PEDIATRIC SURGERY 0039-6109/92 $0.00 + .

20

FLUID AND ELECTROLYTE

MANAGEMENT IN THE
PEDIATRIC SURGICAL PATIENT
Howard C. Filston, MD

One would think that the management of fluid and electrolytes would be
a subject that has long since been defined and quantified, but this is not the
case. Mismanagement of fluid and electrolytes still contributes to morbidity
and death for infants and young children undergoing even the simplest
procedures such as herniorrhaphies. Failure to resuscitate young patients from
hypovolemic deficiency states leads to multisystem failure and, often, to their
demise. Inappropriate overhydration of the young infant during the stresses of
anesthesia and surgery can produce interstitial fluid shifts and pulmonary
edema leading to respiratory difficulties in situations in which one would not
expect to find them. If this occurs in a hospital or ambulatory care setting that
is ill-equipped to deal with critical care management of infants and children,
unexpected morbidity and death can result.
Therefore, it is essential that the surgeon assess the fluid needs of the
patient throughout the entire period of surgical management: in the preopera-
tive resuscitation phase, if there is one; during intraoperative management,
tailoring the fluid and electrolyte administration to the degree of deficiencies
created by the surgical procedure and the underlying disease state; and
throughout the postoperative recovery phase until the patient has regained the
normovolemic state and is back to baseline maintenance fluid requirements.
This assessment must take account, not only of the volume requirement and
the rate of administration and resuscitation, but also of the composition of the
fluid with regard to electrolyte, osmotic, and colloid requirements. Considera-
tions such as the need for additional oxygen-carrying capacity and clotting
factors will influence the choice of some of the fluids in the program. A
thorough understanding of the role of such regulatory factors as the renin-
angiotensin mechanism and, most particularly, antidiuretic hormone is essential

From the Division of Pediatric Surgery and Pediatric Trauma, Department of Surgery,
University of Tennessee Medical Center, Knoxville, Tennessee

SURGICAL CLINICS OF NORTH AMERICA

VOLUME 72 • NUMBER 6 • DECEMBER 1992 1189

1190 FILSTON

to the development of an intelligent fluid and electrolyte program. To this end,

I begin by looking at a simple procedure, a normal child undergoing an inguinal
herniorrhaphy, and assess the fluid requirements to maintain that child through
his anesthetic and surgical procedure. These requirements will be the basis on
which additional fluid requirements are developed and elaborated.

MAINTENANCE FLUID REQUIREMENTS:

THE REPLACEMENT OF INSENSIBLE LOSSES
FOR ALL PATIENTS

The common practice in adult patients of restricting oral intake after

midnight the night before surgery can, in the neonate and young infant, result
in delivering to the operating room a patient who has been significantly fluid
restricted for as much as half a day. For the neonate or young infant who is
accustomed to having fluid volume replaced on an every-3-hour basis and
whose maintenance volume requirement, at 100 mLlkg, is significantly greater
than the volume requirements of an adult at 25 mLlkg, this type of restriction
can lead to significant dehydration. Infants require no more than 3 or at the
most 4 hours to empty their stomachs of clear liquids, so with careful planning,
the infant can be awakened and given a clear-liquid oral feeding within 3 to 4
hours of the scheduled procedure. The vagaries of the operating schedule
demand that if the infant is not given the first slot on the schedule, as
unfortunately is too often the case, this delay must be recognized or anticipated
and additional oral fluids or an intravenous administration be utilized to
compensate for it.
The other extreme of this issue is the overhydration of the infant during a
short operative procedure such as a herniorrhaphy in an attempt to make up
for a supposedly long period of dehydration. The rapid administration of dilute
maintenance-type solutions to a slightly dehydrated patient may, if antidiuretic
hormone is being secreted, lead to retention of large volumes of dilute
electrolyte solutions or dextrose and water, with corresponding shifts of this
fluid along a concentration gradient into the interstitial tissues, including those
of the lungs. The anesthesiologist must pay careful attention to the time of the
infant's last feeding, assume dehydration only from that point, and replace
only those fluids that would have been ingested during the normal feedings in
that time span.
Frequently, the effects of anesthetic agents reduce the cardiac output and
produce peripheral vasodilatation, resulting in hypotension to which the
anesthesiologist may respond with the administration of large volumes of fluid.
Although a bolus of isotonic fluid is an appropriate initial response to this
hypotension, the continued administration of excessive amounts of fluid-
beyond those that can be calculated and anticipated for the degree of fluid
restriction and the losses of the operative case-are inappropriate. Having lived
through an era in my professional career when those of us who believed in
volume restoration had to fight strongly and constantly for the use of volume
resuscitation rather than pressor agents, I hesitate to suggest that a pressor
agent may be appropriate, but in some instances, the addition of a beta-
adrenergic drug such as dopamine for the hypotensive patient who has failed to
respond to what should be an adequate fluid volume restoration may be appropriate,
and some balance of fluid resuscitation and control of vasomotor tone in a
thoughtful program is probably most appropriate.
For the patient undergoing a procedure in which there is little anticipated
FLUID AND ELECTROLYTE MANAGEMENT IN THE PEDIATRIC SURGICAL PATIENT 1191

