid hypotension resulting from anesthetic-induced vasodila-
tion to preserve uteroplacental flow, maintain normoglycemia,
avoid decreasing the COP further, and prevent pulmonary and
cerebral edema. Restoration of depleted plasma volume to normal
is crucial for control of hypertension and for administration of
anesthesia. Plasma volume is decreased in preeclamptic patients
because of elevated lower body venous pressure from uterine
compression, pressure-induced natriuresis as seen in hyperten-
sion, severe vasoconstriction, and hypoproteinemia owing to pro-
teinuria. After parturition, the fluid retained during pregnancy is
usually quickly mobilized and excreted. The preeclamptic patient
may have compromised hepatic, renal, or cardiac function,
however, impeding the timely excretion of this fluid. These
patients are at risk for pulmonary edema because of cardiac
failure secondary to severe hypertension and lowered COP.
Cardiac filling pressures should be monitored, and a PAOP of less
than 12 to 15mmHg should be maintained. Plasma tonicity must
be maintained by administration of isotonic fluid because these
patients are at risk for cerebral edema.
Patient Undergoing Liposuction
Patients undergoing liposuction pose unique challenges in regard
to intraoperative fluid management (see Chapter 64). A major
factor determining the amount of intravascular volume changes
during surgery is the type of liposuction being performed. Lipo-
suction can be an office-based procedure using minimal sedation
and subcutaneous infiltration of large volumes of dilute lidocaine
technique and tends to be used to remove smaller volumes of fat
(<3000mL). Infiltration of large fluid volumes is used to make
of the smaller volumes of fat removed and less damage to subcu-
taneous tissue, the tumescent technique has less risk of large fluid
rine solution, and more sedation. When a volume of more than
2000 to 3000mL of fat is removed, the patient may require general
anesthesia. These cases may lead to life-threatening complications
involving fluid management. Ideally, the epinephrine in the
tumescent solution decreases systemic absorption of the extremely
large volumes of fluid administered subcutaneously. If the fluid is
not removed before the effect of the epinephrine has worn off,
however, the patient can absorb a significant amount of the
administered fluid. Case reports120-122 have described the develop-
ment of pulmonary edema after liposuction.
Complicating the management of liposuction patients
further are the large doses of lidocaine administered. During this
procedure, patients receive 70 to 80mg/kg of lidocaine. Vasocon-
striction and removal of most of the infused solution is thought
to prevent lidocaine toxicity.123 One report124 suggested, however,
that lidocaine might impair alveolar epithelial fluid clearance.
This impaired clearance of pulmonary fluid in conjunction with
increased intravascular volume from the absorbed solution may
explain some of the adverse occurrences, including death, reported
after liposuction. There are significant data to support the increase
in the maximum dose of lidocaine from the standard 7mg/kg,
but the maximum safe dosage of lidocaine for tumescent liposuc-
tion is still controversial.
The American Academy of Dermatology has published
guidelines for liposuction.125,126 The use of significant amounts of
intravenous sedation or analgesic administration was considered
to be potentially dangerous, particularly when conducted in an
office-based setting. The recommended maximum dose of lido-
caine was 55mg/kg. This increased dosage mandates proper
preoperative evaluation of patients medications to determine
whether they are using any drugs that inhibit the cytochrome 3A4
or 1A2 system. These are the major pathways by which lidocaine
is metabolized, and even mild inhibition of these enzyme systems
can lead to potentially lethal serum lidocaine levels.
Controlled trials regarding the best method for performing
liposuction are lacking. A large body of practical knowledge sug-
gests, however, that liposuction can be performed safely using
large doses of lidocaine and large volumes of infused solution.
Vigilance regarding potential volume overload and pulmonary
edema is required to perform this procedure safely.
