Preface

Critical care: respiratory focus

Guest Editor

Critical care has become an important facet of many medical and surgical

small animal cases. The specialty continues to grow and to draw practi-
tioners interested in dedicating their careers to intensive case management.
The skilled critical care practitioner can lend an additional perspective to the
management of the seriously ill patient and is often instrumental in aﬀecting
the outcome of these patients.

This issue of the Veterinary Clinics of North America: Small Animal Prac-

tice is the second in a series of two issues covering the topic of Critical Care.
The ﬁrst issue, published in November 2001, focused on topics related to the
cardiovascular system and addressed shock, resuscitation, electrolyte and
acid-base derangements, and patient monitoring. This issue focuses on the
respiratory system and covers respiratory failure, respiratory monitoring,
respiratory pharmacotherapy, and pulmonary support with oxygen and
mechanical ventilation. In addition to the respiratory topics, we have also
included general critical care topics such as nosocomial infection, bacterial
translocation, analgesia, and parenteral nutrition. My hope is that these
two issues will together provide a comprehensive and updated view of small
animal critical care.

Once again, I would like to thank the authors for their excellent contri-

butions. Their willingness to share their knowledge will beneﬁt many critical
patients. I would also like to thank my husband for his consistent support
and encouragement; I could not accommodate projects such as this into

Vet Clin Small Anim 32 (2002) xi–xii

Nishi Dhupa, BVM, MRCVS

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 0 5 3 - 0

my life so gracefully without him. Finally, thanks to John Vassallo at WB
Saunders for his help and editorial expertise.

Nishi Dhupa, BVM, MRCVS

College of Veterinary Medicine

Cornell University

Box31

Ithaca, NY 14853–6401, USA

E-mail address: nd46@cornell.edu

xii

N. Dhupa / Vet Clin Small Anim 32 (2002) xi–xii

Oxygen therapy and toxicity

Ann Marie Manning, DVM

Emergency and Critical Care, The Angell Memorial Animal Hospital,

350 South Huntington Street, Boston, MA 02130, USA

Oxygen (O

) supplementation increases the O

content of blood, increa-

ses the partial pressure of oxygen (P

) in the capillary blood, and improves

tissue delivery of O

. In addition to improving tissue oxygenation, the

administration of O

may improve the function of O

-dependent cellular

systems, such as the cytochrome P

450

system, which is important to drug

and toxin metabolism; nitric oxide synthase, which regulates vasodilation;
and host defense systems. Improved tissue oxygenation is also beneﬁcial for
wound healing. Given the important contributions that supplemental O

can make, it is no wonder that O

is one of the most common drugs admin-

istered in the emergency and intensive care settings.

The physiology of oxygenation

The important steps of oxygenation are O

uptake, diﬀusion, delivery,

and consumption (metabolism). Because not all diseases that cause O

dep-

rivation are responsive to O

therapy, the clinician must possess a basic

understanding of O

physiology, the mechanisms that control O

delivery

to the tissues, and the various circumstances that produce hypoxia. This
understanding enables the clinician to determine which patients are deprived
of O

and how best to correct the situation. The following section is a brief

review of O

physiology followed by the various causes of hypoxia and their

responsiveness to O

supplementation.

Oxygen uptake and diﬀusion

uptake begins with the extraction of O

from the environment during

respiration followed by the movement of O

into the lungs, which serve as

part of the O

delivery system. With inhalation of atmospheric air (room

air), the fraction of inspired O

) is equal to 21%. The P

of inhaled

Vet Clin Small Anim 32 (2002) 1005–1020

E-mail address: amanning@mspca.org (A.M. Manning).

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 0 4 3 - 8

at sea level is equal to 159 mm Hg. The P

decreases progressively as

it traverses the lung, blood, and tissues, and this process is known as the
O

cascade [1]. The P

initially decreases from 159 mm Hg to 149 mm

Hg when inspired air is warmed and humidiﬁed by the mucous membranes
in the nasal passages and upper airways. As air passes into terminal bron-
chioles and alveoli, O

is diluted by carbon dioxide (CO

) and the P

decreases to 99 mm Hg. From the alveoli, O

must diﬀuse across surfactant,

alveolar epithelium (pneumocytes I and II), basement membrane, capillary
basement membrane, and endothelial membrane of pulmonary capillaries
before reaching the erythrocyte [2]. The P

in the alveoli is equal to 99

mm Hg compared with a P

of 40 mm Hg in pulmonary venous blood

[3]. The diﬀerence of 59 mm Hg between alveolar air and venous blood
provides the driving pressure that enables O

diﬀusion to occur.

Oxygen delivery

diﬀuses into the plasma of pulmonary capillaries and then into red

blood cells, where it combines reversibly with the iron atom of hemoglobin
(Hb) and converts deoxyhemoglobin to oxyhemoglobin. Each Hb binds
four O

molecules, and each gram of Hb can transport 1.36 mL of O

when

fully saturated. With this information, it is possible for the clinician to cal-
culate the arterial oxygen content (Ca

), which is the sum of O

dissolved in

plasma and O

chemically bound to Hb. Ca

can be calculated using the

following equation:

Cao

¼ Bound O

þ Dissolved O

¼ ðHb 1:36 Sao

Þ þ ðPao

0:0031Þ

where Sa

is arterial blood oxygen saturation, Pa

is the partial pressure

of arterial oxygen, and 0.0031 is the Bunsen coeﬃcient.

Under optimal conditions, arterial blood with a Pa

of 100 mm Hg and

complete saturation of Hb at a concentration of 15 g/dL would contain 200
mL of O

per liter of blood [1]:

Cao

¼ ð15 1:36 10Þ þ ð100 0:0031Þ

Cao

¼ 200 mL

From this equation, it is clear that the O

dissolved in plasma

(Pa

· 0.0031) contributes little to O

content unless there is increased

cardiac output in the presence of severe anemia or unless hyperbaric O

therapy is used, in which case, the Pa

can be raised substantially. Total

content is thus mainly dependent on Hb concentration and Sa

Hemoglobin and the oxyhemoglobin dissociation curve

Hb is the primary O

carrier in the circulating blood, and myoglobin is

the primary O

reservoir in muscle tissue. Hb has a high aﬃnity for O

1006

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

high O

tension in the lungs and is able to release O

at low O

tension in the

tissues. The oxyhemoglobin dissociation curve (Fig. 1) illustrates the rela-
tion between O

and Hb at varying O

tensions or, more accurately, des-

cribes the saturation of Hb with O

as determined by the concentration of

. The shape of the curve is sigmoidal and results from the conformational

changes in the Hb molecules that occur during uptake and release of O

[3,4]. The ﬂat portion of the curve is an area of high P

(greater than

100 mm Hg). The steep portion of the curve is the physiologic range of
P

(30–100 mm Hg).

The aﬃnity of O

for Hb is inﬂuenced by many physiologic changes.

Decreased aﬃnity of Hb for O

(a right shift in the curve) facilitates the

release of O

to the tissues. A right shift is seen with acidosis, hyperthermia,

exercise, increased partial pressure of CO

), and increased 2,3-diphos-

phoglycerate (2,3-DPG) concentrations in the red cells, as occurs when there
is insuﬃcient O

delivery to the tissues (e.g. anemia, altitude) lasting longer

than 3 to 4 hours. Increased aﬃnity of Hb for O

(left shift) facilitates O

loading into the blood in the alveoli. A left shift is seen with alkalosis, hypo-
thermia, decreased 2,3-DPG levels, carbon monoxide (CO) poisoning, and
hypocarbia [3,4]. Under normal physiologic conditions, near maximum
Hb saturation is achieved at a P

of 75 to 80 mm Hg.

Total oxygen delivery

Total O

delivery (D

) may also be calculated as the product of cardiac

output and Ca

[5]:

¼ Cao

Cardiac Output

¼ 200 mL=L 5 L=min
¼ 1000 mL

Fig. 1. The oxyhemoglobin dissociation curve. Point a represents the ﬂat portion of the curve,
where hemoglobin is maximally bound with oxygen. Point b represents the physiologic range of
the partial pressure of oxygen.

1007

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

Approximately 1000 mL of O

leaves the left ventricle each minute and is

distributed to regional vascular beds. When supply is adequate, O

con-

sumption is a function of metabolic rate and there is a wide margin between
supply and demand. Normal systemic O

consumption is 250 mL/min (25%

of O

transport), and the remaining 750 mL of O

returns to the right heart

in venous blood [3]. Only a small fraction of O

is extracted from the

capillary blood under normal conditions. In situations of low blood ﬂow,
the tissues have the ability to compensate by increasing O

extraction.

Oxygen consumption

Once O

enters the cell, it diﬀuses down a gradient into the mitochondria.

The main location for O

unloading is in the tissue capillaries, and O

de-

livery maintains an interstitial P

of 20 to 40 mm Hg [3]. The mitochon-

dria consume 80% to 90% of the O

, and 10% to 20% is consumed by

subcellular organs. The mitochondria use O

to produce energy through

oxidative phosphorylation.

Oxygen deprivation

There are four physiologic situations that cause cells to convert from

aerobic to anaerobic metabolism with the production of lactic acid:

1. Increased metabolic demand or O

consumption generated by fever,

shivering, seizures, or strenuous exercise may produce acidosis if the
O

supply is unable to meet demand.

2. Hypoxia, cardiac pump failure, and severe anemia create a deﬁcient

supply of O

, resulting in conversion from aerobic to anaerobic metab-

olism.

3. Uncoupling of oxidative phosphorylation, as occurs with cyanide poi-

soning, blocks cellular metabolism and stops O

utilization.

4. Loss of microcirculatory autoregulation, as occurs in sepsis, leaves tis-

sues with inadequate O

to meet metabolic demands.

Central and local mechanisms protect tissue oxygenation via shifts in the

dissociation curve and regulation of local blood ﬂow. Hypoxia results if

these mechanisms fail. Tissue hypoxia occurs if the intracellular P

is less

than10 mm Hg or if the mitochondrial P

is less than 6 to 7mm Hg [6].

Insuﬃcient O

delivery is secondary to hypoxemia or inadequate blood ﬂow.

Indications for oxygen therapy

Clinical signs of oxygen deprivation

The decision to administer supplemental O

is often based on several fac-

tors, including clinical signs, results of diagnostic tests (e.g. arterial blood

1008

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

gas and pulse oximeter measurements), and the patient’s clinical disease.
Clinical signs of hypoxia include cyanosis, dyspnea, tachypnea, tachycardia,
anxiety, postural changes, and central nervous system (CNS) depression [7–
9]. Greater than 5 g/dL of deoxygenated Hb must be present in the circulat-
ing blood before cyanosis can be detected [7,9,10]. Therefore, a patient may
be hypoxic without obvious cyanosis. Tachycardia is a common ﬁnding as
cardiac output is increased through stimulation of carotid body chemore-
ceptors and activation of the sympathetic nervous system in an attempt to
circulate the available pool of Hb more quickly. Postural changes include
abduction of the elbows (orthopneic stance), extension of the head and neck,
recruitment of abdominal muscles, and open-mouth breathing. Patients with
severe hypoxia may experience collapse or mental confusion or may become
comatose.

There are many disease conditions that can produce hypoxia, and recog-

nition of these conditions should prompt the clinician to consider the need
for supplemental O

therapy. The response to O

therapy is variable or non-

existent depending on the cause of hypoxia, however. The following is a dis-
cussion of the types of hypoxia, associated disease conditions, and their
responsiveness to O

supplementation.

Categories of hypoxia

Anoxic hypoxia

Anoxic hypoxia is the inadequate delivery of O

from the lungs to the

blood, resulting in reduced arterial O

content. This form of hypoxia may

result from ﬁve diﬀerent pathophysiologic causes, which include a low F

hypoventilation, diﬀusion impairment, ventilation/perfusion (V/Q) mis-
match, and pulmonary shunt. A P

in inspired gas/air (low F

) causes

diminished arterial O

content and may be caused by inadequate O

delivery

to patients receiving anesthesia or mechanical ventilation, excessive
rebreathing of dead space gas, or high-altitude sickness [1]. This form of
hypoxia can be corrected easily with administration of supplemental O

Alveolar hypoventilation occurs when chest wall excursions are insuﬃ-

cient to achieve adequate lung inﬂation and gas exchange. CO

replaces

in poorly ventilated alveoli, and the Pa

decreases. Alveolar hypoventi-

lation may result from peripheral neuromuscular disease, CNS disorders,
cervical spinal cord lesions, oversedation, thoracic wall defects, rib fractures,
or pleural space disease [8,11]. Supplemental O

administration helps tem-

porarily in this situation, but if the underlying problem cannot be remedied
in a timely fashion, many of these patients require mechanical ventilation to
reduce severe hypoventilation.

Diﬀusion impairment occurs when the alveolar-capillary membrane inter-

face thickens or the surface area available for diﬀusion decreases, as may
occur with interstitial pulmonary edema, pulmonary interstitial ﬁbrosis,
chronic emphysema, and adult respiratory distress syndrome (ARDS) [2].

1009

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

This form of hypoxia is responsive to supplemental O

administration,

because O

administration increases the driving pressure for O

V/Q mismatch occurs when alveolar ventilation and pulmonary blood

ﬂow are not uniform. The degree of V/Q mismatch is dependent on whether
ventilation or perfusion predominates. Pulmonary thromboembolism is an
example of high V/Q mismatch, where regions of the lung are ventilated but
are not perfused, increasing physiologic dead space. This form of V/Q mis-
match responds well to supplemental O

administration. Low V/Q mismatch

is an inadequate supply of O

to alveoli in the presence of normal perfusion

[12]. This type of V/Q mismatch is seen commonly with pulmonary edema,
pulmonary contusions, pneumonia, asthma, and pulmonary neoplasia.
Response to O

supplementation is generally poor to fair in cases of low

V/Q mismatch.

An extreme form of low V/Q mismatch exists with shunts in which per-

fusion occurs in the absence of ventilation. This type of V/Q mismatch
occurs in several disease states, such as severe cardiogenic or noncardio-
genic pulmonary edema, severe pneumonia, lung atelectasis, lung lobe tor-
sion, and right-to-left cardiac shunts. Because the shunted blood makes
no contact with ventilated lung units, supplemental O

administration

provides no beneﬁt. In both types of low V/Q mismatch, however, if mechan-
ical ventilation with positive end-expiratory pressure (PEEP) is provided,
loss of alveolar gas volume may be relieved and supplemental O

may be

eﬃcacious.

Anemic hypoxia and dyshemoglobinemias

Anemic hypoxia results when an inadequate quantity of Hb (<7 g/dL) is

available to transport a suﬃcient supply of O

for metabolism, Hb is present

but rendered nonfunctional by conformational changes in the Hb molecule,
or O

binding to the Hb molecule is blocked. Anemic patients can generally

tolerate Hb concentrations as low as 7 g/dL by increasing cardiac output to
maintain O

delivery. If myocardial function is limited or if O

consumption

exceeds a patient’s ability to increase cardiac output, hypoxemia develops.
Because of limited cardiac reserves, patients with myocardial disease gener-
ally cannot tolerate an Hb concentration below 10 to 11 g/dL. Patients with
anemic hypoxia beneﬁt minimally from supplemental O

therapy, because

their Hb is already fully saturated, and any increase from O

dissolved in

plasma (Pa

· 0.0031) is minor.

Methemoglobinemia (metHb), such as that occurring with acetamino-

phen toxicity, is an example of a conformational change that renders Hb
nonfunctional. Methemoglobin causes the iron component of the heme mol-
ecule to be oxidized from its normal ferrous state (Fe

) to the ferric state

(Fe

), rendering it incapable of O

binding. Patients can generally tolerate

methemoglobin levels up to 15% to 20%, but levels greater than 30%
produce signs of cyanosis, dyspnea, and discoloration of the blood [6].
Methemoglobin levels greater than 80% are often associated with death.

1010

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

A methemoglobin level of 1.5 g/dL creates the same eﬀect as decreasing Hb
by 5 g/dL. O

administration in cases of metHb is generally ineﬀective

because O

binding to Hb is blocked, but some improvement may be seen

as a result of the slight increase in the fraction of O

dissolved in the plasma.

CO poisoning demonstrates how access to the heme molecule may be

blocked. CO has 218 times the aﬃnity for Hb that O

has, and once bound,

carboxyhemoglobin (HbCO) is quite stable [3,4]. The presence of HbCO
shifts the O

Hb dissociation curve to the left, which increases the aﬃnity

of O

for Hb in the lungs but decreases its ability to be released to the tis-

sues, and tissue hypoxia results [4]. Clinical symptoms appear when HbCO
levels exceed 20% and become severe when levels reach 30% to 40% [6]. In-
creasing the concentration of inspired O

adds little O

to Hb but increases

the amount of O

dissolved in plasma. Recommendations in human patients

with CO poisoning include provision of 100% oxygen for 20 to 30 minutes
and/or hyperbaric O

therapy to increase the P

, and thus the driving pres-

sure for O

, which displaces CO from the heme molecule.

Stagnant hypoxia

Stagnant hypoxia results from inadequate O

delivery to tissue because of

low blood ﬂow. Common causes of stagnant hypoxia include pump failure
(cardiogenic shock) and loss of circulating volume as seen with hemorrhage
or signiﬁcant intravascular dehydration. Although O

therapy may be

slightly beneﬁcial in these patients, this form of hypoxia is best corrected
by intravascular ﬂuid resuscitation or improved cardiac function rather than
by O

administration alone [13].

Histiocytic hypoxia

Histiocytic hypoxia results from the cell’s inability to use O

. Prolonged

use of nitroprusside, a balanced vasodilator used in the intensive care set-
ting, may lead to formation of cyanide as a byproduct of its metabolism.
Cyanide poisons the electron transport system so that cells are unable to uti-
lize O

, and metabolic acidosis results. O

administration is of no beneﬁt in

this situation. Rather, therapy should be directed at cyanide removal to cor-
rect the problem.

Modes and techniques for oxygen delivery

A number of devices are available for O

delivery, and the method chosen

should be based on the desired F

, equipment availability, and anticipated

treatment duration as well as the patient’s clinical condition, size, and tem-
perament. Any O

delivery system that causes patient resistance may create

the undesirable side eﬀect of increasing O

demand and consumption.

Table 1 summarizes the various O

delivery systems, the F

that may be

achieved with each system, and the O

ﬂow rate required to reach the target

1011

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

. Each of these systems is discussed in further detail in the following

section.

Flow-by oxygen

This method of O

supplementation is perhaps the easiest in an emer-

gency situation while attempts are being made to stabilize the patient. The
O

line is placed within 1 to 3 cm of the patient’s nose and mouth, creating

a small area of increased F

. An F

of approximately 0.25 to 0.45 may be

achieved in this way with little stress to the patient. The drawbacks to this
method are that it requires a care provider to be present to hold the delivery
system and to ensure that the patient does not move away from the O

source, it requires a high O

ﬂow rate that is potentially wasteful, and it

creates rapid airﬂow that disturbs some patients so that they avoid the O

source.

Face mask

delivery by face mask is a useful short-term method for O

supplemen-

tation in emergency situations. With a well-ﬁtted mask and the O

ﬂow rate

set at 6 to 10 L/min, an F

of 0.35 to 0.55 may be achieved [1,10]. The pres-

ence of a reservoir bag increases the amount of O

available for inhalation

and may raise the F

as high as 0.5 to 0.8 at an O

ﬂow rate of 8 to 10 L/

min [1,5,10]. The disadvantages of this method include gas leakage from
poorly ﬁtted masks (particularly in cats and brachycephalic breeds), poor
elimination of CO

, and lack of patient cooperation.

Nasal catheter

The use of a nasal catheter allows for more prolonged O

delivery and

permits access to the patient for examination and treatment purposes with-
out loss of the O

-rich environment as can occur with use of an O

cage.

A potential drawback to this method of supplementation is that the F

cannot always be determined accurately [11]. O

delivery through a nasal

catheter can achieve an F

of 0.3 to 0.5 with an O

ﬂow rate of 100 to

Table 1
Fraction of inspired oxygen (F

) achieved with various oxygen delivery systems

Target F

Device

Oxygen ﬂow (liters/minute)

0.24–0.45

Flow-by

6–8

0.30–0.50

Nasal catheter

1–6

0.35–0.55

Face mask

6–10

0.40–0.50

Oxygen cage

0.50–0.90

Reservoir mask

5–15

0.21–1.00

Continuous positive airway pressure

10–15

1012

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

150 mL/kg/min (1–6 L/min) [14]. Use of bilateral nasal catheters may
increase the target F

to 0.7. The appropriate O

ﬂow rate is based on

patient size, respiratory rate, respiratory pattern, and degree of open-mouth
breathing. In general, a higher F

is achieved in patients with tachypnea

and low tidal volumes than in those patients with a normal respiratory rate
at the same ﬂow rate. Additionally, each liter per minute increase in O

ﬂow

raises the F

by 3% to 4% [1–5,8,11,13–18].

Placement of the nasal O

catheter is relatively easy. The clinician should

premeasure the catheter the distance from the nostril to the medial canthus
of the eye and clearly mark this point. A small amount of topical anesthetic
(2% lidocaine jelly [Proparacaine]) is introduced into one nostril, and a well-
lubricated, soft, rubber catheter (5–10 French) is passed ventromedially into
the nostril and into the ventral meatus. The catheter is advanced to the pre-
determined level of the medial canthus and secured to the skin at the nares
with adhesive glue, suture, or a single staple. A second site on the forehead
or ventral to the ear is chosen to secure the length of the catheter [14,19]. An
adaptor allows connection of the nasal catheter to the tubing from a humidi-
ﬁed O

source. Placement of an Elizabethan collar is often necessary to pre-

vent the patient from removing the nasal catheter.

Problems with this method of O

delivery include poor patient tolerance,

jet damage to the nasal mucosa, desiccation of the nasal mucosa, and gastric
dilatation [14]. When a unilateral nasal catheter is used, the catheter should
be replaced with a new catheter in the opposite nares every 48 hours so as to
reduce damage to the airway.

Oxygen cage

An O

cage provides a sealed environment, where the F

, humidity, and

ambient temperature can be manipulated in a predictable manner and CO

can be removed eﬃciently [9]. Most commercially available cages are capa-
ble of providing a maximal F

of 0.4 to 0.5 and allow the ambient temper-

ature to be maintained optimally at 22

C (70F) with a relative humidity of

40% to 50% [7,9,10]. Desirable O

cages would have a Plexiglas front to

allow complete visualization of the patient and access ports for entry and
exit of intravenous lines and monitoring leads. Cages that have doors within
doors or plastic sleeves allow manipulation of the patient without creating
enormous O

loss from the cage.

The major advantage of the O

cage is that it is a noninvasive means of

providing O

support to a critically ill animal. The disadvantage of the O

cage is isolation of the patient from the clinician, which precludes frequent
hands-on evaluation. Each time the cage door is opened to access the
patient, loss of the O

-rich environment occurs and the patient may decom-

pensate [14]. The amount of O

required to ﬁll the cage and the amount of

lost with each entry into the cage make this form of O

supplementation

relatively wasteful and expensive.

1013

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

Elizabethan collar canopy

Utilization of an Elizabethan collar covered with plastic wrap may create

a miniature O

-enriched environment. The Elizabethan collar is secured

around the patient’s neck, and the end of an O

line is fed into the collar and

secured. A small hole in the top corner of the plastic facing is created to
allow elimination of CO

and to help modulate the temperature within. This

is a temporary method of O

delivery that can achieve an F

of 0.3 to 0.4

with an O

ﬂow rate of 0.2 to 0.5 L/min [7,8]. Although this method of O

delivery is easy to use in an emergency situation, disadvantages, such as O

leakage, high humidity, hyperthermia, and patient intolerance, preclude its
use in the long term.

Intratracheal catheter

Intratracheal or transtracheal catheters improve O

delivery by bypassing

anatomical dead space and allow for continuous O

delivery at low O

ﬂow

rates [18,20]. This form of O

supplementation can achieve an F

of 0.4 to

0.6 at an O

ﬂow rate of 50 mL/kg/min [7,9,10,18]. Placement of an intratra-

cheal catheter is similar to performing a transtracheal wash. The selected
insertion site is clipped and surgically prepared, and a small bleb of local
anesthetic (2% lidocaine) is instilled at the site. A long, large-gauge, ﬂexible
catheter is used, the end of which is fenestrated before insertion to reduce jet
damage to the trachea. The needle of the catheter is introduced percutane-
ously through the cricothyroid ligament in cats and small dogs or between
two tracheal rings in larger dogs, and the catheter tip is fed to the level of
the ﬁfth intercostal rib space (just cranial to the carina) [19]. The needle is
withdrawn, covered with a needle guard, and secured in place with a neck
wrap, taking care to avoid kinking of the catheter. The end of the catheter
is then attached to a humidiﬁed O

source.

Use of an intratracheal catheter for O

supplementation is inexpensive,

generally well tolerated by the patient, and allows easy access to the patient.
Potential disadvantages include catheter kinking at the insertion site, subcu-
taneous emphysema, jet damage to airway mucosa, tracheitis, broncho-
spasm, infection at the insertion site, and airway obstruction as a result of
excessive mucus accumulation. This is also a more invasive procedure than
placement of a nasal catheter and may require sedation of the patient to
allow placement.

Mechanical ventilation

Mechanical ventilation becomes necessary when a patient is unable to

sustain a Pa

greater than 60 mm Hg through its own eﬀorts, despite con-

ventional O

supplementation [9]. Patient failure may result from respira-

tory fatigue, respiratory arrest, intracranial disease, or hypoventilation of
any cause [8]. Likewise, any patient that is struggling to maintain a Pa

1014

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

between 60 and 65 mm Hg and cannot rest because of respiratory eﬀort
would beneﬁt from mechanical ventilation. Additionally, patients that
require O

supplementation at an F

greater than 0.6 for more than 24

hours are at risk for O

toxicity. Mechanical ventilation with application

of PEEP may allow the F

to be reduced to a safer level, thereby decreas-

ing the risk for toxicity.

Monitoring oxygen therapy

Improvement in the clinical condition of the patient is a suﬃcient means

of determining response to O

therapy in lieu of other diagnostics. Patients

that respond favorably to O

supplementation generally show improved

mucous membrane color, decreased respiratory rate and eﬀort, and reduced
anxiety. Diagnostic tests like arterial blood gas measurements and pulse
oximetry provide more objective means to monitor the eﬃcacy of O

supplementation.

An arterial blood gas measures the amount of oxygen (Pa

) and carbon

dioxide (Pa

) in the arterial blood. This information allows the clinician to

determine the eﬀectiveness of gas exchange and eﬃcacy of supplemental O

therapy. A Pa

less than 70 mm Hg and/or a Pa

greater than 45 mm Hg

signals the need for supplemental O

, whereas a Pa

less than 60 mm Hg

and/or a Pa

greater than 50 mm Hg indicates respiratory failure and the

need for ventilatory support. During O

supplementation, the expected

should be ﬁve times the F

(eg, at an F

¼ 40%, the expected

¼ 40 · 5 ¼ 200 mm Hg). Any value lower than the expected value indi-

cates a problem with gas exchange. An arterial blood gas may be obtained
from the femoral artery, the dorsal metatarsal artery, or, in the anesthetized
patient, the lingual artery. The major drawback to this method of assess-
ment is that arterial blood gases can be diﬃcult to obtain and may be stress-
ful to critically ill patients.

Determination of the alveolar-arterial gradient (A-a gradient) is another

useful calculation that can be made with the information provided by the
arterial blood gas. Calculation of the A-a gradient assesses the severity of
V/Q mismatch. The A-a gradient may be calculated using the following
formula:

A-a Gradient

¼ Pao

Pao

¼ ðFio

½P

1:2 Paco

Þ Pao

where F

is the fraction of inspired oxygen, P

is the barometric pres-

sure, PH

O is the vapor pressure of water, Pa

is the alveolar partial pres-

sure of oxygen, Pa

is the arterial partial pressure of oxygen, and Pa

the arterial partial pressure of carbon dioxide [7,21]. In the normal patient,
where ventilation and perfusion are matched, the A-a gradient is less than
10 mm Hg.

1015

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

Pulse oximetry is a simple noninvasive method for continuous or inter-

mittent monitoring of oxygen saturation (Sa

) or the percentage of oxy-

hemoglobin in the blood. An Sa

less than or equal to 93% indicates the

need for O

supplementation. Measurements are obtained by attaching a

transducer to the lip, ear, and digit or, in the anesthetized patient, via a
rectal or esophageal probe. The accuracy of pulse oximetry may be limited
in many instances, and its use is discussed in further detail elsewhere in this
issue.

Complications of oxygen therapy

Denitrogenation (absorption) atelectasis

In areas of poor oxygenation (with adequate perfusion), alveoli are held

open primarily by nitrogen rather than by O

. Because the body is saturated

with nitrogen, there is no eﬀective nitrogen gradient between mixed venous
(pulmonary arterial) blood and the alveolus. Therefore, nitrogen remains in
the alveolus, preventing its collapse. As O

is administered to the patient, it

displaces nitrogen in the alveolus so that O

volume becomes the main fac-

tor holding the alveolus open. Now, as mixed venous blood ﬂows past the
same alveolus, O

rapidly diﬀuses down its concentration gradient and

enters the blood, leaving behind inadequate amounts of gas to hold the
alveolus open, and the alveolus collapses. Thus, an adequately perfused but
poorly ventilated lung unit becomes both poorly ventilated and poorly per-
fused after O

administration [5,6]. This event is known as denitrogenation

(absorption) atelectasis.

Miscellaneous complications

In patients with chronic respiratory disease and chronic CO

retention,

hypoxemia becomes the main stimulus for ventilation, because the sensitiv-
ity of central chemoreceptors is lost [12]. Supplemental O

administration in

these patients may cause acute deterioration because of suppression of their
respiratory drive. These patients require positive-pressure ventilation rather
than O

supplementation alone.

Other potentially deleterious eﬀects of O

supplementation include de-

creased erythropoiesis, reduced cardiac output, pulmonary vasodilation,
and systemic arteriolar vasoconstriction [16,17,22]. O

toxicity represents

the most serious complication associated with supplemental O

therapy.

Oxygen toxicity

When the administration of O

occurs at levels that exceed biotrans-

formation and clearance, toxicity occurs. O

is a potent drug because it serves

as an eﬃcient electron acceptor in respiration and has a powerful oxidizing
eﬀect. When the mitochondrial and nonmitochondrial metabolism of O

1016

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

saturated, clearance is limited and there is excessive accumulation of toxic
O

intermediates. Toxic O

metabolites include superoxide (O

), hydrogen

peroxide (HOOH), hydroperoxide (ROOH), and hydroxyl radical (HO)
[15,23]. O

radicals are produced at a low level during normal metabolism.

These radicals are removed by the normal defense mechanisms, including
superoxide dismutase, which acts to converts O

to H

, and catalase

and glutathione peroxidase, which clear H

by sequential degradation

to water. Glutathione peroxidase is also able to remove products of lipid
peroxidation [6,23]. Other nonenzymatic antioxidants include vitamin A,
vitamin C, alpha-tocopherol (vitamin E), N-acetylcysteine, b-carotene, urate,
bilirubin, and hemoglobin [1,15,23].

During hyperoxia, the normal defense mechanisms are overwhelmed and

intermediates accumulate. The consequences of O

toxicity include lipid

peroxidation of cell membranes with loss of cell integrity, oxidation of sulf-
hydryl groups, alteration of enzyme function, protein structural damage,
and impairment of transcription and replication of RNA resulting in defects
of DNA cross-linking and nucleic acid damage [5,23]. The lungs are partic-
ularly sensitive to the eﬀects of lipid peroxidation, and signs of pulmonary
damage predominate in O

toxicity.

Pathophysiology of pulmonary oxygen toxicity

Initially, hyperoxia causes endothelial cell damage and destruction of

alveolar lining cells, increasing microvascular permeability [5]. Damage to
the endothelium allows inﬂammatory precursors to enter the pulmonary
interstitium, leading to alveolar edema, hemorrhage, and congestion. Poly-
morphonuclear cells adhere to the endothelial cells and generate chemotac-
tic factors to attract more inﬂammatory cells. These early changes represent
an exudative phase of pulmonary O

toxicity. The early stages of pulmonary

damage are characterized by proliferation of alveolar type I epithelial cells, a
ﬁbrin exudate, and a prominent alveolar membrane [17]. Blood is shunted
away from areas of the lung where oxygenation is inadequate, and V/Q mis-
match develops.

In the late stages of O

toxicity, alveolar type I epithelial cells are lost,

alveolar type II epithelial cells and ﬁbroblasts proliferate, the basement
membrane is denuded, and ﬁbrosis results. These changes represent a prolif-
erative phase of pulmonary O

toxicity [5,17]. Absorption atelectasis occurs

secondary to inactivation of surfactant when O

is rapidly taken into the

pulmonary blood, thus worsening V/Q mismatch.

Diagnosis of oxygen toxicity

Clinical signs of O

toxicity include tachypnea, dyspnea, nasal conges-

tion, and a cough resulting from tracheobronchitis. Human patients experi-
ence chest pain, paresthesias, and anorexia in the early stages of O

toxicity

[17]. Diagnosis of O

toxicity is diﬃcult and is based on ﬁndings of edema or

1017

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

an inﬁltrative lung pattern on radiographs, a worsening gas exchange man-
ifested as a progressively decreasing Pa

, and evidence of V/Q mismatch.

Treatment of oxygen toxicity

There is no eﬀective treatment for O

toxicity; therefore, prevention is

most important. To decrease the risk for O

toxicity, the clinician should

adhere to the following guidelines:

1. Use a Pa

of 70 as an endpoint for O

therapy.

2. Use the lowest F

possible to achieve a Pa

of 70.

3. Do not use an F

greater than 0.6 for longer than 24 hours.

4. Use PEEP to help decrease the F

as necessary. The use of PEEP

maximizes the Pa

at lower F

levels.

Most injury is sustained with exposure to 24 to 48 hours of an F

at 0.6

to 1.0 [5,19]. ARDS patients are more susceptible to O

toxicity, because

toxicity accelerates lung ﬁbrosis; however, any patient with underlying lung
injury is at higher risk for O

toxicity than is a patient with normal lungs. In

these patients, O

toxicity may manifest at an F

less than 0.6 and/or when

supplemental O

therapy has been provided for less than 12 to 24 hours [5].

Certain drugs may oﬀer protection against O

toxicity, such as vitamin E,

vitamin C, b-carotene, mannitol, and N-acetylcysteine, all of which act as
antioxidants [15]. Deferoxamine, an iron chelator, inhibits production of
hydroxyl radicals by preventing activation of the Haber-Weiss reaction.
Conversely, certain drugs have been shown to enhance O

toxicity through

their ability to increase tissue O

consumption, cause increased production

of free radicals through their metabolism, or decrease protective antioxidant
systems. Such drugs include epinephrine, norepinephrine, steroids, cyclo-
phosphamide, thyroid hormone, and nitrofurantoin. The eﬀect of steroids
on the development of O

toxicity and their use in the treatment of O

tox-

icity is controversial. Some reports indicate that steroids like dexamethasone
decrease the levels of antioxidant enzymes and therefore potentiate injury,
whereas other reports suggest that high doses of steroids are therapeutic
in the late stages of O

toxicity [6]. Based on this limited information,

steroids should be avoided except in the late stages of toxicity.

Future directions

Future directions for O

therapy include the use of O

-carrying Hb solu-

tions. Hb-based O

carriers administered intravascularly bind O

in the

lungs and deliver O

to the tissues in exchange for CO

. Hb solutions have

been shown to improve tissue O

tension in cases of anemic hypoxia and

stagnant hypoxia at room air (F

¼ 0.21). The use of Hb-based O

carriers

may reduce or completely eliminate the need for supplemental O

at high

during conditions of anemic and stagnant hypoxemia, thus reducing

the risk of O

toxicity.

1018

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

Partial liquid ventilation with the use of perﬂuorocarbons is also an area

of intensive ongoing investigation. Perﬂuorocarbons are inert liquids that
are used to ﬁll the pulmonary airspace and are employed in conjunction with
mechanical ventilation or tracheal gas insuﬄation to improve oxygenation
[24,25]. The advantages of perﬂuorocarbons are numerous and include
facilitated opening of collapsed and noncompliant lung segments, reduced
oxidative damage in acute lung injury, and diminished shear forces acting
on lung parenchyma [24–26]. Perﬂuorocarbons may also function as ‘‘liquid
PEEP’’ by preventing complete collapse of unstable alveoli at low airway
pressures [24]. Use of partial liquid ventilation would also allow for lower
F

in patients at risk for O

toxicity. Although partial liquid ventilation

is still under investigation, it represents an exciting new direction for O

therapy.

References

[1] ErmakovS, Hoyt JW. Oxygen therapy and pharmacologic modulation of respiratory drive.

In: Chernow B, editor. The pharmacologic approach to the critically ill patient. 3rd edition.
Baltimore: Williams & Wilkins; 1994. p. 579–604.

[2] Guyton AC, Hall JE. Physical principles of gas exchanges; diﬀusion of oxygen and carbon

dioxide through the respiratory membrane. In: Textbook of medical physiology. 10th
edition. Philadelphia: WB Saunders; 2000. p. 452–62.

[3] Guyton AC, Hall JE. Transport of oxygen and carbon dioxide in the blood and body

ﬂuids. In: Textbook of medical physiology. 10th edition. Philadelphia: WB Saunders;
2000. p. 463–73.

[4] Hsia C. Respiratory function of hemoglobin. N Engl J Med 1998;38:239–47.
[5] O’Connor BS, Vender JS. Oxygen therapy. Crit Care Clin 1995;11:67–78.
[6] Silverman HJ. Pharmacologic approach in patients with pulmonary failure. In: Chernow B,

editor. The pharmacologic approach to the critically ill patient. 3rd edition. Baltimore:
Williams & Wilkins; 1994. p. 114–38.

[7] Camps-Palau AM, Marks SL, Cornick JL. Small animal oxygen therapy. Compend Contin

Educ Pract Vet 1999;21:587–98.

[8] Murtaugh RJ. Acute respiratory distress. Vet Clin North Am Small Anim Pract 1994;

24:1041–55.

[9] Pascoe PJ. Oxygen and ventilatory support for the critical patient. Semin Vet Med Surg

Small Anim 1998;3:202–9.

[10] Drobatz KJ, Hackner S, Powell S. Oxygen supplementation. In: Bonagura JD, editor.

Kirk’s current veterinary therapy XII. Philadelphia: WB Saunders; 1995. p. 175–9.

[11] March PA. Neural regulation of respiration physiology and pathophysiology. Probl Vet

Med 1992;4:387–404.

[12] West JB, editor. Ventilation-perfusion relationships. In: Respiratory physiology: the

essentials. 4th edition. Baltimore: Williams & Wilkins; 1995. p. 51–68.

[13] Hayes MA, Timmins AC, Yau EHS, et al. Elevation of systemic oxygen delivery in the

treatment of critically ill patients. N Engl J Med 1994;330:1717–22.

[14] Fitzpatrick RK, Crowe DT. Nasal oxygen administration in dogs and cats: experimental

and clinical investigations. J Am Anim Hosp Assoc 1986;22:293–300.

[15] Halliwell B, Gutteridge JM. Oxygen toxicity, oxygen radicals, transition metals and

disease. J Biochem 1984;219:1–14.