blood loss, in which the peritoneal cavity is not significantly invaded, and in
which the disease produces neither inflammation, edema, nor ascites, fluid
administration should consist primarily of replacing insensible losses. Insensible
losses are maintenance losses and reflect the constant evaporation of water and
a very small amount of electrolyte from the surface of the body as well as the
considerable volumes of water lost during breathing as the respiratory passages
hydrate the inspired air to a partial pressure of water vapor of approximately
47 mm Hg. Because only a small amount of oxygen is extracted from the
inspired air and the rest of the oxygen and all of the nitrogen and pollutants
that have been hydrated during inspiration are then expired, most of this water
of hydration is lost to the atmosphere.
The average adult (defined as one who has a body surface area of
approximately 1. 73 m2 and weighs approximately 70 kg) requires 1000 mL of
water per square meter or approximately 25 mLlkg. This volume will replace
.the free water losses of evaporation from the skin and the losses from hydration
of inspired gases that are subsequently exhaled. In addition, it will supply an
adequate volume of free water as a solvent for the day's metabolic wastes,
which are excreted as urine. I have generally rounded off the typical adult's
maintenance fluid requirement to 1500 mL of free water, to which must be
added 5% dextrose to provide a minimal caloric load to operate the Krebs cycle
and provide an osmotic content to avoid lysis of red cells at the site of fluid
administration. The insensible loss of electrolytes for an adult is primarily in
the form of urinary and fecal excretion, with a small amount of loss by
evaporation from the skin surface. For a typical adult, this requirement amounts
to 45 to 60 mEq of sodium chloride per 24 hours and approximately 40 to 45
mEq of potassium per 24 hours.
To define the insensible or maintemmce requirements for infants and
children ranging from the extremely premature low-birth-weight neonate to
the child who is approaching adult weight and body habitus requires one or
more formulas that relate these requirements to the patient's size and the
increased volume requirement per weight for the tiny infant. The most com-
monly applied formula is that which states that the free water requirement per
day is equal to:
100 mLlkg up to 10 kg
50 mLlkg from 11 to 20 kg and
20 mLlkg beyond 20 kg
with a maximum cut-off at adult maintenance, which we have defined for our
purposes as 1500 to 2000 mLiday. A second formula '2 that approximates the
first over a wide range of sizes and ages defines the maintenance requirement
as being equal to:
(100 mL - 3 x age [years]) x weight (kg)
When this formula yields 1500 mL, the maximum has been reached. Notice
that these are essentially reduction formulas that account for the fact that the
newborn infant requires 100 mL of fluid per kilogram per day whereas the
adult requires only 25 mLlkg daily.
Similarly, electrolytes can be defined for the infant by a formula which
states that the need for sodium, chloride, and potassium is equal to 3 mEq/kg
per 24 hours. However, we have already stated that the adult requires 45 to 60
mEq of sodium chloride and 40 to 45 mEq of potassium per day. Were we to
utilize a formula of 3 mEqlkg throughout the entire age spectrum leading to
adulthood, the adult would require 3 mEqlkg, so the average adult at 70 kg
1192 FILSTON

would require 210 mEq/day, which is far in excess of the requirement previously
stated. On the other hand, by adding 3 mEqlkg to the 100 mLlkg free-water
requirement for the infant, we would produce a solution equivalent to one-
fifth (20%) normal saline. Normal saline has 154 mEq/L or 15.4 mEq/dL, and
one-fifth normal saline would have 3 mEq/dL. If we use one-fifth normal saline
as a standard maintenance solution and apply appropriately the formulas that
we know reduce the volume of fluid administration throughout the childhood
age groups, we would find that when we reach the 1500-mL maximum that
we defined for the typical adult, we would be giving 45 mEq of sodium chloride
per day in our 1500 mL of fluid (3 mEq/dL = 30 mEq/L = 45 mEqll.5 L).
We have now ascertained the insensible losses for children as an amount
calculated by a formula that defines the volume and electrolyte needs in relation
to size throughout a wide range up to a maximum for the typical adult. If we
take an example, then, of a 6-kg 4-month-old infant undergoing an inguinal
herniorrhaphy scheduled for 10 AM, we should first of all let him have his
regular feeding schedule up until 7 AM the morning of surgery, at which time,
he should get a clear liquid feeding. After he is anesthetized for the surgical
procedure, an intravenous line is started, and, considering that a herniorrhaphy
usually lasts approximately 1 hour (not actually operating time), we would
plan to give him half of his insensible loss requirement for a 4-hour period
during the operating time. His 24-hour requirement, 100 mL - 3 x his age in
years, means that he would require 98 to 100 mLlkg or 600 mL in 24 hours. In
4 hours, he would, therefore, require one sixth of this amount, or 100 mL of a
solution containing 3 mEq of sodium chloride per 100 mL, or 100 mL of one-
fifth normal saline containing 5% dextrose. We could add to this fluid 3 mEq
of potassium chloride (or potassium phosphate) to provide him with his
potassium maintenance requirement. An acceptable rate of fluid administration
for this child would be 50 mL during his hour in the operating room and 50
mL in the hour following surgery, at which point, he would be returned to his
maintenance rate of 25 mLlhour until he is taking clear liquid adequately by
mouth. Unless the child for some reason was dehydrated prior to being fluid
restricted to assure an empty stomach, there is little to substantiate a need for
any additional significant volumes of fluid; therefore, hypotension in the
operating room must be addressed in terms other than the administration of
additional volumes of dilute fluid. Obviously, if the patient has been fluid
restricted for a longer period of time, additional volume considerations are
appropriate.

MEASURED LOSSES

The second category of fluids is an easy one to understand and calculate.

These are the fluids that are lost directly from the body in volumes significant
enough to be collected and measured. The typical loss of this type is that from
a nasogastric tube and consists of gastric aspirate. Because these amounts can
be directly measured, they can be replaced precisely, volume for volume. The
composition of such fluid varies somewhat depending on whether it is primarily
a gastric aspirate or fluid from beyond the pylorus. Gastric juice contains
sodium, potassium, chloride, and hydrochloric acid or hydrogen ions in
amounts that are similar to a solution of 5% dextrose in half-normal saline to
which has been added 30 mEq of potassium chloride per liter. Table 1 shows a
comparison between gastric aspirate and this half-normal saline-potassium
solution. Losses from beyond the pylorus are essentially ultra filtrates of the
FLUID AND ELECTROLYTE MANAGEMENT IN THE PEDIATRIC SURGICAL PATIENT 1193

Table 1. SOLUTION FOR REPLACING LOSSES OF GASTRIC SECRETIONS

Gastric Juice Replacement Fluid
(mEq/L) Ds%/half NS + 30 mEq/KCl/L
Sodium 60-75 77
Chloride 105-130 107
Potassium 5-30 30
H+ 0-65 o

plasma or serum and are best replaced by balanced salt solutions such as
lactated Ringer's solution (D5%/LR).
Most patients undergoing elective surgery will not have a measured loss
consideration, at least until the postoperative period. However, when a meas-
urable loss is recognized, particularly those of significant volume from the
gastrointestinal tract, a timely replacement (at least an every-4-hour schedule)
should be ensured.