Intravascular Volume Resuscitation of
Patient in Hemorrhagic Shock
The urgent priorities for the resuscitation of a seriously bleeding
patient are (1) ensuring an adequate airway, (2) ensuring adequate
respiration, and (3) supporting the circulation. Stopping blood
loss is most often achieved through manual pressure on the bleed-
ing site or operative intervention (rarely arterial tourniquet appli-
cation), or both. Formerly, intravenous crystalloid (United States)
or colloid (Europe) solution was administered to normalize blood

pressure. More recently, such intravenous fluids have been admin-
istered at rates just sufficient to maintain arterial systolic blood
pressure between 80 and 100mmHg (with some centers limiting
total infused volume during emergency medical services trans-
port to 1 to 2L of balanced salt solution [permissive hypoten-
sion]).86 Benefits of this approach (achieved at the expense of
nonvital organ hypoperfusion and lactic acidosis) include mini-
mizing (1) the rate of additional blood loss, (2) hypothermia from
Intravascular Fluid and Electrolyte Physiology 1735 54
the more histotoxic of the two in the event of extravasation)
should be administered to counteract the calcium-binding effects
of the citrate in blood products if (1) laboratory or bedside moni-
toring reveals a low ionized calcium level, (2) the patient receives
blood products at a rate of 10 combined units of PRBCs and fresh
frozen plasma per hour, (3) hepatic function (cirrhosis) or hepatic
blood flow (high aortic cross-clamp) is compromised, or (4)
hypotension remains unresponsive to apparently adequate volume

Section IV Anesthesia Management

the rapid intravenous administration of room temperature fluids,
and (3) the dilution of blood hemoglobin and plasma coagulation
factors.
When the patient is on the operating table, blood volume
expansion is best achieved with fresh whole blood (now rarely
available except at forward military trauma centers using a
walking blood bank of pretyped soldiers) or the next best thing:
combined infusion of warmed PRBCs, fresh frozen plasma, cryo-
precipitate, and platelets in ratios sufficient to maintain normal
coagulation status (e.g., PRBC-to-fresh frozen plasma ratio = 1:1
after the one blood volume exchange; 10U of cryoprecipitate and
platelets per 10U of PRBCs for coagulopathic bleeding). More
recently, there have been case reports of the administration of
recombinant activated factor VII (NovoSeven) arresting exsan-
guinating hemorrhage owing to dilutional coagulopathy, but the
precise use and dosing of this agent in trauma resuscitation have
yet to be fully elucidated (see Chapter 72).
During massive transfusion of a severely hemorrhaging
patient undergoing operative intervention, circulating volume
can repeatedly alternate between hypovolemia and normovo-
lemia, depending on the progress of surgery and the speed at
which the hospital blood bank can provide blood components to
the anesthesia team. Especially if balanced salt solutions are given
in large quantities, overhydration (especially with mildly hypo
tonic Ringers lactate solution) should be avoided in patients with
head injury because this increases cerebral edema, increases
intracranial pressure, and reduces cerebral perfusion pressure
and brain oxygenation.
The patients acid-base status should be corrected primarily
through ensuring adequate ventilation and normalizing blood
volume and composition. Metabolic acidosis usually can be
rapidly reversed by intravenous administration of sodium bicar-
bonate, but at the expense of temporary elevation in sodium
(which may be beneficial in head-injured patients). Ten percent
calcium gluconate or calcium chloride solution (the latter being
administration.
In addition to ensuring that all crystalloid, colloid, and
blood product infusions are warmed (preferably to 40C [104F]),
surface heating should be implemented (though use of forced-air
heating blankets and, if necessary, warming the operating room
to 26.6C [80F]). Hypotensive patients also should be evaluated
for nonhemorrhagic mechanisms of shock, including cardiac/
obstructive (tamponade, tension pneumothorax, air embolism)
and distributive (head and spinal cord injury, anaphylaxis, trans-
fusion reaction).