[16] Haque WA, Boehmer J, Clemson BS, et al. Hemodynamic eﬀects of supplemental oxygen

administration in congestive heart failure. J Am Coll Cardiol 1996;27:353–7.

1019

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

[17] Jackson RM. Oxygen therapy and toxicity. In: Shoemaker WC, Ayres SM, Grenvik A,

et al., editors. Textbook of critical care. 3rd edition. Philadelphia: WB Saunders; 1995.
p. 784–9.

[18] Mann FA, Wagner-Mann C, Allert JA, et al. Comparison of intranasal and intratracheal

oxygen administration in healthy awake dogs. Am J Vet Res 1992;53:856–60.

[19] Court MH. Respiratory support of the critically ill small animal patient. In: Murtaugh RJ,

Kaplan PM, editors. Veterinary emergency and critical care medicine. St. Louis: Mosby
Year Book; 1992. p. 575–92.

[20] Tarpy SP, Celli BR. Long term oxygen therapy. N Engl J Med 1995;333:710–4.
[21] West JB. Gas exchange. In: Pulmonary pathophysiology: the essentials. 4th edition.

Baltimore: Williams & Wilkins; 1995. p. 18–40.

[22] Rubanyi GM, Vanhoutte PM. Superoxide anions and hyperoxia inactivate endothelium

derived relaxing factor. Am J Physiol 1986;250:H22–7.

[23] Otto CM. Oxidative stress: too much oxygen. Proceedings of International Veterinary

Emergency Critical Care 1998;6:523–8.

[24] Blanch L, Van der Kloot TE, Youngblood AM, et al. Selective tracheal gas insuﬄation

during partial liquid ventilation improves lung function in an animal model of unilateral
acute lung injury. Crit Care Med 2001;29:2251–7.

[25] Steinhorn DM, Papo MC, Rotta AT, et al. Liquid ventilation attenuates pulmonary

oxidative damage. J Crit Care 1999;14:20–8.

[26] Overbeck MC, Pranikoﬀ T, Hirschl RB. Partial liquid ventilation provides eﬀective gas

exchange in a large animal model. J Crit Care 1996;11:37–42.

1020

A.M. Manning / Vet Clin Small Anim 32 (2002) 1005–1020

Pulse oximetry and end tidal carbon

dioxide monitoring

Tim B. Hackett, DVM, MS

Emergency and Critical Care Medicine, Department of Clinical Sciences,

Colorado State University, Fort Collins, CO 80523, USA

Critical Care Unit, Veterinary Teaching Hospital, 300 West Drake Road,

Colorado State University, Fort Collins, CO 80523, USA

Advanced, noninvasive, patient-side monitors are providing clinicians

with more ways to follow physiologic variables when managing complex
cases. Pulse oximeters and capnographs provide information on two sepa-
rate aspects of the respiratory system, gas exchange and ventilation. With
newer and less expensive electronics, these monitors are becoming increas-
ingly popular with veterinarians who are eager to provide the best care for
their patients and clients. Clinicians preparing to use the machines should
have a complete understanding of the mechanics and physiology behind the
monitors and a thorough understanding of their limitations.

The tremendous variations in color visible all around us occur because

light reﬂected by diﬀerent atoms and molecules have diﬀering wavelengths.
Spectrophotometry uses the light reﬂection properties of molecules to meas-
ure the concentration of chemical species in liquid or gaseous environments
[1]. Applied to the relative oxygenation of hemoglobin molecules, the pro-
cess is termed oximetry. When oximetry is applied to pulsatile (arterial)
blood, the term is pulse oximetry. Similarly, when spectrophotometry is
applied to the concentration of carbon dioxide (CO

) in gas, it is called cap-

nometry. When applied to a single exhaled breath, it is called capnography [1].

Pulse oximetry, an estimate of arterial oxygen saturation (SaO

), provides

a safe and simple method of assessing patient oxygenation in a continuous,
noninvasive manner [2,3]. End-tidal carbon dioxide (ETCO

) measured by

capnography can reliably estimate ventilation [4,5]. Continuous monitoring
of ETCO

and SaO

has been considered the minimal standard of care in

human anesthesia since 1985 [6,7]. While these monitors provide continuous
data and can detect sudden changes in respiratory function, they are most

Vet Clin Small Anim 32 (2002) 1021–1029

E-mail address: tim.hackett@colostate.edu (T.B. Hackett).

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 -5 6 1 6 ( 0 2 ) 0 0 0 4 2 -6

eﬀective when combined with careful observation and invasive crosschecks
like arterial blood gas analysis [5,8,9]. Despite the numbers and sophistica-
tion of the equipment, there is no substitute for frequent careful physical
examination.

Pulse oximetry

The pulse oximeter calculates the saturation of hemoglobin (SaO

) using

the principle of spectrophotometry. An oxygenated hemoglobin molecule
(oxyhemoglobin) and a reduced or deoxygenated hemoglobin (deoxyhemo-
globin) molecule absorb light diﬀerently. Oxyhemoglobin absorbs infrared
light whereas visible red light passes through the molecule. Deoxyhemo-
globin absorbs visible red light whereas the infrared light passes through.
The spectrophotometer shines red light (wavelength 660 nm) and infrared
light (wavelength 920 nm) through an arterial bed. A photodetector oppo-
site the two light-emitting diodes produces a current that is directly propor-
tional to the intensity of the two wavelengths of light reaching it. A
microprocessor computes the arterial hemoglobin oxygen saturation by
comparing the relative amounts of light of both wavelengths reaching the
photodetector. The pulse oximeter uses the two-wavelength technology to
calculate functional oxygen saturation [1]. By limiting readings to regions
of pulsatile ﬂow (plethysmography), pulse oximeters report functional oxy-
gen saturation of arterial blood (SaO

). To perform this task accurately, the

machine needs a good waveform [10]. A pulse oximeter with a pulse rate and
waveform display allows the operator to judge the quality of the signal. The
reported pulse rate should correspond to the patient’s heart rate and the
waveform should look like a normal arterial waveform.

Oxygen binds reversibly with hemoglobin using this blood protein as an

eﬃcient vehicle for transporting oxygen to the tissues. A hemoglobin mole-
cule is considered ‘‘saturated’’ when it has bound four oxygen molecules.
When the ﬁrst oxygen molecule is bound, the remaining sites ﬁll rapidly. The
hemoglobin is then referred to as oxyhemoglobin. A hemoglobin molecule
that is not carrying anything on its four binding sites is referred to as
reduced or deoxyhemoglobin. Saturation of hemoglobin in arterial blood
(Sa) is the ratio of the number of oxyhemoglobin molecules to the total
number of hemoglobin molecules available to bind oxygen. More precisely,
this percent is referred to as functional saturation.

Functional Saturation

¼ ½oxyhemoglobin=½deoxyhemoglobin

þ ½oxyhemoglobin:

The aﬃnity of hemoglobin for oxygen produces a sigmoid curve that can

be described as follows. The ﬂat upper portion describes oxygen loading of

1022

T.B. Hackett / Vet Clin Small Anim 32 (2002) 1021–1029

hemoglobin as blood passes through the lungs. When oxygen tensions are
high in the blood, additional loading of oxygen onto hemoglobin is not sig-
niﬁcantly increased because most of the hemoglobin molecules are already
saturated. The steep lower portion of the sigmoid curve describes the rela-
tion at the tissue level. Hemoglobin molecules in this portion are not as well
saturated because they have already given up some of their oxygen to tis-
sues. Large amounts of oxygen are being oﬄoaded from hemoglobin mole-
cules, even with small changes in blood oxygen tensions. The shape of the
oxyhemoglobin dissociation curve presents physiologic advantages. When
deoxygenated blood reaches the lungs, the hemoglobin molecules are
quickly saturated and the curve plateaus where moderate lung pathology
or decreased inspired oxygen concentrations (high altitude) will not alter
normal arterial oxygen content. This plateau is responsible for one of the
most serious limitations of pulse oximetry. When the SaO

exceeds 90%

(corresponding to a PaO

of >60 mm Hg) the oxyhemoglobin curve is rela-

tively ﬂat, and large changes in PaO

are associated with small changes in

SaO

. Thus, SaO

is not a sensitive marker of changes in pulmonary gas

exchange in the SaO

normal range [1]. It is, however, a good, continuous

indicator of rapidly developing hypoxemia [7].

Studies have shown that pulse oximeters can produce information that is

useful for managing critical small animal patients [8,11]. Pulse oximeters
correlate well with arterial saturation measured in arterial blood samples
(r

¼ 0.84) [10]. The accuracy of pulse oximetry becomes unreliable at SaO

values below 85% [1,12]. Despite this limitation, pulse oximetry is a stand-
ard-of-care method for recognizing early hypoxemia [6]. The SaO

is nor-

mally above 94% and should always be maintained above 90%. An SaO

of 90% at normal pH corresponds to a PaO

of 60 mmHg. As the SaO

drops below 90%, the PaO

drops precipitously and the patient develops

life-threatening hypoxemia.

In the anesthetized or critically ill patient, a higher SaO

(closer to 100%)

is more desirable. Along with cardiac output and hemoglobin concentration,
SaO

is a critical variable in determining total oxygen delivery to the tissues.

Because tissue oxygen delivery is the ultimate goal when managing shock or
impaired perfusion, the clinician must optimize these variables with ﬂuids,
hemoglobin sources, and supplemental oxygen.

There are distinct advantages to the use of pulse oximetry over arterial

blood gas analysis for monitoring oxygen saturation: it is non-invasive and
continuous, it carries less morbidity and patient discomfort, it increases the
detection of hypoxemic episodes, and it is less expensive [2,7,13]. SaO

meas-

ured by pulse oximetry is actually more accurate than the nomogram-
derived values generated from arterial blood gas analysis. There are several
clinically important limitations to pulse oximetry (Table 1). Some of these
can be changed (modiﬁable limitations) while others cannot (non-modiﬁable
limitations).

1023

T.B. Hackett / Vet Clin Small Anim 32 (2002) 1021–1029

Capnography

A major product of tissue metabolism, CO

is transported to the lungs as

a bicarbonate ion where it is eliminated from the body as CO

. CO

is an

extremely soluble gas and freely diﬀuses from the blood into the alveoli
along a partial pressure gradient. If alveolar ventilation is inadequate,
CO

accumulates in the lungs. As the pressure gradient between blood and

alveolar gas equalizes, CO

accumulates in the blood. Prolonged hypercap-

nia (PCO

50 mm Hg) can result in severe respiratory acidosis and serious

organ dysfunction. CO

is the most potent cerebral vasodilator [14].

Increased intracranial pressure with resulting brain damage is the most seri-
ous complication of prolonged hypercapnia. Ventilator patients, anesthe-
tized patients, and animals that have neuromuscular weakness must have
their CO

levels followed closely to correct serious hypoventilation before

irreversible complications develop [7].

A hyperventilating patient will increase the amount of fresh air in the

alveoli, which increases the pressure gradient to CO

, resulting in a drop

in blood CO

. The partial pressure of CO

(PCO

), measured directly by

Table 1
Modiﬁable and non-modiﬁable limitations of pulse oximetry

Modiﬁable

Non-modiﬁable

Optical interference: bright environmental

light sources (surgical lamps, ﬁberoptic
lights, direct sunlight)

Dyshemoglobinemias: carboxyhemoglobin

and methemoglobin will falsely elevate the
SpO

and not reﬂect true arterial oxygen

saturation. The degree of error will be
proportionate to the degree of
dyshemoglobinemia. These
dyshemoglobinemias can be quantitated
using a co-oximeter (device transmitting
four wavelengths of light through a blood
sample, capable of quantitating
methhemoglobin and carboxyhemoglobin).

Optical shunting: Portion of the LED light

reaches the photodetector without passing
through the pulsating vascular bed

Intravenous dyes: methylene blue and

ﬂuorescein for example can cause falsely
low SpO

Low perfusion signals: hypotension (blood

pressure <30 mm Hg), hypothermia, or
elevated systemic vascular resistance

Hyperbilirubinemia: has been reported to

cause falsely low and high values.

Motion artifact: shivering, agitation,

seizures, tremors

Severe anemia: hemodilution (Hb

concentration <3 g/dL) might limit pulse
oximetry with inadequate hemoglobin
concentrations.

Multiple pulsatile states: right-sided heart

failure, tricuspid insuﬃciency, pericardial
eﬀusion, high levels of positive end-
expiratory pressure ventilation, and
arteriovenous malformations

1024

T.B. Hackett / Vet Clin Small Anim 32 (2002) 1021–1029

blood gas analysis, gives important information about the adequacy of
alveolar ventilation.

Capnography uses spectrophotometry to measure the maximal concen-

tration of CO

in expired gas. CO

will absorb certain wavelengths of infra-

red light. By channeling a small sample of gas from the conducting airways
between an infrared light source and a photodetector, the capnograph
directly measures the concentration of CO

in the gas sample [1]. At the end

of expiration, when the gas reaching the capnograph is coming from all the
alveoli, the concentration of CO

is reported as the ETCO

(Fig. 1). As with

complete arterial waveforms displayed on quality pulse oximeters, the better
capnographs have a graphic display showing the entire ventilatory cycle. By
measuring and graphing CO

concentrations throughout a ventilatory cycle,

the clinician can decide if the reported ETCO

closely approximates PaCO

Changes in the capnograph waveform can signal a variety of problems
including circulatory shock (Fig. 2), rebreathing CO

(Fig. 3), a leak in the

breathing circuit (Fig. 4), or airway obstruction (Fig. 5).

Most capnographs today are called sidestream monitors. Tubing runs

from a sampling port that is positioned near the opening of the endotracheal
tube to the monitor that houses the infrared light source and photodetector.
The sidestream monitor draws 125 to 500 mL/min

. This volume gas is neg-

ligible in most patients, but it can cause hypoxia in small patients who are
anesthetized using a low-ﬂow closed circuit.

End-tidal CO

pressure is the partial pressure of CO

in the expired air at

the end of expiration. This value approximates arterial CO

tension assum-

ing CO

in pulmonary capillaries is in equilibrium with alveolar gas and

tidal volumes are large enough to displace dead space. ETCO

depends

Fig. 1. A normal single-breath capnogram. At the beginning of expiration (A), the graph begins
at zero and quickly plateaus (B,C). The plateau phase represents the sum of all alveolar gas,
assuming similar ventilation–perfusion relationships, and it is normally relatively ﬂat. The
measured PETCO

is determined at the end of the plateau phase (C). It is represented on the

monitor in this example as 37 mm Hg. As inspiration begins (C) the graph quickly returns to
zero (D).

1025

T.B. Hackett / Vet Clin Small Anim 32 (2002) 1021–1029

on the rate of CO

production, the eﬃciency of CO

exchange at the alveolar

capillaries, and the rate of CO

removal by alveolar ventilation. The

PaCO

–ETCO

diﬀerence increases with increasing alveolar dead space and

states of poor perfusion [1]. In low-ﬂow states like circulatory shock and car-
diac arrest, inadequate perfusion results in less CO

carried to the alveoli for

elimination [5,15]. With assisted ventilation, the alveoli are quickly cleared
of CO

and the ETCO

drops.

With a basic understanding of the physiology, clinicians can estimate

metabolism, circulation, and ventilation using ETCO

. ETCO

generally

requires patients to be intubated. For this reason, capnography is typically

Fig. 2. With a normal single-breath capnogram waveform but progressively dropping PETCO

(12 mm Hg in this example), the patient is either severely hyperventilating or has extremely poor
perfusion. In the ventilated patient, the capnograph will quickly drop to zero if the patient
suﬀers complete cardiac arrest.

Fig. 3. Single-breath capnogram. The patient is rebreathing CO

with a PETCO

reported as 43

mm Hg. Notice that at baseline (during inspiration through the beginning of expiration) the
graph is greater than zero. Rebreathing carbon dioxide could be the result of low fresh gas ﬂow
or a faulty/exhausted CO

scavenging system.

1026

T.B. Hackett / Vet Clin Small Anim 32 (2002) 1021–1029

used in anesthetized patients. Sampling tubes have been inserted into the
nasal passages of awake dogs and have been show to reliably measure
ETCO

when animals’ mouths were held shut [5].

Clinical applications of capnography

There are ﬁve major clinical applications of capnography.

1) Ensuring that an endotracheal tube or mask ventilates the lungs. CO

concentrations of gastric gas and inspired gas are negligible. CO

eliminated in large quantities only through the lungs. An end-tidal mon-
itor attached to an endotracheal tube correctly placed will therefore de-
tect exhaled CO

(assuming the lungs are being perfused) [7].

Fig. 4. The gradual rate of increase in the PETCO

graph suggests an obstruction in the airway

caused by partial airway obstruction or a kinked endotracheal tube.

Fig. 5. In this example the lack of alveolar plateau could be due to tachypnea or a leak in the
breathing circuit. In any event, the lack of plateau should alert the clinician that the reported
PETCO

(25 mm Hg) is probably lower than the actual PCO

1027

T.B. Hackett / Vet Clin Small Anim 32 (2002) 1021–1029

2) Estimating of PaCO

in stable patients with normal systemic and pul-

monary perfusion. Capnography is especially useful in continuously
monitoring patients who are being ventilated or in assessing the need
for mechanical ventilation in patients with neuromuscular disease [5].

3) Reﬂecting change in pulmonary blood ﬂow or dead space ventilation [1].
4) Detecting addition of excess CO

to systemic circulation, such as in ma-

lignant hyperthermia [1].

5) Monitoring eﬃcacy of cardiopulmonary resuscitation. As CPR results

in forward blood ﬂow and reperfusion of ischemic tissues, CO

will be

returned to the alveolar-capillary membrane. With adequate ventilation,
the capnograph will show a progressive rise in PETCO

[15].

Summary

Noninvasive monitoring of cardiopulmonary function through pulse oxi-

metry and capnography provides immediate and important information for
the clinician. These monitors are not a replacement for vigilant attention to
the patient, however; they should be used in conjunction with arterial blood
gas analysis and serial physical examinations to ensure that the continuous
readings are accurate and make clinical sense.

References

[1] Marino PL. Oximetry and capnography. In: The ICU book, 2nd edition. Baltimore, MD:

Williams & Wilkins; 1998. p. 355–70.

[2] Fairman N. Evaluation of pulse oximetry as a continuous monitoring technique in

critically ill dogs in the small animal intensive care unit. J Vet Emerg Crit Care 1992;2:50–6.

[3] Wahr JA, Tremper KK. Noninvasive oxygen monitoring techniques. Crit Care Clin 1995;

11:199–217.

[4] Grosenbauch DA, Muir WW. Using end-tidal carbon dioxide to monitor patients. Vet

Med 1998;93:67–74.

[5] Hendricks JC, King LG. Practicality, usefulness, and limits of end-tidal carbon dioxide

monitoring in critical small animal patients. J Vet Emerg Crit Care 1994;4:29–39.

[6] American Society of Anesthesiologists. Standards for intra-operative monitoring. Park

Ridge (IL): American Society of Anesthesiologists; 1991.

[7] Wright B, Hellyer PW. Respiratory monitoring during anesthesia: pulse oximetry and

capnography. Comp Cont Ed Pract Vet 1996;18:1083–97.

[8] Hendricks JC, King LG. Practicality, usefulness, and limits of pulse oximetry in critical

small animal patients. J Vet Emerg Crit Care 1993;3:5–12.

[9] Proulx J. Respiratory monitoring: arterial blood gas analysis, pulse oximetry, and end-tidal

carbon dioxide analysis. Clin Tech Small Anim Pract 1999;14:227–30.

[10] Severinghaus JW, Spellman MJ. Pulse oximeter failure thresholds in hypotension and

vasoconstriction. Anesthesiology 1990;73:532–7.

[11] Jacobsen JD, Miller MW, Mathews NS, et al. Evaluation of accuracy of pulse oximetry in

dogs. Am J Vet Res 1992;53:537–40.

[12] Shoemaker WC, Belsberg H, Wo CCJ, et al. Multicenter study of noninvasive monitoring

systems as alternatives to invasive monitoring of acutely ill emergency patients. Chest 1998;
114:1643–52.

1028

T.B. Hackett / Vet Clin Small Anim 32 (2002) 1021–1029

[13] Grossenbaugh DA, Muir WW. Pulse oximetry: a practical, eﬃcient monitoring method.

Vet Med 1998;93:60–6.

[14] Van Poznak A. Special considerations for veterinary neuroanesthesia. In: Short CE, editor.

Principles and practice of veterinary anesthesia. Baltimore, MD: Williams & Wilkins; 1987.
p. 180–2.

[15] Levine RL, Wayne MA, Miller CC. End-tidal carbon dioxide and outcome of out-of-hos-

pital cardiac arrest. N Engl J Med 1997;337:301–6.

1029

T.B. Hackett / Vet Clin Small Anim 32 (2002) 1021–1029

Blood gas analysis

Thomas K. Day, DVM, MS

Louisville Veterinary Specialty and Emergency Services, 12905 Shelbyville Road,

Suite 3, Louisville, KY 40243, USA

Analysis of blood gases from an artery or central vein can provide valua-

ble information on the cardiopulmonary and acid-base status of a critically
ill veterinary patient. Arterial blood samples provide information primarily
regarding pulmonary function, whereas a jugular or mixed venous sample
can provide information on overall cardiac performance (perfusion) and
whole-body acid-base status. Traditionally, arterial samples have been used
primarily to assess blood gases and make judgments regarding perfusion
and overall metabolic status of the tissues. There has been recent interest
and evidence suggesting that central venous blood (jugular or pulmonary
artery) can be used to monitor trends in overall perfusion and acid-base sta-
tus in small animals with cardiovascular compromise. There also is evidence
in human medicine that jugular venous oxygen saturation (Sj

) can pro-

vide evidence regarding cerebral oxygenation when combined with other
oxygen delivery variables.

Sample collection

The ﬁrst step of blood gas analysis is obtaining an appropriate sample.

Equipment that can be used consists of a 1- or 3-mL syringe with a small-
gauge needle (23–29 gauge). A small amount of heparin should be collected
to coat the hub of the needle. Too much heparin can result in dilution of the
sample and the possibility of inaccurate results, such as lowpH, lowP

and lowbicarbonate (HCO

) [1]. The sample needs to be stored in an anae-

robic environment as soon as collection is completed. Any air bubbles with-
in the sample should be evacuated, and the needle should be occluded by
placing a rubber stopper over the needle. Another method to maintain an
anaerobic environment would be to remove the needle after sample collec-
tion and place a syringe plug over the tip of the syringe. There is evidence

Vet Clin Small Anim 32 (2002) 1031–1048

E-mail address: lvses@earthlink.net (T.K. Day).

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 0 3 5 - 9

that a tight-ﬁtting syringe cap is superior to a rubber stopper over the needle
in maintaining an anaerobic environment [2]. Air bubbles that are contained
within the blood sample can alter the blood gas values. The P

of room air

is extremely low, and the P

of room air is much higher compared with the

blood sample. Therefore, prolonged exposure to an air bubble, especially
one that occupies more than 10% of the sample volume, can result in a
decrease in P

and an increase in P

[2].

The sample should be analyzed as soon as possible, usually within 15 to

20 minutes of collection. The sample should be placed on a rocker or be con-
stantly mixed if there is a delay in analysis. Blood that is permitted to stand
before analysis results in an increased P

, decreased pH, decreased glu-

cose, and increased lactate. These changes are attributed to glycolysis by
white blood cells, red blood cells, and platelets, and they occur much more
slowly at 4

°C versus 25°C (room temperature) [3]. A delay in analysis neces-

sitates that the sample be placed in the refrigerator or on ice for no more
than 2 hours before analysis [3].

Arterial blood samples can be collected via occasional percutaneous

punctures or via an indwelling catheter. Arteries that can be used in the dog
include the femoral, dorsal pedal, and dorsal auricular artery in large
breeds. Indwelling catheters can be maintained most eﬃciently in the dorsal
pedal and dorsal auricular arteries. Obtaining an arterial sample in cats can
be more diﬃcult, and the only site that can be used without diﬃculty is the
femoral artery. The dorsal pedal artery can be used in some instances. An
indwelling catheter is much more diﬃcult to insert and maintain in cats and
usually requires a cutdown procedure of the femoral artery.

Equipment

Blood gas analyzers

The technologic advances in bedside blood gas analyzers have resulted in

an accurate and cost-eﬀective manner of evaluating blood gases and acid-
base status in veterinary patients. One of the more popular and aﬀordable
units is the i-STAT (Sensor Devices Incorporated; Waukesha, WI). The
i-STAT has been evaluated in dogs and horses and corresponds accurately
with more expensive laboratory-based units that can cost tens of thousands
of dollars [4,5]. Other bedside units that are available and have been eval-
uated include the Stat-Pal II (PPG Industries, La Jolla, CA). Development
of an accurate and reliable bedside monitor can decrease the amount of time
from sample collection to sample analysis, resulting in the most accurate
results. The blood gas variables P

and P

are measured directly as is

blood pH. Oxygen saturation (S

), total blood carbon dioxide (T

), base

excess, and HCO

are all calculated variables. Recently, blood lactate

determination (see details below) has been incorporated as a routine vari-
able that is determined concurrently with the blood gas analysis.

1032

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

Co-oximeters

is dependent on several variables, including Pa

, temperature of the

blood, the concentration of hydrogen ion (pH), and the quantity of 2,3-
diphosphoglycerate. The bedside blood gas analyzers do not take these var-
iables into consideration when reporting S

. The analyzer also does not

consider abnormal concentrations of dysfunctional hemoglobin (methemo-
globin, carboxyhemoglobin, and sulfhemoglobin).

A co-oximeter directly measures percentages of carboxy- and methemo-

globin, and S

. Fortunately, clinical conditions that result in increased

methemoglobin are uncommon in veterinary emergency medicine. Methe-
moglobinemia results in poor arterial blood oxygen saturation (Sa

although Pa

can be normal. The Pa

is of primary interest in veterinary

patients rather than the associated Sa

. Therefore, purchasing a co-oxi-

meter for routine use in veterinary medicine is not necessary.

Pulse oximetry

Pulse oximeters are common noninvasive devices used to monitor veteri-

nary patients while they are under anesthesia and during advanced respira-
tory support (mechanical ventilation). The principle of pulse oximetry is
that the saturation of pulsed arterial blood (Sp

) can be determined on a

beat by beat basis. Beer’s lawis used to determine the amount of absorption
of two wavelengths of light by hemoglobin to diﬀerentiate oxygenated from
deoxygenated hemoglobin [6].

Pulse oximetry must be used with caution when determining the Sa

critically ill veterinary patients, especially those with cardiovascular com-
promise and poor perfusion. A detectable pulsation at the site of probe
placement is required for a pulse oximeter to report Sp

. Poor perfusion

of the site of probe placement (usually the tongue) results in lowSp

values

despite the possibility of adequate Sp

. Rising Pa

values can occur even

with normal Pa

, Sa

, and Sp

values. Other factors that can result in

poor Sp

values in spite of normal Sa

include excessive room light, dry

area at the site of probe placement, and excessive movement. Unfortunately,
in some critically ill veterinary patients, an arterial blood sample should be
collected to determine if a lowSp

value is real.

Capnometry

Capnometers are noninvasive devices used to measure expired carbon

dioxide. The end-tidal carbon dioxide (ET

) equates closely to Pa

[7].

The primary use of capnometry is to monitor the adequacy of ventilation.
Hypoventilation (see below) is deﬁned as an increase in Pa

. Any increase

in ET

correlates with an increase in Pa

in patients with normal cardio-

vascular function. An increase in the diﬀerence between Pa

and ET

(Pa

) usually is a sign of increased dead space ventilation and

1033

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

altered ventilation-perfusion of the lung. The most common cause of an
increase in Pa

is poor perfusion of the lungs (ie, poor cardiac

output) [8]. Therefore, the cardiovascular status of a patient should be inves-
tigated when there is a sudden decrease in ET

during anesthesia or mon-

itoring of ventilator patients. Pulmonary thromboembolism is another
example of a clinical scenario that results in an increase in the Pa

value. Monitoring of ET

during cardiopulmonary resuscitation (CPR)

can aid in assessing lung perfusion and cardiac output and can predict CPR
outcome [9].

Arterial or venous sample?

The ﬁrst question to ask before collection of a blood sample for analysis

is the goal of the information to be obtained. Analysis of pulmonary func-
tion requires an arterial blood sample. Proper assessment of the Pa

value

obtained from the arterial blood sample also requires knowledge of the frac-
tion of inspired oxygen (Fi

) and the partial pressure of oxygen in the

alveoli (P

). Calculation of the alveolar-arterial diﬀerence (P

)

and other indices of oxygenation (Pa

/Fi

) helps to determine states of

hypoxemia (see below).

Analysis of tissue perfusion and overall acid-base status, conversely,

requires evaluation of a mixed venous (pulmonary artery) blood sample
[10]. The value of venous blood gas analysis was not considered strongly
until a large disparity between arterial and venous blood gases was deter-
mined in human patients during CPR [11]. The partial pressure of oxygen
in venous blood (P

) is also an important variable that is used to help

determine global tissue perfusion. It is possible to have normal or higher
than normal Pa

and severely compromised P

in dogs with normal pul-

monary function and poor tissue perfusion [10].

A true mixed venous blood sample is collected from the pulmonary artery.

Catheterization of the pulmonary artery is not a routine technique in veteri-
nary patients, although catheterization of the jugular vein is technically easier
and less expensive to perform. Arterial, jugular, and mixed venous oxygen
values have been determined in normal dogs [12]. A comparison of jugular
and mixed venous oxygen did not correlate well in a swine model of endotoxic
shock, although the acid-base variables did correlate well between the two
samples [13]. The use of jugular venous blood to assess venous oxygenation
has been advocated in veterinary critical care, however [14,15].

The recent addition of lactate analysis with blood gas values using the

i-STAT bedside analyzer has provided information on variables that aﬀect
lactate production. Two types of lactic acidosis (production of lactate
exceeds utilization by kidneys and liver) occur [16,17]. Type A (hypoxic) lac-
tic acidosis occurs when mitochondrial function is normal but oxygen deliv-
ery to tissues is inadequate. Type B (nonhypoxic) lactic acidosis occurs when

1034

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

oxygen delivery is adequate but there is a defect in mitochondrial function.
Type A is most common in veterinary patients, especially those with cardio-
vascular disease. Type B is not discussed here.

Hypoxic lactic acidosis occurs whenever oxygen delivery is impaired

because of poor tissue perfusion or reduced oxygen content. Clinical exam-
ples include CPR, shock, cardiac failure, hypovolemia, hypoxemia, and
severe anemia. Lactate concentrations greater than 2 to 2.5 mEq/L are con-
sidered elevated in dogs and cats [16]. Venous lactate concentrations are of
greater value than arterial lactate concentrations, because the venous sample
implies tissue perfusion.

Arterial blood gas analysis

Arterial blood samples provide information primarily related to the pul-

monary system using the true blood gas variables of Pa

, Sa

, and Pa

Proper evaluation of the blood gas variables can only occur with an arterial
sample, although, traditionally, overall acid-base status and decisions on
therapy have also been based on arterial samples. Recent information, espe-
cially that related to the critically ill patient, has directed overall acid-base
status to central venous samples and not to arterial samples because of the
potential for large disparities between the two [11]. The following discussion
on arterial blood gas analysis has two sections. The ﬁrst discusses the blood
gases. The second discusses the acid-base disturbances. A separate section
discusses venous blood gas analysis. The reader is reminded of the disparity
between arterial and venous samples in the critically ill patient. For continu-
ity purposes, however, all the acid-base disturbances are discussed in the
arterial blood gas analysis section. Pertinent emphasis is applied where nec-
essary to alert the reader to the proper interpretation of arterial and venous
samples.

Blood gas analysis

The only blood gases that are included in the blood gas analysis are Pa

and Pa

. The Sa

is a derived value that may not be as reliable based on

the method of calculation by the blood gas analyzer. A co-oximeter is
required to measure Sa

directly. The Pa

provides information on ven-

tilation whereas the Pa

provides information regarding oxygenation. A

complete description and explanation of the physiology of Pa

and Pa

are found elsewhere [18].

Ventilation

Alveolar ventilation is indirectly proportional to Pa

based on the

alveolar ventilation equation:

Paco

¼ Vco

0:863=VA

1035

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

where V

is the amount of carbon dioxide produced by the metabolism

and delivered to the lungs, VA is alveolar ventilation, and 0.863 is a constant
that equates dissimilar units for V

and VA.

Hyperventilation is deﬁned as a decrease in Pa

(hypocapnia), and hypo-

ventilation is deﬁned as an increase in Pa

(hypercapnia). The respiratory

rate and eﬀort determine alveolar ventilation; however, a clinician is not able
to describe hyperventilation by observing an increased respiratory rate. Anal-
ysis of Pa

is required to determine hyperventilation. Common causes of

hyperventilation in critically ill veterinary patients include pain, hypoxemia
of any cause, pulmonary disease, and central nervous system (CNS)–medi-
ated hypocapnia (liver disease, sepsis, heatstroke, and CNS disease).

Oxygenation

Measurement of Pa

is the basis to determine oxygenation. Two terms

have been used interchangeably that are somewhat diﬀerent in deﬁnition:
hypoxemia and hypoxia. Hypoxemia refers to a decrease in Pa

(<60 mm

Hg), whereas hypoxia refers to a general reduction in oxygen delivery
whether by hypoxemia or decreased cardiac output [7]. Pa

is primarily

determined by P

, which is described by the alveolar gas equation:

Pao

¼ Pio

1:2 ðPaco

where P

is the partial pressure of inspired oxygen and 1.2 represents the

respiratory quotient. The P

is determined by the following equation:

Pio

¼ Fio

ðP

47Þ

where F

is the fraction of inspired oxygen, P

is the barometric pressure,

and 47 represents water vapor pressure in millimeters of mercury, which is
subtracted as dry gases are measured.

Any evaluation of Pa

must include other variables to describe

adequately the hypoxemia. The F

must be presented (especially when

supplemental oxygen is administered), P

should be calculated, and the

diﬀerence between P

and Pa

must also be calculated. The P

is an important variable that helps to determine the severity of arterial

hypoxemia (described below).

There are ﬁve common causes of hypoxemia: hypoventilation, decreased

, diﬀusion impairment, ventilation-perfusion (V-Q) mismatch, and

right-to-left pulmonary shunt. Hypoventilation, by deﬁnition, is hypoxemia
because of an increased Pa

and is usually resolved with ventilation.

Hypoxemia caused by lowF

is resolved by providing supplemental oxy-

gen. Diﬀusion impairment in veterinary patients can be caused by diﬀuse
interstitial disease, severe emphysema, or vasculitis. Supplemental oxygen
may help to resolve hypoxemia in addition to treatment of the primary dis-
order. V-Q mismatch implies that ventilation and perfusion of alveoli are
not proportional. An increase in ventilation without appropriate perfusion
results in a V-Q ratio greater than 1, commonly referred to as increased

1036

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

alveolar dead space. Clinical examples include pulmonary thromboembo-
lism and severe emphysema. A decrease in ventilation to alveoli that are still
perfused results in a V-Q ratio less than 1. Examples include pulmonary
edema, pneumonia, and pulmonary contusions. Supplemental oxygen may
improve hypoxemia depending on the severity of the V-Q mismatch. A
V-Q ratio that approaches zero or little to no ventilation in areas of the lung
with adequate perfusion is referred to as right-to-left pulmonary shunt.
There is no ventilation of a large area of lung, but perfusion still occurs
(ie, a more severe V-Q mismatch). Pulmonary arterial blood delivered to the
lung to be oxygenated returns to the pulmonary veins and left atrium unoxy-
genated, resulting in severe arterial hypoxemia. Common causes of right-to-
left shunt are similar to those of V-Q mismatch but include more severe pre-
sentations, such as severe pulmonary edema, consolidation of a lung lobe
due to pneumonia, atelectasis, and extrapulmonary shunts (eg, reverse pat-
ent ductus arteriosis, severe ventricular septal defects). Supplemental oxygen
does not resolve hypoxemia caused by right-to-left shunts.

The P

diﬀerence is used to determine whether or not a V-Q

mismatch is present. The P

diﬀerence is a variable that should

be calculated whenever Pa

is evaluated. Large P

values repre-

sent the severity of hypoxemia. The normal P

is 5 to 15 mm

Hg depending on whether or not supplemental oxygen is administered [7].
The variability of P

during supplemental oxygen administration

may necessitate calculation of another variable to help determine the
severity of hypoxemia. The Pa

ratio is a more reliable method of

determining the severity of hypoxemia when F

is greater than 0.5 and

is less than 100 mm Hg [19]. In other words, the Pa

ratio indi-

cates the responsiveness of hypoxemia to oxygen supplementation and may
also indicate howmuch of the hypoxemia is a result of right-to-left intrapul-
monary shunt. The normal Pa

value while breathing room air is 100/

0.2

¼ 500 mm Hg. Pa

ratios less than 300 indicate acute lung injury,

and values less than 200 indicate acute respiratory distress syndrome [20].
Calculation of these variables may help to determine whether or not the
patient requires mechanical ventilation [21].

The P

has been used and found to correlate with outcome in

veterinary patients [22]. Although P

is primarily a pulmonary

variable, cardiovascular function can aﬀect the value obtained. Also, sur-
vival in patients with a large P

was not inﬂuenced by the disease

process. Therefore, P

and Pa

can be extremely valuable

calculations to determine the severity of hypoxemia.

One deﬁnition of hypoxemia also includes calculation of the content of

oxygen in arterial blood (Ca

) [7]. The Ca

equation is as follows:

Cao

¼ ðHb Sao

1:34Þ þ ðPao

0:003Þ

where Hb is the hemoglobin concentration, 1.34 is the amount of oxygen
carried by 1 g of hemoglobin, and 0.003 is the percentage of dissolved

1037

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

oxygen in plasma (0.3%) expressed as a decimal. The ﬁnal units of Ca

are

expressed as milliliters of oxygen per deciliter of blood. The equation shows
that the bulk of the Ca

calculation is based on the amount of hemoglobin.

Anemia aﬀects Ca

more than hypoxemia alone.

The importance of Pa

, Sa

, and Ca

revolves around one of the pri-

mary cardiovascular variables called oxygen delivery (D

). The equation

for D

is as follows:

¼ CO Cao

where CO is cardiac output. Therefore, a ‘‘simple’’ blood gas variable such
as Pa

can lead to the calculation of many variables that provide valuable

information about the critically ill patient.

Acid-base status

A complete discussion of acid-base balance is beyond the scope of this

review. The reader is referred elsewhere for an in-depth discussion of acid-
base balance [23]. The discussion that follows concentrates on major points
regarding acid-base status so that important variables can be obtained and
appropriate decisions can be made in the critically ill veterinary patient.

The basis for the traditional approach to acid-base balance revolves

around hydrogen ion concentration (expressed as pH), P

, and HCO

Carbon dioxide combines with water to form carbonic acid, which further
dissociates to hydrogen ion and bicarbonate. The Henderson-Hasselbalch
(H-H) equation describes the interactions:

þ pK þ log ½HCO

=½H

This equation can be substituted with a clinically applicable index of

HCO

, which results in the more useful form of the equation:

þ pK þ log ½HCO

=0:03ðPco

Þ:

Simpliﬁed, changes in HCO

and P

dictate the four primary acid-

base disturbances as well as the direction of compensation in an attempt
to maintain pH (see details below). Compensation can be acute, involving
the lungs, liver, and extracellular ﬂuid, and chronic, primarily involving the
kidneys.