RESUSCITATION FROM HYPOVOLEMIA

Whenever the patient has experienced losses beyond the standard insen-
sible losses, or when measurable losses of body fluids have not been replaced
in a timely fashion, the hypovolemic state exists; and fluid considerations
become not only more significant, but more sophisticated. Other than loss of
airway integrity, there is no more urgent state than that of the patient who is
in hypovolemic shock or near-shock from volume deficits, as in this state, at
least some of the body's tissues are being underperfused and therefore made
ischemic. The longer such a condition persists, the more damage there will be
to vital tissues. Eventually, not only is there irreversible hypoxic injury to the
tissues, but there is release of destructive subcellular elements that may
deleteriously alter the function of enzyme systems in other organs and bring
on a state of respiratory failure. 7 It also leaves the patient with depressed
immune function and open to the further destructive effects of this septic state.
Therefore, assessment of the patient's history and physical findings, together
with a judgment about the effects of the patient's illness on the state of volume
integrity, must be achieved rapidly.
If there is any doubt about the patient's state of volume integrity, a rapid
infusion or "push" of an osmotically active solution such as balanced salt
solution (D5%/LR) should be administered. It has been shown experimentally,
and extensively confirmed clinically, that a fully volume-repleted patient can
tolerate a rapid expansion of the blood volume of 25% without any deleterious
circulatory effects. Using 8% as the blood volume for infants and children, the
blood volume would be 80 mLlkg, and the 25% "push" would equal 20 mLi
kg. If hypovolemia is suspected, therefore, a rapid infusion of D5%ILR, 20 mLi
kg, should improve circulatory dynamics to the point of allowing a small
urinary response, which would confirm the hypovolemic state, while at the
same time such a push will not harm the fully repleted patient who has a
normal circulatory system. Obviously, there may be an occasional patient in
incipient heart failure who is volume repleted and who will be tipped into
heart failure by such a volume expansion. Nevertheless, this risk is a better
one to take than to leave a patient volume depleted.
Although one would hope that a thorough understanding of the principles
1194 FILSTON

and guidelines presented in this article would result in appropriate volume

maintenance and repletion in every patient, the fall-back position of it is better
to overload than underload should be understood. Failure to restore circulatory
integrity, thereby leaving the patient in hypovolemic shock, has greater con-
sequences for the eventual survival of the patient than does the state of volume
overload, which should be easily recognizable and generally treatable by
subsequent volume restriction, the use of diuretic agents, and short-term
pulmonary support. This type of "error" should cause no long-term harm to
the patient's cellular integrity.
Depending on the degree of dehydration and volume depletion with which
the patient presents, the 20-mLlkg push may cause only brief improvement in
the patient's circulatory dynamics and a small urinary increment, or it may re-
establish circulatory dynamics and produce a sustained urinary output. Sub-
sequent to the push, the balanced salt solution should be continued at a rate
adequate to maintain the urine output at appropriate levels for the child's age
and size, as discussed later in this article. Also, a scheme for assessing volume
requirements for various intra-abdominal lesions and operative procedures is
presented, and when this scheme is understood, it can be used to help assess
the preoperative volume requirements.
The solution chosen to restore volume deficits must always be osmotically
active. When a patient is hypovolemic, volume receptors in the great vessels
leading to the heart signal the pituitary to secrete antidiuretic hormone, the
effect of which is to make permeable the collecting duct in the kidney, which
passes in juxtaposition to a capillary that has become hypertonic in its osmotic
state by the active resorption processes of the renal tubule. 3 With the collecting
duct made permeable by antidiuretic hormone, it is impossible to excrete dilute
urine, because the urine in the collecting duct will be made hyperconcentrated
by the resorption of water into the hypertoniC juxtatubular capillary. The
retention of hypotonic fluids administered in the face of hypovolemia dilutes
the osmotic activity within the vascular space, thus leading to a fluid shift from
the vascular to the extravascular space by diffusion along simple concentration
lines. Administration of hypotonic fluids containing no colloid will dilute not
only the colloid osmotic pressure within the vascular space but the crystalloid
osmotic pressure as well. The use of osmotically active fluid in the form of
balanced salt solution (D5%/LR) will ensure that at least the crystalloid osmotic
pressure is maintained. There is still some dilutional effect when colloid is not
present in the resuscitation fluid, and it has been shown that the resuscitation
of a volume-depleted patient with balanced salt solution will result in a three
to eight fold shift of fluid from the vascular to the extravascular space compared
with resuscitation of the same volume deficit with fluids containing not only
balanced salt solution but also a physiologic level of albumin in the form of a
5% solution. 1, 5, 6, 8 Clearly, if a balanced salt solution containing an appropriate
level of crystalloid osmotic activity results in such a significant inefficiency of
volume restitution, a solution such as half-normal saline (D5%/half NS) or
quarter-normal saline (D5%/quarter NS), which gives up a significant amount
of crystalloid osmotic pressure, will cause an even greater shift of fluid from
the vascular to the extravascular space. Markedly excessive volumes of such
fluids would be required to restore the vascular volume to normal, resulting in
massive interstitial edema, just the result we do not want to produce.
When the patient is known to be colloid depleted or when the risk of
producing any degree of interstitial edema is great, for instance, when pul-
monary dynamics are already compromised, colloid-containing solutions such
as 5% albumin in balanced salt solution should be the choice for volume
FLUID AND ELECTROLYTE MANAGEMENT IN THE PEDIATRIC SURGICAL PATIENT 1195

restitution. Blood products such as plasma and whole blood have even greater
holding power as far as their ability to restore vascular volume deficits without
significant shifts to the extravascular space. However, the relative risks of the
use of these products limit them to situations in which either red cell carrying
capacity is severely depleted, in which case, the use of packed red blood-cell
transfusions combined with balanced salt solution would be appropriate, or in
which clotting factor deficiencies exist, in which case, the use of plasma will
not only give excellent restoration of volume but will also provide the needed
clotting factors. Plasma can also be used where the need for excellent volume
restoration with minimal edema formation makes the risk of transfusion
reactions and the introduction of infectious agents worth taking.
Whenever volume depletion is recognized or anticipated, some type of
osmotically active fluid (e.g., DS%/LR) should be administered to the patient
until the volume depletion is overcome. Only then should the more hypotonic
solutions that are generally used to replace measured losses (such as gastric
aspirates) or the even more dilute quarter-normal or one-fifth normal saline
solutions used for replacement of insensible losses be introduced into the fluid
program. Preoperative volume restoration should continue at a rapid pace until
volume normalcy is confirmed by the resumption of an appropriate urine
output.
Blood pressure alone is not an adequate measure of volume restoration,
because the ability of the vascular system to constrict and shunt blood from
less vital tissues means that normal central blood pressure will be maintained
until quite late in the hypovolemic shock state. In a young, healthy individual
with highly responsive cardiac function and the ability to constrict the peripheral
arterioles, more than 30% of the blood volume may be lost before a significant
fall in central blood pressure is appreciated. 2 By the time that state exists,
severe restriction of flow to many vital tissues has already occurred.