In addition to providing life support, the anesthesia team
should administer central nervous system depressant medica-
tions as tolerated to provide unconsciousness (e.g., administer
scopolamine for systolic blood pressure <60mmHg; administer
ketamine or midazolam, or both, for systolic blood pressure 60
to 90mmHg; add fentanyl for systolic blood pressure 80 to
100mmHg; and add an inhaled agent when systolic blood pres-
sure is sustained >100mmHg, with normal urine flow and
minimal arterial pressure variation with mechanical ventilation).
Complete muscle relaxation often must be achieved whether or
not fully anesthetizing drug doses are tolerated, and brain wave
monitors can be helpful beyond standard monitors, an arterial
line, and a central venous catheter.
Acknowledgments
The authors thank Dr. Alan W. Grogono, former Chairman and
Professor, Tulane Medical Center, and Dr. Ian Kucera, former
Assistant Professor, Texas Tech Health Sciences Center, Lubbock,
Texas, for their past contributions to earlier versions of this
chapter. The authors also thank Dr. Casey Daste, Dr. Alex Hellman,
and Dr. John Vu for their help in the preparation of the references
in this chapter.

References
1. Rhoades RA, Tanner GA: Medical Physiology.
Boston, Little, Brown, 1995.
2. Fall PJ: Hyponatremia and hypernatremia: A sys-
tematic approach to causes and their correction.
Postgrad Med 107:75, 2000.
3. Coran AG: Perioperative care of the pediatric
patient. Surg Ann 23(Pt 1):31, 1991.
4. Barajas L, Powers K: The structure of the juxta-
glomerular apparatus (JGA) and the control of
renin secretion: An update. J Hypertens 2:3,
1984.
5. Herrmann H, Dzau V: The feedback regulation of
angiotensinogen production by components of
the renin-angiotensin system. Circ Res 52:328,
1983.
6. Clauser E, Bouhnik J, Coezzy E, et al: Synthesis and
release of immunoreactive angiotensinogen by rat
liver slices. Endocrinology 112:1188, 1983.
7. Campbell D: The site of angiotensin production. J
Hypertens 3:199, 1985.
8. Ganong W: The brain renin-angiotensin system.
Annu Rev Physiol 46:17, 1984.
9. Reid I: The renin-angiotensin system and body
function. Arch Intern Med 145:1475, 1985.
10. Reid I: Actions of angiotensin II on the brain:
Mechanisms and physiologic role. Am J Physiol
246:F533, 1984.
11. Kost C: Effect of angiotensin II on plasma adeno
sine concentrations in the rat. J Cardiovasc Phar-
macol 17:838, 1991.
12. Ferrario C, Santos R, Brosnihan K, et al: A hypoth-
esis regarding the function of angiotensin peptides.
Clin Exp Hypertens 10A:107, 1988.
IV 1736 Anesthesia Management
13. Zubrow AB, Daniel SS, Stark RI, et al: Plasma renin,
catecholamine, and vasopressin during nitroprus-
side-induced hypotension in ewes. Anesthesiology
58:245, 1983.
14. Agre P: Aquaporin water channels in kidney. J Am
Soc Nephrol 11:764, 2000.
15. Nielsen S, Kwon TH, Christensen BM, et al: Physi-
ology and pathophysiology of renal aquaporins. J
Am Soc Nephrol 19:647, 1999.
16. Jensen V: The TURP syndrome. Can J Anaesth
38:90, 1991.
17. Pasini A, Belloni C: Intraoperative complications of
697 consecutive operative hysteroscopies. Min
Ginecol 53:13, 2001.
18. Moses AM, Blumenthal SA, Streeten DH: Drugs
and water metabolism. In Arieff AI, Defronzo FA
(eds): Fluid, Electrolyte and Acid-Base Disorders,
vol 1. New York, Churchill Livingstone, 1985, p
1145.
19. Zaloga GP, Prough DS: Fluids and electrolytes. In
Barash PG, Cullen BF, Stoelting RK (eds): Clinical
Anesthesia, 2nd ed. Philadelphia, JB Lippincott,
1992, pp 214-236.