Normal values of the arterial blood gas can followthe ‘‘rule of 4’’: pH of

7.40, Pa

of 80 to 100 mm Hg (based on F

¼ 0.21), Pa

of 40 mm Hg,

and HCO

of 24 mEq/L.

Acid-base analysis should followa logical series to determine abnormal-

ities that actually followthe previous order of presentation. First, examine
the pH; second, the Pa

; and third, the HCO

. Then, based on the

expected compensatory changes (see below), determine if you have a pri-
mary disorder with an appropriate compensatory response, a mixed dis-
order (actually two or three primary disorders occurring simultaneously), or
a compensated disorder. A compensated disorder can only be identiﬁed by a

1038

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

normal pH with abnormal Pa

and HCO

values that have changed

appropriately based on the primary disorder. Performing an analysis out
of this order can introduce confusion. One other important point is that
electrolytes, physical examination, discovery of an underlying cause, and
other diagnostic techniques must be evaluated concurrently with the blood
gas analysis to determine the appropriate deﬁnitive therapy for the disorder.
Like any diagnostic tool, blood gas analysis should be used as one ‘‘piece of
the puzzle’’ and should not be relied on as an exclusive diagnostic test.

Metabolic acidosis

Metabolic acidosis is a common acid-base disturbance in critically ill

patients. The primary event in metabolic acidosis is a decrease in the HCO

concentration. The H-H equation states that a decrease in HCO

results in

a compensatory decrease in Pa

or hyperventilation. The amount of

decrease in P

in relation to a decrease in HCO

is a 1.2-mm Hg decrease

in Pa

for every 1.0-mEq/L decrease in HCO

The anion gap can be used to help diﬀerentiate potential causes of meta-

bolic acidosis. The calculation of anion gap is most practical with the fol-
lowing equation:

ðNa

þ K

Þ ðCl

þ HCO

Þ:

There are two broad categories of metabolic acidosis that combine the

concept of anion gap. An increased anion gap metabolic acidosis usually
involves normochloridemia. The cause of the increased anion gap, although
the chloride is normal, is attributed to unmeasured anions (those that do not
contain chloride). Common causes of increased anion gap (normochlore-
mic) metabolic acidosis include ethylene glycol toxicosis, diabetic ketoacido-
sis, uremic acidosis, lactic acidosis, and salicylate intoxication [24].

Metabolic acidosis characterized by a normal anion gap usually has

hyperchloridemia as a component. In most instances, HCO

is lost in

excess of chloride, resulting in hyperchloridemia. Common causes of normal
anion gap (hyperchloremic) metabolic acidosis include acute small bowel
diarrhea, dilutional acidosis (rapid administration of 0.9% sodium chloride),
and posthypocapnic metabolic acidosis [24].

Treatment of metabolic acidosis primarily involves identiﬁcation and

treatment of the underlying cause. In many instances, treatment of the pri-
mary cause corrects the metabolic acidosis. Administration of sodium bicar-
bonate has been a controversial issue in veterinary medicine. Most agree
that if the pH is below7.1 or 7.2, bicarbonate therapy to raise the pH higher
than 7.2 is indicated along with therapy of the underlying cause of the met-
abolic acidosis. Blood pH less than 7.1 or 7.2 may lead to life-threatening
cardiovascular complications, including impaired cardiac contractility, poor
pressor response to catecholamines, and sensitization to ventricular arrhyth-
mias [24]. A common formula that is used to calculate total HCO

deﬁcit is

1039

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

as follows:

HCO

dose

ðmEqÞ ¼ 0:3 body weight ðkgÞ base excess

ðdeficitÞ ðmEqÞ:

An arbitrary administration of one fourth to one third of the total dose

over 20 to 30 minutes should correct the pH to higher than the desired 7.2.
Repeat blood gas analysis is required to determine the desired pH.

Sodium bicarbonate is not an innocuous drug, and there are potential

complications of administration. Volume overload can occur as a result of
the amount of sodium, decreased serum ionized calcium, increased aﬃnity
of hemoglobin for oxygen with potential to decrease oxygen delivery to tis-
sues, paradoxic cerebrospinal ﬂuid acidosis, and increased Pa

Respiratory acidosis

The primary event in respiratory acidosis is an increase in the Pa

con-

centration (also called hypercapnia). The H-H equation states that an
increase in Pa

results in a compensatory increase in HCO

. The amount

of increase in HCO

in relation to an increase in Pa

is a 0.15-mEq/L

increase in HCO

for every 1-mm Hg increase in Pa

in an acute respi-

ratory acidosis. The amount of increase in HCO

is higher in chronic res-

piratory acidosis, because the kidneys have had time to provide a
compensatory eﬀect. The amount of increase in HCO

in relation to an

increase in Pa

is a 0.35-mEq/L increase in HCO

for every 1-mm Hg

increase in Pa

in chronic respiratory acidosis [25].

Common causes of respiratory acidosis in critically ill veterinary patients

include any pulmonary disease (acute or chronic), neurologic disease, drugs
(eg, anesthetic agents, opioid analgesics), neuromuscular disease, and pleu-
ral disease [25]. Acute hypercapnia increases cerebral blood ﬂow(CBF) and
intracranial pressure, which can have detrimental eﬀects in the patient with
brain injury. When the Pa

concentrations reach approximately 100 mm

Hg, a state of severe cerebral depression results. The sympathetic nervous
system is also stimulated by hypercapnia, predisposing patients to arrhyth-
mias. Finally, elevations in Pa

cause vasodilation that is signiﬁed

by ‘‘brick red’’ mucous membranes. Extreme vasodilation can result in
hypotension.

Therapy for respiratory acidosis in nonanesthetized patients usually is

related to providing mechanical ventilation. Analysis of arterial blood gases
can help to determine the deﬁnition of respiratory failure, especially in rela-
tion to concurrent hypoxemia. One deﬁnition of respiratory failure is Pa

greater than 50 mm Hg in the nonsedated and nonanesthetized patient [21].
Other deﬁnitions of respiratory failure include Pa

less than 50 mm Hg

with F

of 0.21 or a poor response of Pa

less than 50 mm Hg with F

greater than 0.5 [21]. Hypercapnia and hypoxemia concurrently should be
treated aggressively with mechanical ventilation.

1040

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

Metabolic alkalosis

The primary event in metabolic alkalosis is an increase in the HCO

concentration. The H-H equation states that an increase in HCO

results

in a compensatory increase in the Pa

or hypoventilation. The amount

of increase in P

in relation to an increase in HCO

is a 0.6-mm Hg

increase in Pa

for every 1.0-mEq/L increase in HCO

[24]. Note that the

response to metabolic alkalosis is usually not as marked as in meta-

bolic acidosis.

Metabolic alkalosis is usually the result of the hydrogen ion (H

) loss that

occurs with pure gastric vomiting. Two other causes of metabolic alkalosis
are diuretic therapy and judicious administration of sodium bicarbonate.
Loss of H

in pure gastric vomiting (pyloric obstruction) also involves loss

of Cl

. Renal regulation of electrolytes is important in the creation and

maintenance of excess HCO

. Loss of hydrogen chloride in pure gastric

vomiting usually results in excessive loss of extracellular volume as well. The
kidney attempts to retain as much sodium as possible, usually retaining Cl

with Na

. The concurrent loss of chloride results in the retention of the next

most abundant negative ion in the body, HCO

. The result is an excess of

HCO

and metabolic alkalosis with pH values commonly greater than

7.50. Hypochloridemia and hypokalemia can also occur with excessive
hydrogen chloride loss. In many instances, ‘‘hypoelectrolytemia’’ can result
(decreased Na

, K

, Cl

) in patients with metabolic alkalosis caused by gas-

tric hydrogen chloride loss.

Clinical signs of metabolic acidosis usually involve the underlying cause

and the clinical eﬀects of electrolyte disturbances. Severe hypokalemia
(<2.5 mEq/L) can result in weakness and changes on the ECG. A decrease
in ionized calcium as a result of metabolic alkalosis can contribute to weak-
ness and muscle twitching [24].

Metabolic alkalosis is a secondary event, and therapy must be directed to

the primary cause. The acid-base changes in metabolic alkalosis are not cor-
rected until adequate chloride is replaced. Concurrently, deﬁcits in sodium
and potassium should be replaced as well. Commonly, administration of
0.9% sodium chloride with potassium supplementation is the ﬂuid of choice
to replace electrolyte deﬁcits and help correct metabolic alkalosis. Antiemet-
ics, surgery for pyloric obstruction, and other speciﬁc therapy to correct the
underlying cause help to reduce further losses of Cl

and H

Respiratory alkalosis

The primary event in respiratory alkalosis is a decrease in the Pa

con-

centration (also called hypocapnia). The H-H equation states that a decrease
in Pa

results in a compensatory decrease in HCO

. The amount of

decrease in HCO

in relation to a decrease in Pa

is a 0.25-mEq/L

decrease in HCO

for every 1-mm Hg decrease in Pa

in acute respira-

tory alkalosis. The amount of decrease in HCO

is higher in chronic respi-

ratory alkalosis, because the kidneys have had time to provide a

1041

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

compensatory eﬀect. The amount of decrease in HCO

in relation to a

decrease in Pa

is a 0.55-mEq/L decrease in HCO

for every 1-mm Hg

decrease in Pa

in chronic respiratory acidosis [25].

Common causes of respiratory alkalosis in critically ill patients include

stimulation of peripheral chemoreceptors in response to hypoxemia, any
pulmonary disease, and direct stimulation of the respiratory center (eg, heat-
stroke, drugs, sepsis, CNS disease). Clinical signs of respiratory alkalosis in
veterinary patients have not been recognized aside from obvious signs of the
underlying disease process. Anesthetized patients can have decreases in car-
diac output and blood pressure in response to hypocapnia, but this eﬀect has
not been observed in awake veterinary patients.

Therapy is directed to treatment of the underlying disease process,

because no other therapy is eﬀective.

Mixed acid-base disturbances

Discussion of mixed acid-base disturbances has involved entire chapters,

because the concepts of multiple acid-base disorders occurring simultane-
ously are complex. The reader is referred elsewhere for a detailed description
of mixed acid-base disorders [26,27]. Mixed acid-base disorders are ex-
tremely common in critically ill veterinary patients.

Diagnosis of a mixed acid-base disorder can be diﬃcult. Therefore, guide-

lines exist that can help the clinician to establish a logical approach to blood
gas analysis [27]. The nontraditional approach to acid-base balance can be
helpful in a mixed disorder. There is a diﬀerence between a mixed disorder
and expected compensation for a simple disorder. First, followthe diagnosis
of simple disorders as described previously. If the calculations are not com-
patible with a simple disorder, the following steps may be helpful in deter-
mining the presence of a mixed disorder [27]:

1. The presence of a normal pH and abnormal Pa

and HCO

implies a

mixed disorder.

2. A change in pH in the opposite direction than that predicted for a pri-

mary disorder requires a diagnosis of a mixed disorder.

3. When Pa

and HCO

change in opposite directions than those pre-

dicted, there is a mixed disorder.

4. Values that are obtained within the expected compensatory eﬀect do not

prove that a simple disorder is present; they just suggest the possibility if
supported by the clinical data.

Laboratory or collection errors could result in the diagnosis of a mixed

disorder. Analysis of a venous blood gas for pulmonary values (Pa

)

results in miscalculation of compensation. Too much heparin in the syringe,
storage of the blood sample for longer than 15 to 20 minutes at room tem-
perature, and miscalculation of HCO

could result in the diagnosis of a

mixed disorder when one is not present. Mixed disorders fall into the follow-

1042

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

ing categories: mixed disorders with neutralizing eﬀects on pH, mixed disor-
ders with additive eﬀects on pH, and triple disorders [27]. Clinical examples
of these three categories are brieﬂy described belowwith emphasis only on
common problems in critically ill patients. Every possibility of a mixed dis-
order is not presented.

Neutralizing eﬀect on pH

A mixed respiratory alkalosis and metabolic acidosis can occur in many

clinical situations. Most of these patients have acute respiratory alkalosis in
the presence of a high anion gap metabolic acidosis. The blood gas analysis
is characterized by lowPa

and lowHCO

in the face of a pH that may

be close to normal. The only simple disorder that can result in these same
ﬁndings is chronic respiratory alkalosis that has completely compensated.
Clinical ﬁndings can help to diﬀerentiate between the two, and direct ther-
apy for the acid-base disturbance is not warranted. Therapy is centered
on correcting the clinical cause of the acid-base disorder.

Gastric dilatation–volvulus (GDV) is a common example. The metabolic

acidosis is usually a result of lactic acidosis secondary to poor perfusion, and
the acute respiratory alkalosis is a result of pain-induced hyperventilation.
Therapy would include gastric decompression, aggressive ﬂuid therapy, and
analgesia. Dogs or cats with low-output heart failure (lactic acidosis) that
develop acute pulmonary edema (hypoxemia with secondary hyperventila-
tion) is another common clinical presentation of this mixed disorder. Ther-
apy would include increasing cardiac output and providing oxygen support
and diuretic therapy. CPR can result in this mixed disorder as well. Caution
is advised when analyzing arterial or venous samples only, however. Arterial
samples can reveal respiratory alkalosis caused by hyperventilation (usually
by the clinician performing CPR), yet the central venous sample can reveal
increased partial pressure of carbon dioxide in jugular venous blood (P

VCO

)

as a result of poor perfusion (see below).

Metabolic acidosis and metabolic alkalosis can occur simultaneously.

Most of these patients have a chronic disorder, resulting in metabolic acido-
sis (eg, renal failure, diarrhea, uncomplicated ketoacidosis), and they begin
vomiting pure gastric contents (metabolic alkalosis). The importance of
making the diagnosis of this mixed disorder is that if one disorder is treated
without recognition of the second, the second can appear unopposed. The
opposite can also occur when metabolic alkalosis caused by pure gastric loss
leads to severe volume depletion, resulting in lactic acidosis. Vomiting and
diarrhea are important causes of hyperchloremic metabolic acidosis and
metabolic alkalosis (parvovirus enteritis). Caution is advised in this sce-
nario; serum electrolytes and results of blood gas analysis may be within
the normal range in this setting, because these disturbances have opposite ef-
fects on HCO

and Cl

. A physical examination and history are important

factors in making this diagnosis.

1043

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

Additive eﬀect on pH

Disorders with an additive eﬀect on pH can have dramatically high or

lowvalues. A mixed acidosis (metabolic and respiratory) can occur in many
clinical situations, including cardiopulmonary arrest and severe pulmonary
edema secondary to low-output cardiac failure and lactic acidosis as a result
of poor tissue perfusion. The Pa

is usually normal or high, the HCO

low, and the pH can be dangerously low. Therapy is directed to improving
tissue perfusion with the possibility of providing mechanical ventilation,
especially if hypoxemia is present.

A mixed alkalosis (metabolic and respiratory) is uncommon in veterinary

patients. The Pa

is low, the HCO

is low, and the pH is high. This type

of disorder can occur in the GDV complex (hydrogen chloride loss and
pain) and in patients with congestive heart failure that have been treated
with diuretics. Therapy consists of providing a source of chloride to correct
the metabolic component and treating the underlying cause of the respira-
tory component.

Triple disorders

Triple disorders usually occur when a mixed metabolic disorder (acidosis

and alkalosis) is complicated by an acute respiratory disturbance [27]. The
pH and HCO

may be increased, decreased, or normal. The Pa

is increased

when the mixed metabolic disturbance is complicated by acute respiratory
acidosis, and Pa

is decreased when acute respiratory alkalosis occurs.

Applying the nontraditional approach to blood gas analysis (see below) may
help in clarifying a triple disorder, although the calculation can be tedious.

Two examples of triple acid-base disorders are low-output heart failure

[28] and GDV syndrome [29]. Low-output heart failure patients treated with
diuretics can have metabolic acidosis secondary to lactic acidosis and meta-
bolic alkalosis caused by hypochloridemia. Acute pulmonary edema can
result in increased ventilation and decreased Pa

(respiratory alkalosis).

Conversely, if the edema becomes severe, hypoventilation and increased
Pa

(respiratory acidosis) can occur. GDV syndrome is one of the com-

mon disorders that can have any acid-base disturbance and a triple disturb-
ance. Metabolic acidosis occurs because of poor perfusion, metabolic
alkalosis can occur secondary to hypochloridemia, and distention of the
torsed stomach can decrease ventilation (respiratory acidosis). Patients with
minimal distention can have pain-induced hyperventilation and decreased
Pa

(respiratory alkalosis) in addition to the two metabolic disturbances.

Venous blood gas analysis

The ability to adequately assess oxygenation is limited and usually

involves measuring some aspect of global oxygenation compared with indi-
vidual organ oxygenation. The ability to examine a measure of global

1044

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

oxygenation is limited in some situations, such as sepsis, in which global mea-
sures of oxygenation may be normal, yet individual organs become hypoxic.
Evidence of poor oxygenation of organs usually is recognized after the
organ has been damaged and begins to fail. Analysis of venous blood gas
can provide global information on tissue perfusion and tissue acid-base bal-
ance but provides no information on pulmonary variables (Pa

and Pa

Tissue perfusion

There are similarities in arterial blood gas values and mixed (pulmonary

artery) venous blood gas values, with the exception of oxygen and S

normal dogs [12]. An investigation in people seems to have ﬁrst brought
to attention the concept of using venous oxygen and venous oxygen satura-
tion (S

) as measures of global tissue perfusion during cardiopulmonary

arrest [11]. Pulmonary artery catheterization is limited in veterinary medi-
cine, although a jugular venous sample can provide similar information
regarding perfusion.

The P

and associated S

can provide base information to calculate the

content of oxygen in venous blood (C

) similar to that of the calculation of

. The diﬀerence in the two values, referred to as the a-v O

diﬀerence, can

be used as a measure of total body oxygen oxygenation. The P

from a jug-

ular venous or pulmonary artery catheter can provide evidence of tissue per-
fusion. A lowP

indicates that the maximum amount of oxygen is being

extracted from the tissues, thus widening the a-v O

diﬀerence. Normal P

values range from 35 to 50 mm Hg. When the P

falls between 28 and 35 mm

Hg, there is limited oxygen reserve in the tissues and a state of anaerobic
metabolism is pending [30]. Aggressive therapy with colloids, crystalloids,
inotropes, or other methods of increasing perfusion is necessary to prevent
further decreases. A P

value below27 mm Hg indicates the presence of

an anaerobic metabolism and production of lactic acidosis. This value is so
lowthat it is considered a preterminal event. Elevation in P

higher than

60 mm Hg in an animal breathing room air suggests decreased tissue uptake
of oxygen such that the tissues do not receive the oxygen because of shunting
of blood at the capillary level. Sepsis is a common clinical entity in which ele-
vations in P

occur because of shunting of blood away from the tissues [14].

Venous lactate can also provide evidence of poor tissue perfusion,

although the venous value would be expected to be similar to the arterial
value. Lactate concentrations higher than 2 to 2.5 mEq/L indicate abnormal
accumulation in critically ill dogs and cats [14,16]. The i-STAT blood gas
analyzer currently has the capability to provide lactate concentrations with
the blood gas information.

Acid-base status

Venous acid-base analysis can be important in scenarios of poor tissue

perfusion. There was a large disparity between arterial and venous acid-base

1045

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

values during CPR such that the arterial blood indicated an adequate pH
but the venous blood indicated severe acidosis with elevated P

[11]. An

elevation in P

VCO

usually indicates inadequate ventilation in the patient

with normal tissue perfusion, yet during poor tissue perfusion, elevated
P

VCO

indicates poor perfusion. Therapy in a patient with such disparity

between arterial and venous blood consists of providing perfusion with col-
loids, crystalloids, or inotropes, for example.

Cerebral oxygenation

An interesting concept in the monitoring of cerebral oxygenation has

been suggested in human patients with brain injury. Although many var-
iables have been proposed to monitor brain-injured human patients, Sj

has been suggested as measuring the adequacy of CBF [31]. The following
equation has been proposed as providing the most accurate information on
global CBF:

CMRO

¼ CBFðCao

Cjvo

where CMRO

is the cerebral metabolic rate for oxygen and Cj

is the

content of oxygen in jugular venous blood, which is calculated from the par-
tial pressure of oxygen in jugular venous blood (Pj

) and Sj

similar to

the calculations for Ca

and C

noted previously. An abnormally low

(<50% [normal

¼ 65%]) suggests the possibility of cerebral ischemia.

Collection of jugular venous blood and calculation of CBF are within the
realms of veterinary critical care, and the previous formula may be used
to monitor veterinary patients with traumatic brain injury.

Nontraditional approach to acid-base analysis

The traditional approach to acid-base balance gives the impression that

and HCO

are independent variables that aﬀect [H

]. Only Pa

is an independent variable, however. An increase in Pa

results in an

increase in hydrogen ions, which are buﬀered primarily by proteins, and the
HCO

concentration increases secondarily. The traditional approach also

does not take into account the eﬀects of electrolytes (Na

, K

, Cl

) and

plasma proteins on acid-base balance.

Stewart [32] developed a method of acid-base analysis that is directed by

three physical laws: maintenance of electroneutrality, satisfaction of dissoci-
ation equilibrium for incompletely dissociated solutes, and conservation of
mass. The nontraditional approach identiﬁes three independent variables:
Pa

, the strong ion diﬀerence, and the total concentration of weak acids

(proteins).

A complete and extensive explanation of the nontraditional approach to

acid-base balance is provided elsewhere [33]. The Stewart approach was
modiﬁed to extend the traditional assessment in acid-base disturbances and

1046

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

can be useful in the assessment of mixed acid-base disturbances, including
triple disturbances.

Summary

Evaluation of both arterial and central venous blood can be valuable in

monitoring the critically ill veterinary patient. The traditional approach,
which concentrates on arterial blood analysis only, may miss important
aspects of oxygen delivery to tissues, especially in patients with poor perfu-
sion. The advances that have resulted in aﬀordable bedside blood gas ana-
lyzers have created a clinical situation in which blood gas analysis should
be an integral part of critical care monitoring. Following basic principles
of interpretation, blood gas analysis, which has traditionally been viewed
as a complex method of monitoring, should become more useful. Assessing
both the arterial and central venous samples should result in more eﬃcient
and higher quality care for veterinary patients.

References

[1] Haskin SC. Sampling and storage of blood for pH and blood gas analysis. JAVMA 1977;

170:429.

[2] Pruden EL, Siggaard-Anderson O, Tietz W. Blood gases and pH. In: Burtis CA, Ash-

wood ER, editors. Tietz fundamentals of clinical chemistry. Philadelphia: WB Saunders;
1996. p. 506.

[3] Madiedo G, Sciacca R, Hause L. Air bubbles and temperature eﬀect on blood gas analysis.

J Comp Pathol 1980;33:864.

[4] Looney AL, Ludders J, Erb HN, et al. Use of a handheld device for analysis of blood

electrolyte concentration and blood gas partial pressure in dogs and horses. JAVMA 1998;
213:526–31.

[5] Raﬀe MR, Faudskar LS. Point-of-care laboratory testing in the intensive care unit.

In: Bonagura JD, editor. Kirk’s current veterinary therapy XIII, small animal practice.
Philadelphia: WB Saunders; 2000. p. 110–2.

[6] Clark JS, Votteri B, Ariagno R, et al. Noninvasive assessment of blood gases: state of the

art. Am Rev Respir Dis 1992;145:220.

[7] Martin L. Pao

, Sao

, and oxygen content. In: All you really need to knowto interpret

arterial blood gases. 2nd edition. Philadelphia: Lippincott Williams & Wilkins; 2000. p. 79.

[8] Shibutani K, Shirasaki S, Braaz T, et al. Changes in cardiac output aﬀect PETCO2, CO2

transport, and O2 uptake during unsteady state in humans. J Clin Monit 1992;8:170–6.

[9] Callaham M, Barton C. Prediction of outcome of cardiopulmonary resuscitation from end-

tidal carbon dioxide concentration. Crit Care Med 1990;18:358–62.

[10] Mathias DW, Cliﬀord PS, Klopfenstein HS. Mixed venous blood gases are superior to

arterial blood gases in assessing acid-base status and oxygenation during acute pericardial
tamponade in dogs. J Clin Invest 1988;82:833–7.

[11] Weil MH, RackowEC, Trevino R, et al. Diﬀerence in acid-base state between venous and

arterial blood during cardiopulmonary resuscitation. N Engl J Med 1986;315:153–9.

[12] IlkiwJE, Rose RJ, Martin ICA. A comparison of simultaneously collected arterial, mixed

venous and cephalic venous blood samples in the assessment of blood gas and acid base
status in dogs. J Vet Intern Med 1991;5:294–7.

1047

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

[13] Wohl JA, Bags A, Lin JL, et al. Use of jugular venous blood, compared with mixed venous

blood, for measurement of venous oxygenation indices in a porcine model of endotoxic
shock. Am J Vet Res 1997;58:910–4.

[14] Aldrich J, Haskins SC. Monitoring the critically ill patient. In: Bonagura JD, editor. Kirk’s

current veterinary therapy XII, small animal practice. Philadelphia: WB Saunders; 1995.
p. 98–105.

[15] Wohl JA, Murtaugh RJ. Use of catecholamines in critical care patients. In: Bonagura JD,

editor. Kirk’s current veterinary therapy XII, small animal practice. Philadelphia: WB
Saunders; 1995. p. 188–93.

[16] Hughes D. Lactate measurement: diagnostic, therapeutic and prognostic implications. In:

Bonagura JD, editor. Kirk’s current veterinary therapy XIII, small animal practice.
Philadelphia: WB Saunders; 2000. p. 112–6.

[17] Kreisberg RA. Pathogenesis and management of lactic acidosis. Annu Rev Med 1984;

35:181.

[18] West JB. Respiratory physiology: The essentials. 6th edition. Philadelphia: Lippincott

Williams & Wilkins; 2000.

[19] Gowda M, Klocke RA. Variability of indices of hypoxemia in adult respiratory distress

syndrome [abstract]. Crit Care Med 1997;25:41.

[20] Luce JM. Acute lung injury and the acute respiratory distress syndrome. Crit Care Med

1998;26:369–74.

[21] Bateman SW. Ventilating the lung injured patient: What’s new? In: Proceedings of the

American College of Veterinary Surgeons Symposium, Chicago; 2001. p. 562–5.

[22] Van Pelt DR, et al. Oxygen-tension based indices as predictors of survival in critically ill

dogs: clinical observations and review. J Vet Emerg Crit Care 1991;1:19.

[23] DiBartola SP, de Morais HA. Acid-base disorders. In: DiBartola SP, editor. Fluid therapy

in small animal practice. 2nd edition. Philadelphia: WB Saunders; 2000. p. 189–261.

[24] DiBartola SP. Metabolic acid-base disorders. In: Fluid therapy in small animal practice.

2nd edition. Philadelphia: WB Saunders; 2000. p. 213.

[25] de Morais HA. Respiratory acid-base disorders. In: DiBartola SP, editor. Fluid therapy in

small animal practice. 2nd edition. Philadelphia: WB Saunders; 2000. p. 246.

[26] Adams LG, Polzin DJ. Mixed acid-base disorders. Vet Clin North Am Small Anim Pract.

1989;19:307–26.

[27] de Morais HA. Mixed acid-base disorders. In: DiBartola SP, editor. Fluid therapy in small

animal practice. 2nd edition. Philadelphia: WB Saunders; 2000. p. 251–61.

[28] Aberman A, Fulop M. The metabolic and respiratory acidosis of acute pulmonary edema.

Ann Intern Med 1972;76:173–8.

[29] Muir WW III. Acid-base and electrolyte disturbances in dogs with gastric dilatation-

volvulus. JAVMA 1982;181:229–31.

[30] Snyder JV, Carroll GC. Tissue oxygenation: a physiological approach to a clinical

problem. Curr Probl Surg 1982;19:650–719.

[31] Deyo DJ, Yancy V, Prough DS. Brain function monitoring. In: Grevnick A, Shoemaker

WC, Ayres SM, Holbrook PR, editors. Textbook of critical care. 4th edition. Philadelphia:
WB Saunders; 2000. p. 1816–7.

[32] Stewart PA. Modern quantitative acid-base chemistry. Can J Physiol Pharmacol 1983;

61:1444.

[33] DiBartola SP. Introduction to acid-base disorders. In: Fluid therapy in small animal

practice. 2nd edition. Philadelphia: WB Saunders; 2000. p. 202–4.

1048

T.K. Day / Vet Clin Small Anim 32 (2002) 1031–1048

Causes of respiratory failure

Lisa L. Powell, DVM

Department of Small Animal Clinical Sciences, University of Minnesota,

College of Veterinary Medicine, 1352 Boyd Avenue, St.Paul, MN 55108, USA

Respiratory failure is deﬁned as ineﬀective gas exchange in the lungs by

the respiratory system. This can be caused by the inability to deliver
adequate amounts of ambient air to the alveoli (ventilatory disease) or prob-
lems with gas exchange across the alveoli–pulmonary blood vessel (lung
parenchymal disease). Pleural space disease causes respiratory distress
because of the inability of the lungs to expand and ﬁll with ambient air.
Hypoventilation leads to hypercapnia and hypoxemia, whereas interference
with gas exchange usually results in hypoxemia only. This is because of the
increased ability of carbon dioxide (CO

) to diﬀuse into the alveoli from the

pulmonary blood vessels during gas exchange. However, hypercarbia can
develop with severe pulmonary parenchymal disease.

Gas exchange and oxygen transport

The main pulmonary artery enters the lung from the right ventricle car-

rying mixed venous blood. The pulmonary artery branches into smaller arte-
rioles, ending in a meshwork of capillaries surrounded by alveoli, where gas
exchange takes place. The partial pressure of oxygen in pulmonary arteries
is about 40 mm Hg. In normal alveoli, the partial pressure of oxygen is 100
mm Hg. As mixed venous blood enters the pulmonary capillaries from the
pulmonary arteries and arterioles, oxygen diﬀuses down its concentration
gradient from the alveoli into the pulmonary capillary meshwork. The oxy-
gen is then transported to the tissues bound to hemoglobin, where it again
diﬀuses down its concentration gradient. The tissues then use the oxygen as
energy for normal metabolism. Normal tissue oxygenation depends on
adequate ventilation, eﬀective gas exchange in the lungs, suﬃcient amounts
of hemoglobin in the circulation, and normal perfusion.

Vet Clin Small Anim 32 (2002) 1049–1058

E-mail address: powel029@tc.umn.edu (L.L. Powell).

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 0 4 1 - 4

Assessment of ventilation and oxygenation

Objective assessment of ventilation and oxygenation can be achieved

through the evaluation of arterial blood gas measurements, pulse oximetry,
and end-tidal CO

monitors (capnography). Arterial blood gas analysis is the

gold standard, allowing measurement of partial pressures of CO

(PaCO

oxygen (PaO

), blood pH, and bicarbonate levels. Normal mean ranges for

pH in the dog is 7.407

þ/ 0.028, PaCO

36.8

þ/ 3 mm Hg, and PaO

92.1

þ/ 5.6 mm Hg. Normal mean ranges for cats are pH 7.38þ/

0.038, PaCO

þ/ 2.9 mm Hg, and PaO

106.8

þ/ 5.7 [1]. Arterial

blood oxygen levels of 80 to 100 mm Hg are generally considered normal
(see Table 1).

The alveolar–arterial oxygen gradient (P(A

a)O

) is an equation that

estimates the eﬀectiveness of gas exchange in the lungs while removing the
variability of ventilation. Alveolar oxygen tension (PAO

) is estimated by

the use of the alveolar gas equation:

PAO

¼ FiO

ðPb PH

Þ PaCO

where:

FiO

¼ inspired oxygen concentration

¼ barometric pressure

¼ water vapor pressure at body temperature

PaCO

¼ partial pressure of CO

in arterial blood

¼ respiratory quotient ð0:8Þ

At sea level and room air, the alveolar gas equation can be simpliﬁed to:

PAO

¼ 150 ðPaCO

Þ1:25

The arterial oxygen tension (PaO

) measured from blood gas analysis is

subtracted from this equation to yield the alveolar-arterial oxygen gradient
(P(A

a)O

ðA aÞO

¼ ½150 ðPaCO

Þ1:25 PaO

½1

The normal P(A

a)O

is less than or equal to 15 mm Hg. An elevated alveo-

lar–arterial oxygen gradient is supportive of pulmonary parenchymal disease,

Table 1
Oxygen-based indices

Arterial oxygen tension (PaO

)

Pulse oximetry (SaO

)

Alveolar-arterial oxygen gradient: P(A

a)O

¼ [150(PaCO

)1.25]

PaO

Arterial oxygen/inspired oxygen ratio: PaO

/FiO

1050

L.L. Powell / Vet Clin Small Anim 32 (2002) 1049–1058

speciﬁcally ventilation/perfusion inequality. The gradient should be normal in
patients who have pure hypoventilation and no concurrent pulmonary paren-
chymal disease. Serial calculations of the P(A

a)O

allow for comparisons of

oxygenation in patients with variable ventilatory eﬀort (varying PaCO

). Ele-

vations in the gradient have been associated with increased mortality in crit-
ically ill veterinary patients [2]. For accurate comparisons, the sequential
gradients must be calculated from an arterial blood gas while the patient is
breathing room air, or the inspired oxygen concentration must be constant.

Another method of quantifying severity of lung disease and associated

hypoxemia is through calculation of the ratio of arterial oxygen tension to
inspired oxygen concentration (PaO

/FiO

). This method is especially useful

for comparison of PaO

levels while oxygen is being administered, or if the

inspired oxygen concentrations are variable. Normal PaO

should be ﬁve

times the FiO

[3]. For example, the FiO

of room air is 0.21, and normal

PaO

on room air is about ﬁve times that, or 100 mm Hg. Approximate

inspired oxygen concentrations based on the method of oxygen administra-
tion are given in Table 2 [4]. Normal values for PaO

/FiO

are between 250

and 400. Values less than 200 indicate signiﬁcant pulmonary parenchymal
disease, speciﬁcally intrapulmonary shunting involving 20% or more of
functional lung units [5].

Pulse oximetry measures blood hemoglobin that is saturated with oxygen,

giving an indirect measurement of PaO

. Following the oxyhemoglobin dis-

sociation curve, a pulse oximetry measurement of 91% to 100% reﬂects a
normal PaO

(80–100 mm Hg). As the pulse oximetry measurement falls

below 91%, however, the PaO

decreases rapidly (see Fig. 1) [6]. Pulse oxi-

metry is a useful monitoring tool, but drawbacks exist. Ventilation cannot
be assessed through the use of a pulse oximeter because CO

levels are not

measured. Pigmentation of mucous membranes decreases the accuracy of
the pulse oximetry measurement because the reading depends on the diﬀer-
ential light absorption of reduced hemoglobin and saturated hemoglobin
measured at two diﬀerent wavelengths of light [7]. Other factors that aﬀect
the reliability of a pulse oximetry reading include motion artifact, decreased
peripheral perfusion, and the presence of saturated but nonfunctional

Table 2
Approximate FiO

levels for diﬀerent methods of oxygen administration

Method of oxygen
administration

Approximate inspired
oxygen concentration
(FiO

)

cage

40%

Nasal cannula (unilateral)

40–50%

Nasal cannula (bilateral)

50–60%

Tight-ﬁtting face mask

70–90%

Oxygen tent

60–70%

Intubation

100%

1051

L.L. Powell / Vet Clin Small Anim 32 (2002) 1049–1058

hemoglobin (ie, carboxyhemoglobin, methemoglobinemia). A pulse oximeter
is useful, it does not require expensive blood gas machines, and it needs
minimal technical expertise for operation. However, the operator must be
aware of potential factors that aﬀect the pulse oximeter’s accuracy to avoid
over-interpretation of low readings.

End-tidal CO

monitoring is accomplished through the use of capno-

graphy. These devices measure expired CO

levels and reﬂect PaCO

These monitors are useful when assessing adequacy of ventilation during
cardiopulmonary resuscitation, anesthesia or heavy sedation, and mechani-
cal ventilation. These monitors must be attached to an endotracheal tube or
in-line with a tight-ﬁtting mask, making it diﬃcult to assess expired CO

lev-

els in awake, non-intubated patients.

Causes of hypoxemia

There are ﬁve general causes of hypoxemia that can result in respiratory

failure. These include hypoventilation, ventilation-perfusion mismatch (V/Q
mismatch), intrapulmonary shunt, diﬀusion impairment, and a decrease in
inspired oxygen concentration [8].

Hypoventilation

Hypoventilation results in hypercapnia, measured on an arterial blood

gas as a partial pressure of CO

(PaCO

) of greater than 45 mm Hg. Due

to the decreased air ﬂow to the alveoli, hypoxemia also occurs. Providing
an enriched oxygen environment will increase the PaO

to above normal val-

ues because there is no interference to gas exchange in the pulmonary paren-
chyma [8]. Hypercapnia will remain unless ventilation is provided or the
cause of the hypoventilation is corrected. Causes of hypoventilation include

Fig. 1. The oxyhemoglobin saturation curve. As saturation falls below 91%, the PaO

decreases

signiﬁcantly.

1052

L.L. Powell / Vet Clin Small Anim 32 (2002) 1049–1058

diseases of the central nervous system, peripheral neurologic disease,
obstruction or disease of the airways (trachea, bronchial tree), general anes-
thesia and heavy sedation, and pleural space/chest wall pathology.

Control of ventilation by the central nervous system is carried out pri-

marily by the brainstem (medullary centers). Upper motor neurons project
into the spinal cord at the cervical spinal segments C3 to C5, aﬀecting the
phrenic nerves, and to the thoracic segments T1-8, aﬀecting the intercostal
nerves. Other neurons in the medulla control muscles in the laryngeal/pha-
ryngeal area. Diseases aﬀecting the central nervous system or motorneuron
units can result in hypoventilation.

Brain lesions involving the pontomedullary region and caudal brain stem

include neoplasia, vascular abnormalities (thrombosis, hemorrhage), and
inﬂammation/infection of the central nervous system. Increased intracranial
pressure from cerebral edema or trauma can also result in loss of ventilatory
regulation.

Generalized lower motor neuron disease can lead to respiratory muscle

paralysis and hypoventilation. Nerves that are aﬀected include the phrenic
nerve, causing diaphragmatic paralysis, and the nerves to the intercostal
muscles. Examples of generalized lower motor neuron diseases include bot-
ulism, tick paralysis, and polyradiculoneuritis (‘‘coonhound paralysis’’).
Patients presenting with these diseases exhibit ﬂaccid paralysis of all four
limbs, hyporeﬂexia, abdominal breathing, and potential cranial nerve deﬁ-
cits, especially absent palpebral response.

Obstruction of the trachea by foreign bodies, tumors, severe tracheal col-

lapse, inﬂammation, or laryngeal paralysis leads to severe respiratory dis-
tress because of obstruction of air ﬂow, which results in hypercapnia and
severe hypoxemia. These patients often present cyanotic because of the
severity of the hypoxemia and subsequent decrease in hemoglobin satura-
tion. Loud upper airway stridor is often a characteristic sign of airway
obstruction. Hyperthermia results from the inability to expire and release
heat from the body. Patients might require a tracheostomy below the
obstruction if the obstruction cannot be relieved quickly, especially if cyano-
sis, mental depression, and hyperthermia are present.