INTRAOPERATIVE FLUID ADMINISTRATION

During the operative procedure, the fluid choice should be that of the
most dominant fluid loss. For simple elective procedures such as treatment of
hernias and "lumps and bumps," insensible loss is the dominant fluid deficit,
and appropriate hypotonic fluids should be given to replace free water and
minimal electrolytes. For procedures in which significant blood loss occurs or
in which major prolonged invasion of the abdominal or thoracic cavity is
involved, volume-restoration fluids of the balanced salt or blood products
variety should be chosen. It would be best to base this volume restoration on
a combination of the assessment of the preoperative volume restoration state
and the intraoperative and postoperative fluid requirements. If the patient
arrives in the operating room fully volume restored, this estimate can be based
on the intraoperative and postoperative requirements alone.

SCIENTIFIC BASIS FOR THE FLUID AND

ELECTROLYTE PROGRAM

Until the early 1960s, the guiding principles for fluid and electrolyte
management of the postoperative patient in general, and the child patient
specifically, were fluid restriction and absolute electrolyte restriction. High
levels of aldosterone and antidiuretic hormone were noted to be present in the
1196 FILSTON

postoperative patient. At that time, it was not understood that antidiuretic

hormone's major role was that of volume protection; it was thought, on the
contrary, to be a minor modulator of osmolality or tonicity that would ordinarily
be secreted only when the serum electrolytes were hypertonic. In the hypertonic
state, antidiuretic hormone would make the collecting duct permeable, allowing
water to be passively resorbed and resulting in dilution of the hypertonic
serum. The presence of antidiuretic hormone in postoperative and post-trauma
patients whose serum electrolytes were either normal or hypotonic was deemed
"inappropriate" at that time. As a result of this view, the principles of fluid
restriction and salt restriction held sway until the work of Shires9 - 11 and many
others proved that the postoperative patient was indeed in a state of hypovo-
lemia and that antidiuretic hormone was the principal volume protector of the
body.
One of the original works illustrating the effects of the hypovolemic state
was that of Wiggers. 13, 14 In a series of experiments in dogs, sufficient blood
was shed to produce a fall of the systolic pressure from the normal 100 mm
Hg to a severely hypovolemic 30 mm Hg. When this state was maintained for
more than 2 hours followed by the restoration of all the shed blood, the
mortality rate exceeded 80%. From this work came a concept of irreversible
shock, which contended that if the patient was hypovolemic long enough,
even complete restoration of the shed blood failed to save the animal. In time,
it was appreciated that what occurred with the hypovolemic patient was a state
of vasoconstriction and shunting in which more and more tissues were deprived
of circulation and thus made hypoxic and acidotic with a build-up of waste
products that were detrimental to the functioning of enzyme systems. With
the realization that the initial injury involved the endothelial cell of the capillary,
turning it from a semipermeable to a highly permeable membrane, it became
apparent that when the circulation was restored by reinfusing the shed blood,
the reopening of these damaged capillary beds resulted in massive fluid shifts
from the vascular space to the extravascular space, a shift that might even be
accompanied by albumin molecules, which then acted osmotically to pull
additional fluid from the vascular to the extravascular space. Thus, although
all of the shed blood was restored to the circulation, the dog resumed the
hypovolemic state when the massive fluid shifts took place through the
damaged capillary beds. This fact could be confirmed by the rise in hematocrit
from the preoperative state to a state of hyperconcentration following restora-
tion of the shed blood.
This work was the origin of the concept of third-space shifting, in which
it is postulated that fluids shift from the vascular to the extravascular space
when there is ischemic damage to the capillaries. Third-space shifts also occur
when osmotic imbalances are created, such as when there is inflammation of
tissues, resulting in leakage of albumin into the extravascular space, and when
there is a hydrostatic pressure increase, such as in the face of bowel obstruction
or obstruction of the sinusoidal capillary system in the liver, creating ascites.
Thus, any edema or ascites and any shift of fluid into the bowel lumen or wall
in obstructive or inflammatory states represents a third-space shift.
The "third space" can be defined as that which ordinarily would act as a
second space or that would restore volume to a depleted vascular space
secondary to the fall of hydrostatic pressure created by volume depletion.
However, in the state of pathophysiology, the fluid cannot be restored because
of an osmotic or hydrostatic pressure imbalance favoring diffusion of fluid out
of the vascular into the extravascular tissues. The third space is thus a
dysfunctional second space.
FLUID AND ELECTROLYTE MANAGEMENT IN THE PEDIATRIC SURGICAL PATIENT 1197

The correction of the third-space deficit was confirmed by Shires and

associatesl l in their studies in the early 1960s repeating the Wiggers experiments
in dogs with one modification. When these investigators had shed the blood,
maintained the hypovolemic state for more than 2 hours, and restored the shed
blood, they added increasing volumes of balanced salt solution. In so doing,
they showed that they could indeed reverse the seemingly irreversible shock
state and salvage the dogs. It was soon understood that this administration of
additional balanced salt solution compensated for that volume of fluid that had
been lost from the vascular space through the damaged capillaries after
reinstitution of the circulation, thus maintaining the restored volume in a way
that simply replacing the shed blood had been unable to do.
Although there was initial opposition to the concept of giving large volumes
of salt-containing solution to postoperative patients, and there was an initial
over-reaction resulting in fluid overload in many patients, understanding of
these concepts allows the development of a rational fluid administration
program for the postoperative, post-trauma patient. This program is based on
calculation of insensible losses, measurement of external losses, and a reasoned
estimation of fluids required to restore internal third-space shifts, together with
careful monitoring of urine output to guide the fluid administration program
and correct errors in the initial "guess."