20. Gravenstein D: Transurethral resection of prostate
(TURP) syndrome: A review of the pathology and
management. Anesth Analg 84:438, 1997.
21. Ayus JC, Wheeler J, Arieff AI: Postoperative
hyponatremic encephalopathy in menstruant
women. Ann Intern Med 117:891, 1992.
22. Fraser CL, Sarnacki P: Na+/K+ ATPase pump func-
tion in male rat brain synaptosomes is different
from that of females. Am J Physiol 257:E284,
1989.
23. Black RM: Disorders of serum sodium and serum
potassium. In Rippe JM, Irwin RS, Alpert JS, et al
(eds): Intensive Care Medicine. Boston, Little,
Brown, 1985.
24. Sarnaik AP, Meert K, Hackbarth R, et al: Manage-
ment of hyponatremic seizures in children with
hypertonic saline: A safe and effective strategy. Crit
Care Med 19:758, 1991.
25. Oh M, Carrol H: Disorders of sodium metabolism:
Hypernatremia and hyponatremia. Crit Care Med
20:94, 1992.
26. Perez GO, Oster JR, Robertson GL: Severe hyper-
natremia with impaired thirst. Am J Nephrol 9:421,
1989.
27. Illner HP, Cunningham JN, Shires GT: Red blood
cell sodium content and permeability changes in
hemorrhagic shock. Am J Surg 143:349, 1982.
28. Coran AG: Perioperative care of the pediatric
patient. Surg Ann 23(Pt 1):31, 1991.
29. Brown MJ, Brown DL, Murphy MB: Hypokalemia
from 2 receptor stimulation by circulating epine-
phrine. N Engl J Med 309:1414, 1983.
30. Vincent HH, Boomsma F, Man int Veld AJ, et al:
Effects of selective and nonselective beta agonists
on plasma potassium and norepinephrine. J Cardio-
vasc Pharmacol 6:107, 1984.
31. Papademetrious V, Fletcher R, Khatri IM, et al:
Diuretic-induced hypokalemia in uncomplicated
systemic hypertension: Effect of plasma potassium
correction on cardiac arrhythmias. Am J Cardiol
52:1017, 1983.
32. Papademetrious V, Burris J, Kukish SIM, et al:
Effectiveness of potassium chloride or triamterene
in thiazide hypokalemia. Arch Intern Med 145:1986,
1985.
33. DeFronzo RA: Intravenous potassium chloride
therapy. JAMA 245:2446, 1981.
34. Solomon RJ, Katz JD: Disturbances of potassium
homeostasis. In Stoelting RK (ed): Advances in
Anesthesia, vol 3. Chicago, Year Book Medical Pub-
lishers, 1986.
35. Allon M, Takeshian A, Shanklin N: Effect of insu-
lin-plus-glucose infusion with or without epine-
phrine or fasting hyperkalemia. Kidney Int 43:212,
1993.
36. Holick MF, Krane SM, Potts JT: Calcium, phospho-
rus, and bone metabolism: Calcium-regulating hor-
mones. In Isselbacher KJ, Braunwald E, Wilson JD,
et al (eds): Harrisons Principles of Internal
Medicine, 13th ed. New York, McGraw-Hill, 1994,
p 2139.
37. Vivien B, Langeron O, Morrell E, et al: Early hypo
calcemia in severe trauma. Crit Care Med 33:1946,
2005.
38. Harris MNE, Crowther A, Jupp RA, et al: Magne-
sium and coronary revascularization. Br J Anesth
60:779, 1988.
39. Kamalanathan S, Vijayan A: Fluid and electrolyte
management. In Cooper DH, Krainik AJ, Lubner SJ,
et al (eds): The Washington Manual of Medical
Therapeutics, 32nd ed. Philadelphia, Lippincott
Williams & Wilkins, 2007, p 89.