General anesthesia inhibits the body’s normal homeostatic mechanisms,

allowing for a decreased response to the presence of hypercapnia. Normally,
elevations in blood CO

levels stimulate the respiratory center in the medulla

by way of chemoreceptors located in the carotid arteries and aortic arch.
With increasing depth of general anesthesia there is a decreased response
of the medulla to rising CO

levels, resulting in hypercapnia. Assuring ven-

tilation by placing anesthetized patients on ventilators or by administering
intermittent positive pressure ventilation helps to maintain normal blood
CO

levels. CO

levels can be measured by using an end-tidal CO

monitor

or by measuring arterial or venous blood gases. Arterial blood gas analysis
is preferred if assessment of both blood oxygen levels and CO

levels are

desired.

1053

L.L. Powell / Vet Clin Small Anim 32 (2002) 1049–1058

Severe pleural space disease can cause hypoventilation due to the inability

of the lungs to expand against air or ﬂuid in the thoracic cavity. Normally,
the chest wall expands on inspiration, generating negative pressure and
allowing the lungs to inﬂate and ﬁll with air. When air or ﬂuid within the
pleural cavity prevents lung inﬂation, collapse and atalectasis of the lung
lobes occur, preventing normal gas exchange and resulting primarily in
hypoxemia, but with severe pleural space disease (tension pneumothorax,
open/sucking chest wounds) a rise in blood CO

levels can be seen. Chest

wall disease such as ﬂail chest can cause hypoventilation due to the inability
of the chest wall to expand normally on inspiration, causing a decrease in air
ﬂow into the lungs.

Ventilation/perfusion mismatch

Ventilation/perfusion (V/Q) abnormalities occur when ventilation and

blood ﬂow in the lung are unequal, resulting in ineﬃcient transfer of oxygen
and CO

and subsequent hypoxemia [8]. This is a common cause of hypo-

xemia and respiratory failure. Examples of diseases that cause a decrease
in alveolar ventilation with preserved perfusion (low V/Q ratio) include
alveolar edema, pneumonia, alveolar hemorrhage, and interstitial lung dis-
ease. A high V/Q ratio occurs with decreased perfusion to the lung units
with preservation of ventilation. Pulmonary thromboembolism, heartworm
emboli, or other pulmonary vascular emboli cause decreased perfusion to
the lung and V/Q mismatch. An important feature of V/Q mismatch is that
the PaO

will increase when an enriched oxygen environment is provided

(increased inspired oxygen concentration, or FiO

Pulmonary edema is either cardiogenic or noncardiogenic in origin. Left-

sided heart failure results in increased intrapulmonary capillary hydrostatic
pressure, causing transudation of ﬂuid into the pulmonary interstitium and,
with increasing severity, into the alveoli. The ﬂuid is characterized by a low-
protein transudate. In contrast, noncardiogenic edema is a high-protein ﬂu-
id that enters the pulmonary interstitium and alveoli because of increased
vascular permeability. The mechanism of development of noncardiogenic
edema is not well understood, but it has been found to be associated with
an increase in intracranial pressure and an activation of the sympathetic
nervous system. Vasculitis from any cause can increase pulmonary vascular
permeability, predisposing to edema. Noncardiogenic edema might be
caused by severe primary lung injury or a systemic illness. Primary pulmo-
nary diseases that can lead to noncardiogenic edema include blunt trauma,
aspiration pneumonia, submersion injury (near-drowning), smoke inhala-
tion injury, pulmonary thromboembolism, or oxygen toxicity. Systemic dis-
eases that can result in the development of noncardiogenic edema include
neurogenic causes (increased intracranial pressure, seizure activity), electro-
cution, sepsis, systemic inﬂammatory response syndrome (leading to the

1054

L.L. Powell / Vet Clin Small Anim 32 (2002) 1049–1058

development of ARDS, or the acute respiratory distress syndrome), and
pancreatitis [9].

Pneumonia results in interstitial and alveolar accumulations of inﬂamma-

tory cells, ﬂuid, and bacteria. Causes of pneumonia include aspiration,
septicemia (viral or bacterial), immunosuppression, and secondary to acute
lung injury (submersion injury, smoke inhalation, mechanical ventilation,
oxygen toxicity).

Alveolar hemorrhage resulting in V/Q inequality can develop as a result

of blunt trauma (pulmonary contusion), severe thrombocytopenia (platelet
number <20,000/UL), or coagulopathy. Anticoagulant rodenticide toxicity
is a common cause of intrapulmonary hemorrhage. Other coagulopathies
include disseminated intravascular coagulation from various causes, hepatic
failure, and factor deﬁciencies.

Accumulation of edema ﬂuid, blood, inﬂammatory cells, or neoplastic

cells in the pulmonary interstitium is another cause of V/Q inequality. Inter-
stitial edema can be caused by left-sided heart failure or severe hypoprotei-
nemia. Coagulopathies, blunt trauma, and severe thrombocytopenia can
result in interstitial hemorrhage. Interstitial pneumonia causes an exudative
barrier to gas exchange. Metastatic or primary pulmonary neoplasia results
in an accumulation of neoplastic cells in the pulmonary interstitium, result-
ing in hypoxemia and potential respiratory failure.

Intrapulmonary shunt

An intrapulmonary shunt is deﬁned as a portion of the cardiac output

that enters the left side of the heart without being oxygenated in the lungs
because of pathology of the lung parenchyma [10]. There are three diﬀerent
types of intrapulmonary shunts described: anatomic shunts, capillary
shunts, and venous admixture.

An anatomic shunt describes blood that enters the left side of the heart

without ﬁrst passing through the lungs. Normally, about 2% to 5% of the
cardiac output directly enters the left side of the heart without being oxygen-
ated in the lungs through the bronchial, pleural, and thebesian veins [10]. An
example of a pathologic shunt would be a right-to-left intracardiac shunt.

Capillary shunts describe blood that passes through the lungs but does

not get oxygenated because it does not respire with alveolar gas. Atalectasis
and consolidation of a lung lobe are examples of capillary shunts.

Venous admixture occurs when blood traverses the pulmonary capillaries

and respires with alveoli that have a low oxygen tension. This occurs primar-
ily because of V/Q inequality.

Intrapulmonary shunting secondary to venous admixture is a common

cause of hypoxemia in veterinary patients. Any disease causing alveolar
pathology contributes to shunting and subsequent hypoxemia. Alveoli ﬁlled
with edema ﬂuid, inﬂammatory cells, bacteria, or blood impairs gas ex-
change, resulting in shunt and hypoxemia.

1055

L.L. Powell / Vet Clin Small Anim 32 (2002) 1049–1058

Oxygen tension indices such as arterial blood oxygen levels, the alveolar–

arterial oxygen gradient, and the PaO

/FiO

ratio can be used to measure

and monitor the extent of intrapulmonary shunt. Resistance to oxygen
administration characterizes true intrapulmonary shunts because gas ex-
change theoretically does not occur. Not all of the pulmonary parenchyma
is diseased, however, so oxygen therapy is recommended and might improve
hypoxemia, depending on the severity of disease.

Diﬀusion impairment

Diﬀusion impairment occurs when the blood–gas barrier is thickened so

that oxygen cannot equilibrate as red blood cells traverse the pulmonary
vasculature. With exercise, hypoxemia is especially severe because there is
a combined decrease in diﬀusion and contact time between the red blood cell
and oxygen within the alveoli [8]. CO

levels might be slightly increased, but

because CO

diﬀuses more readily than oxygen, blood levels are usually nor-

mal or low. Diﬀusion impairment as the sole cause of respiratory failure is
rare in veterinary patients. Some diseases causing diﬀusion impairment
include pulmonary interstitial pathology such as ﬁbrosis, interstitial pneu-
monia, and severe interstitial hemorrhage. Patients who have diﬀusion
impairment respond readily to an enriched oxygen environment because the
oxygen tension within the alveoli is elevated, increasing the amount of oxy-
gen available for gas exchange across the thickened diﬀusion barrier.

Low inspired oxygen concentration

Low inspired oxygen concentrations cause hypoxemia because of the low

oxygen tension in the inspired medium. Examples of situations in which low
inspired oxygen concentrations might be experienced include high altitudes,
anesthesia with inadvertent lack of oxygen administration (ie, equipment
failure, empty oxygen reservoir), and inhalation of altered air in the environ-
ment (ie, smoke inhalation, nitrous oxide administration). An enriched oxy-
gen environment immediately reverses hypoxemia in these cases.

Respiratory muscle fatigue

Respiratory muscle fatigue as a result of severe, prolonged tachypnea,

and dyspnea can contribute ultimately to respiratory failure. Increased ven-
tilatory work consumes excess oxygen, requires nutrients, and can result in
production of lactic acidosis from anaerobic metabolism in a hypoxemic
state [11]. Hypoxemia, poor nutrition, systemic illness, and metabolic
derangements contribute to the possible development of respiratory muscle
fatigue. Treatment includes rest of the respiratory muscles attained through

1056

L.L. Powell / Vet Clin Small Anim 32 (2002) 1049–1058

mechanical ventilation. With positive-pressure ventilation and associated
neuromuscular blockade or sedation, unloading of the respiratory muscles
occurs and fatigue is reversed.

Consequences of hypoxemia

Cellular respiration in mitochondria requires oxygen to sustain life. Tis-

sues exposed to low oxygen tension (hypoxia) must use anaerobic pathways
to provide metabolic energy. Lactic acid, a metabolite of anaerobic metab-
olism, increases in circulation, lowering blood pH (acidemia). Acidemia is
detrimental to many physiologic processes including normal cardiac and
neurologic function. Tissue hypoxia can result in end-organ damage includ-
ing hepatic, renal, cardiac, and cerebral failure, ultimately resulting in death.

Summary

There are many causes of respiratory failure in veterinary patients.

Assessment of oxygenation is imperative for the diagnosis and monitoring
of these patients. Oxygen therapy should be instituted when hypoxemia is
diagnosed to prevent tissue hypoxia, end-organ damage, and death. Meth-
ods of administering oxygen include commercial oxygen cages, mask oxy-
gen, nasal cannulation (for dogs), and intubation. Mechanical ventilation
is an option in many referral hospitals for patients who are severely hypo-
xemic and are not responding to inspired oxygen concentrations achieved
with other methods of oxygen administration. One rule of thumbused to
assess need for mechanical ventilation is a PaO

of less than 50 mm Hg

despite aggressive oxygen therapy, or a PaCO

of greater than 50 mm Hg

despite treatment for causes of hypoventilation. A mechanical ventilator has
the ability to vary the FiO

by increments of one, from 21% to 100% (0.21–1)

oxygen in inspired gas. Positive end-expiratory pressure (PEEP) is also
available on most ventilators. PEEP allows the alveoli to remain open on
expiration, allowing gas exchange to occur in both inspiration and expira-
tion. PEEP also helps diseased alveoli to inﬂate, increasing the available sur-
face area for gas exchange and improving arterial blood oxygen tension.
Because patients requiring mechanical ventilation have severe respiratory
failure that did not respond to conventional oxygen therapy, the prognosis
is guarded for most of these patients unless ventilation is instituted due to
primary hypoventilation and lung parenchyma is normal.

Hypoxemia caused by respiratory failure is a common problem in small

animal veterinary patients. Assessment of blood oxygenation and continual
monitoring of respiratory rate and eﬀort are essential in management of
these patients. Oxygen therapy should be instituted if hypoxemia is diag-
nosed. The prognosis depends on the underlying disease process and
response to treatment with an enriched oxygen environment.

1057

L.L. Powell / Vet Clin Small Anim 32 (2002) 1049–1058

References

[1] King LG, Hendricks JC. Clinical pulmonary function tests. In: Ettinger SJ, Feldman EC,

editors. Textbook of veterinary internal medicine, 4th edition, Vol. 1. Philadelphia, PA:
WB Saunders; 1995. p. 738–54.

[2] Van Pelt DR, Wingﬁeld WE, Wheeler SL, et al. Oxygen–tension based indices as predictors

of survival in critically ill dogs: clinical observations and review. JVECCS 1991;1:19–25.

[3] Shapiro BA, Peruzzi WT, Templin R. Normal ranges and interpretive guidelines. In:

Shapiro BA, editor. Clinical application of blood gases, 5th edition. St. Louis, MO:
Mosby Year Book; 1994. p. 57–67.

[4] Mensack S, Murtaugh R. Oxygen toxicity. Comp Contin Ed Pract Vet 1999;21:341–51.
[5] Covelli HD, Nessan VJ, Tuttle WK. Oxygen derived variables in acute respiratory failure.

Crit Care Med 1983;11:646–9.

[6] West JB. Gas transport to the periphery. In: Coryell PA, editor. Respiratory physiol-

ogy—the essentials, 5th edition. Baltimore, MD: Williams & Wilkins; 1995. p. 71–88.

[7] Barton LJ, Devey JJ, Gorski S, et al. Evaluation of transmittance and reﬂectance pulse

oximetry in a canine model of hypotension and desaturation. JVECCS 1996;6:21–8.

[8] West JB. Gas exchange. In: Kelly PJ, editor. Pulmonary pathophysiology—the essentials,

5th edition. Baltimore, MD: Williams & Wilkins; 1998. p. 17–34.

[9] Drobatz KJ, Saunders HM. Noncardiogenic pulmonary edema. In: Bonagura JD,

editor. Kirk’s current veterinary therapy XIII: small animal practice. Philadelphia, PA:
WB Saunders; 2000. p. 810–2.

[10] Shapiro BA, Peruzzi WT, Templin R. Assessment of the lung as an oxygenator. In: Shapiro

BA, editor. Clinical application of blood gases, 5th edition. St. Louis, MO: Mosby Year
Book; 1994. p. 85–112.

[11] MacIntyre NR. Ventilatory muscles and mechanical ventilatory support. Crit Care Med

1997;25:1106–7.

1058

L.L. Powell / Vet Clin Small Anim 32 (2002) 1049–1058

Respiratory muscle fatigue

Linda Barton, DVM

Emergency and Critical Care, The Animal Medical Center, 510 East 62nd Street,

New York, NY 10021, USA

The contribution of respiratory muscle fatigue to the development of ven-

tilatory failure has been the subject of considerable interest and has stimu-
lated much research. Experimental studies in dogs have shown respiratory
muscle fatigue to be a cause of ventilatory failure in both cardiogenic and
septic shock models [1,2]. In clinical conditions resulting in acute or chronic
hypercapnia, respiratory muscle fatigue is believed to occur; however, the
speciﬁc role of fatigue has been diﬃcult to prove.

As deﬁned by the National Heart, Lung, and Blood Institute–sponsored

Respiratory Muscle Fatigue Workshop Group, respiratory muscle fatigue is
a condition in which there is a loss in the capacity for developing force and/
or velocity of a muscle, resulting from muscle activity under load and which
is reversible by rest [3]. Muscle fatigue is distinguished from muscle weak-
ness as a reduction in force generation that is ﬁxed and not reversible by
rest, although muscle weakness may be a predisposition to muscle fatigue.
Muscle fatigue should not be considered in dichotomous terms (present or
absent) but rather as a continuum [4]. Fatigue is a process that begins when-
ever a muscle is subjected to an unsustainable load and may ultimately result
in exhaustion or task failure. Fatigue of the respiratory muscles progressing
to task failure results in ventilatory failure. Hypercapnia is the hallmark of
ventilatory failure. Ventilatory failure is characterized by hypoxia and
hypercapnia in contrast to failure of gas exchange, which is characterized
by hypoxia with normo- or hypocapnia. Although discussed as separate
entities, there are interrelations between failure of gas exchange and ventila-
tory failure. Most of the lung diseases that lead to hypoxia also increase the
work of breathing (WOB) and therefore the energy demands of the respira-
tory muscles. Hypoxia itself decreases the amount of energy available to the
respiratory muscles, predisposing them to fatigue [5].

Vet Clin Small Anim 32 (2002) 1059–1071

E-mail address: linda.barton@amcny.org (L. Barton).

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 0 3 6 - 0

Despite considerable research eﬀorts, the site and mechanism of the

decreased function produced by respiratory muscle fatigue have not been
fully elucidated. Theoretically, fatigue may occur at any point along the
extensive chain of command involved in voluntary muscle contrac-
tion, beginning with the brain and ending with the contractile machinery
(brain, spinal cord, nerve, neuromuscular junction, muscle cell membrane,
transverse tubular system, calcium release, actin-myosin activation, and
cross-bridge formation) [4]. Fatigue is generally considered in two broad
categories: failure to generate force because of reduced central motor output
(central fatigue) and failure to generate force because of fatigue either at the
neuromuscular junction or within the muscle machinery (peripheral fatigue)
[4]. Current evidence suggests that the decline in force seen during diaphrag-
matic fatigue can be attributed to both central and peripheral fatigue [4,6–8].
Studies in experimental animals and in healthy human volunteers and
patients suggest that the muscle is the primary site of fatigue and that
changes in central respiratory drive occur to protect the muscle [8,9]. Recent
experimental studies have shown that oxygen-derived free radicals generated
during strenuous contraction can modify respiratory muscle contractile
function and contribute to the development of muscle fatigue [10]. When
inspiratory muscles perform fatiguing work, the central controllers may
reﬂexively reduce inspiratory time, frequency, the duty cycle (the fraction
of the total respiratory cycle duration spent in inspiration), or inspiratory
drive, a strategy that may serve to save energy and avoid exhaustion at the
expense of hypoventilation [9]. Central fatigue may be an inescapable con-
sequence of the imposition of fatiguing loads to breathing and may repre-
sent an important protective mechanism that avoids the adverse eﬀects of
prolonged forceful contraction on the respiratory muscles [6,11].

Physiology of respiratory muscle fatigue

The respiratory muscles, the centers in the central nervous system con-

trolling them, the intervening neural connections, and the structures they
displace (the ribcage and the abdomen) form a pump, which performs the
vital function of ventilating the lungs [12]. Each time a spontaneous breath
is taken, the inspiratory muscles must generate a force suﬃcient to overcome
the elastic and ﬂow-resistive load imposed by the lungs and chest wall. The
elastic load represents the work performed on the tissue of the lung and
chest wall when a change in volume occurs. The ﬂow-resistive load is the
work performed to overcome airway, tissue, and viscous resistance to gas
ﬂow. The ability to take a breath is schematically represented in Figure 1
as a balance between the load imposed on the inspiratory muscles and neu-
romuscular competence. The load imposed on the respiratory muscles
equals the pressure developed by the inspiratory muscles (P

). The maximum

inspiratory pressure (P

I, max

) is a measure of neuromuscular competence.

Normally, the balance is weighed heavily in favor of neuromuscular

1060

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

competence (P

I, max

). It can be seen that the value of P

I, max

is determined

by the balance between load and competence. Increased inspiratory load or
decreased neuromuscular competence causes an increase in P

I, max

. Ven-

tilatory failure results when P

I, max

reaches a critical value. Table 1 lists

clinical conditions resulting in an increased P

I, max

In health, there are reserves in neuromuscular competence that permit

considerable increases in inspiratory load. For spontaneous ventilation to
continue, however, the inspiratory muscles must be capable of sustaining the
increased load over time. The ability of the respiratory muscles to sustain an
increased load without the appearance of fatigue is called endurance and is
determined by the balance between energy supply and energy demand [12].
Figure 2 illustrates the variables aﬀecting respiratory muscle endurance.
Energy supplies depend on the inspiratory muscle blood ﬂow, the concentra-
tions of oxygen and blood substrate concentrations, the muscle’s ability to
extract and utilize energy sources, and the muscle’s energy stores. Energy
demands increase proportionally with the mean tidal pressure developed
by the inspiratory muscles (P

), which is expressed as a fraction of the max-

imal inspiratory pressure (P

I, max

), the minute ventilation (V

), the

inspiratory duty cycle (t

tot

), and the mean inspiratory ﬂow rate (V

Energy demands are inversely related to the eﬃciency of the muscles
[5,9,12]. Under normal conditions, energy supplies are adequate to meet
demands and a large recruitable reserve exists. Fatigue develops when the
mean rate of energy demands exceeds the mean rate of energy.

It can be seen that the value of P

I, max

is determined by the balance

between inspiratory muscle load and neuromuscular competence. P

I, max

is also one of the determinants of respiratory muscle energy demands;

Fig. 1. The ability to take a spontaneous breath is determined by the balance between the load
imposed on the respiratory muscles (P

) and the neuromuscular competence of the ventilatory

pump (P

I, max

1061

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

therefore, the two balances (between load and competence and energy sup-
ply and demand) are linked. In rat studies, P

I, max

has been directly

related to diaphragm endurance time. Roussos et al [13] demonstrated that
the critical value of P

I, max

that could be generated indeﬁnitely at func-

tional residual capacity was around 0.06. Greater values of P

I, max

were

inversely related to the endurance time in a curvilinear fashion.

Factors predisposing to respiratory muscle fatigue

From the previous discussion, it can be seen that fatigue of respiratory

muscles can occur when there is an unfavorable balance between the factors
aﬀecting energy supply, energy demand, and neuromuscular competence.

Energy supply

Important factors aﬀecting the supply of energy to the respiratory

muscles are the, oxygen concentration of the blood, cardiac output, and
blood substrate concentration (ie, glucose, free fatty acids). Arterial oxygen
concentration is decreased with anemia, decreased hemoglobin oxygen

Table 1
Clinical conditions causing an increase in P

I, max

Increased load (P

)

Decreased neuromuscular competence (P

I, max

)

Increased restrictive load

Decreased central drive

Bronchospasm

Drug overdose

Airway edema/increased secretions

Brain stem lesion

Upper airway obstruction

Hypothyroidism

Ventilatory circuit resistance

(ventilated patients)

Malnutrition
Metabolic alkalosis

Endotracheal tube kinking

(ventilated patients)

Increased lung elastic load

Muscle weakness

Hyperinﬂation

(auto–positive end-expiratory pressure)

Electrolyte derangement
Malnutrition

Alveolar edema

Myopathy

Infection

Hyperinﬂation

Atelectasis

Corticosteroids

Interstitial inﬂammation/edema

Disuse atrophy

Lung tumor

Sepsis

Increased chest wall elastic load

Impaired nerve/neuromuscular transmission

Pleural eﬀusion

Phrenic nerve injury

Pneumothorax

Spinal cord lesion

Flail chest

Neuromuscular blockers

Tumor

Myasthenia gravis

Obesity

Aminoglycosides

Ascites

Botulism

Abdominal distention

1062

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

binding capacity, and hypoxemia. In addition to adequate oxygen delivery,
the muscle must be able to extract and utilize energy from the blood. In sep-
sis, increases in respiratory muscle oxygen consumption out of proportion to
load suggest that the processes of oxygenation and phosphorylation are
uncoupled [12,14]. Blood ﬂow to muscles may be decreased in low cardiac
output states. In experimental models of cardiogenic and septic shock, it has
been shown that blood ﬂow to the inspiratory muscles remained high, rep-
resenting a substantial percentage of cardiac output. The amount of blood
ﬂow was insuﬃcient, however, and led to fatigue of the respiratory muscles
and inability to maintain alveolar ventilation [1,2,12]. Blood ﬂow to muscles
can also be decreased during strenuous inspiratory eﬀorts. Forceful muscle
contractions cause compression of intramuscular vessels, limiting nutrient
blood ﬂow. Because unimpeded ﬂow blood occurs only during expiration,
increases in the duty cycle also decrease blood ﬂow to the muscles. In severe

Fig. 2. Respiratory muscles ultimately fatigue if the energy demands exceed the energy supplied
to the muscles. P

I, max

¼ inspiratory pressure/maximum inspiratory pressure; _V

¼ minute

ventilation; t

tot

¼ duty cycle (fraction of inspiration to total breathing cycle duration);

¼ mean inspiratory ﬂow (tidal volume/inspiratory time).

1063

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

asthma, inspiratory muscles may continue to contract during expiration,
further limiting blood ﬂow and increasing the vulnerability of the respiratory
muscles to fatigue [9]. Malnutrition and catabolic states can cause depletion
of glycogen and other energy stores, predisposing to muscle fatigue.

Energy demands

The clinical conditions causing increased inspiratory load are listed in

Table 1. Intrinsic positive end-expiratory pressure (PEEPi) refers to positive
pressure in the alveoli at the end of expiration. PEEPi develops from airﬂow
obstruction or decreased elastic recoil of the lung and has been detected in
patients with chronic obstructive pulmonary disease (COPD), cardiogenic
pulmonary edema, chest trauma, and pneumonia [12]. PEEPi increases the
lung elastic load, because the inspiratory muscles have to develop a pressure
equal to the level of PEEPi before airﬂow can begin.

Energy demands increase proportionally with increases in minute ventila-

tion. For Pa

to remain at its normal value, minute ventilation must

increase whenever there is an increase in carbon dioxide production or an
increase in dead space ventilation. Carbon dioxide production may increase
as a result of the following:

1. Fever or sepsis: carbon dioxide production increases during hyperther-

mia by approximately 9% to 14% for each degree Centigrade rise in tem-
perature.

2. Shivering: an increase in either physiologic or pathologic (ie, seizures)

muscle tone increases the metabolism of the muscles and thus carbon
dioxide production.

3. Agitation: carbon dioxide production is increased secondary to in-

creased muscle activity.

4. Severe burns or trauma: being catabolic states, these conditions elevate

carbon dioxide production.

5. Hyperalimentation: intravenous hyperalimentation in excess of caloric

requirement increases carbon dioxide production [12,15].

It has previously been suggested that diets high in carbohydrates were

harmful to patients with ventilatory compromise. To elucidate the relative
importance of excess carbohydrates versus excess total calories in carbon
dioxide production, Talpers et al [16] compared three isocaloric regimens
containing 40%, 60%, and 75% carbohydrates and found no diﬀerence in the
amount of carbon dioxide produced. In contrast, when carbohydrates were
held constant but total calories were increased, carbon dioxide production
increased from 181 mL/min

when calories were equivalent to the calcu-

lated resting energy requirement (REE) to 211 mL/min

·1.5 REE and

244 mL/min

·2.0 REE [12].

Physiologic dead space is increased in virtually all the diverse processes

that aﬀect the lung parenchyma and the distribution of airﬂow. Alveolar dead

1064

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

space increases if lung perfusion is reduced. Decreased perfusion is seen sec-
ondary to pulmonary embolism and hypovolemia. During positive-pressure
ventilation, there is alveolar wall distention and compression of the capilla-
ries of the well-ventilated alveoli, causing increased dead space [12].

Eﬃciency is deﬁned as the ratio of mechanical work to the oxygen cost of

breathing. Therefore, energy demands increase when muscle eﬃciency is
decreased. Hyperinﬂation reduces the eﬃciency of the respiratory muscles
[5,9,12]. As lung volume increases, shortening of the inspiratory muscles and
alterations in their geometry require greater excitation and energy consump-
tion to perform a given amount of work. Like other skeletal muscles, respi-
ratory muscles obey the length-tension relation. At any given level of
activation, changes in muscle ﬁber length alter active and passive tension,
modifying actin-myosin interaction. At a speciﬁc ﬁber length (Lo), active
tension is maximal, whereas it declines below and above the Lo [12,17].

Respiratory muscle length depends largely on lung volume and, to a lesser

extent, on thoracoabdominal conﬁguration. Animal experiments have shown
that the Lo for inspiratory muscles (diaphragm and intercostals) is near
residual volume [12]. Respiratory muscle eﬃciency is also reduced when
the inspiratory load is increased [5].

Decreased neuromuscular competence

Decreased respiratory drive or altered neural transmission may cause a

decrease in P

I, max

. The most common cause of decreased competence in crit-

ically ill patients is muscle weakness. Mechanically ventilated patients may
develop muscle weakness secondary to disuse atrophy. Electrolyte imbalances
(hypocalcemia, hypokalemia, hypophosphatemia, and both hypo- and hyper-
magnesemia) can adversely aﬀect muscle strength. Molloy et al [17] reported
an improvement in all measured parameters of respiratory muscle power
when 17 hypomagnesemic patients were treated intravenously with magne-
sium. Administration of corticosteroids has been shown to cause respiratory
muscle weakness. Decramer et al [19] reported a signiﬁcant relation between
maximal inspiratory pressure measured 10 days after admission and the aver-
age daily dose of corticosteroid during the previous 6 months in patients with
an exacerbation of COPD and asthma. Patients with no underlying pulmo-
nary disease developed reversible inspiratory muscle weakness as a result of
high-dose steroids administered over several weeks [20]. Fluorinated steroids,
such as triamcinolone and dexamethasone, lead to more marked myopathy
than nonﬂuorinated steroids, such as hydrocortisone, prednisone, and corti-
sone acetate [15]. Muscle strength is also aﬀected by malnutrition.

Detection of respiratory muscle fatigue

The diagnosis of fatigue requires demonstration of a decrease in force

generation. Such measurements can be made in experimental settings but

1065

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

have proven more diﬃcult in patient populations. In the clinical setting, it is
diﬃcult to control other variables (eg, changes in lung volume, changes in
chest wall geometry, patient cooperation) that may aﬀect the result
[4,5,17]. Bedside measurement is also diﬃcult because of lack of baseline
measurement before fatigue.

Analysis of the electromyographic (EMG) power spectrum has been used

to detect inspiratory muscle fatigue. It has been shown that during fatigue,
the power of the low-frequency components of the EMG power spectrum
increases, whereas the power of the high-frequency components decreases.
The shift in the power spectrum occurs before the loss of force generation,
making it a useful test to monitor for the development of fatigue. Use of this
analysis is limited, however, because measurement of the power spectrum at a
single point in time is inadequate to detect fatigue; the power shift must be
observed as the muscle passes from the rested to the fatigued state. The tech-
nique has been used to detect the development of fatigue during weaning
from mechanical ventilation [4,5,17,18,21]. Recently, some doubt has been
cast on the validity of this technique. The measurement of electric activity
of the inspiratory muscles can be inﬂuenced by changes in the spatial relations
between the recording electrodes and the muscle. Secondly, the cellular mech-
anisms responsible for the shifts in the power spectrum are unknown [17].

The rate of relaxation of the diaphragm has been used to detect muscle

fatigue. An early physiologic event in the progress of a fatiguing contraction
is the slowing of the muscle relaxation rate. Maximum relaxation rate
(MRR) also requires serial measurements for detection of fatigue. The wide
range of normal values for MRR makes it diﬃcult to obtain useful informa-
tion from a single measurement [4]. Goldstone et al [22] measured MRR in a
group of intubated patients before and during weaning. Serial measure-
ments of MRR remained unchanged in patients who weaned successfully
and slowed in patients failing to wean.

A sequence of changes in breathing pattern suggestive of respiratory

muscle fatigue has been described. First, there is an early stage of rapid shal-
low breathing. There is then an inward displacement of the abdomen accom-
panied by a decrease in abdominal pressure during inspiration (abdominal
paradox) and uncoordinated chest wall movements characterized by altera-
tion between predominantly abdominal and ribcage displacements during
inspiration (respiratory alternans). These changes are generally seen before
increases in carbon dioxide. Subjects become bradycardic as ventilatory fail-
ure develops [9,13,21]. Cohen et al [21] compared changes in the EMG
power spectrum to observed physical examination parameters in 12 patients
who experienced diﬃculty during weaning. Seven patients developed
changes in the EMG power spectrum indicative of fatigue. Physical exami-
nation changes were noted in these patients, including increased respiratory
rate (6/7), abdominal paradox (6/7), and respiratory alternans (4/7). In all
instances, the shift in the EMG power spectrum was seen in advance of the
changes in respiratory pattern. Changes in the respiratory pattern were not

1066

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

seen in any of the patients without EMC evidence of diaphragmatic fatigue
[21]. Several studies performed subsequently have refuted these ﬁndings
[23,24]. In a study of healthy volunteers breathing against restrictive loads,
Tobin et al [24] found that fatigue was neither necessary nor suﬃcient for
the induction of abnormal ribcage-abdominal motion.

As previously stated, fatigue is a continuous process that begins whenever

the respiratory muscles are subjected to an unsustainable load; therefore, it
would be clinically useful to measure the demands placed on the inspiratory
muscles. WOB is described as the work necessary to overcome the ‘‘after-
load’’ imposed on the respiratory system [25]. It is used as an index of the
energy demands on the respiratory muscles. WOB can be measured in one
of two ways in mechanically ventilated patients The oxygen cost of breath-
ing is the diﬀerence between total body oxygen uptake (V

) during con-

trolled ventilation with the muscles at rest and V

during spontaneous

breathing. Bedside measurements of oxygen consumption are reasonable,
but measurements made in critically ill patients have proven unreliable.
WOB can be more accurately determined by measuring the mechanical
WOB. Measurements of mechanical work have been neglected in the past
because of the complex methodology required to obtain accurate and reli-
able results. Computerized bedside monitoring of WOB has recently become
available in the human market, stimulating increased interest in using WOB
measurements to identifying fatiguing loads on the respiratory muscles,
especially during weaning from mechanical ventilation [25–27].

Clinical conditions associated with inspiratory muscle fatigue

Fatigue of respiratory muscles can be demonstrated in experimental ani-

mals and healthy human volunteers when the muscles are subjected to an
increased load. Respiratory muscle fatigue as a cause of ventilatory failure
has been proven in experimental models of cardiogenic and septic shock
in the dog [1,2]. There is growing evidence that respiratory insuﬃciency in
patients with septic shock, traditionally attributed to acute lung injury, may
also be associated with respiratory muscle fatigue and injury [14]. Respira-
tory muscle fatigue has been demonstrated in patients failing to wean from
mechanical ventilation. Brochard et al [26] demonstrated changes in the
EMG power spectrum consistent with diaphragmatic fatigue in seven of
eight patients who met the usual criteria for weaning but failed to wean. The
contribution of muscle fatigue to other causes of acute and chronic hyper-
capnia is likely but has not yet been proven.

Many of the clinical studies in this area have focused on respiratory

muscle fatigue in patients undergoing weaning from mechanical ventilation.
Mechanical ventilation and respiratory muscle rest are recommended for
patients with ventilatory failure. Respiratory muscle fatigue can result from
mechanical ventilation, however. Based on the mode of ventilation selected,
the workload of the respiratory muscles is in one of three states: totally

1067

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

unloaded, partially loaded, or fully loaded [25]. During controlled ventila-
tion, no inspiratory eﬀort is made by the patient; the ventilator provides all
the necessary work. Controlled ventilation causes respiratory muscles to be
unloaded, predisposing to atrophy. Muscles that are used most often, such
as the inspiratory muscles, atrophy the fastest [4]. In contrast, increased
muscle loading leading to muscle fatigue can occur from insuﬃcient ventila-
tory support. Increased WOB may occur with ventilatory modes tradition-
ally thought to provide respiratory muscle rest, such as assist-control (AC)
and synchronized intermittent mandatory ventilation (SIMV) [3,25,27,29].
Studies have shown that substantial patient work is performed during SIMV
and AC [3,29]. In one study, patient work was 33% to 50% of the work
required to passively inﬂate the chest and, on average, accounted for 63%
of the total work during spontaneous breathing [29]. With either fatigue
or atrophy, the respiratory muscles are weak and incapable of generat-
ing suﬃcient force to maintain alveolar ventilation and allow weaning [26].
Civetta [30] has suggested the term nosocomial respiratory failure or iatro-
genic ventilator dependency to describe the inappropriate prolongation of
ventilatory support caused by either respiratory muscle atrophy or fatigue.
Newer modes of ventilation, such as pressure support, allow partial loading
of respiratory muscles to prevent nosocomial respiratory failure. Brochard
et al [26] reported that EMG power spectrum changes consistent with dia-
phragmatic fatigue in patients failing to wean could be reversed with the
addition of 15 cm H

O of pressure support.

With the recent availability of bedside monitoring of WOB, there is grow-

ing interest in the estimation of WOB in the management of mechanically
ventilated patients, especially during weaning. It is recommended that ven-
tilatory support be titrated to maintain normal physiologic work [25,26,28].
Ventilatory therapy guided by WOB estimates has been advocated to
prevent muscle fatigue and allow for more rapid weaning, resulting in a
reduction in length of stay in the intensive care unit and hospital costs
[25,26]. WOB measurements have been used to guide weaning [11,31]. In
each of these studies, WOB was measured if patients failed a 20-minute
spontaneous breathing trial. If the physiologic WOB was not excessive, the
patient was extubated despite tachypnea. In the study by Kirton et al [11], 20
of 21 patients were successfully weaned. In the study by DeHaven et al [31],
97 of 105 patients were successfully weaned. A weaning protocol based on
WOB estimation was shown to result in more aggressive weaning [32]. On
average, patients were weaned 1.68 days faster on the WOB protocol than
patients weaned on a conventional protocol.

Treatment

Energy supply, energy demand, and neuromuscular competence are

closely linked, and imbalances in these parameters can lead to respiratory
muscle fatigue. Therefore, it is rational to direct therapeutic eﬀorts toward

1068

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

increasing energy supply to the respiratory muscles, minimizing energy
demands, and maximizing neuromuscular competence by improving con-
tractility and optimizing respiratory drive.

Increasing energy supply

Energy supply to the respiratory muscles can be increased by improving

cardiac output. Decreases in hemoglobin concentration or Pa

cause a

reduction in the oxygen content of arterial blood and therefore decreased
oxygen delivery. Anemia or hypoxia should be identiﬁed and corrected to
improve oxygen delivery to the respiratory muscles.

Minimizing energy demand

Therapy directed at reducing airway resistance by bronchodilation or

increasing pulmonary compliance by treating pulmonary edema reduces the
load and the energy demands of the inspiratory muscles.

Respiratory muscle energy demands can be substantially decreased by

mechanical ventilation. Controlled ventilation to provide total rest is advo-
cated for fatigued muscles. A 24-hour period of complete rest has been
advocated as a reasonable time to allow muscle recovery from fatigue
[25]. Mechanical ventilation is also recommended in circumstances where
cardiac output is inadequate. In conditions like cardiogenic shock, when
total body oxygen delivery is reduced, delivery of blood to the working res-
piratory muscles may ‘‘steal’’ oxygen from other tissues, predisposing them
to dysfunction [5,9,12,33]. Viires et al [33] have demonstrated this ‘‘stealing
eﬀect’’ in experimental models of cardiogenic shock. They have shown that
the respiratory muscles of spontaneously breathing dogs with a low cardiac
output produced by pericardial tamponade received more than 20% of the
cardiac output compared with 3% when the animals were paralyzed and
ventilated. The large fraction of the cardiac output taken up by the respira-
tory muscles in the spontaneously breathing animals resulted in reduced
blood ﬂow to the brain, liver, and other skeletal muscles compared with the
mechanically ventilated animals with a similar reduction in cardiac output.

Maximizing neuromuscular competence

Muscle strength is an important component of neuromuscular com-

petence. Treatable or avoidable causes of muscle weakness, including hy-
percapnia, acidosis, hypocalcemia, hypokalemia, hypomagnesemia, and
hypophosphatemia, should not be ignored. Inadequate nutrition adversely
aﬀects muscle strength. The strength of the ventilatory pump can be im-
proved with nutritional repletion.