ESTIMATING INTRAOPERATIVE AND POSTOPERATIVE

THIRD-SPACE REQUIREMENTS

In contradistinction to the formulas that can closely approximate the needs

consequent to insensible losses and the ability to measure exactly those losses
that occur outside the body, such as by gastric aspirate through a nasogastric
tube, those fluid volumes that shift from the vascular to the extravascular space
as a result of either ischemic injury to the capillaries or changes in hydrostatic
and osmotic pressure dynamics during intra-abdominal surgery cannot be
measured, only estimated. Fluids shift from the vascular to the extravascular
space during and after intra-abdominal surgery as a result of bowel obstruction
with sequestration of fluids within the bowel lumen, traumatic and inflamma-
tory changes in the bowel wall tissues themselves that lead to edema formation,
and changes in the pressure relations through the sinusoidal systems of the
liver, resulting in the formation of ascites. Ascites (and pleural fluid accumu-
lations) can also result from fluid shifts from the bowel or the peritoneal
surfaces into the abdominal or thoracic cavity as the result of either trauma to
the peritoneal and serosal surfaces or inflammation or infection causing damage
with weeping of protein-containing fluids into the peritoneal cavity.
The patient's fluid requirements during and after operation can be arrived
at by calculating the insensible losses to be replaced by maintenance fluids,
then by measuring the intraoperative and continuing postoperative measurable
losses from the nasogastric tube or any ostomies or fistulas that are resulting
in significant external losses, and then by guessing the amount of third-space
losses that will result from the disease and procedure for the particular patient.
In those patients in whom a significant volume shift is likely, a sizable portion
of the estimate of the fluid requirement is open to error, inasmuch as a guess
can result in either a correct or an incorrect figure. With this in mind, it is
essential that careful hourly monitoring assess the results of the volume
restoration, assuring that an adequate urine output has been achieved, which
will demonstrate adequate volume restoration in the patient. By the same
1198 FILSTON

token, an excessive urine output demonstrates an overabundance of volume

restoration and requires a reduction in the volume of fluid being administered.
Once insensible loss, measured loss, and a guess for third-space require-
ments are available, these incremental volumes are totalled to give a 24-hour
volume for fluid administration, and an hourly rate is calculated by dividing
this volume by 24. The most osmotically effective solution is then chosen to be
run into the patient at this rate, and careful monitoring of urine output tells
one whether the appropriate volume is being used. Too little or no urine output
should result in an increase in the fluid administration rate. Additional hourly
increases in the rate based on urine output will eventually restore the volume
deficit and result in an appropriate urine output. However, the greater rate of
fluid administration will result in consumption of the originally calculated
volume before the end of the 24 hours. Thus, additional osmotically active fluid
will have to be administered to complete the 24 hours: this volume represents
the difference between the estimated ("guessed") third-space fluid volume and
the required or "correct" volume.
Although it is impossible prospectively always to guess correctly the
volume required for third-space restitution, the correct volume can be defined
as that amount of third-space osmotically active fluid that, when added to the
calculated insensible loss and the measured losses and administered over a 24-
hour period, will result in a urine output that is appropriate for the individual
patient. Knowing this, the goal of the prospective program is to work toward
this hourly urine output by adjusting the rate of fluid administration until the
urine output is within the appropriate range.

APPROPRIATE URINE OUTPUT

For infants and young children, a urine output of 40 mLlkg per 24 hours
reflects adequate volume restitution. For an adult, on the other hand, 1200 mLi
day, which represents about twice the amount of urine needed for adequate
clearing of the day's metabolic solutes, should be sufficient. The older child
does not need a greater urine output than an adult, so the formula 40 mLlkg
per day should be used only until the child weighs 30 kg, at which time, the
calculation equals 1200 mL. For the young infant, 40 mLlkg per day is
approximately 1.5 to 2.0 mLlkg per hour. This hourly formula can be utilized
until the child weighs 25 to 30 kg, at which point, the calculation becomes 50
to 60 mLihour (1200 mLlday).

REFINING THE GUESS FOR PATIENTS WITH INTRA-

ABDOMINAL DISEASE AND SURGICAL PROCEDURES

We have shown elsewhere that the guess can be refined by relating the
degree of intra-abdominal trauma or obstructive or inflammatory disease to the
size of the patient. 4 The abdominal cavity is divided into quadrants. An
additional volume of osmotically active solutions will be required that is equal
to one fourth of the maintenance volume for each quadrant that is affected by
either an obstructive or inflammatory disease or a surgical intervention. With
this system, the patient could require a maximum of two times the maintenance
volume in the form of balanced salt solution or more active osmotic solution
(red cells, plasma) if the disease affects the entire peritoneal cavity (four
quadrants) and the surgeon explores the entire abdominal cavity (four quad-
rants).
FLUID AND ELECTROLYTE MANAGEMENT IN THE PEDIATRIC SURGICAL PATIENT 1199

For an individual with simple appendicitis, at most, two quadrants or one

half of the maintenance volume would be required because the disease affects
one quadrant and the surgeon affects one quadrant. On the other hand, a
patient with generalized peritonitis from a perforated appendix would require
four quadrants for the disease (because the entire peritoneal cavity is inflamed)
and, pOSSibly, one quadrant for a simple appendectomy. However, an addi-
tional three quadrants, or a total of four quadrants for the surgical intervention,
would be estimated if the surgeon explores the entire abdomen, runs the
bowel, takes down adhesions, and irrigates out the inflammatory exudates.
Remember that this volume is in addition to the calculated maintenance and
the measured losses and that the solution chosen for this volume restitution
must be an osmotically active one: at a minimum, balanced salt solution (DS%/
LR).
It is essential to realize that although this scheme attempts to refine the
guess, it still yields only a guess. There are many variables, including the state
of the patient's preoperative volume restitution, the degree of inflammation
throughout the peritoneal cavity, the extent of obstruction of the bowel
secondary to the inflammatory component, and the extent and degree of
trauma resulting from the surgical exploration. The amount of time the
inflammatory process has been active and any degree of ischemic injury to the
tissues resulting in loss of capillary integrity and intrinsic volume depletion
will also markedly affect the volumes required for restitution. Thus, although
the scheme is offered as an attempt to refine the guess, it still must be
recognized that careful monitoring and adjustments in the rate of fluid admin-
istration are essential to provide the patient with adequate volume restoration
in a timely fashion, as well as to ensure that an overload does not take place.
An example may help to clarify some of these points.