40. James MFM: Clinical use of magnesium infusions
in anesthesia. Anesth Analg 74:129, 1992.
41. Cross HS, Debiec H, Peterlik M: Mechanism and
regulation of intestinal phosphate reabsorption.
Miner Electrolyte Metab 16:115, 1990.
42. Rhoades RA, Tanner GA: Medical Physiology.
Boston, Little, Brown, 1995.
43. Cohena J, Koganb A, Saharb G, et al: Hypophos-
phatemia following open heart surgery: Incidence
and consequences. Eur J Cardiothorac Surg 26:306,
2004.
44. Levinsky NG: Acidosis and alkalosis. In Isselbacher
KJ, Braunwald E, Wilson JD, et al (eds): Harrisons
Principles of Internal Medicine, 13th ed. New York,
McGraw-Hill, 1994, p 255.
45. Drummond JC, Moore SS: The influence of dex-
trose administration on neurologic outcome after
temporary spinal cord ischemia in the rabbit.
Anesthesiology 70:64, 1989.
46. Sieber FE, Traystman RJ: Special issues: Glucose
and the brain. Crit Care Med 20:104, 1992.
47. Stagnaro-Green A: Perioperative glucose control:
Does it really matter? Mt Sinai J Med 58:299, 1991.
48. McMurry JF Jr: Wound healing with diabetes mel-
litus: Better glucose control for better wound
healing in diabetes. Surg Clin North Am 64:769,
1984.
49. Welborn LG, McGill WA, Hannallah RS, et al: Peri-
operative blood glucose concentrations in pediatric
outpatients. Anesthesiology 65:543, 1986.
50. Welborn LG, McGill WA, Hannallah RS, et al:
Glucose concentrations for routine intravenous
infusion in pediatric outpatient surgery. Anesthesi-
ology 67:427, 1987.
51. Degoute CS, Ray MJ, Manchon M, et al: Intraopera-
tive glucose infusion and blood lactate: Endocrine
and metabolic relationships during abdominal
aortic surgery. Anesthesiology 71:355, 1989.
52. The Carter Center of Emory University: Closing the
gap: The problem of diabetes mellitus in the United
States. Diabetes Care 8:391, 1985.
53. Nesmeth J: Type 2 diabetes in children and adoles-
cents. Pediatr Rev 105:671, 2000.
54. Foster DW: Endocrinology and metabolism. In
Wilson JD, Braunwald E, Fauci AS, et al (eds): Har-
risons Principles of Internal Medicine, 12th ed. New
York, McGraw-Hill, 1991, p 1743.
55. Engelgau MM, Herman WH, Smith PJ, et al: The
epidemiology of diabetes and pregnancy in the U.S.,
1988. Diabetes Care 18:1029, 1995.
56. Stewart WJ, McSweeney SM, Kellet MA, et al:
Increased risk of severe protamine reactions in
NPH insulin-dependent diabetics undergoing
cardiac catheterization. Circulation 70:788,
1984.
57. Humalog (insulin lispro) package insert. Indiana-
polis, IN, Eli Lilly, 2001.
58. Kucera ML, Graham JP: Insulin lispro, a new
insulin analog. Pharmacotherapy 18:526, 1998.
59. Dimarchi RD, Chance RE, Lon HB, et al: Prepara-
tion of an insulin with improved pharmacokinetics
relative to human insulin through consideration of
structure homology with insulin like growth factor
1. Horm Res 41(Suppl 2):93, 1994.
60. Howey DC, Bowsher RR, Brunelle RL, Woodworth
JR: Lys(B28) Pro(B29) human insulin: A rapidly
absorbed analog of human insulin. Diabetes 43:396,
1994.
61. Ciszak E, Beals JM, Frank BH, et al: Role of C-ter-
minal B-chain residues in insulin assembly: The
structure of Hexameric Lys(B28) Pro(B29) human
insulin. Structure 3:615, 1995.