Attention has been given to the use of pharmacologic agents to improve

the contractility and endurance of the respiratory muscles. A number of

1069

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

drugs, including xanthines, glycosides, catecholamines, and phosphodiester-
ase inhibitors, have been investigated [4,5,9,34]. Theophylline has been
shown to have a positive inotropic eﬀect on respiratory muscles at therapeu-
tic doses. The eﬀects of theophylline seem to be greater on fatigued muscle
than on rested muscle. The mechanism of action is not clear, but it is thought
to facilitate the inﬂux of calcium through the slow channels and by activation
of a calcium-induced calcium release from the sarcoplasmic reticulum [4].

Speciﬁc training of the respiratory muscles has been shown to enhance

strength and endurance in human patients with chronically increased
inspiratory loads; however, this requires a level of patient cooperation not
available with veterinary patients.

References

[1] Aubier M, Trippenbach T, Roussos C. Respiratory muscle fatigue during cardiogenic

shock. J Appl Physiol 1981;51:499–508.

[2] Hussain SN, Magder AC, Roussos C. Respiratory muscle fatigue: a cause of ventilatory

failure in septic shock. J Appl Physiol 1985;58:2033–40.

[3] NHLBI Workshop on Respiratory Muscle Fatigue. Report of the Respiratory Muscle

Fatigue Workshop Group. Am Rev Respir Dis 1990;142:474–80.

[4] Roussos C, Zakynthinos S. Fatigue of the respiratory muscles. Intensive Care Med

1996;22:134–55.

[5] Roussos C, Macklem PT. The respiratory muscles. N Engl J Med 1982;307:786–97.
[6] DeTroyer A. Respiratory muscle function. In: Shoemaker WC, Ayres SM, Grenvik A, et al,

editors. Textbook of critical care. 4th edition. Philadelphia: WB Saunders; 2000. p. 1172–84.

[7] Gandevia SC, Allen GM, Butler JE, et al. Human respiratory muscles: sensations, reﬂexes

and fatiguability. Clin Exp Pharmacol Physiol 1998;25:757–63.

[8] McKenzie DK, Bellemare F. Respiratory muscle fatigue. Adv Exp Med Biol 1995;384:

401–14.

[9] Roussos C, Macklem PT. Clinical implications of respiratory muscle fatigue. In: Fishman

AP, editor. Pulmonary disease and disorders. 2nd edition. New York: McGraw-Hill; 1988.
p. 2275–86.

[10] Supinski G. Free radical induced respiratory muscle dysfunction. Mol Cell Biochem 1998;

179:125–34.

[11] Kirton OC, DeHaven B, Morgan JP, et al. Elevated imposed work of breathing

masquerading as ventilator weaning intolerance. Chest 1995;108:1021–5.

[12] Vassilakopoulos T, Zakynthinos S, Roussos C. Respiratory muscles and weaning failure.

Eur Respir J 1996;9:2383–400.

[13] Roussos C, Fixley M, Gross D, et al. Fatigue of inspiratory muscles and their synergic

behaviour. J Appl Physiol 1979;46:897–904.

[14] Hussain SN. Respiratory muscle dysfunction in sepsis. Mol Cell Biochem 1998;179:125–34.
[15] Reid WD, MacGowan NA. Respiratory muscle injury in animal models and humans. Mol

Cell Biochem 1998;179:63–80.

[16] Talpers SS, Romberger DJ, Bunce SB, et al. Nutritionally associated increased carbon

dioxide production (excess total calories vs high proportion of carbohydrate calories).
Chest 1992;102:551–5.

[17] Molloy DW, Dhingra S, Solven F, et al. Hypomagnesemia and respiratory muscle power.

Am Rev Respir Dis 1984;129:497–8.

[18] Marini JJ, Rodriguez RM, Lamb V. The inspiratory workload of patient initiated

mechanical ventilation. Am Rev Respir Dis 1986;134:902–9.

1070

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

[19] Decramer M, Lacquet LM, Fagard R, et al. Corticosteroids contribute to muscle weakness

in chronic airﬂow obstruction. Am J Respir Crit Care Med 1994;150:11–6.

[20] Weiner P, Azgad Y, Weiner M. Inspiratory muscle training during treatment with cor-

ticosteroids in humans. Chest 1995;107:1041–4.

[21] Cohen CA, Zagelbaum G, Gross D, et al. Clinical manifestations of inspiratory muscle

fatigue. Am J Med 1982;73:308–16.

[22] Goldstone JC, Green M, Moxham J. Maximum relaxation rate of the diaphragm during

weaning from mechanical ventilation. Thorax 1994;49:54–60.

[23] Tobin MJ, Guenther SM, Perez W, et al. Konno-Mead analysis of ribcage-abdominal

motion during successful and unsuccessful trials of weaning from mechanical ventilation.
Am Rev Respir Dis 1987;135:1320–8.

[24] Tobin MJ, Guenther SM, Perez W, et al. Does rib cage-abdominal paradox signify

respiratory muscle fatigue. Am J Med 1982;73:308–16.

[25] French CJ. Work of breathing measurements in the critically ill patient. Anaesth Intensive

Care 1999;27:561–73.

[26] Brochard L, Harf A, Lorino H, et al. Inspiratory pressure support prevents diaphragmatic

fatigue during weaning from mechanical ventilation. Am Rev Respir Dis 1989;139:513–21.

[27] Tobin MJ, Laghi F, Jubran A. Respiratory muscle dysfunction in mechanically-ventilated

patients. Mol Cell Biochem 1998;179:87–98.

[28] Banner MJ. Respiratory muscle loading and the work of breathing. J Cardiothor Vasc

Anesth 1995;9:192–204.

[29] Marini JJ, Capps JS, Culver BH. The inspiratory work of breathing during assisted

mechanical ventilation. Chest 1985;87:612–8.

[30] Civetta JM. Nosocomial respiratory failure or iatrogenic ventilator dependency. Crit Care

Med 1993;21:171–3.

[31] DeHaven CB, Kirton OC, Morgan JP, et al. Breathing measurements reduced false-

negative classiﬁcation of tachypneic preextubation trial failures. Crit Care Med 1996;24:
976–80.

[32] Gluck EH, Barkoviak MJ, Balk RA, et al. Medical eﬀectiveness of esophageal balloon

pressure manometry in weaning patients from mechanical ventilation. Crit Care Med
1995;23:504–9.

[33] Viires N, Sillye G, Aubier M, et al. Regional blood ﬂow distribution in dogs during induced

hypotension and low cardiac output: spontaneous breathing versus artiﬁcial ventilation.
J Clin Invest 1983;72:935–47.

[34] Fujii Y, Hoshi T, Toyooka H. Colforsin daropate improves contractility in fatigued canine

diaphragm. Anesth Analg 2001;92:762–6.

1071

L. Barton / Vet Clin Small Anim 32 (2002) 1059–1071

Respiratory pharmacotherapy

in emergency and critical care medicine

Elizabeth A. Rozanski, DVM

Mark P. Rondeau, DVM

Section of Critical Care, Department of Clinical Sciences, Tufts University,

200 Westboro Road, North Grafton, MA 01536, USA

Section of Internal Medicine, Department of Clinical Studies, 3900 Delaney,

University of Pennsylvania–Philadelphia, Philadelphia, PA 19104, USA

Diagnosis and management of respiratory conditions in critically ill ani-

mals may be challenging. Animals with respiratory impairment may be dif-
ﬁcult to evaluate completely because of stress and the potential for
subsequent worsening of respiratory distress and hypoxemia. Additionally,
because of the signiﬁcant reserve of the respiratory system, the underlying
condition may be quite advanced by the time that respiratory distress
becomes apparent. The focus of this article is on pharmacotherapeutics of
respiratory diseases aﬀecting critically ill small animal patients. Conditions
that often aﬀect the respiratory function of the critically ill dog include
upper airway obstruction, trauma, pneumonia, acute lung injury (ALI) or
acute respiratory distress syndrome (ARDS), pulmonary thromboembolism
(PTE), and pulmonary edema (cardiogenic and noncardiogenic). Conditions
that frequently aﬀect the cat include asthma/chronic bronchitis, pleural eﬀu-
sion, pulmonary edema, and, rarely, pneumonia. Neoplasia may initially
present in both species as respiratory distress, but it is not the focus of this
article. Additionally, the interested reader is referred to a multitude of excel-
lent reviews for more information on the management of the patient with
chronic pulmonary disease [1–3].

Emergency management of respiratory distress

Initial management of the patient with respiratory distress should focus

on minimizing patient stress and identifying diﬀerentials based on historical

Vet Clin Small Anim 32 (2002) 1073–1086

* Corresponding author.

E-mail address: Elizabeth.rozanski@tufts.edu (E.A. Rozanski).

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 -5 6 1 6 ( 0 2 ) 0 0 0 3 9 -6

and brief physical examination ﬁndings. Supplemental oxygen may be
administered either via a facemask, oxygen cage, or as ‘‘ﬂow-by’’ with oxy-
gen tubing held up close to the mouth and nares. Nasal oxygen may be too
stressful to be placed on patient arrival. Rarely, a patient with severe distress
requires emergency induction of anesthesia and manual ventilation with an
Ambu bag (Harrell Medical, Lake Oswego, OR), anesthesia machine, or
mechanical ventilator (Fig. 1). The clinician is urged to remember that in
animals with marked distress or respiratory embarrassment, the doses of
anesthetic agents required to permit endotracheal intubation are reduced.

Upper airway obstruction

Upper airway obstructions can usually be rapidly appreciated as a result

of loud and stridorous breathing [4]. The two most common upper airway
obstructions identiﬁed in a critical care setting are laryngeal paralysis and
brachycephalic airway syndrome. Most animals aﬀected with laryngeal
paralysis are older large-breed (eg, retrievers and setters) dogs with a more
chronic history of loud or harsh breathing and perhaps a change in bark.
Dogs are often presented to an emergency facility in moderate to severe res-
piratory distress with an acute exacerbation. Dogs are often hyperthermic,
sometimes in excess of 105

°F. Successful pharmacologic management of

aﬀected dogs includes supplemental oxygen, sedation, anti-inﬂammatory
agents, intravenous ﬂuids, and, possibly, general anesthesia/tracheostomy
(Table 1). The most commonly used drug for sedation is low-dose acepro-
mazine. Typically, many dogs improve if the anxiety associated with upper
airway obstruction is relieved. The clinician should recall that laryngeal
paralysis is a slowly progressive condition and that a dog may breathe

Fig. 1. A bulldog with a temporary tracheostomy for treatment of severe brachycephalic airway
syndrome.

1074

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

adequately through a small laryngeal aperture. Because of the increased air-
ﬂow rates across a paralyzed larynx, edema and erythema of the arytenoids
are commonly observed. Occasionally, everted laryngeal saccules develop
and further occlude the ﬂow of air. Most dogs beneﬁt from some anti-
inﬂammatory drugs, such as dexamethasone or prednisone. Long-term ther-
apy may include medical management or surgical palliative procedures.
Dogs with laryngeal paralysis are prone to aspiration pneumonia, particu-
larly after surgery because of loss of the ability to guard the airway [5]. Dogs
with acute worsening of brachycephalic airway syndrome are typically
treated with similar pharmaceutic agents and, eventually, further medical
or surgical interventions, such as soft palate resection or even permanent
tracheostomy (see Fig. 1).

Severe upper airway disease in cats is uncommon but may occur. Naso-

pharyngeal polyps may result in partial upper airway obstruction, as can
laryngeal tumors, laryngospasm, or profound airway swelling. Laryngeal
paralysis is rare in the cat but has been reported [6]. Although some cats
have a brachycephalic conﬁrmation (eg, Persians), it is uncommon for sur-
gical or medical management to be required.

Trauma

Dogs frequently experience thoracic trauma. The most common mecha-

nism of injury is being hit by a car, although bite wounds, falls, and kicks
also occur. Cats are much less commonly presented with thoracic trauma.
Unfortunately, this most likely represents a high immediate death rate in the
cat with major thoracic wounds. Thoracic trauma is typically treated with
supportive care, including rest, supplemental oxygen, thoracocentesis for
pneumothorax, and intravenous ﬂuids as needed to maintain perfusion. A
recent retrospective study concluded that antibiotics were unwarranted in
isolated canine pulmonary contusions [7]. Some controversy exists regarding
ﬂuid therapy in pulmonary contusions [9,34]. Some authors worry that large
volumes of crystalloids may leak more easily across a damaged endothelium
and magnify extravascular lung water, whereas other investigators believe
that extravasation of colloids may be even more detrimental, because

Table 1
Pharmaceutic agents used in the medical management of upper airway disease

Pharmaceutic agent

Dose

Route

Comments

Acepromazine

0.01–0.05 mg/kg

IV, IM

Sedative, do not exceed

3 mg per dog

Dexamethasone

0.01–0.25 mg/kg

Anti-inﬂammatory

Prednisone

0.25–1.00 mg/kg

SC, PO

Anti-inﬂammatory

Butorphanol

0.1–0.4 mg/kg

IV, IM

Analgesic, sedative

Abbreviations: IV, intravenous; IM, intramuscularly; SC, subcutaneously; PO, orally.

Diﬀerent medications may be combined in an individual patient.

1075

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

colloids may persist in the interstitial space for long periods as a result of
their greater size. The best current recommendations include maintaining
adequate but not excessive intravascular volume and carefully monitoring
the animal for progression of respiratory diﬃculty [7–9]. Diuretics, such as
furosemide, should be avoided in an animal with hypovolemia, despite the
presence of adventitial lung sounds. The clinician is urged to remember that
pulmonary crackles signify ﬂuid rather than being speciﬁc for pulmonary
edema caused by vascular overload.

Pneumonia

Pneumonia is both a common presenting complaint and a common devel-

opment in hospitalized dogs. Pneumonia is quite rare in cats. Most frequently,
pneumonia in dogs develops as a result of aspiration, although community-
acquired fungal and hematogenous causes are also possible. Clinical ﬁnd-
ings often include alveolar inﬁltrates on thoracic radiographs, cough, fever,
and lethargy. Some dogs have concurrent megaesophagus, laryngeal paraly-
sis, or a history of extreme weakness or vomiting. Cytologic evaluation of air-
way secretions, which may be obtained by performing a tracheal wash, can
document an inﬂammatory response. Treatment of pneumonia should
include eliminating the underlying cause if possible, using appropriate antibi-
otics, providing supplemental oxygen as needed, and performing physiother-
apy (eg, nebulization with coupage) [10,11]. Antibiotic therapy is ideally
chosen based on bacterial culture and sensitivity testing of ﬂuid samples from
tracheal wash. Broad-spectrum antibiotics should be initiated as soon as
pneumonia is identiﬁed and appropriate samples for culture have been
obtained while awaiting receipt of culture results (Table 2). For a commun-
ity-acquired pneumonia thought to be due to Bordetella bronchiseptica, a
recent microbiologic survey suggests that most isolates are susceptible to
tetracycline, doxycycline, enroﬂoxacin, and amoxicillin/clavulanic acid [12].
In endemic areas, fungal pneumonias (blastomycosis, histoplasmosis, and
coccidiomycosis) may result in moderate to severe pulmonary disease.

Table 2
Antibiotics used in treatment of pneumonia

Antibiotic

Dose
(mg/kg)

Route

Gram-
positive

Gram-
negative

Anaerobes

Ampicillin

22 every 6 hours

þþþ

Cefazolin

22 every 8 hours

þþþ

þþ

Gentimicin

5–6 every 24 hours

þþþ

Enroﬂoxacin

2.5–10 every 12–24

hours

IM, IV, PO

þþþ

Metronidazole

10 every 8 hours

þþþ

Abbreviations: IV, intravenous; IM, intramuscular; PO, orally;

þ, positive.

Many other individual antibiotics and combinations may also be used. Ideal antimicrobial

therapy is based on bacterial culture and sensitivity data.

1076

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

In critically ill hospitalized dogs, therapy is often begun pending bacterial

culture and sensitivity results with intravenous combinations like ampicillin
and enroﬂoxacin or cefazolin, gentamicin, and metronidazole. In hospital-
acquired infections expected to be resistant, antibiotic choices may include
imipenem, amikacin, or cefoxitin. Nebulization and coupage are often used
as adjuvants in therapy of pneumonia, although no controlled trials exist in
the veterinary literature reporting their beneﬁt. Cats with suspected pneu-
monia should also be treated with intravenous antibiotics and supplemental
oxygen as required. It is not uncommon for cats that are suspected radio-
graphically of having pneumonia to have congestive heart failure and pul-
monary edema.

Acute lung injury and acute respiratory distress syndrome

Acute lung injury and acute respiratory distress syndrome are terms that

are applied to human beings with respiratory distress. These terms have
recently been applied to animals [13,14,33]. These conditions are deﬁned as
clinical syndromes of respiratory distress characterized by the presence of
an antecedent event (eg, trauma/sepsis), the development of bilateral alveo-
lar inﬁltrates on thoracic radiographs, decreased pulmonary compliance,
and the absence of congestive heart failure (as evidenced by pulmonary
capillary wedge pressure of <18 mm Hg) [15]. Diﬀerentiation of ALI and
ARDS in people is based on the presence of these criteria and assessment
of the ratio of Pa

to the fraction of inspired oxygen (Fi

), with a value

of <300 equal to ALI and a value of <200 equal to ARDS. No consensus
statement has yet been produced by the American College of Veterinary
Emergency and Critical Care to apply to animals, although it seems likely
that similar criteria will be proposed in dogs. It is less clear as to whether
cats develop ALI or ARDS. Certainly, respiratory distress does develop
in the critically ill cat, but it seems that this species is more prone to volume
overload/congestive heart failure or the development of pleural eﬀusion
rather than to ALI or ARDS.

Pharmacologically, no speciﬁc therapies have been found to be helpful in

the treatment of human beings with ARDS. Altered mechanical ventilatory
strategies seem to be beneﬁcial, and many active research programs exist to
try to elucidate both the pathogenesis and therapy for ARDS [15]. Current
veterinary recommendations include careful monitoring of the patient at
risk, limiting peak pressures in ventilated animals, and treating aggressively
for the underlying cause (Fig. 2).

Pulmonary thromboembolism

PTE is a condition that has been recognized with increasing frequency in

critically ill dogs. Cats have been identiﬁed with PTE, but it is exceedingly

1077

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

rare [16,17]. PTE may be a challenging condition to identify successfully in
dogs as well as in people, because the clinical signs may vary from vague to
peracute death. In practice, anticipation of the patient at risk for PTE is
important. Risk factors that have been documented in dogs with PTE
include immune-mediated hemolytic anemia, Cushing’s syndrome, sepsis,
and neoplasia [18,19].

Pharmacologic interventions used in the treatment of PTE include anti-

coagulant therapy and thrombolytic therapies (Table 3). All animals with
PTE also beneﬁt from standard hemodynamic and respiratory support.

Anticoagulant therapies used in animals include unfractionated heparin

(UFH), low-molecular-weight heparin (LMWH), and warfarin. UFH has
been the most widely used anticoagulant in veterinary patients because of
its widespread availability and low cost. Both forms of heparin act primarily
to limit the eventual conversion of ﬁbrinogen to ﬁbrin by accelerating the
action of antithrombin III in inhibiting activated coagulation factors (II,
IX, X, XI, and XII). In people, heparin has been associated with the devel-
opment of heparin-induced thrombocytopenia (HIT). This is an acute onset
of marked thrombocytopenia in an individual receiving heparin therapy.
Aﬀected patients have a paradoxic tendency for thrombosis instead of hem-
orrhage. It is thought to be an immune-mediated reaction, with antibodies
formed against heparin and complexed with platelet factor 4 [20]. LMWH
diﬀers from UFH by being more speciﬁc for factor X, having a more pre-
dictable anticoagulant eﬀect at a given dosage, and requiring less routine
monitoring [21,22]. LMWH also seems less likely to result in the develop-
ment of HIT. The most often used LMWHs are enoxaparin (Lovenox) and
dalteparin (Fragmin) LMWHs are not yet widely used in veterinary medi-
cine because of higher cost and less clinical experience, although a recent

Fig. 2. A Labrador Retriever receiving mechanical ventilation for support of acute lung injury
after being hit by a car.

1078

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

study reported on the pharmacokinetics of a LMWH in dogs [23]. UFH
given as a treatment for PTE should be dosed to result in prolongation of the
activated partial thromboplastin time (aPTT) of 1.5 times to twice baseline
values. In hypercoagulable animals, this often requires high doses (250–500
IU/kg administered subcutaneously every 6 hours) of UFH. As a result of
the properties of the LMWH preparations, prolongations of the aPTT do
not occur at therapeutic doses, but patients may be monitored by following
clinical signs or by assessing for anti-Xa activity.

Warfarin acts to prevent the formation of vitamin K–dependent coagula-

tion factors (II, VII, IX, and X) [24]. Warfarin also inhibits the production
of the anticoagulant protein C. The half-life of protein C is shorter than that
of the procoagulant factors, so initial overlap with heparin is recommended
to prevent a hypercoagulable state. In people, a warfarin-associated skin
necrosis has been described, but this has not been documented in animals.
Warfarin therapy is challenging in small animals. The prothrombin time
(PT) should be followed every 48 hours until stable for several days, then
every 3 days for 1 month, and then once a week or every other week. The
desired end point is a PT of 1.5 times to twice the baseline value. It is com-
mon for the PT to prolong or shorten in dogs without a clear-cut explana-
tion, occasionally severely enough to result in life-threatening hemorrhage.
In human medicine, diﬀerences in PT reagents are adjusted for via calcula-
tion of the international normalized ratio (INR), but this is not frequently
performed in animals. It is prudent to use the same machine or laboratory
for patient monitoring to try to improve the likelihood of a successful
outcome.

Thrombolytic agents, such as streptokinase or tissue plasminogen activa-

tor (TPA), have been used in people with PTE, but no clinical reports yet
exist in dogs. Thrombolytics act to accelerate the ﬁbrinolytic pathways by

Table 3
Pharmaceutic agents used in the medical management of pulmonary thromboembolism

Drug

Dose

Mechanism of action

Unfractionated heparin

10–20 IU/kg/h CRI to 300

IU/kg SC every 6 hours

Enhance the activity of AT in

inhibiting activated clotting
factors

Low-molecular-weight

heparin

100 IU/kg every 12 hours

Same, more predictable

eﬃcacy

Aspirin

0.5–2.0 mg/kg every 24 hours

Antiplatelet

Streptokinase

25,000 IU for 1 hour, then

10,000 IU/h

Thrombolytic

Warfarin

0.05–0.2 mg/kg

Anticoagulant, requires

careful monitoring

Abbreviations: CRI, continuous rate infusion; AT, antithrombin; SC, subcutaneous.

Pulmonary thromboembolism is a notoriously diﬃcult disease to identify and treat.

Dosage are current best recommendations but are not derived from controlled trials.

1079

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

either directly or indirectly activating plasminogen to plasmin. Streptoki-
nase has been most widely evaluated in animals [25–27]. Its action results
from binding directly to plasminogen and creating a component that stim-
ulates the conversion of other molecules of plasminogen to plasmin. Strep-
tokinase is not speciﬁc for plasminogen bound to ﬁbrin and also acts to
deplete ﬁbrinogen and other procoagulant factors; thus, at least theoreti-
cally, it may result in an increased hemorrhagic potential. TPA is more spe-
ciﬁc and has an increased ability to activate the plasminogen that is
associated with a thrombus. Indications and guidelines for the use of throm-
bolytics in dogs with PTE are not yet deﬁned. It seems prudent to consider
their use in animals that are hemodynamically unstable from large PTE. At
our hospital, the use of streptokinase for large PTE in dogs has not yet been
successful; however, this may reﬂect patient selection and the time frame of
intervention rather than failure of the thrombolytic agent. Surgical removal
of the clot (thromboembolectomy) has also been recommend in some ani-
mals, but, again, clear-cut successful outcomes are rare.

Pulmonary edema

Pulmonary edema is deﬁned as excessive lung water. Pulmonary edema

may develop in dogs as a result of either cardiogenic or noncardiogenic
causes. Cardiogenic edema (caused by left-sided congestive heart failure)
is more common. Heart failure results in the accumulation of pulmonary
edema because of increased hydrostatic pressure when pulmonary venous
hypertension develops as a result of increases in left atrial pressure. Left
atrial pressure may increase for a variety of reasons, including congenital
conditions (eg, patent ductus arteriosus) or acquired conditions (eg, chronic
valvular disease or cardiomyopathy). Cardiogenic pulmonary edema is usu-
ally recognized radiographically as a heavy interstitial to alveolar pattern,
often in association with cardiomegaly and pulmonary venous congestion.
Dogs often ﬁrst develop edema in the perihilar region, and as edema devel-
ops, it becomes diﬀuse. Cats often develop patchy alveolar inﬁltrates rather
than edema in speciﬁc locations. Treatment of cardiogenic pulmonary
edema includes rest, supplemental oxygen, vasodilators. and diuretic therapy
(Table 4).

Furosemide is the initial diuretic of choice, although other diuretics often

must be added, such as spironolactone or hydrochlorothiazide. In an emer-
gency situation, furosemide is administered either intravenously (preferable)
or intramuscularly at 2 to 4 mg/kg. This can be repeated up to once an hour
for 4 hours and then decreased to two to four times a day. Excessive use of
diuretics may result in marked prerenal azotemia and subsequent anorexia
and lethargy.

Vasodilators are useful in the treatment of heart failure by reducing both

afterload and preload. Commonly used vasodilators in the emergency
setting include topical nitroglycerin paste and a continuous rate infusion

1080

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

of nitroprusside. Nitroprusside is a potent vasodilator, so its use requires an
infusion pump and close monitoring. Angiotensin converting enzyme (ACE)
inhibitors are also useful for the long-term management of congestive heart
failure. Occasionally, intravenous enalaprilat has been used in severe acute
cases, but there is little information available in critically ill animals. The ulti-
mate long-term therapy for heart disease depends on the underlying cause
but typically include diuretics and vasodilators. Medications to control rate,
such as b-blockers, calcium-channel antagonists, or digoxin, are often used
as well.

Noncardiogenic pulmonary edema may also result in pulmonary inﬁl-

trates and respiratory distress. Noncardiogenic pulmonary edema is a
high-protein ﬂuid that typically develops after another insult. Common
causes of noncardiogenic pulmonary edema include seizures, upper airway
obstruction, and electrocution, although a number of diseases have been
linked to the formation of noncardiogenic edema [28]. ALI and ARDS are
forms of noncardiogenic edema. Thoracic radiographs from aﬀected dogs
typically have a dorsal caudal distribution to the pulmonary inﬁltrates,
although in severe cases, the alveolar inﬁltrates may aﬀect all lung lobes.
Therapy of noncardiogenic edema is directed primarily at supportive care
and treatment if needed for the underlying disease.

Pleural eﬀusion

Pleural eﬀusion may also result in the development of moderate to

marked respiratory distress. Aﬀected animals may be presented as an emer-
gency or may be chronically aﬀected. Pleural eﬀusion is a sign rather than a
speciﬁc diagnosis, so care must be taken to identify fully the underlying
cause. Management of the pet with pleural eﬀusion is directed toward
removing the eﬀusion and then tailoring treatment to the ﬁnal diagnosis.
Samples of the ﬂuid should be saved for cytologic evaluation and bacterial
culture if indicted. Causes of pleural eﬀusion commonly include congestive

Table 4
Pharmaceutic agents used in the medical management of pulmonary edema

Drug

Dosage

Route

Mechanism

Furosemide

1–4 mg/kg

IV, IM, PO

Diuretic

Spironolactone

1–2 mg/kg every 12 hours

Diuretic

Hydrochlorothiazide

1–2 mg/kg every 12 hours

Diuretic

Enalapril

0.5 mg/kg every 12–24 hours

ACE inhibitor

Nitroprusside

0.5–5.0 lg/kg/min

Balanced vasodilator

Nitroglycerin

0.125–0.25 in per 10 kg

Topically

Venodilator

Hydralazine

0.5–1.0 mg/kg every 12 hours

Arterial vasodilator

Abbreviations: IM, intramuscularly; IV, intravenously; PO, orally; ACE, angiotensin con-

verting enzyme.

Often, a balanced approach results in the best patient outcome.

1081

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

heart failure (particularly in cats), neoplasia, pyothorax, pneumothorax,
hemothorax, and chylothorax as well as less common causes, such as lung
lobe torsion, pancreatitis, or diaphragmatic hernia. Pharmaceutic agents
used in the treatment of pleural eﬀusion include diuretics, antibiotics, chemo-
therapy, vitamin K/fresh–frozen plasma, rutin, and others.

Allergic airway disease

Allergic lower airway disease occurs in cats and, less commonly, in dogs.

Mediators of inﬂammation in people (and likely small animals) include mast
cells, eosinophils, and activated T lymphocytes. Clinical signs of cats with
lower airway disease include coughing, which may be productive, and respi-
ratory distress. Some cats never develop overt respiratory diﬃculties beyond
coughing, whereas others are presented as emergencies with extreme respira-
tory distress. Physical examination of cats with severe respiratory distress
from lower airway disease frequently documents the presence of pronounced
crackles on thoracic auscultation and a normal to elevated rectal tempera-
ture. Thoracic radiographs may appear normal, or they may document
hyperinﬂation and a prominent bronchial pattern. Some cats have right
middle lung lobe collapse or signs of cor pulmonale. Typically, cats respond
rapidly to supportive care and anti-inﬂammatory therapy. A transoral tra-
cheal wash is commonly performed when the patient is stabilized to evaluate
for cytologic abnormalities or signs of infectious or parasitic disease. Bacte-
rial colonization and/or infection may be identiﬁed on the basis of a positive
airway culture. Antibiotic therapy is warranted in individual cats based on
sensitivity data, although it should be remembered that the larger airways in
the cats are normally not sterile. Common bacterial species cultured from
the airways of apparently healthy cats include Pasteurella spp, Staphylococ-
cus spp, Streptococcus spp, Escherichia coli, and even Pseudomonas spp
[29]. Mycoplasma spp have the ability to damage the airway epithelium,
so a positive culture for Mycoplasma spp may have more clinical relevance
than other isolates and should be treated with appropriate antibiotics.

Treatment of the asthmatic cat includes glucocorticoids (prednisone or

long-acting reposital preparations) and bronchodilators (Table 5). Cats with
documentation or the potential for parasites should also be dewormed.
Commonly used bronchodilators include theophylline (sustained release
product at night) and b

-agonists like terbutaline or inhaled albuterol.

Inhaled b

-agonists may be given via an Aerochamber and associated mask.

For most cats, the second smallest size works well. Albuterol (100 lg per
puﬀ [200-dose vial]) has immediate eﬀect and typically lasts 4 hours. The
usual dose is two puﬀs twice a day, but it may be given every 30 minutes for
up to 4 hours in a crisis. For cats that seem to require daily prednisone,
inhaled steroids maybe beneﬁcial. Fluticasone (Flovent, 220 lg per puﬀ)
has been used in cats. Theoretically, there is no systemic absorption, so

1082

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

long-term prednisone side eﬀects are unlikely. Recommended doses include
1 puﬀ per day if asthma is judged to be mild or twice daily in more severe
cases. Inhaled steroids seem to take 7 to 10 days to reach eﬃcacy, so a short
course of oral prednisone (7–14 days) is required when beginning therapy.
Always administer inhaled steroids after bronchodilators so that broncho-
constriction does not interfere with lung distribution of the drug.

The procedure for administering inhaled medications includes priming

the spacer before ﬁrst use and after each cleaning by injecting four to ﬁve
puﬀs into the chamber. This allows particles to adsorb to the inside of the
chamber so that the dose gets to the patient. Wait 15 minutes after priming
before using. Apply the required dose into the chamber away from the cat.
Apply the mask to the cat’s face gently for 10 breaths or 30 seconds. There is
a one-way valve, so the drug only comes out if the cat breathes it. Wait
5 minutes between administering doses. Try to avoid treating the eyes
because of the potential for corneal ulceration. The chamber must be cleaned
once a week with warm soapy water. Most cats tolerate the face mask well,
and it is often easier to use than oral medications.

Proposed treatments include leukotriene antagonists (eg, Accolate),

cyclosporine A, anti-interleukin 5 antibodies, or cyproheptadine. The prog-
nosis for asthma is usually good, although some cats have recurrent bouts
and require frequent medications.

Allergic lung disease is less common in dogs than in cats. Pulmonary inﬁl-

trates with eosinophilia (PIE), including eosinophilic bronchitis or pneumo-
nitis and eosinophilic pulmonary granulomatosis, have been reported in dogs
[30,35]. Clinical signs of aﬀected dogs include cough, respiratory distress,

Table 5
Pharmaceutic agents used in the medical management of allergic airway disease

Drug

Dosage

Mechanism of action/
comments

Prednisone

5 mg PO per cat every 12 hours,

taper

Anti-inﬂammatory, eﬀective,

inexpensive

Methylprednisolone

acetate

10–20 mg per cat SC every

2–4 weeks

Good for the ‘‘hard-to-pill’’

cat, but long-term eﬀects
possible

Terbutaline

0.625–1.25 mg per cat PO every

8–24 hours

-agonist

0.01 mg/kg IV or SC every

8 hours

Theophylline

5 mg/kg PO every 8–12 hours

Methylxanthine long-lasting

products occasionally
available

Zaﬁrlukast (Accolate)

5 mg per cat PO every

12–24 hours

Leukotriene receptor

antagonist

Abbreviations: PO, orally; SC, subcutaneously; IV, intravenously.

Also see text for description of aerosol medications.

1083

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

and lethargy. Substantial pulmonary inﬁltrates are usually observed on
thoracic radiographs. Diﬀerential diagnoses typically include bacterial
pneumonia, heartworm, and neoplasia. Aﬀected dogs often have a history
of prior antibiotic therapy with little or no improvement. Collection of
respiratory samples for cytologic evaluation may be made via a transoral
tracheal wash (TTW) or via a transtracheal aspirate (TTA).

When a speciﬁc cause is identiﬁed, eosinophilic pneumonitis has most

commonly been associated with occult heartworm infection in dogs. It is
proposed that microﬁlaria become trapped in the pulmonary circulation and
then become surrounded by neutrophils and eosinophils. Aﬀected dogs in
one report had a 1-week to 6-month history of progressive coughing, dysp-
nea, and exercise intolerance. All dogs had a peripheral eosinophilia and a
preponderance of eosinophils (60%–80%) on TTA cytology [31]. Pulmonary
inﬁltrates consisting of diﬀuse homogeneous interstitial, peribronchial, and
alveolar patterns were detected on thoracic radiography. There was a rapid
resolution of clinical (within 24 hours) and radiographic (3–5 days) signs
after therapy with prednisone (1–2 mg/kg once a day).

Some dogs with developing (early) heartworm infections acquire eosino-

philic pneumonitis, and repeat heartworm testing in 3 to 6 months is required
to conﬁrm heartworm disease as the cause. Eosinophilic pneumonitis has also
been associated with other parasitic infections, such as ascarid migration or
lungworm. Eosinophilic pulmonary granulomatosis is the most severe form
of pulmonary hypersensitivity. A large study by Calvert et al [32] reported
on 11 dogs over an 8-year period. In this report, the dogs ranged in age from
1 to 11 years. Clinical signs included coughing, dyspnea, and lethargy. Seven
dogs were identiﬁed as having Diroﬁlaria immitis infections. Radiographic
changes included bronchial, interstitial, alveolar, and nodular patterns. Hilar
lymphadenopathy was present in some dogs. Dogs were treated with a variety
of immunosuppressant agents, including glucocorticoids, azathioprine, and
cyclophosphamide. Most dogs initially showed improvement but then usu-
ally died or were euthanized within 1 year. Heartworm infection is a clear
trigger for allergic lung disease and should be excluded in aﬀected dogs. Most
dogs respond to treatment with doses of glucocorticoids that are in an anti-
inﬂammatory to immunosuppressive range. Some dogs may require intermit-
tent therapy to remain asymptomatic. Dogs aﬀected with granulomatous
changes have a much worse prognosis for long-term survival.

Summary

Successful pharmacologic management of most respiratory diseases is

possible. All moderately to severely aﬀected animals beneﬁt from rest and
supplemental oxygen. Careful identiﬁcation of the underlying cause as well
as an understanding of the pathophysiology behind various diseases is essen-
tial to successful patient outcome.

1084

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

References

[1] Ford RB. Chronic lung disease in old dogs and cats. Vet Rec 1990;126:399–402.
[2] McKiernan BC. Diagnosis and treatment of canine chronic bronchitis. Twenty years of

experience. Vet Clin North Am Small Anim Pract 2000;30:1267–78.

[3] Padrid P, Amis TC. Chronic tracheobronchial disease in the dog. Vet Clin North Am Small

Anim Pract 1992;22:1203–29.

[4] Holt D, Brockman D. Diagnosis and management of laryngeal disease in the dog and cat.

Vet Clin North Am Small Anim Pract 1994;24:855–71.

[5] MacPhail CM, Monnet E. Outcome of and postoperative complications in dogs under-

going surgical treatment of laryngeal paralysis: 140 cases (1985–1998). JAVMA 2001;218:
1949–56.

[6] Schachter S, Norris CR. Laryngeal paralysis in cats: 16 cases (1990–1999). JAVMA

2000;216:1100–3.

[7] Powell LL, Rozanski EA, Tidwell AS, Rush JE. A retrospective analysis of pulmonary

contusions secondary to motor vehicular accidents in 143 dogs: 1994–1997. J Vet Emerg
Crit Care 1999;9:127–36.

[8] Cohn SM, Fisher BT, Rosenﬁeld AT, Cross JH. Resuscitation of pulmonary contusion:

hypertonic saline is not beneﬁcial. Shock 1997;8:292–9.

[9] Richardson JD, Franz JL, Grover FL, Trinkle JK. Pulmonary contusion and hemorrhage–

crystalloid versus colloid replacement. J Surg Res 1974;16:330–6.

[10] Jameson PH, King LA, Lappin MR, Jones RL. Comparison of clinical signs, diagnostic

ﬁndings, organisms isolated, and clinical outcome in dogs with bacterial pneumonia: 93
cases (1986–1991). JAVMA 1995;206:206–9.

[11] Murphy ST, Ellison GW, McKiernan BC, Mathews KG, Kubilis PS. Pulmonary lobectomy

in the management of pneumonia in dogs: 59 cases. (1972–1994). JAVMA 1997;210:235–9.

[12] Speakman AJ, Dawson S, Corkill JE, et al. Antibiotic susceptibility of canine Bordetella

bronchiseptica isolates. Vet Microbiol 2000;71:193–200.

[13] Parent C, King LG, Walker LM, Van Winkle TJ. Clinical and clinicopathologic ﬁndings in

dogs with acute respiratory distress syndrome: 19 cases (1985–1993). JAVMA 1996;
208:1419–27.

[14] Parent C, King LG, Van Winkle TJ, Walker LM. Respiratory function and treatment in dogs

with acute respiratory distress syndrome: 19 cases (1985–1993). JAVMA 1996;208:1428–33.

[15] The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes

for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000;
342:1301–8.

[16] Sottiaux J, Franck M. Pulmonary embolism and cor pulmonale in a cat. J Small Anim

Pract 1999;40:88–91.

[17] Whitley NT, Stepien RL. Defaecation syncope and pulmonary thromboembolism in a cat.

Aust Vet J 2001;79:403–5.