EXAMPLE

A 4-year-old, 20-kg child has had several days of crampy abdominal pain
that has become more severe and less intermittent. For the last 2 days, he has
been vomiting repeatedly, and the vomitus, which was originally made up of
ingested food and liquids, has become first bile stained and then greenish
brown. His fever has gradually inched up to the 38° to 39°C range, and his
urine output has dropped significantly, with his urine darkening over the past
24 hours. He has become increasingly lethargic and is dizzy when he arises
from bed.
On physical examination, he is found to have a temperature of 38.3°C, his
abdomen is quite tender with guarding throughout, and bowel activity is
depressed. He was thought to have a ruptured appendix and was hydrated
with lactated Ringer's solution until his urine output came up to 38 mLihour
for 2 hours, at which point, he was taken to the operating room and his
abdomen explored. A ruptured Meckel's diverticulum with inflammatory adhe-
sions of bowel loops and a high-grade distal partial small-bowel obstruction
was encountered. The diverticulum was wedged out with a transverse closure
of the bowel, inflammatory adhesions were taken down, and the peritoneal
cavity was lavaged with warm saline and closed without drainage. An incidental
appendectomy was performed.
Question: Do we have evidence that his volume status was reasonably reconstituted
prior to the operative procedure?
Assuming that his 38 mLihour urine output represented adequate initial
1200 FlLSTON

volume restoration, but also assuming that he has both uncorrected long-term
losses and an ongoing inflammatory state that will result in additional fluid
losses, we would calculate his 24-hour volume requirements first by the formula
for maintenance-IOO mLlkg to 10 kg and SO mLlkg to 20 kg, which reaches
our maximum volume of lS00 mL, which we will eventually give as DS%I
quarter NS with 20 to 30 mEq of potassium chloride added per liter. Assuming
he had 100 mL of loss through his nasogastric tube before and during the
operation, he will need an additional 100 mL of fluid in the form of DS%/half
NS plus 30 mEq of potassium chloride per liter. His nasogastric losses must be
monitored continually on at least a 4-hour basis and his measured-loss resto-
ration increased as indicated. Applying the quadrant scheme, his generalized
peritonitis would mandate four quadrants or one additional maintenance
volume. The extent of surgical trauma is open to some debate, but with a
general exploration of the abdominal cavity, running of the bowel, and resection
of the Meckel's diverticulum, at least three or four quadrants would probably
be an appropriate guess. Using four quadrants, we would then have eight
quadrants or two additional maintenance volumes, which we will give in the
form of DS%/LR to this patient who is otherwise in good health and should be
able to tolerate the fluid shifts that will occur when his volume deficit is
restored with a fluid that is deficient in colloid osmotic pressure. He has no
requirement for additional red cell carrying capacity, and we have no infor-
mation that he has any clotting deficit. We could certainly use an albumin-
containing solution, but the question of expense versus benefit must be factored
in.
This child's total fluid administration for the 24 hours following surgery
would then look like this: volume restitution-3000 mL of DS%/LR, mainte-
nance fluid-1S00 mL of DS%/quarter NS plus 30 mEq of KCl, and measured
10ss-100 mL of DS%/half NS plus 3 mEq of potassium chloride with additional
volumes added every 4 hours depending on nasogastric losses. This totals 4600
mL, or approximately 190 mLihour. We should, therefore, begin by adminis-
tering 190 mL of DS%/LR per hour and monitor his urine output. If it is
significantly less than 30 mLihour (1.S mLlkg per hour), we should increase
the rate of volume administration; if it is significantly greater than 40 mLihour
(2 mLlkg per hour), we should decrease the rate of volume administration.
Careful monitoring of urine output and adjustments of rate should allow us to
bring this child into normovolemia and maintain him there in the postoperative
period.
Once a rate is achieved that results in the appropriate hourly urine output,
we can assume that his volume is restored and that he will tolerate the less
osmotically efficient fluids that are reqUired for replacement of insensible losses.
We can then administer fluid in a variety of ways, anyone of which will ensure
that the patient continues to receive the volumes of balanced salt solution for
the remainder of the 24 hours that will maintain his urine output in the
appropriate range. One way is to continue to give only DS%/LR until the rate
of fluid required is down to the maintenance rate, at which point, the fluid is
changed to maintenance fluid (DS%/quarter NS). Another plan is to give a
combination of the calculated insensible losses (the maintenance fluids), the
measured losses, and whatever additional balanced salt solution is required to
maintain the appropriate hourly urine output for the balance of the 24-hour
postoperative period.
In this example of a 4-year-old, 20-kg patient, we started running balanced
salt solution at 190 mLihour and found that after 3 hours, his urine output was
averaging SO mLihour. When the fluid administration rate was slowed to ISO
FLUID AND ELECTROLYTE MANAGEMENT IN THE PEDIATRIC SURGICAL PATIENT 1201

mLihour, his urine output gradually carne down to 38 mLihour and stayed
there. Because this is within the appropriate range, the rate of fluid adminis-
tration for the remainder of the postoperative period should be 150 mLlhour
as long as the urine output remains within the appropriate range. If 190 mL of
D5%/LR were run each hour for 3 hours for a total of 570 mL, and if 150 mL
were run in per hour for the next 3 hours, this would be an additional 450 mL,
or a total of 1020 mL of D5%/LR that had been administered out of our original
"guess" of 3000 mL based on the quadrant system.
However, we have demonstrated that his volume deficit has been corrected,
inasmuch as he is maintaining appropriate urine output at this fluid adminis-
tration rate. We now have 18 hours (24 - 6) during which we will need to
administer his maintenance fluid of D5%/quarter NS plus potassium, which we
calculated at 1500 mL, the 100 mL of nasogastric losses in the form of D5'[o/half
NS, and whatever additional D5%/LR is required to maintain a rate of 150 mLi
hour for the 18 hours. A volume of 150 mLihour for 18 hours is 2700 mL, of
which 1500 mL will be D5%/quarter NS and 100 mL will be D5%/half NS plus
potassium chloride 30 mEq/L. This leaves 1100 mL (2700 - 1600) to be
administered in the form of D5%1LR. The original guess of 3000 mL was off
880 mL (3000 - 2120).
Maintenance fluid of 1500 mL for this child would require a rate of
approximately 65 mLihour. If his urine output is not being maintained with a
rate of 65 mLlhour by the end of the first 24 hours, the child will need additional
third-space fluid in the second 24 hours until the fluid administration rate can
be reduced to 65 mLihour while still maintaining a urine output in the
appropriate range of 30 to 40 mLihour. In all probability, this will be possible,
and over the next 18 hours, we will probably be able to reduce the rate of
administration and thus eliminate more of the D5%/LR.
Should we choose to use balanced salt solution exclusively during the first
24 hours after the operation, this would be perfectly appropriate, and we could
then adjust the rate until it is down to 65 mLihour, the rate required to
administer maintenance fluid. At that point, the patient should be switched to
D5%/quarter NS with added potassium for his maintenance fluid. Failure to do
so, with a continued administration of balanced salt solution over the next 24
hours or more when the deficit has been made up, will result in a relative
deficit of free water as the child's primary loss becomes insensible loss.
Continued administration of isotonic electrolyte solutions will eventually
result in raising his serum osmolality in the form of hypertonic sodium and
chloride balances so that eventually antidiuretic hormone will be secreted, not
because the patient is hypovolemic, but because he is hypertonic or hyperosmotic.
Failure to recognize this would lead to the inaccurate conclusion that the patient
is again hypovolemic, and if this supposed hypovolemia is corrected by
administration of additional volumes of osmotically active solution (D5%/LR),
the situation will be worsened as the patient becomes more hyperosmotic and
continues to secrete antidiuretic hormone and retain fluid. Under these circum-
stances, the patient can become fluid overloaded, resulting in pulmonary
edema. He can be extracted from such a state only by volume restriction and
the administration, at maintenance rates, of D5%/water. This will allow the
slow excretion of his excess electrolytes and the appropriate replacement thereof
with water, bringing him back into balance. Consult the case report in the
article in this issue on "Surgical Management of Children with Hemoglobin-
opathies" for a graphic example of the dire consequences of failure to recognize
these factors in the postoperative patient.
1202 FlLSTON