62. Lantus (insulin glargine) package insert. Kansas
City, MO, Aventis Pharmaceuticals, April 2000.
63. Heinemann L, Linkeschova R, Rave K, et al: Time
action profile of the long-acting insulin analog
insulin glargine (HOE 901) in comparison with
those of NPH insulin and placebo. Diabetes Care
23:644, 2000.
64. Madsen H, Ditzel J: Changes in red blood cell
oxygen transport in diabetic pregnancy. Am J
Obstet Gynecol 143:421, 1982.
65. Kitamura A, Hoshino T: Patients with diabetic neu-
ropathy are at risk of a greater intraoperative reduc-
tion in core temperature. Anesthesiology 92:1311,
2000.
66. Tsueda K, Huang KC, Dumond SW, et al: Cardiac
sympathetic tone in anaesthetized diabetics. Can J
Anaesth 38:20, 1991.
67. Kannel W, McGee D: Diabetes and cardiovascular
disease: The Framingham Study. JAMA 241:2035,
1979.
68. Haffner SM, Lehto S, Ronnemaa T, et al: Mortality
from coronary heart disease in subjects with type 2
diabetes and in nondiabetic subjects with and
without prior myocardial infarction. N Engl J Med
339:229, 1998.
69. Schaade DS: Surgery and diabetes. Med Clin North
Am 72:1531, 1988.
70. Gavin LA: Management of diabetes mellitus during
surgery. West J Med 151:525, 1989.
71. Reissel E, Orko R, Maunuksela E, et al: Predictabil-
ity of difficult laryngoscopy in patients with long-
term diabetes mellitus. Anaesthesia 43:1024, 1990.
72. Hogan K, Rusty D, Springman SR: Difficult laryn-
goscopy and diabetes mellitus. Anesth Analg
67:1162, 1988.
73. Davidson MB, Schriger DC, Peters AL, et al: A
clinical approach for the diagnosis of diabetes mel-
litus: An analysis using glycosylated hemoglobin
levels. JAMA 276:1246, 1996.
74. Nathan DM: Hemoglobin A1c: Infatuation or the
real thing? N Engl J Med 323:1062, 1990.
75. American Hospital Formulary Service: Miscellane-
ous Antidiabetic Agents. Bethesda, MD, ASHP,
2000, 68.20.92 3032-41.
76. Glucophage R (metformin) product information.
Princeton, NJ, Bristol-Myers-Squibb, October
2000.
77. Watson BG, Pay DA, Williamson M, et al: Practical
closed-loop insulin infusion. Life Support Syst
3(Suppl 1):519, 1985.
78. Hirsch IB, Magill JB, Cryer PE, et al: Perioperative
management of surgical patients with diabetes mel-
litus. Anesthesiology 74:346, 1991.
79. Nutt LK, ONeil RG: Effect of elevated glucose on
endothelin-induced store-operated and non-store-

operated calcium influx in renal mesangial cells. J
Am Soc Nephrol 11:1225, 2000.
80. Marhoffer W, Stein M, Maeser D, et al: Impairment
of PMN leukocyte function and metabolic control
of diabetes. Diabetes Care 15:256, 1992.
81. Furnary AP, Zerr KJ, Grunkemeier GL, Starr A:
Continuous intravenous insulin infusion reduces
the incidence of deep sternal wound infection in
diabetic patients after cardiac surgical procedures.
Ann Thorac Surg 67:352, 1999.
82. Rassias AJ, Marrin CA, Arruda J, et al: Insulin infu-
sion improves neutrophil function in diabetic
cardiac surgery patients. Anesth Analg 88:1011,
1999.
83. Khaodhiar L, McCowen K, Bistrian B: Perioperative
hyperglycemia, infections and risk. Curr Opin Clin
Nutr Metab Care 2:79, 1999.