[18] Johnson LR, Lappin MR, Baker DC. Pulmonary thromboembolism in 29 dogs: 1985–

1995. J Vet Intern Med 1999;13:338–45.

[19] LaRue MJ, Murtaugh RJ. Pulmonary thromboembolism in dogs: 47 cases (1986–1987).

JAVMA 1990;197:1368–72.

[20] Deitcher SR, Carman TL. Heparin-induced thrombocytopenia: natural history, diagnosis,

and management. Vascular Medicine 2001;2:113–9.

[21] Kleinschmidt K, Charles R. Pharmacology of low molecular weight heparins. Emerg Med

Clin North Am 2001;19:1025–49.

[22] Mammen EF. Low molecular weight heparins and heparin-induced thrombocytopenia.

Clinical and Applied Thrombosis and Hemostasis 1999;1(Suppl):S72–5.

[23] Mischke R, Grebe S, Jacobs C, et al. Aminolytic heparin activity and values for several

hemostatic variables after repeated subcutaneous administration of high doses of a low
molecular weight heparin in healthy dogs. Am J Vet Res 2001;62:595–8.

1085

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

[24] Monnet E, Morgan MR. Eﬀect of three loading doses of warfarin on the international

normalized ration for dogs. Am J Vet Res 2000;61:48–50.

[25] Killingsworth CR, Eyster GE, Adams T, Bartlett PC, Bell TG. Streptokinase treatment of

cats with experimentally induced aortic thrombosis. Am J Vet Res 1986;47:1351–9.

[26] Moore K, Dhupa N, Murtaugh RJ, Rush JE. Retrospective study of streptokinase

administration in 46 cats with arterial thromboembolism. J Vet Emerg Crit Care 2000;
10:245–57.

[27] Ramsey CC, Burney DP, Macintire DK, Finn-Bodner S. Use of streptokinase in four dogs

with thrombosis. JAVMA 1996;209:780–5.

[28] Drobatz KJ, Saunders HM, Pugh CR, Hendricks JC. Noncardiogenic pulmonary edema in

dogs and cats: 26 cases (1987–1993). JAVMA 1995;206:1732–6.

[29] Padrid PA, Feldman BF, Funk K, Samitz EM, Reil D, Cross CE. Cytologic, microbiologic,

and biochemical analysis of bronchoalveolar lavage ﬂuid obtained from 24 healthy cats.
Am J Vet Res 1991;52:1300–7.

[30] Clercx C, Peeters D, Snaps F, Hansen P, McEntee K, Detilleux J, et al. Eosinophilic

bronchopneumopathy in dogs. J Vet Intern Med 2000;14:282–91.

[31] Corcoran BM, Thoday KL, Henfrey JI, et al. Pulmonary inﬁltration with eosinophils in 14

dogs. J Small Anim Pract 1991;32:494–502.

[32] Calvert CA, Mahaﬀey MB, Lappin MR, et al. Pulmonary and disseminated eosinophilic

granulomatosis in dogs. J Am Anim Hosp Assoc 1987;24:311–20.

[33] Frevert CW, Warner AE. Respiratory distress resulting from acute lung injury in the

veterinary patient. J Vet Intern Med 1992;6:154–65.

[34] Wisner DH, Sturm JA. Controversies in the ﬂuid management of post-traumatic lung

disease. Injury 1986;17:295–300.

[35] Neer MT, Waldron DR, Miller RI. Eosinophilic pulmonary granulomatosis in two dogs

and literature review. J Am Anim Hosp Assoc 1986;22:593–9.

1086

E.A. Rozanski, M.P. Rondeau / Vet Clin Small Anim 32 (2002) 1073–1086

Principles of mechanical ventilation

Sharon Drellich, DVM

Emergency and Critical Care Service, Angell Memorial Animal Hospital,

350 South Huntington Avenue, Boston, MA 02130, USA

Successful therapeutic mechanical ventilation of veterinary patients is one

of the most rewarding aspects of critical care medicine. It requires an enor-
mous commitment on the part of the hospital, including human and
mechanical resources, as well as on the part of the owners or agents for the
patient. The commitment of the latter is emotional as well as ﬁnancial, and
appropriate client education and communication are essential to a successful
outcome. In this article, it is not the author’s intention to give exhaustive
details regarding ventilator settings, drug dosages and regimens, or thera-
peutic strategies. Generally covered are indications for ventilation, the nec-
essary setup and equipment for successful ventilation, some of the pitfalls
and complications, and approaches to weaning ventilator patients. Each
of these topics has generated complete texts and countless papers providing
much greater detail. The goal of this article is to familiarize the practitioner
with basic concepts behind all aspects of therapeutic ventilation. The list of
references should provide a beginning for more detailed information.

The basic cost per day for a patient on the ventilator at Angell Memorial

Animal Hospital (AMAH) is $550.00–650.00. This covers the wear and tear
and maintenance on the machine, some basic monitoring (pulse oximetry),
and the oxygen supplied. The costs of blood gas measurement, blood pres-
sure monitoring, catheter placements, medications, and ﬂuids are all added
on an individual basis. Not accounting for diagnostics and special proce-
dures, the cost to the client can run $500.00 to $1000.00 per day.

Is it worth it? As yet unpublished data from the Veterinary Medical

Teaching Hospital at the University of California at Davis show that over
a period of 7 years, 55% (22/40) of patients that were ventilated because
of hypoventilation survived to be weaned. Twenty (50%) were discharged
from the hospital alive. Twenty percent (9/45) of patients being ventilated
because of hypoxemic respiratory failure were successfully weaned from the
ventilator. Five (11%) of those weaned were discharged alive. Twelve ani-
mals were ventilated because of both hypercapnia and hypoxemia; of these,
4 (33%) were weaned and 2 (17%) were discharged alive. Overall survival to

Vet Clin Small Anim 32 (2002) 1087–1100

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 0 3 4 - 7

discharge for patients ventilated for any reason was 28%. The Davis data
include only patients who were ventilated for longer than 24 hours (S.C.
Haskins, DVM, personal communication, 2002).

Numbers were slightly higher overall in a review of cases from the Veteri-

nary Hospital of the University of Pennsylvania. The overall survival rate
in that review was 39% (16/41) of patients, with a similar trend in better sur-
vival among hypoventilating patients than among hypoxemic patients [1].
That study included patients ventilated for less than 24 hours. Keep in mind
as well that cessation of therapy may have to do with owner ﬁnances or logis-
tics in some cases and may not be related to the patient doing poorly or hav-
ing a poor prognosis. Those patients with hypoventilatory problems seem to
be successfully weaned more often than those with hypoxemic disease.
When data on greater numbers are available, it may be possible to prognos-
ticate that animals with some types of hypoxemic disease (ie, traumatic ver-
sus inﬂammatory) may have a better chance of successful weaning.

What is the signiﬁcance of the 24-hour distinction? Shorter term ventila-

tion implies less severe disease. This may be lung disease or patients that
have become intoxicated and have been treated with gastrointestinal decon-
tamination and antidotes and are sleeping oﬀ their therapy. The shorter
period allows less time for ventilator-induced lung injury and may reduce
the likelihood of ventilator-induced pneumonia. Anything that reduces the
secondary consequences of therapy leads to reduced morbidity and mortal-
ity and greater ease of weaning.

Indications for mechanical ventilation

Critically ill patients with and without pulmonary compromise may

require mechanical ventilation. Hypoventilation is an inadequate minute
ventilation (the volume of fresh gas inspired over 1 minute) and leads to
hypercapnia (Pa

[

60 mm Hg) and acidosis (pH < 7.2). Hypoventilation

can be a result of drugs administered therapeutically (anesthetic or paralytic
agents) or disease of the central (CNS) or peripheral nervous system.
Obstructive upper and lower airway diseases as well as pleural space prob-
lems can lead to hypoventilation. If hypoventilation is allowed to persist, it
can lead to hypoxemia. This occurs because the volume of fresh gas deliv-
ered to the alveoli is inadequate to provide the necessary oxygen to maintain
the Pa

. In patients that do not have CNS disease, the Pa

may be

allowed to rise to 60 mm Hg as long as the pH is greater than 7.2 [2]. This
is called permissive hypercapnia. In patients with CNS disease, permissive
hypercapnia is inappropriate and the subsequent cerebral acidosis may lead
to further deterioration.

If supplemental oxygen fails to improve hypoxemia (Pa

65 mm Hg) or

if the patient is exhausting itself to maintain a normal Pa

, that patient

should be ventilated. In patients that require high peak pressures to main-
tain a Pa

at 100 mm Hg, permissive hypercapnia can again be used to

1088

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

prevent greater potential damage to the lungs from barotrauma. Mechani-
cal ventilation may be useful as a salvage procedure when owners want to
say good-bye to a pet that is dying and require time to travel to the hospital.
Although ventilation during anesthesia is not considered here, keep in mind
that if a patient requires ventilation during thoracic or neurologic surgery,
there is always the possibility that ventilation might be required for some
period after the procedure is complete and anesthesia is over.

It is important to establish early that a patient is likely to beneﬁt from

ventilation. By waiting too long and allowing a hypercapnic animal to
become hypoxemic or a hypoxemic animal to exhaust itself, we put that ani-
mal at risk for further organ dysfunction and greater diﬃculty in weaning.
One author has said, ‘‘the indication for intubation and mechanical ventila-
tion is thinking of it’’ [3]. There may be great improvements one day, which
slow to a static crawl the next. Owners need to be informed and committed.
A lack of improvement does not necessarily indicate failure; it simply means
that progress has slowed. We sometimes invoke the phrase, ‘‘if you give it a
day, you should give it a week; if you give it a week, you should give it two.’’
This means that it takes time for tissues to heal and regain function. Venti-
lation provides a means to support life while the tissues are healing just as
dialysis provides time for the kidneys to heal while homeostasis is being
maintained and ﬂuids support hydration while a patient may be unable to
take in water.

Setting up the critical care area for mechanical ventilation

There are several basic and minimum requirements for any hospital

embarking on therapeutic mechanical ventilation. This is a team eﬀort
requiring clear and open communication between all team members. The
team includes nurses and doctors on all shifts, receptionists, kennel assis-
tants, and the family members of the patient. Trained personnel must be
present 24 hours a day to monitor and nurse the patient. A veterinarian
should be on the premises at all times. Highly accomplished nurses can han-
dle the moment-to-moment care and concerns, but a veterinarian needs to
be able to assess acute changes in condition and the need for therapeutic
adjustments. The nursing staﬀ should have a basic understanding of respira-
tory physiology and how the ventilator works to replace or augment phys-
iologic breathing.

The setup of a ventilator and associated monitoring and support equip-

ment requires a large amount of physical space. The treatment area must
be able to accommodate all the equipment and personnel. An area can be
designated where the equipment can be stored compactly between patients
and then rearranged when needed to facilitate patient care. This should
include a raised table or platform so that the patient can be placed in a
recumbent position at a height that is comfortable for the staﬀ to work at
for long periods. Some nearby or mobile seating should be available for the

1089

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

comfort of those visiting and working with the patient. Monitors and mea-
suring equipment for ECG, blood pressure, temperature, and other para-
meters; ﬂuid pumps; suction equipment; and nursing supplies all need to be
near at hand. An emergency airway tray and cardiopulmonary resuscitation
cart should be easily accessible.

The hospital undertaking mechanical ventilation should have a backup

power source and alternative means of ventilation in case of a power or
mechanical failure. A Bain circuit, anesthesia machine, or ambubag may
suﬃce for the short term while alternative mechanical means are arranged.
Ventilators are available that are pneumatically driven by the oxygen source
and require no electricity. An adequate long-term oxygen supply with emer-
gency backup is required as well.

The facility must have means for providing nutritional support to

patients that are anesthetized for long periods of time and may or may not
have gastrointestinal dysfunction as well. This would include appropriate
tubes and diets for enteral feeding as well as options for parenteral nutri-
tional support.

Because ventilation is not an everyday occurrence, a diagram of the set-

up of the system and a checklist for monitoring and nursing parameters
can be helpful. This is useful in properly setting up the system as well as
for teaching new personnel how to arrange the setup. Scheduled drills
and refresher discussions are useful for keeping the team familiar with the
equipment.

Whether the patient is being ventilated because of pulmonary disease or

hypoventilation, the need for ventilation and associated therapies implies
dysfunction of multiple organ systems. For this reason, monitoring of many
systems is essential. Many of these patients are going be under general anes-
thesia for long periods. This factor alone can aﬀect therapy and outcome.
The following list shows what the author believes to be the minimum essen-
tial parameters to monitor in the anesthetized ventilated patient (a more
detailed discussion of monitoring parameters follows later in the text).

• Volume of all ﬂuids going into the patient (eg, replacement and mainte-

nance ﬂuids, ﬂushes, medications, nutrition, blood products)

• Volume of all ﬂuids coming out of the patient (eg, urine, feces, vomit,

gastric contents suctioned, blood collection for laboratory testing, col-
lections from drains, abscesses, eﬀusions)

• Arterial blood pressure by direct or indirect methods
• Arterial blood gases (not always possible in tiny patients on a repeated

basis)

• Other point-of-care testing:

Packed cell volume
Total solids
Blood glucose
Urine speciﬁc gravity

1090

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

• Pulse oximetry
• ECG
• Respiratory rate
• Body weight
• Body temperature
• Level of consciousness
• Airway pressures and minute volume

In the ideal situation, the following parameters are also monitored:

• End tidal carbon dioxide
• Blood lactate
• Osmolality
• Central venous pressure
• Serum electrolytes
• Direct arterial blood pressure

Detailed notations of changes in subjective and objective parameters help

the clinician to assess patient condition and can guide therapeutic changes.
A ﬂow chart form for noting physiologic parameters, ventilator settings, and
laboratory ﬁndings is essential to the monitoring of the ventilated patient.
There should be space to note any and all monitored parameters, laboratory
ﬁndings, ventilator settings, and changes in condition.

Ventilator

Many factors must be considered when choosing a ventilator for thera-

peutic purposes. Ideally, the ventilator in a veterinary practice should be
able to accommodate patients that are tiny (eg, kittens, small dogs, rabbits,
ferrets) and those that are large (eg, Saint Bernard dogs, Irish Wolfhounds).
In some cases, it may be a better option to have two ventilators that can
service the extremes. There are machines on the market for use on labora-
tory rodents that can be used in clinical patients. In the critical care unit
at AMAH, we have a Seimens Servo 300 (Siemens, Solna, Sweden), which
can provide minute volumes as low as 39 mL or as high as 3.99 L. The
machine can be set for neonatal, pediatric, or adult (human) patients so that
the alarm limits are adjusted accordingly.

Ventilators produced for the veterinary market are not subject to the

same rigorous quality control and speciﬁcations as those made for the
human medical market. If the descriptive literature makes it sound too good
to be true, it probably is.

A therapeutic ventilator must be able to blend the oxygen supply. Venti-

lating a patient on 100% oxygen for the short term (12–18 hours) may be
tolerated, but longer than that puts the patient at risk for developing oxygen
toxicity.

1091

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

The ventilator used for long-term ventilation should have a humidiﬁca-

tion system for the inspired air. This helps to prevent drying of the airway
epithelium and to maintain a thermally normal environment.

The ventilator should be capable of variable pressure, volume, and ﬂow

delivery. Especially in the setting of diseased lungs, higher pressures and vol-
umes are necessary for adequate therapy. The ability to provide positive
end-expiratory pressure (PEEP) can improve outcome as well. The machine
should be reliable and have appropriate alarms that are adjustable to the
patient’s size and physiology. It should be simple to use overall, with all
knobs, settings, and adjustments clearly identiﬁed. A display of some sort
(aneroid or digital) should let the operator know what kind of pressures are
being reached at peak. One should also be able to identify by the same
device or an adjunctive one whether and when the patient is triggering a
breath by itself.

When a patient has been on the ventilator for more than 24 hours, wean-

ing can be diﬃcult, complicated, and demanding. A ventilator with modes
that assist weaning is essential for longer term patients. Assist and weaning
modes include synchronized intermittent mandatory ventilation (SIMV),
pressure support (PS), and combinations of SIMV and PS. More detailed
discussion of weaning follows.

Ventilators range from models driven by pneumatic pressure with simple

settings and modes to ‘‘high-tech’’, electrically powered, microprocessor-con-
trolled models with digital display screens and teaching modules. Regardless
of how they are driven, ventilators can be classiﬁed in several ways. One
classiﬁcation is deﬁned by what ends the inspiratory phase of ventilation
or what cycles the ventilator. Cycling may be limited by pressure, volume,
ﬂow, or time. Some ventilators can be cycled by several means, but others
are cycled only by one or another means. Which type and mode of ventila-
tion are best for any one patient is hard to predict. In some cases, several
modes must be tried to ﬁnd the one that works best for a particular patient.

Pressure-cycled ventilators end inspiration when a preset pressure is

reached in the ventilator circuit. The pressure measured is that of the gas
in the system opposing the thoracic and lung compliance and the airway
resistance. Stiﬀer and less compliant lungs reach peak pressures at lower vol-
umes than more compliant ones.

Volume-cycled ventilators end inspiration when a preset tidal volume is

reached. This may be set directly or by setting the minute volume and res-
piratory rate. In the latter case, it is important to remember to check both
settings if a change is made to ensure appropriate volumes and rates. Vol-
ume modes may be diﬃcult to use in patients with severe lung pathologic
changes and increased airway resistance. The machine delivers a set volume
regardless of whether that volume is in the patient or in the ventilator
tubing. Higher pressures from increased airway resistance may mean that a
lower volume of fresh gas is getting to the lungs and that more is building up
in the circuit.

1092

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

Time can be used to cycle the ventilator by adjusting the inspiratory-to-

expiratory ratio. This may be a useful secondary parameter to adjust if your
pressure- or volume-cycled approach is not adequately oxygenating the
patient.

Flow refers to the rate at which the volume of gas leaves the ventilator

and is delivered to the patient. On some ventilators, setting ﬂow is a function
of setting time and volume rather than setting ﬂow directly.

Ventilation therapy is well documented to induce injury to the lungs while

helping the patient heal and oxygenate. A more versatile ventilator reduces
the likelihood of damaging the lungs in the process of supporting them.

Initiating ventilation and modes of ventilation

Veterinary patients do not tolerate intubation and ventilation without

some form of sedation or anesthesia. CNS or peripheral nervous system dis-
ease leading to the need for ventilation implies that the patient is either
obtunded or has some degree of paresis or paralysis. These animals may
be managed with a temporary tracheostomy tube and mild sedation or an-
xiolytic agents. The advantages of avoiding endotracheal intubation and
anesthesia in these patients are that changes in neurologic status can be
easily observed and serial neurologic examinations are not impeded. The
tracheostomy tube placement should be a planned sterile procedure rather
than an emergency placement so as to limit the possible complications.
Patients that have hypoxemic disease and no neurologic impairment require
anesthesia to varying depths so as to keep them comfortable on the ventila-
tor. The appropriate plane of anesthesia is the lightest plane that keeps the
patient comfortable and allows the pressures required for adequate oxygen-
ation. Anesthesia can be maintained nicely with pentobarbital (1–2 mg/kg/h
constant rate infusion [CRI] titrated to patient needs). Other drugs that have
been used include propofol and combinations of benzodiazepines and fen-
tanyl. Some of these regimens are expensive, especially for larger dogs. Each
has advantages and disadvantages. It is best to decide on an anesthetic plan
that is tailored to the individual patient.

Using pentobarbital alone provides excellent anesthesia, but with use for

longer than 24 hours, the wake up can be rocky and can include seizures.
The author has found that a background CRI of diazepam at 0.5 to 1 mg/
kg/h (if a central intravenous line is present) or midazolam helps to make
emergence easier.

Gas anesthesia is problematic because of environmental contamination

and venting of gases. It is also questionable whether it is a good idea to
depend on the diseased organ you are supporting to provide the anesthesia
required to support it.

When setting the ventilator for initiating therapy, the inspired oxygen

concentration (Fi

) is usually set at 60% if the patient has hypoxemic dis-

ease and at 40% to 50% if it is hypercapnic. The goal is then to decrease the

1093

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

oxygen and support settings to the lowest level required to keep the patient’s
Pa

between 35 and 45 mm Hg and the Pa

at 100 mm Hg (Sa

at 98%

or higher). In the case of the hypoventilating patient, once the carbon di-
oxide has come down to the normal range, the settings are usually all the
normal numbers for tidal and minute volumes, and the inspired oxygen can
usually be at 21%. Listed below are baseline numbers for normal lungs.

• Tidal volume: 10 to 20 mL/kg
• Respiratory rate: 8 to 15 breaths per minute
• Minute volume:150 to 200 mL/kg/min
• Proximal airway pressure: 10 to 20 cm H

• Inspiratory time: 1 second
• Inspiratory-to-expiratory ratio of 1:2
• Inspired oxygen: 21%
• End-expiratory pressure is zero when breathing normally

If the baseline settings are inadequate to reduce the Pa

, the tidal vol-

ume (and thus the minute volume) can be increased, the respiratory rate can
be increased, the inspiratory time can be increased, or the proximal airway
pressure can be increased. One change should be made at a time. Blood
gases should be rechecked after a short wait (5–10 minutes) for equilibra-
tion. If the Pa

is still too high, another change is made.

When initially ventilating the hypoxemic patient, the starting inspired

oxygen should be high enough to keep the Pa

above 65 mm Hg. It can

be started at 100%, for example, and dialed back to the lowest percentage
that keeps the Pa

at 90 to 100 mm Hg. Because of the potential for oxygen

toxicity, the inspired oxygen should be kept greater than 60% for no longer
than 12 hours. A Pa

of 65 mm Hg is compatible with life if the oxygen

required to keep it higher is in the toxic range. The PEEP can be increased
to 15 cm H

O if necessary to improve oxygenation. It is increased in a step-

wise fashion by 2 to 4 cm H

O per change. Increasing the end-expiratory

pressure keeps small airways and alveoli from collapsing at the end of expi-
ration. Higher pressures put the patient at greater risk of iatrogenic pneu-
mothorax.

If the patient seems to be panting, uncomfortable, or is not achieving

your therapeutic goals, careful evaluation of the situation should uncover
a cause. Check that what you have set the ventilator to deliver is indeed
being delivered. A respirometer placed in the circuit can help you to deter-
mine this. Some ventilators have digital readouts of volumes in and out of
the system. Is the Fi

adequate, and is what you set being delivered? Is the

oxygen source connected? Is the ﬂow adequate to meet the patient’s need for
volume in the set time for the respiratory cycle? Check the endotracheal tube
for ﬁt, size, and cuﬀ inﬂation (a leak here can cause inadequate delivery of
gas to the patient). Is the tube still in the airway? Excessive moving about of
equipment can cause the tube to slip or come out inadvertently. Is there any
other problem that may be making the patient uncomfortable, such as a

1094

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

fever, full bladder or colon, abdominal distention from other causes, or
pain? Is the patient’s level of sedation or anesthesia adequate for the proce-
dure you are performing? Those patients with more diseased lungs requiring
higher pressures need a deeper anesthetic plane to be kept comfortable while
on the ventilator. Those that have neuromuscular disease and tracheostomy
tubes may require mild sedation only.

Nursing considerations for the patient on a mechanical ventilator

Any manipulations of the airway tubing, catheters (venous, arterial, uri-

nary, or other), or drainage apparatus must be performed aseptically.
Recumbent ventilated patients lack the usual defenses of coughing, self-
cleaning, and shaking for removal of debris from their airways and skin.
They are prone to nosocomial infections from many ports, and careful tech-
nique can help to prevent this. Suctioning the airway (tracheostomy or endo-
tracheal tubing) is necessary every 4 to 6 hours to remove the accumulated
airway material. The cuﬀ on the airway tube should be deﬂated and reposi-
tioned every 4 hours to prevent pressure injury to the mucosa and sloughing
or stenosis of the trachea. Tracheostomy tubes are available with two cuﬀs;
each cuﬀ is inﬂated alternately to prevent prolonged exposure of any one site
to the cuﬀ pressure. Low-pressure cuﬀs are preferable on any tracheal or
endotracheal tube.

Ventilated patients should be weighed daily, and careful documentation

of all ﬂuids in and out should be recorded. Ventilated patients are critically
ill; like many critically ill and stressed patients, they may retain sodium and
water via the syndrome of inappropriate antidiuretic hormone secretion and
the eﬀects of PEEP on sodium retention [4,5].

Changes in urine output may be the ﬁrst indication of alterations in ﬂuid

retention or renal function. Urine speciﬁc gravity should be checked several
times a day, and urinalysis should be performed every few days. Serum and,
ideally, urine electrolytes should be monitored for inappropriate changes.
The colon should be checked daily for accumulation of feces; enemas may
be periodically needed to evacuate the colon.

Body temperature must be monitored regularly if not continuously. Heat

and cooling should be provided as needed to maintain a normal body
temperature.

Oropharyngeal toilet must be attended to regularly. The mouth should be

ﬂushed with normal saline every 4 to 6 hours, and excessive secretions should
be suctioned from the oropharynx and observed for the development of
ulcers. Ulcers can be prevented by keeping the tissues moist and supporting
the endotracheal and ventilator tubing to prevent their placing pressure on
the mouth and tongue. An atraumatic mouth gag can prevent the teeth from
resting on the tongue and inducing pressure ulcers. The mouth gag should not
contact any soft tissue structures of the mouth. Wrapping the tongue in gauze

1095

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

soaked in a saline-diluted glycerine solution keeps it moist. The tongue posi-
tion should be changed on a regular basis. The sites where any tubing or
instruments are resting on tissues should be moved every few hours to redis-
tribute the pressure. A dilute chlorhexidine solution should be used to ﬂush
out the mouth four to six times a day so as to reduce the resident bacterial
population [6].

Humidiﬁcation of the airways is necessary to prevent undue drying of the

tissues. If there is no means available to supply moisture to the circuit, a few
milliliters of saline can be aseptically instilled down the endotracheal or
tracheostomy tube every few hours.

The eyes need to be kept lubricated and closed to prevent corneal ulcers.

The conjunctiva can be ﬂushed with saline every 4 to 6 hours, and lubricat-
ing ointment can be applied. If the lids are not staying closed on their own,
the eyes can be taped shut to prevent ulcer formation. The eyes should be
examined several times a day to look for corneal changes. It would not be
inappropriate to stain the eyes every few days to ensure that they are free
of ulcers.

Positioning of the patient on the ventilator is important. Often, the dis-

tribution of the pulmonary disease is unequal. There is one side or the other
on which the patient seems to ‘‘do better.’’ The patient should be turned
every 4 hours regardless of which is the ‘‘good side’’ to prevent pressure
sores and decubital ulcers as well as atelectasis of the lungs. Turning and
coupage also help to break up respiratory secretions, which can then be suc-
tioned. The trachea should be aseptically suctioned every 4 hours to remove
secretions that accumulate in the airways. Range-of-motion exercise is
important for patients that are recumbent for long periods regardless of
whether they are anesthetized or have neuromuscular disease. Moving and
massaging the limbs helps to promote venous and lymphatic ﬂow and
prevents the development of dependant edema.

Nutritional needs can often be addressed with nasogastric feedings. This

allows the checking of the stomach for residual food volume. Gastrointesti-
nal stasis in anesthetized recumbent patients is common. Metoclopramide,
cisapride, and bethanecol have all been used successfully by the author to
promote gastrointestinal motility and allow enteral feeding. If the gut is not
functioning, parenteral nutrition must be employed (the reader is referred to
the article in this issue on nutritional support of the critically ill patient).

Because of the need for multiple continuous infusions of drugs, repeated

blood sampling to monitor therapy, and, possibly, parenteral nutrition,
most patients on long-term ventilation are best managed with multilumen
central venous catheters and an arterial catheter. The arterial line provides
the means for repeated arterial blood gas sampling and direct monitoring of
arterial blood pressure. Blood pressure and venous return can be adversely
aﬀected by positive-pressure ventilation. The arterial and central venous
access allows us to monitor those parameters and quickly solve problems
by addressing the ventilator settings, ﬂuid therapy, medical therapy, or vaso-

1096

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

pressor support as indicated. In small dogs and cats, it may not be possible
to place an arterial line safely. The author has seen cats and small dogs lose
blood ﬂow to a limb with an arterial catheter. This has occurred with dorsal
pedal (placed percutaneously) and femoral (placed by a cutdown) artery
catheters. Removing the catheter and providing warmth and range-of-
motion exercise to the leg can salvage it. It is an important part of the nurs-
ing care of these patients that all catheter sites and their distal extremities be
checked regularly for inﬂammation, occlusion, and temperature. A Doppler
probe can be placed on the site if blood ﬂow is uncertain.

Problems associated with positive-pressure ventilation

The anesthetic and sedative drugs used to keep patients comfortable on

the ventilator all have systemic eﬀects. Depending on the drug regimen used,
we can expect reduced mean arterial pressure, bradycardia and cardiac
depression, hypothermia, gastrointestinal ileus and reﬂux, diarrhea or con-
stipation, and changes in blood volume and concentration caused by splenic
engorgement or contraction. By providing appropriate ﬂuid therapy and
medication adjustments as needed, these things can usually be managed sim-
ply. Adding in the eﬀects of positive-pressure ventilation may confuse and
confound the observer.

The positive pressure instilled into the lungs by the ventilator is distrib-

uted throughout the chest cavity and across the pleural space. When placed
on the great vessels of the thorax, that same positive pressure impedes
venous return and thoracic blood ﬂow. This results in a reduction in cardiac
output and arterial blood pressure. More aggressive ventilator settings cause
greater reductions in cardiac output and blood pressure. If the lungs are less
compliant because of disease, less of the pressure is transmitted to the non-
lung thorax and the changes in circulation are reduced. This problem is
addressed in the patient by ensuring adequate vascular volume and using
ventilator settings that are adequate to maintain Pa

and Pa

at desired

levels without exceeding the goals.

If the volume and pressure required to meet the goals of Pa

and Pa

are high, pneumothorax, pneumomediastinum, and alveolar rupture are
likely to occur. Less compliant lungs are at greater risk for rupturing
[1,7,8]. If a pneumothorax occurs in a patient undergoing positive-pressure
ventilation, a chest tube must be placed and continuous suction applied. In a
closed thorax, this situation can quickly lead to complete lung collapse, ten-
sion pneumothorax, and death. Pneumothorax is not an indication that
therapy has been unsuccessful; rather, it is a complication.

Barotrauma (injury induced by pressure) and volutrauma occur at the

alveolar level as well as in the airways. Alterations occur in alveolar and vas-
cular permeability as well as in the components of the StarlingÕs forces in
particular. When small airways and alveoli are diseased because of changes
in surfactant, inﬂammation of the tissues, and the presence of edema, they

1097

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

close at the end of expiration. The force required for the ventilator to open
those airways is greater than that required to ventilate already opened air-
ways. The shearing force of the tissues snapping open and closed is also inju-
rious because it causes cell rupture, ﬂuid accumulation, inﬂammatory cell
recruitment, and the activation of inﬂammatory cascades [9].

Pneumonia is a common sequelae to mechanical ventilation in all species

[1,10]. It has many causes, and strategies have been developed to help avoid
it. Bacterial colonization of the airways must occur initially for pneumonia
to develop. The bacteria can come from the nasal cavity, sinuses, oral cavity,
pharynx, or upper gastrointestinal tract [10]. For various reasons, ventilated
patients may already be on antibiotics, which could predispose them to
resistant bacterial infections. The patients may also be on medications that
alter gastric pH (eg, histamine receptor antagonists, proton pump inhibi-
tors) and therefore alter gastrointestinal ﬂora. The underlying disease or
subsequent development of multiple organ dysfunction may lead to bacte-
rial translocation and the development of hematogenous pneumonia.

Manipulations of the airway, handling of any body ﬂuids, and instillation

of any substance into any oriﬁce or catheter of the patient must be carried
out aseptically. Attention to oropharyngeal care, urine and fecal retention,
and patient positioning all help to prevent nosocomial pneumonia. Careful
attention to body temperature, blood count, physical condition, and any
changes in patient condition can alert the clinician to the potential for the
development of pneumonia. If and when it occurs, samples should be col-
lected for culture, and antibiotic regimens should be rational based on the
presumed site of bacterial origin initially and culture results ultimately. In
the University of Pennsylvania study, all patients that developed pneumonia
except one had multiple organisms cultured from airway secretions. The
organisms included Escherichia coli, Enterobacter cloacae, Pseudomonas aer-
uginosa, Klebsiella pneumonia, Acinetobacter species, Staphylococcus species,
and Streptococcus species [1].

Weaning ventilatory support

When to wean a patient from a ventilator must be considered before the

patient goes on the ventilator. The clinician must have a goal in mind with
objective parameters for when this animal should be able to sustain life with-
out mechanical support. This does not imply that the animal does not still
require some supplemental oxygen by means of nasal insuﬄation or an oxy-
gen cage. First and foremost, the disease process that initiated the need
for positive-pressure ventilation should show clear evidence of resolving.
Weaning requires that the patient is metabolically stable. If there are imbal-
ances in electrolytes, nutrient utilization, or cardiovascular function,
increasing the work of breathing is likely to fail. For this author, in general,
when metabolic and cardiovascular parameters are as close to normal as
possible and when the Pa

is maintained above 75 mm Hg with 60% or less

1098

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

inspired oxygen and the PEEP is down to 2 cm H

O, a weaning trial is

appropriate. It is diﬃcult to predict how long it may take to wean a patient
from the ventilator or if a patient may fail to be weaned. There is a degree of
subjectivity in deciding when to wean; however, this subjectivity should be
as limited as possible [11]. Those patients with neuromuscular disease may
have a harder time than others because they are starting to use muscles that
have been aﬀected by their primary disease and not just by medications.
Ventilator neuromyopathy is a problem in weaning people oﬀ the ventilator
who have been on various sedative and paralytic agents for long periods
[12]. Veterinary patients probably experience the same thing to some extent.

To successfully wean, the work of breathing of the patient is gradually

increased as the animal shows us that it can breathe, oxygenate, get rid of
carbon dioxide, and not fatigue. The simplest method, which can be used
with ventilators that do not have weaning modes, is to stop mechanical ven-
tilation and let the patient breathe. This can become tricky depending on
which anesthetics are used and what drugs the patient may require as it
fatigues, to ventilate it again. This method is fairly labor-intensive because
it requires constant watching of the patient. Several cycles of going oﬀ and
back on to the ventilator for gradually longer periods of time are usually
necessary. In the author’s experience, this method works well when patients
have been on the ventilator for shorter periods (<48 hours).

The weaning modes mentioned earlier, SIMV and PS, make weaning of

the long-term patient easier. With SIMV, the number of breaths per minute
is set by the clinician. The patient can trigger the breaths and, if needed, the
machine supplements the breath to the set volume or pressure. If the number
of triggered breaths falls short of the set rate, the ventilator supplies the dif-
ference. As the patient is able to do more of the work of breathing, it triggers
more breaths. This mode can be used with or without PS. Pressure support
lets the patient initiate and generate the breath but ensures that the set pres-
sure is met by supplementing the patient’s eﬀort. The patient can set the rate.
The pressure can be increased if higher tidal volumes are needed.

The clinician gradually reduces the pressure support provided to increase

the patient’s work of breathing until support is no longer needed. In both PS
and SIMV, the patient is required to work against the resistance in the ven-
tilator system. This is a drawback that cannot be avoided and requires the
patient to work harder for each breath.

Successful weaning is a delicate and constantly changing balance between

the patient factors (level of anesthesia, neuromuscular function, and physio-
logic status), the ventilator factors (modes available, resistance, and limits),
and the observation and actions of the clinician.

Summary

Mechanical ventilation is an enormous undertaking for a veterinary hos-

pital in general and for any patient in particular. It is a team eﬀort requiring

1099

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

large amounts of space, supplies, labor, and time. It requires committed
owners and clinicians who communicate clearly with each other. It also
requires a signiﬁcant ﬁnancial commitment initially from the hospital to
obtain the equipment and expertise and then from the owner to maintain the
patient.

All members of the patient care team should have a basic understanding

of respiratory physiology and ventilator mechanics. Clear goals for therapy
and end points should be established. If they cannot be met, the goals
should be reassessed in light of changes in patient condition. Weaning may
be diﬃcult and long, but once successful, it is most rewarding for the
patient, family, clinician, and team.

References

[1] King LG, Hendricks JC. Use of positive-pressure ventilation in dogs and cats: 41 cases

(1990–1992). JAVMA 1994;204:1045–52.

[2] Artigas A, Bernard GR, Carlet J, et al. The American-European Consensus Conference on

ARDS Part 2. Am J Respir Crit Care Med 1998;157:1332–47.

[3] Marino PL. Principles of mechanical ventilation. In: The ICU book. Baltimore: Williams &

Wilkins, 1999. p. 421–33.

[4] Christensen G, Bugge JF, Ostensen J, et al. Atrial natriuretic factor and renal sodium

excretion during ventilation with PEEP in hypervolemic dogs. J Appl Physiol 1992;72:993–7.

[5] Hauptman JG, Richter MA, Wood SL, et al. Eﬀects of anesthesia, surgery, and intra-

venous administration of ﬂuids on plasma antidiuretic hormone concentrations in healthy
dogs. Am J Vet Res 2000;61:1273–6.

[6] Fudge M, Anderson JG, Aldreich J, et al. Oral lesions associated with orotracheal

administered mechanical ventilation in critically ill dogs. J Vet Emerg Crit Care 1997;7:
79–87.

[7] Brown DC, Holt D. Subcutaneous emphysema, pneumothorax, pneumomediastinum, and

pneumopericardium associated with positive-pressure ventilation in a cat. JAVMA 1995;
206:997–9.

[8] Gammon RB, Shin MS, Buchalter SE. Pulmonary barotrauma in mechanical ventilation.

Chest 1992;102:568–72.

[9] Dreyfuss D, Saumon G. Ventilator induced lung injury. Am J Respir Crit Care Med

1998;157:294–323.

[10] Kollef MH. The prevention of ventilator associated pneumonia. N Engl J Med 1999;340:

627–34.

[11] Ely EW, Baker AM, Dunagan DP, et al. Eﬀect on the duration of mechanical ventilation of

identifying patients capable of breathing spontaneously. N Engl J Med 1996;335:1864–6.

[12] Hund EF, Fogel W, Krieger D, et al. Critical illness polyneuropathy: clinical ﬁndings and

outcomes of a frequent cause of neuromuscular weaning failure. Crit Care Med 1996;24:
1282–3.

1100

S. Drellich / Vet Clin Small Anim 32 (2002) 1087–1100

Nosocomial infections

Justine A. Johnson, DVM

Emergency and Critical Care, Ocean State Veterinary Specialists,

1480 South County Trail, East Greenwich, RI01818, USA

Nosocomial infections are infections caused by bacteria or other infec-

tious organisms that are acquired by the patient during hospitalization. Bac-
teria associated with nosocomial infections are often resistant to antibiotics,
particularly the antibiotics most frequently used in that particular hospital
setting. Nosocomial infections aﬀect approximately 5% to 10% of all hos-
pitalized human patients [1], causing signiﬁcant increases in morbidity,
mortality, and hospital costs. Common nosocomial infections include
bloodstream infections (BSIs), urinary tract infections, pneumonia, surgical
wound infections, and infectious diarrhea. The incidence of nosocomial
infections in veterinary hospitals is not well established, but the factors
contributing to the increase in nosocomial infections in human hospitals are
becoming more common in veterinary medicine. These include the increas-
ing use of invasive devices (eg, intravenous and urinary catheters), the
increase in duration of hospitalization, the increase in intensive care practi-
ces, and the increased use of antimicrobial drugs.