SPECIAL CONSIDERATIONS: ELECTROLYTE

DEFICIENCY STATES
In the discussion so far, it has been assumed that the patient presents in
electrolyte and osmotic neutrality. It is under these circumstances that the
patient's insensible loss requirements are primarily free water with a minimal
amount of electrolytes required for anticipated losses in the urine and stool. If,
however, the patient presents in a state of hypo-osmolality or hypoelectroly-
temia, that is, with a serum sodium concentration, for instance, of 120 mEq/L
and a serum chloride of 75 mEq/L, additional electrolytes will be required to
correct this deficiency state, and the usual administration for the infant and
child of 3 mEq/kg per day of sodium and chloride per 100 mL of required
maintenance fluid will be insufficient. Failure to utilize the insensible loss-
maintenance fluid volumes to carry in the additional required electrolytes
would necessitate the administration of additional volumes of fluid beyond
those necessary for restoration of losses and anticipated losses and the use of
hypertonic electrolyte solutions with the attendant risk of rapid fluid shifts in
the brain that could result in damage to the brain cells. If presented with a
patient similar to the one noted above, who is 4 years old and weighs 20 kg
and who has electrolytes at the levels noted of sodium 120 mEqlL and chloride
75 mEq/L, appropriate correction would require that the chloride, which is a
total body-fluid ion, be corrected. This could be done by utilizing a combination
of sodium chloride and potassium chloride. For simplicity, we will simply
correct it all here with sodium chloride. If we take the deficit in chloride of
105 - 75 or 30 mEq/L and recognize that chloride is distributed through the
total body water, which is 60% of the body weight, the correction requires:
30 mEq of NaCl x 60% x 20 kg = 30 x 0.6 x 20 = 360 mEq
In addition, he still requires the 3 mEq/lOO mL that would be in his
maintenance fluid for replacement of insensible losses (= 45 mEq/day), giving
a required total of 405 mEq of sodium chloride.
To provide 405 mEq of sodium chloride in a maintenance fluid volume of
1500 mL would necessitate using a hypertonic sodium chloride solution. Giving
1500 mL of normal saline (15.4 mEq/dL) as maintenance for 2 days would
provide 231 mEq each day or 462 mEq of sodium chloride. This would replace
the 360-mEq deficit plus the 45-mEq/day maintenance for 2 days with a slight
excess. If the patient is symptomatic at these hypotonic levels of serum
electrolytes, it would be better to correct about one half of the deficit in 8 hours
using a hypertonic electrolyte solution such as 3% sodium chloride. On the
other hand, if the patient presents in a hypertonic state and is also dehydrated,
the use of dilute electrolyte solutions or D5%/water could produce rapid shifts
of water into the brain cells, which are hyperconcentrated, resulting in brain
edema. This is avoided by restoring the volume deficit with balanced salt
solution and then correcting the residual hypertonicity with hypotonic main-
tenance solutions.
If, on the other hand, the patient is normovolemic and hypertonic (hyper-
osmotic), the fluid given is restricted to maintenance volumes, and these are
administered as electrolyte-free solutions to allow the normal insensible excre-
tion of electrolytes to correct the hypertonic state.

SUMMARY
The following is a quick guide to the perioperative fluid program discussed
in this article.
FLUID AND ELECTROLYTE MANAGEMENT IN THE PEDIATRIC SURGICAL PATIENT 1203

1. Always assess the state of fluid repletion in any patient presenting for surgical
management (Note: This does not necessarily mean operative management).
2. If the patient is hypovolemic or if there is the possibility of hypovolemia and you
are uncertain, restore volumes equal to 25% of the patient's blood volume with a fluid
push made up of an osmotically active electrolyte solution modified for the additional
requirements of red cell carrying capacity or clotting factors. If this results in a urine
output and correction of hypoperfusion or hypotension, maintain an increased fluid
administration program until a stable urine output and good perfusion are achieved. If
the patient is normovolemic at the time of presentation, particularly if the patient is
having an elective operative procedure and does not have an intravenous line in place,
calculate the insensible losses that will occur during the time of fluid restriction before
surgery and correct at least 50% of these during the operative procedure.
3. Develop the postoperative fluid program as a combination of 24-hour insensible
loss replacement (maintenance fluid), restoration of measured losses, and an estimate
(guess) as to the volume'requirements for third-space fluid shifts. Restore blood loss~s if
appropriate or administer additional volumes of balanced electrolyte solution at a 3-to-1
ratio to replace measured blood loss.
4. Total the insensible loss measurement, the measured losses, and the estimate of
third-space requirement and divide this volume by 24 to get an initial hourly fluid
administration rate.
5. Select the most osmotically active fluid that you intend to use and administer it
first at the calculated rate. Carefully monitor the patient's urine output.
6. Increase or decrease the fluid administration rate to bring the hourly urine output
within the guidelines for the appropriate hourly urine output (milliliters) for the particular
patient based on size (kilograms).
7. When the urine output falls within the appropriate range, maintain that rate of
fluid administration, and recalculate the volumes required because of insensible loss,
measured loss, and third-space shifts by subtracting the amount of fluid already
administered from the volume that will be required in the remainder of the 24 hours;
this will yield the volumes of additional maintenance, measured loss, and third-space
fluids that will make up the remainder of the fluids needed for the 24 hours.
8. Alternatively, administer balanced salt solution for the remainder of the 24 hours
or until the rate of administration can be decreased to the rate of insensible-loss
administration for maintenance fluid and then shift to the more dilute maintenance-type
fluids.
9. Do not administer osmotically active fluids (balanced salt solution or blood
products) beyond the point at which the volume requirement falls to the rate of
maintenance fluid requirement except when these products are being given for their
primary use, namely, restoration of oxygen-carrying capacity or clotting factors. In the
latter cases, osmotically active fluids would be administered as additional fluids, beyond
those needed for the restoration of insensible losses, which would still be given as
hypotonic fluids.
10. Recognize that if, inadvertently, osmotically active fluids are given beyond the
appropriate period and the patient's osmolality increases above normal to a hyperelectro-
lyte state, antidiuretic hormone will be secreted. This problem can be corrected only by
the administration of extremely hypotonic fluids, particularly D5%/water.