84. Boyle PJ, Schwartz NS, Shah SD, et al: Plasma
glucose concentrations at the onset of hypogly
cemic symptoms in patients with poorly controlled
diabetes and in nondiabetics. N Engl J Med
318:1487, 1988.
85. Wasserstrum N: Issues in fluid management during
labor: General considerations. Clin Obstet Gynecol
35:505, 1992.
86. Dutton R: Current concepts in hemorrhagic shock.
Anesthesiol Clin 25:23-34, 2007.
87. Alam HB, Rhee P: New developments in fluid
resuscitation. Surg Clin North Am 87:55, 2007.
88. Shoemaker WC, Appel PL, Kram HB, et al: Pro-
spective trial of supranormal values of survivors as
therapeutic goals in high risk surgical patients.
Chest 94:1176, 1988.
89. Hayes MA, Timmins AC, Yau E, et al: Elevation of
systemic oxygen delivery in the treatment of criti-
cally ill patients. N Engl J Med 330:1712, 1994.
90. Gattinoni L, Brazzi L, Pelosi P, et al: A trial of goal-
oriented hemodynamic therapy in critically ill
patients. N Engl J Med 333:1025, 1995.
91. Boyd O, Grounds M, Bennett ED: A randomized
clinical trial of the effect of deliberate peri-operative
increase of oxygen delivery on mortality in high
risk surgical patients. JAMA 270:2699, 1993.
92. Sinclair S, James S, Singer M: Intraoperative intra-
vascular volume optimisation and length of hospital
stay after repair of proximal femoral fracture: Ran-
domised control trial. BMJ 315:909, 1997.
93. Lobo SMA, Salgado PF, Castillo VGT, et al: Effects
of maximizing oxygen delivery on morbidity and
mortality in high risk surgical patients. Crit Care
Med 28:3396, 2000.
94. Schakford SR, Fortlage DA, Peters RM, et al: Serum
osmolar and electrolyte changes associated with
large infusions of hypertonic sodium lactate for
intravascular volume expansion of patients under-
Intravascular Fluid and Electrolyte Physiology 1737
going aortic reconstruction. Surg Gynecol Obstet
164:127, 1987.
95. Rocha e Silva M, Velasco IT, Portirio MMF: Hyper-
tonic saline resuscitation: Saturated salt-dextran
solutions are equally effective, but induce hemolysis
in dogs. Crit Care Med 18:203, 1990.
96. Velasco IT, Pontieri V, Rocha e Silva M, et al: Hyper-
osmotic NaCl and severe hemorrhagic shock. Am J
Physiol 239:H664-H673, 1980.
97. Bunn F, Roberts I, Tasker R, et al: Hypertonic versus
isotonic crystalloid for fluid resuscitation in criti-
cally ill patients. Cochrane Database Syst Rev 4:
CD002045, 2000; update in 1:CD002045, 2002.
98. Kreimeier U, Bruckner UB, Niemczyk S, et al:
Hyperosmotic saline dextran for resuscitation from
traumatic-hemorrhagic hypotension: Effect on
regional blood flow. Circ Shock 32:83, 1990.
99. Leppaniemi A, Soltero R, Burris D, et al: Early
resuscitations with low-volume polyDClHb is effec-
tive in the treatment of shock induced by penetrat-
ing vascular injury. J Trauma 40:242, 1996.
100. Moss GS, Gould SA: Plasma expanders: An update.
Am J Surg 155:425, 1988.
101. Velanovich V: Crystalloid versus colloid fluid resus-
citation: A meta-analysis of mortality. Surgery
105:65, 1989.
102. Kaufman CR: Usefulness of thromboelastography
in assessment of trauma patient coagulation. J
Trauma 42:716, 1997.
103. Martin G, Bennett-Guerrero E, Wakeling H, et al:
A prospective, randomized comparison of thrombo
elastographic coagulation profile in patients receiv-
ing lactated Ringers solution, 6% hetastarch in a
balanced-saline vehicle, or 6% hetastarch in saline
during major surgery. J Cardiothorac Vasc Anesth
16:441, 2002.