A variety of types of organisms have been implicated in nosocomial infec-

tions. Some organisms are considered to be intrinsically pathogenic in that a
signiﬁcant number of patients that are exposed to the organism develop clini-
cal signs. An example would be Salmonella spp, which can exist in a small
number of carrier animals but can cause clinical illness, especially in stressed
patients. Other organisms that might be considered normal resident ﬂora on
skin, in the upper respiratory tract, or in the gastrointestinal tract can
become pathogens if they gain access to normally sterile parts of the body,
if they become a dominant organism (as in bacterial overgrowth), or if they
are resistant to antimicrobial drugs.

Although it was previously thought that most nosocomial infections orig-

inated from bacteria that were part of the endogenous ﬂora of the patient,
more recent studies suggest that the organisms causing many nosocomial

Vet Clin Small Anim 32 (2002) 1101–1126

E-mail address: jajicu@aol.com (J.A. Johnson).

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 0 3 8 - 4

infections are acquired from the hospital environment [2]. In hospitals, par-
ticularly intensive care units, there are several factors that promote the
development of nosocomial infections and the propagation of antimicrobial
resistance in nosocomial pathogens. One important factor is the frequent
use of antibiotics. Bacterial populations that make up normal ﬂora in a
patient can become altered during antimicrobial therapy. The population
of normal resident ﬂora may be diminished as a result of antimicrobial use,
whereas organisms resistant to antimicrobials ﬂourish. The patient then
becomes a reservoir for resistant organisms. These resistant organisms may
gain access to the patient’s bloodstream via translocation across compro-
mised intestinal mucosa or may be shed in the patient’s body ﬂuids, thus
contaminating the local environment, including hospital surfaces, medical
equipment, and the hands of hospital personnel. The contaminated environ-
ment serves as a source of bacteria for other body systems in that patient as
well as for other patients in the hospital. Bacterial colonization by endemic
hospital organisms occurs in the upper respiratory tract, gastrointestinal
tract, urogenital tract, and skin of many patients within a few days of hos-
pitalization [3]. The development of infection by these organisms is often
associated with the use of invasive devices such as intravenous or urinary
catheters, endotracheal tubes, and surgical instruments that bypass the
patient’s normal defensive mechanisms and introduce bacteria into normally
sterile body sites.

Nosocomial pathogens have been shown to persist in the hospital envi-

ronment in a variety of locations. Examples of fomites that have been iden-
tiﬁed as reservoirs of nosocomial organisms include a whirlpool bathtub [4],
inadequately processed endoscopes [5,6], laryngoscope handles [7], stetho-
scopes [8], computer keyboards [9], faucet handles [9], and thermometers
[10]. Increasing attention is also being focused on the role of hospital per-
sonnel as reservoirs of nosocomial organisms. Methicillin-resistant Staphy-
lococcus outbreaks have been attributed to carriage of the organism in the
nasal passages of clinicians and hospital workers [11,12]. Nasal swabs are
nowa routine part of hospital surveillance and when an outbreak is identi-
ﬁed, staﬀ members are treated with topical antibiotic (mupirocin) ointment
[11,12]. In two outbreaks of nosocomial infections in a human neonatal
ICU, the organisms were found to have been transmitted from household
pets on the hands of ICU nurses. One involved an epidemic of fungemia
caused by Malassezia pachydermatis from a canine ear infection [13], and the
other involved a case of feline ringworm [14]. The role of hospital staﬀ as a
reservoir may prove important in veterinary medicine as well. In a recent
report of an outbreak of methicillin-resistant Staphylococcus infections in
postoperative equine patients, nasal swabs from staﬀ members in the surgi-
cal and recovery areas suggested that the nasal passages of these people may
have been the source of the infections [15]. Although most veterinary hospi-
tals enforce strict policies to contain highly infectious diseases such as parvo-
virus and feline upper respiratory viruses, these examples demonstrate that

1102

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

any commonly encountered organism can be a pathogen if introduced to a
susceptible patient.

Antibiotic resistance

Increased resistance to antibiotics is a common feature of nosocomial

organisms. There are several mechanisms for the development of antimicro-
bial resistance. Some bacteria have intrinsic resistance to speciﬁc types of
antibiotics, meaning that they lack the speciﬁc target sites or transport
mechanisms that allowthe antibiotic to have an eﬀect [16]. These character-
istics often involve the cell wall structure. For example, most obligate anaer-
obes are resistant to the aminoglycoside antibiotics because they lack the
electron transport mechanism necessary for uptake of this class of drug
[17]. Acquired resistance to antibiotics occurs through mutation of the
genetic makeup of the bacteria, creating genes that encode antibiotic resist-
ance. Spontaneous chromosomal point mutations occur infrequently (1 in
10

) [16] but are then passed down to daughter progeny. Chromosomal

DNA can also be transferred between bacteria in the form of transposons,
which are small segments of DNA that can remove themselves from the
chromosome and be transferred to other organisms [16]. In addition to the
chromosomal DNA present in bacteria, there is also DNA on extrachromo-
somal structures known as plasmids. Plasmids containing genes for antimi-
crobial resistance (R factors) can be transferred between organisms and
therefore provide a means for DNA to be mobile and spread resistance rap-
idly within a population. Mechanisms for antimicrobial resistance that may
be adopted by bacterial organisms include alterations in cell wall permeabil-
ity and target sites for antibiotic binding as well as the ability to enzymati-
cally inactivate antibiotics [18].

In hospitals, where the use of antimicrobial drugs is common, bacteria

with antibiotic resistance have a selective advantage compared with other
bacterial populations. The wide application of antimicrobials in medical and
veterinary practice, the use of antibiotics in agriculture (particularly dairy
and meat production), and the use of antiseptics and disinfectants result
in selective pressure. Genes conferring resistance to some disinfectants may
be present on the same plasmids conferring resistance to antibiotics [19].

Nosocomial organisms in human hospitals frequently carry resistance to

multiple antimicrobials, and the pattern of antimicrobial resistance within
hospitals changes when the use of speciﬁc antibiotics is restricted [20].
Resistance has developed against methicillin, vancomycin, and the ﬂuoro-
quinolones, which are antibiotics that were developed to be used in cases
of resistance to penicillins and the b-lactam antibiotics [21]. Resistance has
been documented in the microbial ﬂora of community-acquired bacteria in
small animals. Resistance to ampicillin, penicillin, sulfonamides, and tetra-
cycline is common and may develop without prior treatment with antibiotics
[22–25]. Multiple drug resistance has been demonstrated in nosocomial

1103

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

organisms in veterinary hospitals and is most common in patients that have
been treated with antibiotics [26–28]. In a report of an outbreak of Klebsiella
spp infections at a small animal teaching hospital, 23 of the 24 patients iden-
tiﬁed had been on antibiotics, and the organisms isolated had greater anti-
biotic resistance than community-acquired Klebsiella spp isolated from
other animals [27]. In another study involving horses at a university teaching
hospital, bacteria isolated from fecal matter on the seventh day of hospital-
ization had a much broader spectrum of antimicrobial resistance than iso-
lates obtained at the time of admission [28]. This study demonstrated that
the increase in resistance in a patient’s endogenous bacterial ﬂora can occur
after a short period of hospitalization or antibiotic therapy. In the author’s
experience, bacterial cultures obtained from patients that have been hospi-
talized for more than a fewdays frequently reveal multiple drug resistance.
Life-threatening infections in these patients may require the use of expensive
and specially ordered antibiotics not commonly used in veterinary medicine,
such as third-generation cephalosporins, methicillin, or piperacillin.

Bloodstream infections

The incidence of nosocomial BSIs in human hospitals has increased two-

to threefold in the past two decades, and 90% of these infections are related
to intravenous catheter use [29]. Catheter-related infections can range in
severity from localized phlebitis to fatal bacteremia and sepsis. BSIs have
been reported to increase mortality in hospitalized patients 14-fold [30]. In
one study, mortality was reported to be 31% when bacterial BSIs occurred
and up to 67% if fungemia developed as a result of Candida infection [31].
Duration of catheterization is the most important risk factor for the develop-
ment of catheter-related BSIs [32]. In human patients, catheter infections are
most common if the catheter has been in place for 4 to 5 days [33]. Intrave-
nous catheters rarely produce bacteremia until they have been in place for at
least 48 hours, and catheters left in place for more than 48 to 72 hours have
been associated with rates of bacteremia ranging between 2% and 5% [34].

Most catheter infections are caused by contamination of the device either

at the time of insertion or during use [35]. The most common organisms
involved in catheter-related infections in human patients are normal skin
ﬂora, including Staphylococcus aureus and Staphylococcus epidermidis [32].
Organisms may be introduced on the hands of the person placing or han-
dling the catheter. These organisms may be part of the health care worker’s
own skin ﬂora, may be part of the patient’s skin ﬂora acquired from a site
that has not been disinfected, or may be fecal or other types of bacteria that
the health care worker has handled. Contamination of the catheter can also
occur during episodes of bacteremia or fungemia or by means of contami-
nation of the infusate during manufacture or preparation [36]. Increased
manipulation of the catheter (including blood sampling), disconnecting or

1104

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

changing of lines, and the use of hypertonic solutions (including parenteral
nutrition and lipid emulsions) increase the potential for catheter infection
[37]. Immune-compromised patients, including neonates and chemotherapy
patients, are at increased risk.

There are fewdata in the veterinary literature documenting the incidence

of catheter-related BSIs in small animal patients. Most reports describe out-
breaks of catheter infections in veterinary hospitals that have been linked to
inadequate skin preparation [38], contaminated antiseptic solutions [39],
contaminated gauze squares [40], and other unidentiﬁed common vehicles
[41]. In these outbreak reports, Serratia spp were a common isolate, and the
organisms were resistant to multiple commonly used antibiotics [38,40].
Despite the lack of documentation in the literature, it is likely that most cath-
eter-related infections involve organisms that are ubiquitous to the patient’s
skin ﬂora [42] and organisms carried on the hands of hospital personnel. In a
study reporting results of a surveillance project in a veterinary hospital, 26%
of jugular catheters sampled had positive bacterial cultures after an average
of only 2.7 days [43]. More than 50% of the organisms isolated were Klebsi-
ella spp and Enterobacter spp. These results are consistent with the author’s
experience when monitoring results of routine catheter cultures at a univer-
sity teaching hospital [44]. Although contamination of the catheter site, hos-
pital equipment, and veterinary care providers’ hands are likely a common
source of catheter-associated BSI (CABSI) infection for veterinary patients,
there is the added challenge that our patients often soil the catheter site dur-
ing use. Catheter dressings routinely become soiled with saliva, food, urine,
feces, blood, and other materials, and changing these dressings requires fur-
ther contact with the potentially contaminated hands of the veterinary care
providers.

The risk of nosocomial BSIs can be minimized by using aseptic technique

during catheter placement. The catheter site should be clipped and cleaned
using a disinfectant solution. Chlorhexidine has been shown to be more
eﬀective than povidone iodine or alcohol as a solitary cleansing agent [45].
Chlorhexidine requires a contact time of at least 30 seconds to provide max-
imum disinfection. One commonly employed preparation technique involves
scrubbing the catheter site three times with chlorhexidine, wiping the site
with 70% alcohol between each scrub, and letting the chlorhexidine sit for
30 seconds in the last cycle. All containers holding disinfectant and sponges
should be cleaned routinely and periodically sterilized.

The technique employed by the person placing the catheter may inﬂuence

the development of catheter infection. In one study, individuals who had
placed fewer than 50 catheters had a positive catheter culture rate of 50%,
whereas individuals who had placed more than 50 catheters had a rate of
25%. The diﬀerence was attributed to better attention to aseptic technique
and decreased tissue trauma induced by the more experienced personnel
[46]. The catheter should be secured in place and covered with a clean dress-
ing. The use of antiseptic and antibiotic ointments at the catheter site is

1105

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

controversial. Triple antibiotic ointment reduces the risk of catheter infec-
tions, especially by S. aureus and gram-negative bacilli, but it has been
shown to increase the risk of fungal infections in human patients [47]. For
this reason, povidone iodine ointment has been recommended for use with
parenteral nutrition catheters, which are at high risk for fungal colonization.

While the catheter is in use, manipulation and disconnection of the infu-

sion lines should be kept to a minimum, and whenever the lines are discon-
nected, minimum preparation should include a 2-minute clinical hand wash
or the use of disposable gloves. Despite the increase in infection risk with
prolonged catheterization, studies in human and canine patients have shown
no beneﬁt to scheduled changing of intravenous catheters every 72 hours
[40,48,49]. These studies have led to the current recommendation in human
medicine that catheters be removed as early as medically indicated but that
routine catheter changes be avoided unless there is evidence of an infection
[50]. Changing the ﬂuid lines every 48 hours may reduce the risk of infection
[34], especially if the lines are frequently disconnected or used for blood
sampling or drug administration. Any poorly functioning catheter should
be removed. A catheter that accepts ﬂuid solutions but does not yield a
blood return (‘‘ﬂashback’’) may be in the process of becoming occluded.
Occlusion may be caused by the development of a ﬁbrin sheath in the inner
lumen, which decreases the catheter’s eﬀectiveness and provides a medium
for bacterial colonization [51].

The use of sterile barriers, including gloves, a gown, a mask, and drapes,

during catheter placement and management has been shown to decrease the
risk of catheter infections [52]. The added cost of such precautions may be jus-
tiﬁed in debilitated or immune-compromised patients. One study has demon-
strated that the use of sterile gloves provides no signiﬁcant beneﬁt compared
with a 2-minute antiseptic hand wash or the use of nonsterile disposable latex
gloves in preventing catheter infections in the general hospital population
[53]. Relying on hand washing to minimize catheter contamination necessi-
tates that staﬀ compliance be high, and this can be diﬃcult to enforce.

Symptoms of a catheter infection include phlebitis or cellulitis at the cath-

eter site [54], which manifests as pain or lameness, swelling, redness, palpa-
tion of a thickened or cordlike vein, or purulent discharge. Leukocytosis and
fever may develop even in the absence of bacteremia. If bacteremia devel-
ops, symptoms include fever or hypothermia, leukocytosis (with left shift)
or neutropenia, and shock. Catheter infection should be suspected in any
patient with an intravenous catheter that develops a fever or leukocytosis
without evidence of infection at any other body site. Often, removal of the
infected catheter results in resolution of symptoms, and no further treatment
is necessary. If a catheter infection is suspected, the catheter should be
removed and the tip of the catheter should be submitted for bacterial culture
and antimicrobial sensitivity testing. If a culture is to be submitted, the cath-
eter dressing should be removed and the skin around the catheter site should
be disinfected (in a similar fashion as when placing a catheter). This reduces

1106

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

the risk of culturing a skin contaminant. The catheter is removed, and sterile
scissors are used to cut the end of the catheter oﬀ. The tip is then placed into
a sterile glass tube (ie, for blood collection), and a small amount of sterile
saline is added to prevent drying. If bacteremia is suspected, blood cultures
should be obtained and empiric antibiotic therapy should be started.
Reports on catheter infections in human hospitals demonstrate that leading
pathogens are coagulase-negative staphylococci, S. aureus, and enterococci
[55]. The limited reports in veterinary medicine suggest that Serratia spp,
Staphylococcus spp, Streptococcus spp, Klebsiella spp, and Enterococcus spp
are the most common isolates [39,41,43]. Initial antibiotic therapy should be
broad spectrum, and if life-threatening bacteremia is suspected, a combina-
tion of drugs may be required. The antibiotic selection can be narrowed
when culture results are available. Candida spp are a common isolate in
human catheter infections and BSIs [31]. Although fungemia is not yet a
commonly reported problem in veterinary medicine, the increased use of
parenteral nutrition products and systemic antibiotics is likely to increase
the risk of this problem in small animal patients. The author has encoun-
tered two cases of fatal fungemia in patients being treated with immuno-
suppressive drug therapy for immune-mediated hemolytic anemia.

A variety of antimicrobial catheter materials and coatings, including

chlorhexidine–silver sulfadiazine and rifampicin-minocycline, are under
investigation, some of which are eﬀective in reducing the rate of catheter-
related bacteremia. Antimicrobial-impregnated and heparin-bonded cathe-
ters reduce infection by preventing the development of a ﬁbrin sheath and
subsequent bacterial colonization [56]. These products may become more
available in the veterinary market when their use in human medicine is more
widespread.

Pneumonia

Pneumonia is the second most common type of nosocomial infection in

human hospitals and is the leading cause of death as a result of nosocomial
infection, with a mortality rate of up to 50% [57]. The risk factors for hos-
pital-acquired pneumonia in human hospitals include prolonged mechanical
ventilation, residence in an ICU, duration of hospital or ICU stay, and
severity of underlying disease(s). The introduction of bacteria into the lower
respiratory tract usually occurs by aspiration of oropharyngeal contents
[58]. Less common routes of infection include direct inoculation (ie, at the
time of intubation), bacteremia, translocation of bacteria from the gastro-
intestinal tract, and inhalation of aerosolized organisms. Endotracheal intuba-
tion increases the risk of aspiration because it interferes with normal glottic
closure, cough reﬂexes, and mucociliary clearance. Aspiration is more likely
to occur when there is retrograde ﬂow of esophageal and gastric ﬂuid into
the oropharynx. This occurs when the patient is recumbent, sedated, or
debilitated; when the patient has decreased gastrointestinal motility; or

1107

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

when orogastric or nasogastric intubation interferes with esophageal sphinc-
ter function.

Most cases of nosocomial pneumonia in human hospitals occur in pa-

tients undergoing mechanical ventilation. Ventilator-associated pneumonia
(VAP) is divided into two categories: early onset and late onset. Early-
onset VAP is usually caused by gram-positive organisms that are considered
to be normal inhabitants of the oropharynx. Studies have demonstrated that
within 24 to 72 hours of admission to a human ICU, gram-negative aerobic
bacteria colonize the stomach and small bowel and compete with anaerobic
commensals of the large intestine [59]. After the patient has been hospital-
ized for 3 to 5 days, the oropharynx becomes colonized with gram-negative
organisms, possibly because of retrograde colonization from gastric con-
tents [60]. Other organisms colonizing the oropharynx include organisms
from the hospital environment (including the hands of hospital personnel)
that may have developed antimicrobial resistance. Late-onset VAP is there-
fore more commonly caused by gram-negative organisms and is more likely
to be resistant to antimicrobial therapy. Infection by gram-negative bacilli
accounts for 60% to 80% of nosocomial pneumonia in human patients [58].

Because gastric contents may serve as a source of organisms for oropha-

ryngeal colonization, signiﬁcant attention has been focused on minimizing
the bacterial activity in the stomach of patients at risk for VAP. The stom-
ach typically has little bacterial activity because of the lowpH of gastric ﬂu-
ids, the presence of mucus and IgE in salivary ﬂow, and the unidirectional
ﬂowof ingesta preventing retrograde movement of intestinal ﬂora. Gastric
ulcer prophylaxis is commonly used in hospitalized human patients because
of the high incidence of stress ulcers. The use of H2 blockers and antacid
therapy leads to alkalinization of the gastric secretions and allows coloniza-
tion of the stomach with Enterobacteriaceae. Studies have demonstrated
that the use of sucralfate as a single agent in the prevention of gastric ulcers
prevents the rise in gastric pH and leads to a lower concentration of gram-
negative bacilli in gastric aspirates. Patients receiving sucralfate alone have a
signiﬁcantly decreased incidence of pneumonia and decreased mortality [61].

Another method for limiting the colonization of the oropharynx by gas-

trointestinal ﬂora is selective decontamination of the digestive tract (SDD).
The most commonly utilized regimen involves the use of three poorly
absorbed agents (polymyxin E, tobramycin, and amphotericin B) applied
topically to the mucosa of the oropharynx and administered enterally four
times a day. Parenteral antibiotics (ie, cephalosporin) are administered for
at least 4 days to treat any community-acquired infections by gram-positive
organisms that might be incubating at the time of hospitalization, and a
high standard of hygiene is employed to prevent the introduction of exo-
genous bacteria [62]. The goal is to sterilize the oropharynx and upper gastro-
intestinal tract in patients at risk for nosocomial pneumonia, speciﬁcally
patients requiring mechanical ventilation. Selective decontamination of the
digestive tract has been shown to decrease the incidence of nosocomial

1108

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

pneumonia and urinary tract infections, but fewstudies demonstrate a
decrease in mortality [63,64].

The use of gastric motility modiﬁers has been suggested as a means of

preventing retrograde colonization of the upper gastrointestinal tract with
intestinal ﬂora. In one study, metoclopramide delayed the development of
nosocomial pneumonia in critically ill human patients receiving enteral
nutrition but had no eﬀect on the overall frequency of pneumonia or the
mortality rate [65]. A semirecumbent position has been recommended for
patients receiving enteral tube feedings. Elevation of the upper body reduces
the risk of aspiration and may therefore reduce the risk of VAP [66].

Nasogastric intubation and suctioning have been advocated in human

hospitals in patients undergoing mechanical ventilation and in patients
recovering from abdominal surgery. Suctioning of gastric contents is
intended to reduce residual gastric volume in patients with ileus and to
decrease oropharyngeal colonization by gastrointestinal bacteria. Recent
studies have demonstrated that the increase in gastroesophageal reﬂux asso-
ciated with compromising the distal esophageal sphincter with the nasogas-
tric tube may be more signiﬁcant than the risk of retrograde ﬂowof gastric
contents in patients with ileus. Routine nasogastric intubation is therefore
not recommended, and if a tube is used, a small-diameter tube is preferred
[67,68]. In veterinary medicine, some clinicians advocate the use of nasogas-
tric suctioning in patients with severe pancreatitis, postoperative ileus, or
ileus during mechanical ventilation. Nasogastric suctioning should be used
with caution in patients that are recumbent, sedated, or debilitated, because
there may be an increased risk of nosocomial pneumonia.

Mechanical ventilation in veterinary patients is limited mainly to univer-

sity hospitals and large referral centers. Nosocomial pneumonia is common
in veterinary patients undergoing prolonged mechanical ventilation, and the
strategies for the prevention of nosocomial pneumonia in people may be rel-
evant. Aspiration pneumonia is also encountered in patients undergoing
anesthesia and surgery as well as in debilitated patients that are vomiting.
If these patients have been hospitalized for multiple days before aspiration,
it is likely that the oropharynx is colonized with nosocomial organisms, and
the pneumonia may involve organisms resistant to antibiotic therapy. As
stated previously, general hygiene, including hand washing by staﬀ between
patients and the use of disposable gloves, is the most important means of
preventing colonization of patients by nosocomial organisms. Recumbent,
sedated, or debilitated patients may beneﬁt from having the cranial portion
of the body elevated. This may be especially important in patients with mega-
esophagus, a nasogastric tube, or tracheal intubation. If an endotracheal
tube is placed, the cuﬀ should be inﬂated to prevent introduction of aspi-
rated ﬂuids into the lower respiratory tract. If the tube is to be left in place
for more than a fewhours, it should be sterilized and replaced at least every
24 hours. Nebulizers used to aerosolize ﬂuids and medications in the treat-
ment of respiratory disease have been identiﬁed as a potential source of

1109

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

pathogenic organisms and should be cleaned and dried between uses and
disinfected frequently. Oropharyngeal decontamination using Stomadhex
(VRx Products, Harbor City, CA) dex pads or chlorhexidine dental wash
[69] may decrease the rate of oropharyngeal colonization. Removal of sub-
glottic secretions and oropharyngeal ﬂuid that has collected proximal to the
endotracheal tube cuﬀ may help to prevent pneumonia.

The use of enteral versus parenteral nutrition in ventilated patients is con-

troversial. Parenteral nutrition leads to gut mucosal atrophy and an
increased frequency of bacterial translocation. Enteral nutrition provides
support for gut mucosa but may increase the risk of retrograde ﬂowof gas-
tric contents and aspiration. Continuous enteral feeding rather than bolus
feeding may provide the beneﬁts of enteral nutrition with minimal risk
[70]. Elevating the cranial half of the patient’s body may minimize retro-
grade ﬂowand aspiration.

Nosocomial pneumonia should be suspected in any patient that develops

depression, fever, leukocytosis, cough, or dyspnea after periods of vomiting
or intubation. Patients being supported on a mechanical ventilator do not
demonstrate cough or dyspnea because of sedation and intubation. Careful
surveillance in these patients should include monitoring of temperature,
careful auscultation, periodic leukograms, blood gas analysis, and thoracic
radiography.

When pneumonia is suspected, a sample of bronchial secretions should be

obtained via transtracheal aspirate, endotracheal aspirate, percutaneous
lung aspirate, or bronchoalveolar lavage. Initial antibiotic therapy may be
selected based on Gram stain analysis, but the sample should be submitted
for bacterial culture and antibiotic sensitivity testing. This is particularly
important if the patient has been previously treated with broad-spectrum
antibiotics, because the organisms involved in the infection may be resistant
to multiple commonly used antibiotics. In some cases, a patient may not be
considered stable enough to withstand sedation and bronchial ﬂuid sam-
pling. In these cases, empiric therapy may be initiated. Most nosocomial
pneumonias in human beings are caused by gram-negative bacilli or Staph-
ylococcus spp [55]. The organisms vary greatly between institutions, how-
ever, and stored local data on bacterial cultures can be important in
guiding therapy. Some physicians recommend treating all pneumonias ini-
tially as though Pseudomonas is the likely organism because it is the most
virulent pulmonary pathogen [71]. A combination of an aminoglycoside and
a b-lactam antibiotic (ie, ﬁrst-generation cephalosporin) or monotherapy
with a third-generation cephalosporin has been recommended. If there is
no clinical response to antibiotic therapy within 48 hours, bronchial ﬂuid
sampling should be attempted.

Research to improve the safety of mechanical ventilation includes eﬀorts

to alter the composition of the endotracheal tube to reduce the risk of bac-
terial colonization. Also, endotracheal tubes are being developed with spe-
cialized ports that allowcontinuous suction of subglottic ﬂuids.

1110

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

Catheter-associated urinary tract infections

Urinary tract infections are the most common type of nosocomial infec-

tion in many human hospitals, making up roughly 40% of all nosocomial
infections [72]. Most nosocomial urinary tract infections are associated with
the use of urinary catheters, especially indwelling Foley catheters [73,74].
Catheter-associated urinary tract infections (CAUTIs) can lead to complica-
tions, including fever, chronic interstitial nephritis, pyelonephritis, renal fail-
ure, bacteremia, and death [75]. Bacteriuria is the second most common
cause of nosocomial BSIs, leading to bacteremia in 2% to 4% of aﬀected
human patients, and it is associated with a mortality rate three times higher
than that of nonbacteriuric patients [76]. Many patients with nosocomial
bacteriuria are asymptomatic, and these patients are of concern because
they are a major reservoir of antibiotic-resistant organisms that can contam-
inate the hospital and aﬀect other patients [77].

Risk factors for the development of nosocomial urinary tract infections

in human patients include duration of catheterization, female gender,
advanced age, debilitating disease, manipulation of the catheter and collec-
tion system, and trauma to the urethra or bladder [76]. The duration of cath-
eterization is the most important risk factor, with the incidence of urinary
tract infection increasing 5% to 7% with each day of catheterization [78].
The use of closed urinary collection systems has become a standard of care,
because open urinary catheters are associated with a higher incidence of bac-
teriuria. In one study of human patients with indwelling urinary catheters,
all patients with open catheters developed bacteriuria within 4 days, whereas
more than 30 days elapsed before all patients with closed collection systems
developed bacteriuria [75].

Gram-negative enteric bacteria are the most common organisms isolated

in human CAUTIs [55]. The use of systemic antibiotics during urinary cath-
eterization is associated with an increased risk of infection with Pseudo-
monas spp and Serratia spp as well as an increased risk of infection with
organisms resistant to the eﬀects of commonly used antimicrobial drugs [79].

Bacteria can gain access to the bladder through several mechanisms. The

microbes can be transported by the catheter tip at the time of catheteriza-
tion. This risk is minimized if aseptic technique is applied. There is normal
bacterial colonization of the distal portion of the urethral mucosa, however,
and the catheter tip must pass through this area. While the catheter is in
place, bacteria can ascend into the bladder via the catheter lumen or along
the exterior surface of the catheter. If the collection system has been contam-
inated, motile bacteria can ascend into the bladder through the catheter, and
nonmotile bacteria can gain access to the bladder if there is retrograde ﬂow
of urine through the catheter. Retrograde ﬂowoccurs if the collection sys-
tem is elevated above the level of the patient, if the collection lines are
ﬂushed, or if there is obstruction to ﬂowin the collection system. This intra-
lumenal route of bacterial entry is the most common route of bladder

1111

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

contamination in male patients [76]. Some bacteria can create a ‘‘creeping
bioﬁlm’’ along the exterior surface of the catheter, thus gaining access to the
bladder via the urethral lumen [74]. Motile bacteria may also be able to
move within the ﬂuid layer between the catheter and the urethral mucosa.
This extralumenal route of catheter-associated bladder contamination is
most common in female patients [76], and fecal contamination of the cath-
eter is a common source of bacteria.

Catheter-associated bacteriuria has been well documented in dogs and

cats [80]. In a study involving healthy female dogs, 20% of the dogs devel-
oped bacteriuria within 72 hours of a single catheterization [81]. In another
study, the use of open indwelling urinary catheters induced bacteriuria in 20
of 36 healthy cats within a 5-day period [82]. Other studies in small animal
patients have demonstrated that even if a closed collection system is utilized,
bacteriuria develops in 32% to 52% of these animals within a few days
[43,83]. Urethritis or trauma to the urethra and bladder can predispose a
patient to bacterial colonization and urinary tract infections. In a study
involving healthy catheterized male cats, morphologic changes in the ure-
thra, including mild to moderate urethritis and variable loss of mucosal
integrity as well as bacterial colonization, occurred within 1 to 3 days after
catheterization [82]. As in human studies, the use of systemic antibiotics dur-
ing catheterization of small animals may decrease the frequency of urinary
tract infections, but the infections that develop tend to have increased anti-
microbial resistance [82].

Urinary tract infections can be minimized by limiting the use of catheter-

ization. Urinary catheters are indicated when the measurement of urine out-
put is important, when the patient is unable to urinate (ie, urethral
obstruction), and when surgery of the bladder requires that the bladder
remain empty for a period of time. Urinary catheterization may be consid-
ered in recumbent animals to prevent urine scalding. The risk of urinary
tract infection in these already debilitated animals is a concern, however,
and manually expressing the bladder periodically may be a better option.
Catheterization should rarely be used to obtain urine samples. Cystocentesis
is a safe means of obtaining urine in most patients and allows bacterial cul-
ture to be performed. Cystocentesis should be avoided if the patient cannot
be adequately restrained, has an overly distended bladder, or has a coagu-
lopathy. When catheterization is necessary, intermittent catheterization is
recommended compared with a continuous indwelling catheter. The risk
of infection increases with each catheterization; thus, if frequent catheteriza-
tion is anticipated, there may be minimal additional risk to leaving the cath-
eter indwelling. Because the duration of catheterization is the greatest risk
factor for the development of urinary tract infections, once an indwelling
catheter has been placed, the clinician must assess the need for the catheter
on a daily basis and remove it as soon as clinically indicated.

The risk of CAUTIs can also be minimized by strict attention to aseptic

technique and hygiene. The perineal or preputial area should be clipped and

1112

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

cleaned with an antiseptic solution. Povidone iodine solutions may induce
less mucosal reaction and urethritis than chlorhexidine solutions. The anti-
septic should be wiped oﬀ with saline or water. The urinary catheter should
be sterile. Sterile gloves should be worn during catheter placement, although
the ‘‘nontouch’’ technique of handling only the packaging of the catheter
during placement may be adequate in some situations. The catheter should
be no greater in diameter than necessary for the intended purpose and
should be pliable and ﬂexible so as to minimize urethral and bladder trauma.
The catheter should be secured to prevent movement of the nonsterile
exposed portion of catheter into the urethra. A closed collection system
should always be used. Technically, a true closed system consists of a steri-
lized catheter permanently connected to sterilized tubing fused to a sealed
sterilized collection receptacle. The collected urine is removed periodically
from the receptacle through a bottom drain with a one-way valve, which
is clamped closed or capped when not in use [82]. These systems are expen-
sive, and many clinicians instead utilize a combination of a sterile intra-
venous ﬂuid administration set and an empty ﬂuid bag for collection.
Unless such a setup is gas sterilized before use, it is not considered a sterile
closed system. Staﬀ should avoid lifting the collection system above the level
of the patient or ﬂushing the collection lines, because urine in the line and
bag must be considered contaminated. The bag must not be allowed to
become overly full, and the line must not be left clamped or allowed to
become obstructed. Staﬀ should wash their hands and/or wear gloves when-
ever the collection system is opened or the catheter is handled.

Prophylactic antibiotic administration is not recommended. The urine

should be monitored routinely during catheterization for evidence of pyuria
or bacteriuria. If an infection is identiﬁed, antibiotics should be started.
Nosocomial urinary tract infections are frequently polymicrobial and have
an increased spectrum of antimicrobial resistance. It is therefore imperative
that a culture be obtained at the time that infection is identiﬁed or at the
time of catheter removal. The sample can either be urine obtained from the
catheter (not the collection system) before catheter removal or urine
obtained via cystocentesis, or the tip of the catheter itself can be submitted.
If the catheter is to be submitted, the external urethral oriﬁce should be
cleaned before catheter removal, and the tip is submitted as previously
described for intravenous catheter culture. Initial empiric antibiotic therapy
should be based on Gram stain and local data regarding common nosoco-
mial pathogens. Most urinary tract infections are caused by aerobic gram-
negative organisms [43], including resistant Pseudomonas spp. Common
antibiotics selected for urinary tract infections include cephalosporins, tri-
methoprim-sulfa combinations, and b-lactamase–inhibiting penicillin com-
binations, such as amoxicillin with clavulanic acid.

Research is ongoing in human medicine to develop urinary catheters that

reduce the risk of CAUTIs. Novel urinary catheters impregnated with nitro-
furazone or minocycline and rifampin or coated with a silver alloy hydrogel

1113

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

exhibit anti-infective surface activity that signiﬁcantly reduces the risk of
CAUTIs [84]. The reaction of urethral mucosa and submucosa may be
reduced by using Teﬂon-coated or silastic-coated catheters [84,85]. These
products are not likely be readily available in the veterinary market, but
some may be adapted for veterinary use.

Surgical wound infections

Surgical wound infections comprise approximately one quarter of the

reported nosocomial infections in human hospitals, contributing signiﬁ-
cantly to the morbidity and mortality of aﬀected patients and prolonging
hospitalization by an average of 7 days [86]. Factors contributing to the
development of incisional wound infections in human patients include dura-
tion of hospitalization before and after surgery [87], duration of the surgical
procedure, razor preparation of the surgical site, presence of abdominal
drains or active infections in remote regions of the body, and weight loss
or protein depletion of the patient [88]. The duration of the surgical proce-
dure has been cited as the most important contributor to the development of
surgical wound infections in people and animals, because the infection rate
in human beings nearly doubles with every hour the patient spends in sur-
gery [87]. The use of laparoscopic procedures for some types of surgery is
reducing the risk of postoperative infection [89].

Wound infections are the most common type of nosocomial infection

reported in small animals, occurring in 3.5% to 7.6% of all surgical wounds
in dogs and cats [27,90–92]. The risk of infection varies with the contamina-
tion classiﬁcation of the surgical procedure, with the reported incidence of
infections for clean surgeries at 2.5% to 5.7%, clean-contaminated surgeries
at 2.5% to 4.5%, and contaminated surgeries at 5.8% to 21%. Risk factors
that have been speciﬁcally identiﬁed in veterinary patients include prolonged
duration of surgery [93], clipping the surgery site before anesthetic induction
[93], endocrinopathies (particularly diabetes mellitus) [94], infections at sites
remote from the surgical incision [94], and prolonged use of antibiotics after
surgery [93].

The surgical incision disrupts the normal skin or mucous membrane bar-

riers to infection, and the presence of devitalized tissue, such as hematomas,
traumatized soft tissues, dead space, seroma formation, and foreign material
in the wound, limits the ability of normal physiologic defenses to control
bacterial colonization. Endogenous antimicrobial defenses include comple-
ment-mediated, cell-mediated, and humoral mechanisms of the immune sys-
tem but are limited when circulation to the aﬀected tissues is poor. The
eﬀectiveness of antibiotics administered during and after surgery requires
adequate tissue perfusion to maintain eﬀective tissue levels of the antibiotic.
It is essential that the patient be adequately resuscitated from shock or
hypotension before surgical intervention and that measures be taken to
maintain and support perfusion during the postoperative period.

1114

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

Prevention of surgical wound infections requires strict attention to meth-

ods that limit the introduction of exogenous organisms into the wound and
limit the risk of infection by endogenous ﬂora. Surgical sites should be
clipped immediately before surgery, because a recent study demonstrated
that sites clipped before anesthetic induction were three times more likely
to become infected than sites clipped immediately before surgery [93]. Pre-
sumably, the clipping created minor skin abrasion and irritation, which
allowed bacterial colonization to be established by the time the incision was
made. Strict aseptic technique should be followed, including proper prepa-
ration of the surgeon(s), the patient, and all surgical equipment. The sur-
geon(s) should wear a protective cap, mask, and shoe covers as well as a
sterile gown and gloves. It is important to limit traﬃc of personnel in the
surgical suite, and anyone entering the suite should wear clean clothing as
well as a surgical cap, mask, and shoe covers. There should be limited talk-
ing and activity in the proximity of the patient. The surgical suite should be
disinfected routinely, especially after surgeries involving the gastrointestinal
tract or other contaminated surgeries.

The surgical technique is important in that there should be minimal trauma

to soft tissues, and eﬀorts should be taken to remove devitalized tissue and to
limit hemorrhage, seroma formation, and dead space. If contamination is sus-
pected as a result of either leakage from a site within the surgery ﬁeld or a break
in sterile technique, copious lavage should be performed. If a drain is indicated
because of signiﬁcant contamination or unavoidable dead space, closed-suc-
tion drainage is recommended to prevent transport of organisms from the
environment into the wound [88]. The use of monoﬁlament nonabsorbable
suture is recommended in all contaminated wounds. Most importantly, eﬀorts
should be made to minimize the duration of the surgical procedure.

Prophylactic antibiotics are indicated in many surgical procedures but

should be limited to the immediate perioperative period in most cases. Pro-
phylactic antibiotics are recommended for clean surgeries in which implants
are to be left in place, clean surgeries lasting longer than 90 minutes, clean-
contaminated surgeries, and dirty surgeries [88,92]. The ﬁrst 24 hours after
wound contamination is called the ‘‘decisive period’’ during which an
inﬂammatory lesion and subsequent bacterial infection are most likely to
develop. Antibiotics can best prevent the formation of this inﬂammatory
lesion if administered within the ﬁrst 3 hours after contamination, also
called the ‘‘eﬀective period’’ [94]. Antibiotics should be administered 30
minutes before the initial incision to ensure adequate tissue levels of the drug
at the time of contamination. Repeat administration should occur at 2- to 3-
hour intervals during surgery (throughout the eﬀective period), but there is
no beneﬁt to continuing the antibiotic beyond 24 hours (the decisive period)
unless the development of infection is already suspected. In fact, in a study
on small animal patients, the postoperative infection rate was increased
in patients

receiving prolonged postoperative antibiotics compared

with patients receiving only perioperative antibiotics [93]. Perhaps more

1115

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

surprisingly, a recent study in human patients demonstrated no beneﬁt to
prolonging perioperative antibiotics in patients undergoing surgery for
traumatic colonic or hollowviscous injuries [95].