NUTRITIONAL MANAGEMENT OF THE PEDIATRIC

SURGICAL PATIENT

Nutritional support is an essential part of surgical management. The article

by Walter Chwals, M.D. in this issue should be studied for the current scientific
basis for nutritional decisions. The newborn infant is in the most rapid growth
phase of hislher life. This rapid growth phase levels off toward the end of the
first year and is completed by age 2 years. By then, the child should be on
table foods at regular intervals. Therefore, it is the early infancy period that
requires a program of feeding that differs the most from that of the older child
1204 FlLSTON

and adult. The needs of the older infant and child can be extrapolated from
the program for the newborn.
The newborn infant requires about 120 kcallkg/day for normal growth. By
about 3 months of age, the formula (100 kcal - 3 x age [years]) x weight [kg]
= maintenance kcallday becomes useful. '2 Infant formulas are generally based
on breast milk, which has approximately 20 kca1l30 mL; most standard infant
formulas contain 20 kcallounce. Therefore, it requires 50% more volume to
provide a given number of kilocalories. That is, to provide 100 kcal/kg requires
150 mLlkg of a 20-kcallounce formula. Normal infants can easily tolerate enteral
feedings of 100 to 180 mLlkg; larger volumes may cause the ductus arteriosus
to remain patent in some premature infants. Nutritional formulas are available
that are more concentrated, and most products designed for enteral tube
feeding contain 1 kcallmL.
Infants require 2.5 gm of proteinlkg and an optimal amount of 3 gm of fat!
kg to provide essential fatty acids. An additional guideline' is that the fat
calories should not exceed one third of the total calories. Most commercial
formulas are constructed according to these guidelines.
To provide total parenteral support, start with the 120 kcallkg goal to be
provided in 120 mLlkg:
Protein = 2.5 gmlkg = 10 kcallkg (4 kcal/gm)
Fat = 3.0 gmlkg = 27 kcallkg (9 kcal/gm)
Maximum fat = one third of total calories = 40 kcal.
Begin on the low side at 3 gmlkg = 27 kcal. Subtracting the fat and protein
from the total leaves 120 - 37 = 83 kcal/kg to be provided by carbohydrate.
Glucose provides 3.4 kcallgm; therefore, 24.4 gm of glucoselkg (83 divided
by 3.4) in 120 mLlkg = 20% dextrose. Therefore, the nutrient solution would
contain 20% dextrose and 2.1 gm of protein/IOO mL; the fat is provided by
giving 20% fat emulsion. To provide 3 gmlkg requires 15 mL/kg. This fluid can
be administered over 24 hours or more rapidly to allow a fat-free time for
clearing of lipemia.
The other ingredients of the nutrient solution are electrolytes (sodium,
potassium, chloride, calcium, phosphorus, magnesium), trace elements (zinc,
iron, copper, manganese, chromium, and selenium), and multivitamins. The
recommended amounts of these substances vary with the age of the infant or
child and can usually be provided by the hospital pharmacist or nutritionist.
The Committee on Nutrition of the American Academy of Pediatrics regularly
updates recommendations for nutritional support of preterm and term infants
and children and publishes them in a handbook. Standard infant formulas are
made from complex proteins, carbohydrates, and fats. For infants with short-
bowel syndrome, malabsorptive states, specific enzyme deficiencies, biliary
obstruction, or pancreatic insufficiency, specially designed formulas are avail-
able that utilize partially hydrolyzed proteins, simple sugars, and medium-
chain triglycerides as the protein, carbohydrate, and fat sources, respectively.

References

1. Cervera AL, Moss G: Progressive hypovolemia leading to shock after continuous

hemorrhage and 3:1 crystalloid replacement. Am J Surg 129:670, 1975
2. Committee on Trauma, American College of Surgeons: Advanced Trauma Life
Support: 1988 Core Course, Chicago, American College of Surgeons, 1988, p 61
3. Danielson RA: Differential diagnosis and treatment of oliguria in post-traumatic and
post-operative patients. Surg Clin North Am 55:697, 1975
FLUID AND ELECTROLYTE MANAGEMENT IN THE PEDIATRIC SURGICAL PATIENT 1205

4. Filston He, Edwards CH III, Chitwood WR Jr, et al: Estimation of postoperative fluid
requirements in infants and children. Ann Surg 196:76, 1982
5. Moss GS, Siegel De, Cockin MS, et al: Effects of saline and colloid solutions on
pulmonary function in hemorrhagic shock. Surg Gynecol Obstet 133:53, 1971
6. Poole GV, Meredith JW, Pennell T, et al: Comparison of colloids and crystalloids in
resuscitation from hemorrhagic shock. Surg Gynecol Obstet 154:577, 1982
7. Schuster DP, Lefrak SS: Shock. In Civetta JM, Taylor RW, Kirby RR (eds): Critical
Care. Philadelphia, JB Lippincott, 1988, p 903
8. Shires GT III, Peitzman AB, Albert SA, et al: Response of extravascular lung water
to intraoperative fluids. Ann Surg 197:515, 1983
9. Shires T: Fluid therapy in hemorrhagic shock. Arch Surg 88:688, 1964
10. Shires T: The role of sodium-containing solutions in the treatment of oligemic shock.
Surg Clin North Am 45:365, 1965
11. Shires T, Willims J, Brown F: Acute change in ECF associated with major surgical
procedures. Ann Surg 154:803, 1961
12. Wallace WM: Quantitative requirements of the infant and child for water and
electrolyte under varying conditions. Am J Clin Pathol 23:1133, 1953
13. Wiggers CJ: Physiology of Shock. New York, The Commonwealth Fund, Harvard
University Press, 1950, p 137
14. Wiggers CF: Reminiscences and Adventures in Circulation Research. New York,
Grune & Strutton, 1958, p 368

Address reprint requests to

Howard C. Filston, MD
Department of Surgery, Box U-11
University of Tennessee Medical Center
1924 Alcoa Highway
Knoxville, TN 37920