104. Tuullis JL: Albumin: Background and uses. JAMA
237:355, 1977.
105. Mangar D, Gerson JI, Constantine RM, et al: Pul-
monary edema and coagulopathy due to hyskon
(32% dextran-70) administration. Anesth Analg
68:686, 1989.
106. Kozek-Langenecker SA: Effects of hydroxyethyl
starch solutions on hemostasis. Anesthesiology
103:654, 2005.
107. Paut O, Lacroix F: Recent developments in the peri-
operative fluid management for the paediatric
patient. Curr Opin Anaesthesiol 19:268, 2006.
108. Leone BJ, Spahn DR: Anemia, hemodilution, and
oxygen delivery. Anesth Analg 75:651, 1992.
109. Shires T, Williams J, Brown F: Acute change in
extracellular fluids associated with major surgical
procedures. Ann Surg 154:803, 1961.
110. Roth E, Lax L, Maloney J: Ringers lactate solution
and extracellular fluid volume in the surgical
54
patient: A critical analysis. Ann Surg 169:149,
1969.
111. Prien T, Backhaaus N, Pelster F, et al: Effect of intra-
operative fluid administration and colloid osmotic
pressure on the formation of intestinal edema
during gastrointestinal surgery. J Clin Anesth 2:317,
1990.
112. Schrieer RW: Pathogenesis of sodium and water
retention in high-output and low-output cardiac
failure, nephrotic syndrome, cirrhosis and preg-
Section IV Anesthesia Management
nancy. N Engl J Med 319:1127, 1988.
113. Epstein M: Atrial natriuretic factor in patients with
liver disease. Am J Nephrol 9:89, 1989.
114. Salerno F, Badalamenti S, Moser P, et al: Atrial
natriuretic factor in cirrhotic patients with tense
ascites: Effect of large-volume paracentesis. Gastro-
enterology 98:1063, 1990.
115. Wisner D, Busche F, Sturm J, et al: Traumatic shock
and head injury: Effects of fluid resuscitation on the
brain. J Surg Res 46:49, 1989.
116. Jackowski A: Disordered sodium and water in neu-
rosurgery. Br J Neurosurg 6:173, 1992.
117. Dietsch EA: The management of burns. N Engl J
Med 323:1249, 1990.
118. Dietsch EA: Burn management. In Ripe JM, Irwin
RS, Alpert JS, et al (eds): Intensive Care Medicine,
2nd ed. Boston, Little, Brown, 1991, p 1507.
119. Waserstrum N: Issues in fluid management during
labor: Maternal plasma volume status and volume
loading. Clin Obstet Gynecol 35:514, 1992.
120. Gilliland MD, Coates N: Tumescent liposuction
complication by pulmonary edema. Plast Reconstr
Surg 99:215, 1997.
121. Pitman GH: Tumescent liposuction complication
by pulmonary edema. Plast Reconstr Surg 100:1363,
1997.
122. Rao RB, Ely SF, Hoffman RS: Deaths related to lipo-
suction. N Engl J Med 340:1471, 1999.
123. Ostad A, Kageyama N, Moy RL: Tumescent
anesthesia with lidocaine dose of 55mg/kg is safe
for liposuction. Dermatol Surg 22:921, 1996.
124. Laffon M, Jayr C, Barbry P, et al: Lidocaine induces
a reversible decrease in alveolar epithelial fluid
clearance in rats. Anesthesiology 96:393, 2002.
125. Coleman W 3rd, Glogau RG, Klein JA, et al: Guide-
lines of care for liposuction. J Am Acad Dermatol
45:438, 2001.
126. Kucera IJ, Lambert TJ, Klein JA, et al: Liposuction:
Contemporary issues for the anesthesiologist. J Clin
Anesth 18:379-387, 2006.