The antibiotics selected for perioperative use should ideally be of narrow

spectrum, selected on the basis of local data on common hospital wound
infections. Indiscriminate use of broad-spectrum antibiotics may increase
the risk of infection with resistant organisms. Coagulase-positive Staphylo-
coccus spp and Escherichia coli are the most common isolates from surgical
wounds in human patients, and coagulase-positive staphylococci are the
most common organisms isolated in reports on surgical wounds in small
animal patients [92]. First-generation cephalosporin antibiotics are com-
monly selected for perioperative use because they have excellent activity
against both Staphylococcus spp and E. coli and minimal toxicity to the
patient. Some investigators have recommended that the perioperative anti-
biotic protocol for patients undergoing colonic surgery should include the
preoperative administration of oral antibiotics (ie, neomycin and erythro-
mycin or metronidazole) in addition to routine parenteral antibiotics [96].

Trauma victims are particularly susceptible to surgical wound infections,

because open wounds and extensive areas of soft tissue trauma are suscep-
tible to contamination by nosocomial pathogens. All wounds should be cov-
ered at the time of hospital admission and should remain covered until they
can be cleaned and managed appropriately. Open fractures should be cul-
tured at the time of surgery, because most postoperative cases of osteomye-
litis are caused by the same organism cultured at the time of surgery [97].

Postoperative infections in veterinary patients include incisional infec-

tions, body cavity infections (ie, after dehiscence of a gastrointestinal anas-
tomosis), and osteomyelitis or infection of orthopedic implants. Organisms
causing postoperative infections are likely to be resistant to antibiotics
administered at the time of surgery, and it is important to obtain micro-
biologic cultures to guide antimicrobial therapy. Empiric therapy initiated
before culture results are available should be selected based on local data
on nosocomial pathogens and knowledge of the likely organisms given the
body system that is infected and the likely source of the contaminating
organisms.

Nosocomial diarrhea

Nosocomial gastrointestinal infections are usually identiﬁed when an out-

break of infectious diarrhea is identiﬁed in a hospital. The organisms most
commonly isolated during outbreaks of infectious diarrhea in human and
veterinary hospitals are Salmonella spp and Clostridia spp [98–101]. These
organisms have been shown to be present in 10% to 39% of cats and dogs
in a carrier state [101,102]. Risk factors for Clostridium diﬃcile-associated
nosocomial diarrhea in human patients include antimicrobial therapy,
advanced age, chemotherapy, and length of hospital stay [98].

1116

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

In human hospitals, most infections by clostridial organisms are transmit-

ted to patients through contamination of the patient’s local environment or
on the hands of hospital personnel [103,104]. Reports of Salmonella out-
breaks in veterinary hospitals similarly demonstrate that transmission is
generally via a common vehicle in the hospital setting rather than by direct
animal-to-animal transmission [105]. In one study, 2% of hospitalized small
animal patients during a 3-month period developed clinical salmonellosis,
and fecal cultures identiﬁed the same strain in all cases. This suggested a per-
sistent source of the organisms in the hospital environment [99]. In another
veterinary study, analysis of the common environmental exposures in patients
with nosocomial diarrhea demonstrated that these patients were more likely
to have been exposed to the same cages and to the same type of refrigerated
food [106]. Outbreaks of C. diﬃcile have been reported at veterinary teaching
hospitals. C. diﬃcile has been shown to persist in the environment in small
animal hospitals, often in areas of high animal traﬃc as well as in areas in
which roughened surfaces made disinfection diﬃcult [107]. Symptoms
of nosocomial Clostridium spp infection in dogs include mild depression,
anorexia, soft to watery or mucoid feces (with or without blood), mucus,
and tenesmus [98]. An outbreak of cryptosporidiosis has been reported at a
veterinary hospital and involved multiple species, including human beings
[108]. There have been numerous reports of Salmonella spp outbreaks in large
animal teaching hospitals, causing signiﬁcant increases in mortality and mor-
bidity. In most cases, transmission of Salmonella organisms was suspected to
occur via a common environmental source [103,105–107]. Control of such
outbreaks has involved widespread implementation of cleaning and biosecur-
ity measures, including training sessions and manuals for hospital personnel,
extensive modiﬁcations of the physical hospital [109,110], minimizing the
indiscriminate use of certain antibiotics [111], and, in some cases, closure of
the hospital to allowintensive disinfection [112].

There is evidence that the use of multiple antimicrobial agents or pro-

longed antimicrobial therapy in human patients increases the risk of noso-
comial Clostridia spp infection [104]. A surveillance study in canine
patients demonstrated an increase in the isolation of C. diﬃcile in patients
that had been hospitalized and had received antibiotic therapy [113]. In a
report on an outbreak of nosocomial Salmonella infection in dogs, it was
demonstrated that dogs with clinical salmonellosis were 5.6 times more
likely to have been given oral antibiotics, 11.3 times more likely to have been
given parenteral antibiotics, and 37.9 times more likely to have been given
oral and parenteral antibiotics [106].

Prevention of nosocomial gastrointestinal infection should include a pro-

tocol of standardized hygiene practices, limited and judicious use of anti-
microbial therapy, and careful surveillance and rapid identiﬁcation of new
hospital cases. When a case of infectious diarrhea is suspected, the aﬀected
patient(s) should be isolated from other hospitalized patients. Fecal cultures
should be submitted, especially if multiple cases have been identiﬁed, which

1117

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

might suggest the hospital environment as a source. All staﬀ members han-
dling aﬀected patients should wear barrier clothing and gloves and should
wash their hands before handling any other patient in the hospital. All cages
and hospital surfaces should be routinely disinfected, but when a case of
infectious diarrhea is suspected, all potential fomites should be disinfected,
including bedding, food dishes, and thermometers. The administration of
antibiotics for the treatment of diarrhea should be restricted to those
patients with clinical signs of systemic illness or fecal cultures supporting
bacterial pathogens as an underlying cause.

Universal strategies for the prevention of nosocomial infections

Despite consistent evidence that nosocomial pathogens are transmitted

on the hands of hospital personnel [103,104], compliance with hand-washing
protocols in human hospitals is low [114]. Hand-washing with plain soap is
eﬀective against many transient microbial ﬂora, but the use of antiseptics
has been shown to reduce the rate of nosocomial infections [115]. The fre-
quent use of antiseptic soaps can cause drying and irritation to hands, how-
ever, which leads to diminished compliance among hospital staﬀ. Studies
have demonstrated that the use of plain soap in conjunction with an alco-
hol-based hand rub is an eﬀective means of maintaining hand hygiene and
improves compliance [116–118]. The use of nonsterile gloves provides
equal protection to antiseptic hand washing, and the use of gloves may be
a more readily enforceable strategy than a hand-washing protocol [53]. The
use of barrier clothing, including gowns, gloves, masks, and shoe covers, has
been shown to decrease the incidence of nosocomial infection [119].
Although the use of barrier clothing may not be practical in most settings,
patients with increased susceptibility to infection (neonates, chemotherapy
patients) may beneﬁt from the extra precaution. Patients with suspected
infectious disease (especially viral diseases or infectious diarrhea) should
be physically isolated from other patients, and barrier clothing should be
utilized. A high workload has also been associated with increased rates of
nosocomial infections, most likely because of diminished compliance with
hand-washing and infection control measures [120].

Medical equipment can be a source of nosocomial pathogens and must be

kept clean. The level of disinfection required depends on the type of use for
each piece of equipment. Equipment that may contact a sterile body site,
such as surgical instruments and implants, should be sterilized using a steam
autoclave or ethylene oxide gas chamber. The functioning of sterilization
equipment should be checked routinely. Test strips are used to monitor
adequate heat or gas levels in each package of equipment. Commercially
available test kits utilize Bacillus spores that are periodically run through
the sterilization process to determine whether complete kill of the organisms
occurs. Equipment that comes in contact with mucous membranes, in-

1118

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

cluding endoscopy equipment, laryngoscopes, endotracheal tubes, and
orogastric tubes, should undergo routine cleaning and high-level disinfec-
tion with glutaraldehyde or a similar agent. If glutaraldehyde is used, a mini-
mum soak time of 10 minutes is recommended. Hospital surfaces, including
ﬂoors, cages, walls, tables, sinks, tubs, and faucets, should also be routinely
disinfected.

In addition to preventing nosocomial infection, it is important to prevent

the development of multiple resistant bacteria. Clinicians should avoid using
antibiotics when a bacterial infection has not been conﬁrmed. Antibiotics
used in the initial treatment of an infection should be selected based on the
eﬀectiveness against the most likely organisms causing the infection (see
Table 1) as well as on the penetration into the body site aﬀected. The anti-
biotic should be used at an appropriate dose and for an appropriate dura-
tion of time [30]. Whenever possible, a culture should be submitted to
determine the true susceptibility pattern of the bacteria involved. Culture
results from suspected nosocomial infections should be collected and peri-
odically analyzed. Such surveillance for nosocomial infections has been pro-
ven to be an eﬀective strategy for reducing the rate of nosocomial infections
[80]. If it becomes apparent that nosocomial bacteria are resistant to rou-
tinely used antibiotics, cycled changes in antibiotics should be considered.
Antibiotic cycling is a strategy to reduce antimicrobial resistance rates by
withdrawing an antibiotic from use until resistance to the antimicrobial is
diminished and then reintroducing it at a later point in time.

Summary

Nosocomial infections and antimicrobial resistance are topics that have

been intensely studied in human medicine because of their signiﬁcant impact
on human health. In recent years, concerns have been raised that the use of
antibiotics in veterinary medicine, animal husbandry, and agriculture may
be contributing to the development of resistance in common bacterial spe-
cies aﬀecting human beings. Although there is inadequate proof at this time
that the resistance is transmitted from animals to people, if antibiotics con-
tinue to be used indiscriminately in veterinary medicine, veterinarians may
ﬁnd themselves facing regulations restricting the use of some antibiotics.

Nosocomial infections have been reported in veterinary medicine and are

likely to increase in prevalence with the increase in intensive care practices in
many hospitals. Prolonged hospitalization and the use of invasive devices
and procedures increase the risk of nosocomial disease. As in human med-
icine, organisms isolated in the nosocomial infections reported in veterinary
patients have an increasingly broad spectrum of antimicrobial resistance.
Despite these ﬁndings, the use of empiric and prophylactic antibiotic
therapy is still widespread in veterinary medicine. Nosocomial infections
and antimicrobial resistance may have a serious impact on the future of

1119

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

veterinary medicine, because the cost and ability to treat our patients may be
aﬀected by the loss of access to or eﬀectiveness of antimicrobial drugs.

Despite the millions of dollars spent on research to reduce the incidence

of nosocomial infections in human patients, the strategies that have

Table 1
Guide to empiric antibiotic therapy for nosocomial infections

Site of infection

Common organisms
isolated

First-line antibiotic
selection

Intravenous catheter/

bloodstream

Normal skin ﬂora

(Staphylococcus spp,
Streptococcus spp)

First-generation

cephalosporins,
amoxicillin/clavulanic
acid, enroﬂoxacin

Enterobactericeae

(Klebsiella, Escherichia
coli)

Third-generation

cephalosporins,
aminoglycosides,
enroﬂoxacin

Urinary tract

Gram-negative enteric

organisms (E. coli,
Enterobacter spp),
Serratia spp

Aminoglycosides,

enroﬂoxacin

Pseudomonas spp

Third-generation

cephalosporins

Staphylococcus spp,

Streptococcus spp

First-generation

cephalosporins,
amoxicillin/clavulanic
acid, enroﬂoxacin

Lungs (pneumonia)

Gram-negative bacilli

Aminoglycosides combined

with b-lactam
(ie, cephalosporins)

Pseudomonas spp

Third-generation

cephalosporins

Staphylococcus spp

First-generation

cephalosporins,
amoxicillin/clavulanic
acid

Gastrointestinal tract

Clostridia spp

Ampicillin, amoxicillin/

clavulanic acid,
clindamycin

Salmonella spp

Trimethoprim-sulfa,

amoxicillin,
aminoglycosides

Campylobacter spp

Erythromycin,

aminoglycosides,
enroﬂoxacin, clindamycin

Surgical wound

Coagulase-positive

Staphylococcus spp

First-generation

cephalosporins,
amoxicillin/clavulanic
acid, enroﬂoxacin

E. coli

Aminoglycosides,

trimethoprimsulfa

1120

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

consistently proven successful are simple and inexpensive to implement. The
most important factor in preventing nosocomial infections is improving the
hygiene practices of health care providers. Hand-washing or the use of dis-
posable gloves can dramatically reduce the transmission of bacteria between
patients. Aseptic technique should be used in the placement and manage-
ment of all invasive devices. All staﬀ members should be educated on the
risks and symptoms associated with nosocomial infections so that cases can
be detected early and treated appropriately. We in the veterinary profession
have the opportunity to learn from the experiences of the human medical
profession and can take steps to prevent the escalation of nosocomial infec-
tions and their impact on our profession.

References

[1] Wenzel RP, Osterman CA, Townsend TR, et al. Development of a statewide program for

surveillance and reporting of hospital-acquired infections. J Infect Dis 1979;140:741–6.

[2] De Man P, van Der Veeke E, Leemreijze M, et al. Enterobacter species in a pediatric

hospital: horizontal transfer or selection in individual patients? J Infect Dis 2001;184:
211–4.

[3] LeFrock JL, Ellis CA, Weinstein L. The relation between aerobic fecal and oropharyngeal

ﬂora in hospitalized patients. Am J Med Sci 1979;277:275–80.

[4] Berrouane YF, McNutt LA, Zbuschelman BJ, et al. Outbreak of severe Pseudomonas

aeruginosa infections caused by a contaminated drain in a whirlpool bathtub. Clin Infect
Dis 2000;31:1331–7.

[5] Cowen AE. The clinical risks of infection associated with endoscopy. Can J Gastroenterol

2001;15:321–31.

[6] Schelenz S, French G. An outbreak of multidrug-resistant Pseudomonas aeruginosa

infection associated with contamination of bronchoscopes and an endoscope washer-
disinfector. J Hosp Infect 2000;46:23–30.

[7] Simmons SA. Laryngoscope handles: a potential for infection. American Association of

Nurse Anesthetists Journal 2000;68:233–6.

[8] Nunez S, Moreno A, Green K, Villar J. The stethoscope in the emergency department:

a vector of infection? Epidemiol Infect 2000;124:233–7.

[9] Bures S, Fishbain JT, Uyehara CF, et al. Computer keyboards and faucet handles as

reservoirs of nosocomial pathogens in the intensive care unit. Am J Infect Control 2000;
28:465–71.

[10] Van den Berg RW, Claahsen HL, Niessen M, et al. Enterobacter cloacae outbreak in the

NICU related to disinfected thermometers. J Hosp Infect 2000;45:29–34.

[11] Lacey S, Flaxman D, Scales J, Wilson A. The usefulness of masks in preventing transient

carriage of epidemic methicillin-resistant Staphylococcus aureus by healthcare workers.
J Hosp Infect 2001;48:308–11.

[12] Wang JT, Chang SC, Ko WJ, et al. A hospital-acquired outbreak of methicillin-resistant

Staphylococcus aureus infection initiated by a surgeon carrier. J Hosp Infect 2001;47:104–9.

[13] Chang HJ, Miller HH, Watkins N, et al. An epidemic of Malassezia pachydermatis in an

intensive care nursery associated with colonization of health care workers’ pet dogs.
N Engl J Med 1998;338:706–11.

[14] Drusin LM, Ross BG, Rhodes KH, et al. Nosocomial ringworm in a neonatal intensive

care unit: a nurse and her cat. Infect Control Hosp Epidemiol 2000;21:605–7.

[15] Sequin JC, Walker RD, Caron JP, et al. Methicillin-resistant Staphylococcus aureus

outbreak in a veterinary teaching hospital: potential human-to-animal transmission. J Clin
Microbiol 1999;37:1459–63.

1121

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

[16] Hoﬀman SB. Mechanisms of antibiotic resistance. Compend Contin Educ Pract Vet 2001;

23:464–72.

[17] Rasmussen BA, Bush K, Tally FP. Antimicrobial resistance in anaerobes. Clin Infect Dis

1997;24(Suppl):S110–20.

[18] Jenkins SG. Mechanisms of bacterial antibiotic resistance. NewHoriz 1996;4:321–32.
[19] Jeljaszewicz J, Mlynarczyk G, Mlynarczyk A. Antimicrobial resistance in gram-positive

cocci. Int J Antimicrob Agents 2000;16:473–8.

[20] Young EJ, Sewell CM, Koza MA, Clarridge JE. Antibiotic resistance patterns during

aminoglycoside restriction. Am J Med Sci 1985;190:223–7.

[21] Jones RN. Resistance patterns among nosocomial pathogens: trends over the past few

years. Chest 2001;119(Suppl):397S–404S.

[22] Medleau L, Long RE, Brown J, Miller WH. Frequency and antimicrobial susceptibility of

Staphylococcus species isolated from canine pyodermas. Am J Vet Res 1986;47:229–31.

[23] Cox HU, Hoskins JD, Roy AF, et al. Antimicrobial susceptibility of coagulase-positive

staphylococci isolated from Louisiana dogs. Am J Vet Res 1984;45:2039–42.

[24] Lilenbaum W, Nunes ELC, Azeredo MAI. Prevalence and antimicrobial susceptibility of

staphylococci isolated from the skin surface on clinically normal cats. Lett Appl Microbiol
1998;27:224–8.

[25] Schwarz S, Roberts MC, Werckenthin C, et al. Tetracycline resistance in Staphylococcus

spp from domestic animals. Vet Microbiol 1998;63:217–27.

[26] Pellerin JL, Bourdeau P, Sebbag H, Person JM. Epidemiosurveillance of antimicrobial

compound resistance of Staphylococcus intermedius isolates from canine pyodermas.
Comp Immunol Microb Infect Dis 1998;21:115–33.

[27] Glickman LT. Veterinary nosocomial (hospital-acquired) Klebsiella infections. JAVMA

1981;179:1389–92.

[28] Koterba A, Torchia J, Silverthorne C, et al. Nosocomial infections and bacterial antibiotic

resistance in a university equine hospital. JAVMA 1986;189:185–91.

[29] Clemence MA, Walker D, Farr BM. Central venous catheter practices: results of a survey.

Am J Infect Control 1995;23:5–12.

[30] Spengler RG, Greenough WB. Hospital costs and mortality attributed to nosocomial

bacteremias. JAMA 1978;240:2455–8.

[31] Lark RL, Chenoweth C, Saint S, et al. Four year prospective evaluation of nosocomial

bacteremia: epidemiology, microbiology, and patient outcome. Diagn Microbiol Infect
Dis 2000;38:131–40.

[32] Fry DE, Fry RV, Borzotta AP. Nosocomial blood-borne infection secondary to intra-

venous devices. Am J Surg 1994;167:268–72.

[33] Collins RN, Brown PA, Zinner SH, Kass EH. Risk of local and systemic infection with

polyethylene intravenous catheters: a prospective study of 213 catheterizations. N Engl J
Med 1968;279:340–3.

[34] Maki DG. Epidemic nosocomial bacteremias. In: Wenzel RR, editor. Handbook on

hospital-acquired infection. West Palm Beach: CRC Press Inc.; 1981.

[35] Bjornson HS. Pathogenesis, prevention, and management of catheter-associated infec-

tions. NewHoriz 1993;1:271–8.

[36] Raad II, Bodey GP. Infectious complications of indwelling vascular catheters. Clin Infect

Dis 1992;15:197–210.

[37] Freeman J, Goldmann DA, Smith NE, et al. Association of intravenous lipid emulsion

and coagulase-negative staphylococci bacteremia in neonatal intensive care units. N Engl
J Med 1990;323:301–8.

[38] Burrows CF. Inadequate skin preparation as a cause of intravenous catheter-related infec-

tion in the dog. JAVMA 1982;180:747–9.

[39] Fox JG, Beaucage CM, Folta CA, Thornton GW. Nosocomial transmission of Serratia

marcescens in a veterinary hospital due to contamination by benzalkonium chloride.
J Clin Microbiol 1981;14:157–60.

1122

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

[40] Mathews KA, Brooks MJ, Valliant AE. A prospective study of intravenous catheter

contamination. J Vet Emerg Crit Care 1996;6:33–43.

[41] Wilkins RJ. Serratia marcescens septicemia in the dog. J Small Anim Pract 1973;14:

205–15.

[42] Spurlock SL, Spurlock GH. Risk factors of catheter-related complication. Compend

Contin Educ Pract Vet 1990;12:241–8.

[43] Lippert AC, Fulton RB, Parr AM. Nosocomial infection surveillance in a small animal

intensive care unit. J Am Anim Hosp Assoc 1988;24:627–36.

[44] Johnson JJ, Murtaugh RJ. Preventing and treating nosocomial infections. Part 2. Wound,

blood and gastrointestinal infections. Compend Contin Educ Pract Vet 1997;19:693–703.

[45] Maki DG, Ringer M, Alvarado CJ. Prospective randomised trial of povidone-iodine,

alcohol, and chlorhexidine for prevention of infection associated with central venous and
arterial catheters. Lancet 1991;338:339–43.

[46] Tager IB, Ginsberg MB, Ellis SE, et al. An epidemiologic study of the risks associated

with peripheral intravenous catheters. Am J Epidemiol 1983;118:839–51.

[47] Maki DG, Band JD. A comparative study of polyantibiotic and iodophor ointments in

prevention of vascular catheter-related infections. Am J Med 1981;70:739–44.

[48] Timsit JF. Scheduled replacement of central venous catheters is not necessary. Infect

Control Hosp Epidemiol 2000;21:371–4.

[49] Cobb DK, High KP, Sawyer RG, et al. A controlled trial of scheduled replacement of

central venous catheters and pulmonary-artery catheters. N Engl J Med 1992;327:1062–8.

[50] Weinstein RA, Kabins SA. Strategies for prevention and control of multiple drug-

resistant nosocomial infection. Am J Med 1981;70:449–54.

[51] Schmid MW. Risks and complication of peripherally and centrally inserted intravenous

catheters. Crit Care Nurs Clin North Am 2000;12:165–74.

[52] Gross PA, Neu HC, Aswapokee P, et al. Deaths from nosocomial infections: experience in

a university hospital and a community hospital. Am J Med 1980;68:219–23.

[53] Rosoﬀ LJ, Borenstein M, Isenberg HD. Is hand washing really needed in an intensive care

unit? Crit Care Med 1995;23:1211–6.

[54] Druskin M, Siegal JP. Bacterial contamination of indwelling intravenous polyethylene

catheters. JAMA 1963;185:966–8.

[55] Richards MJ, Edwards JR, Culver DH, Gaynes RP. Nosocomial infections in combined

medical-surgical intensive care units in the U.S. Infect Control Hosp Epidemiol 2000;2:
510–5.

[56] Marin MG, Lee JC, Skurnick JH. Prevention of nosocomial bloodstream infections:

eﬀectiveness of antimicrobial-impregnated and heparin-bonded central venous catheters.
Crit Care Med 2000;28:3332–8.

[57] Stevens RM, Teres D, Skillman JJ, et al. Pneumonia in an intensive care unit: a thirty-

month experience. Arch Intern Med 1974;134:106–11.

[58] Toews GB. Southwestern internal medicine conference: nosocomial pneumonia. Am J

Med Sci 1986;291:355–67.

[59] A’Court C, Garrard CS. Nosocomial pneumonia in the intensive care unit: mechanisms

and signiﬁcance. Thorax 1992;47:465–73.

[60] Johanson WG, Pierce AK, Sanford JP, et al. Nosocomial respiratory infections with gram-

negative bacilli: the signiﬁcance of the respiratory tract. Ann Intern Med 1972;77:701–6.

[61] Kappstein I, Schulgen G, Friedrich T, et al. Incidence of pneumonia in mechanically

ventilated patients treated with sucralfate or cimetidine as prophylaxis for stress bleeding:
bacterial colonization of the stomach. Am J Med 1991;91(Suppl 2A):125S–31S.

[62] Craven DE, Steger KA, Barber TW. Preventing nosocomial pneumonia: state of the art

and perspectives for the 1990s. Am J Med 1991;91(Suppl 3B):44S–53S.

[63] Van Saene HKF, Stoutenbeek CC, Stoller JK. Selective decontamination of the digestive

tract in the intensive care unit: current status and future prospects. Crit Care Med
1992;20:691–703.

1123

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

[64] Silvestri L, Mannucci F, van Saene HK. Selective decontamination of the digestive tract:

a life saver. J Hosp Infect 2000;46:106–9.

[65] Yavagal DR, Karnad DR, Oak JL. Metoclopramide for preventing pneumonia in

critically ill patients receiving enteral tube feeding: a randomized controlled trial. Crit
Care Med 2000;28:1408–11.

[66] WoodrowP. Will nursing ICU patients in semi-recumbent positions reduce rates of

nosocomial infections? Nurs Crit Care 2000;5:174–8.

[67] Cheatham ML, Chapman WC, Kep SP, Sawyers JL. A meta-analysis of selective versus

routine nasogastric decompression after elective laparotomy. Ann Surg 1995;221:469–76.

[68] Ibanez J, Penaﬁel A, Marse P, et al. Incidence of gastroesophageal reﬂux and aspiration in

mechanically ventilated patients using small-bore nasogastric tubes. Journal of Parenteral
and Enteral Nutrition 2000;24:103–6.

[69] Fourrier F, Cau-Pottier E, Boutigny H, et al. Eﬀects of dental plaque antiseptic

decontamination on bacterial colonization and nosocomial infections in critically ill
patients. Intensive Care Med 2000;26:1239–47.

[70] Craven DE, Steger KA. Ventilator-associated bacterial pneumonia: challenges in diag-

nosis, treatment, and prevention. NewHoriz 1998;6(Suppl):S30–42.

[71] Cunha BA. Nosocomial pneumonia. Diagnostic and therapeutic considerations. Med Clin

North Am 2001;85:79–114.

[72] Meares EM. Current patterns in nosocomial urinary tract infections. Urology 1991;

37(Suppl):9–12.

[73] Haley RW, Culver DH, Morgan WM, et al. Increased recognition of infectious disease

in U.S. hospitals through increased use of diagnostic tests. Am J Epidemiol 1985;121:
168–81.

[74] Nickel JC. Catheter-associated urinary tract infection: newperspectives on old problems.

Can J Infect Control 1991;6:38–42.

[75] Warren JW, Muncie HL, Bergquist EJ, et al. Sequelae and management of urinary

infection in the patient requiring chronic catheterization. J Urol 1981;125:1–8.

[76] Stamm WE. Catheter-associated urinary tract infections: epidemiology pathogenesis and

prevention. Am J Med 1991;91(Suppl 3B):65S–71S.

[77] Tambyah PA, Maki DG. Catheter-associated urinary tract infection is rarely symp-

tomatic: a prospective study of 1,497 catheterized patients. Arch Intern Med 2000;
160:678–82.

[78] Schaeﬀer AJ. Catheter-associated bacteriuria. Urol Clin North Am 1986;13:735–47.
[79] Asher EF, Oliver BG, Fry DE. Urinary tract infections in the surgical patient. Am Surg

1988;54:466–9.

[80] Wise LA, Jones RL, Reif JS. Nosocomial canine urinary tract infections in a veterinary

teaching hospital (1983–1988). J Am Anim Hosp Assoc 1990;26:148–52.

[81] Biertuempfel PH, Ling GV, Ling GA. Urinary tract infection resulting from catheter-

ization in healthy adult dogs. JAVMA 1981;178:989–91.

[82] Lees GE, Osborne CA, Stevens JB, Ward GE. Adverse eﬀects of open indwelling urethral

catheterization in clinically normal male cats. Am J Vet Res 1981;42:825–33.

[83] Barsanti JA, Blue J, Edmunds J. Urinary tract infection due to indwelling bladder

catheters in dogs and cats. JAVMA 1985;187:384–8.

[84] Maki DG, Tambyah PA. Engineering out the risk for infection with urinary catheters.

Emerg Infect Dis 2001;7:342–7.

[85] Painter MR, Borski AA, Trevino GS, et al. Urethral reactions to foreign objects. J Urol

1971;106:227–30.

[86] Brachman PS, Dan BB, Haley RW, et al. Nosocomial surgical infections: incidence and

cost. Surg Clin North Am 1980;60:15–25.

[87] Cruse PJE, Foord R. A ﬁve-year prospective study of 23,649 surgical wounds. Arch Surg

1973;107:206–10.

[88] Nichols RL. Surgical wound infection. Am J Med 1991;91(Suppl 3B):54S–63S.

1124

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

[89] Gaynes RP, Culver DH, Horan TC, et al. Surgical site infection (SSI) rates in the United

States 1992–1998: The National Nosocomial Infections Surveillance System basic SSI risk
index. Clin Infect Dis 2001;33(Suppl 2):S69–77.

[90] Murtaugh RJ, Mason GD. Antibiotic pressure and nosocomial disease. Vet Clin North

Am Small Anim Pract 1989;19:1259–74.

[91] Romatowski J. Prevention and control of surgical wound infection. JAVMA 1989;194:

107–14.

[92] Vasseur PB, Levy J, Dowd E, Eliot J. Surgical wound infection rates in dogs and cats. Vet

Surg 1988;17:60–4.

[93] Brown DC, Conzemius MG, Shofer F, Swann H. Epidemiologic evaluation of post-

operative wound infections in dogs and cats. JAVMA 1997;210:1302–6.

[94] Van den Bogaard AEJM, Weidema WF. Antimicrobial prophylaxis in canine surgery.

J Small Anim Pract 1985;26:257–66.

[95] Kirton OC, O’Neill PA, Kestner M, Tortella BJ. Perioperative antibiotic use in high-risk

penetrating hollowviscus injury: a prospective randomized, double-blind, placebo-control
trial of 24 hours versus 5 days. J Trauma 2000;49:822–32.

[96] Penwick RC. Perioperative antimicrobial chemoprophylaxis in gastrointestinal surgery.

J Am Anim Hosp Assoc 1988;24:133–45.

[97] Stevenson S, Olmstead ML, Kowalski J. Bacterial culturing for prediction of postoper-

ative complications following open fracture repair in small animals. Vet Surg 1986;
15:99–102.

[98] Kruth SA, Prescott JF, Welch MK, Brodskky MH. Nosocomial diarrhea associated with

enterotoxigenic Clostridium perfringens infection in dogs. JAVMA 1989;195:331–3.

[99] Ketaren K, Brown J, Shotts EB, et al. Canine salmonellosis in a small animal hospital.

JAVMA 1981;179:1017–8.

[100] Barbut F, Petit JC. Epidemiology of Clostridium diﬃcile-associated infections. Clin

Microbiol Infect 2001;7:405–10.

[101] Morse EV, Duncan MA. Canine salmonellosis: prevalence, epizootiology, signs and

public health signiﬁcance. JAVMA 1975;167:817–20.

[102] Riley TV, Adams JE, O’Neill GL, Bowman RA. Gastrointestinal carriage of Clostridium

diﬃcile in cats and dogs attending veterinary clinics. Epidemiol Infect 1991;107:
659–65.

[103] McFarland LV, Mulligan ME, Kwok RY, Stamm WE. Nosocomial acquisition of

Clostridium diﬃcile infection. N Engl J Med 1989;320:203–10.

[104] Streulens MJ, Maas A, Nonhoﬀ C, et al. Epidemiology of diarrhoea caused by

enterotoxigenic Clostridium perfringens. J Med Microbiol 1985;20:363–92.

[105] Calvert CA. Salmonella infections in hospitalized dogs: epizootiology, diagnosis, and

prognosis. J Am Anim Hosp Assoc 1985;21:499–503.

[106] Uhaa IJ, Hird DW, Hirsh DC, Jang SS. Case-control study of risk factors associated with

nosocomial Salmonella krefeld infection in dogs. Am J Vet Res 1988;49:1501–5.

[107] Weese JS, Staempﬂi HR, Prescott JF. Isolation of environmental Clostridium diﬃcile from

a veterinary teaching hospital. J Vet Diagn Invest 2000;12:449–52.

[108] Konkle DM, Nelson KM, Lunn DP. Nosocomial transmission of Cryptosporidium in a

veterinary hospital. J Vet Intern Med 1997;11:340–3.

[109] Tillotson K, Savage CJ, Salman MD, et al. Outbreak of Salmonella infantis infection in a

large animal veterinary teaching hospital. JAVMA 1997;211:1554–7.

[110] House JK, Mainar-Jaime RC, Smith BP, et al. Risk factors for nosocomial Salmonella

infection among hospitalized horses. JAVMA 1999;214:1511–6.

[111] Pare J, Carpenter TE, Thurmond MC. Analysis of spatial and temporal clustering of

horses with Salmonella krefeld in an intensive care unit of a veterinary hospital. JAVMA
1996;209:626–8.

[112] Schott HC II, Ewart SL, Walker RD, et al. An outbreak of salmonellosis among horses at

a veterinary teaching hospital. JAVMA 2001;218:1152–9.

1125

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

[113] Struble AL, Tang YJ, Kass PH, et al. Fecal shedding of Clostridium diﬃcile in dogs: a

period prevalence survey in a veterinary medical teaching hospital. J Vet Diagn Invest
1994;6:342–7.

[114] Graham M. Frequency and duration of handwashing in an intensive care unit. Am J

Infect Control 1990;18:77–81.

[115] Martin MA. Nosocomial infections in intensive care units: an overviewof their epi-

demiology, outcome, and prevention. NewHoriz 1993;1:162–71.

[116] Herruzo-Cabrera R, Garcia-Caballero J, Jesus Fernandez-Acenero M. A newalcohol

solution (N-duopropenide) for hygienic (or routine) hand disinfection is more useful than
classic handwashing: in vitro and in vivo studies in burn and other intensive care units.
Burns 2001;27:747–52.

[117] Pittet D, Hugonnet S, Barbarth S, et al. Eﬀectiveness of a hospital-wide programme to

improve compliance with hand hygiene. Lancet 2000;14:1290–1.

[118] Larson LR, Aiello AE, Bastyr J, et al. Assessment of two hand hygiene regimens for

intensive care unit personnel. Crit Care Med 2001;29:944–51.

[119] Klein BS, Perloﬀ WH, Maki DG. Reduction of nosocomial infection during pediatric

intensive care by protective isolation. N Engl J Med 1989;320:1714–21.

[120] Fierobe L, Llucet JC, Decre D, et al. An outbreak of imipenem-resistant Acinetobacter

baumannii in critically ill surgical patients. Infect Control Hosp Epidemiol 2001;22:35–40.

1126

J.A. Johnson / Vet Clin Small Anim 32 (2002) 1101–1126

Analgesia in critical care

Maria M. Glowaski, DVM

Department of Surgical and Radiological Sciences, School of Veterinary Medicine,

University of California at Davis, 2112 Tupper Hall, Davis, CA95616, USA

Pain is as old as human beings and animals have been in existence. Pain

has been deﬁned as an unpleasant sensory and emotional experience associ-
ated with actual or potential tissue damage or described in terms of such
damage [1]. Critical illness presents as a complicated physiologic condition
that is frequently associated with pain. In addition, chronic pain syndromes
may coexist with acute pain and further confound assessment and treat-
ment. Organ system dysfunction, commonly seen in critically ill patients,
can carry signiﬁcant implications with regard to pain management. Inappro-
priate analgesic dosing is common in the extremely ill patient because of the
clinician’s inability to assess pain severity accurately coupled with concerns
about hemodynamic instability or excess drug accumulation.

Pain mechanisms and pathways

One of the most vital functions of the nervous system is to provide infor-

mation about the occurrence or threat of injury. The sensation of pain, by
its inherent aversive nature, contributes to this function. The peripheral neu-
ral apparatus that responds to noxious (injurious or potentially injurious)
stimuli provides a signal to alert the organism of potential injury. This appa-
ratus must respond to the multiple energy forms that produce injury (eg,
heat, mechanical and chemical stimuli) and provide information to the cen-
tral nervous system (CNS) regarding the location and intensity of noxious
stimuli.

Highly specialized sensory ﬁbers, alone or in concert with other special-

ized ﬁbers, provide information to the CNS not only about the environment
but also about the state of the organism itself. In the case of the sensory
capacity of the skin, cutaneous stimuli may evoke a sense of cooling,
warmth, or touch. Accordingly, there are sensory ﬁbers that are selectively

Vet Clin Small Anim 32 (2002) 1127–1144

E-mail address: drglow@yahoo.com (M.M. Glowaski).

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 -5 6 1 6 ( 0 2 ) 0 0 0 4 4 -X

sensitive to these stimuli. Tissue damage results in a cascade of events
that leads to enhanced pain to natural stimuli, termed hyperalgesia. A
corresponding increase in the responsiveness of nociceptors occurs, termed
sensitization.

An outstanding contribution to the understanding of pain mechanisms

was the development and publication of the gate control theory by Melzack
and Wall in 1965 [2]. The gate control theory emphasizes the importance
of both the ascending and descending modulating systems and lays down a
solid framework for the management of diﬀerent pain syndromes [3]. The
gate control theory is based on the fact that small-diameter nerve ﬁbers
carry pain stimuli through a ‘‘gate mechanism’’ but that larger diameter
nerve ﬁbers going through the same gate can inhibit the transmission of the
smaller nerves carrying the pain signal. Chemicals released as a response to
the pain stimuli also inﬂuence whether the gate is open or closed for the
brain to receive the pain signal. This led to the theory that the pain signals
can be interfered with by stimulating the periphery of the pain site, the appro-
priate signal-carrying nerves at the spinal cord, or particular corresponding
areas in the brain stem or cerebral cortex. It is generally recognized that the
pain gate can be shut by means of the administration of opioid analgesics.

Nociception and antinociception

Pain is a complex mixture of unpleasant sensory, emotional, and mental

experiences as well as certain autonomic (involuntary) responses and psy-
chologic and behavioral reactions provoked by tissue damage. Injury to tis-
sues, whether induced by disease, inﬂammation, or trauma, constitutes a
noxious stimulus and causes cellular breakdown with liberation of biochem-
ical substances. These activate special receptors on nociceptors that can be
sensitive to heat, cold, mechanical stimuli, or chemical mediators. All tissues
are also innervated by polymodal C and A-d receptors. In addition, all
tissues have many aﬀerents that are unresponsive in normal tissues but that
become active during inﬂammation (silent nociceptors) as well as some A-b
ﬁbers that become chemosensitive during inﬂammation and aﬀect CNS pain
pathways. All these diﬀerent peripheral nerve ﬁbers exhibit plasticity that
can be rapid (seconds) or delayed (hours to days) and is dependent on elec-
trophysiologic properties, growth factors, cytokines, or other tissue factors.
Although nociceptive impulses are often called pain impulses, pain is not
experienced until information reaches the brain, where the sensory and emo-
tional experience takes place.

Pain provoked by injury or disease is the net eﬀect of many simultane-

ously interacting biochemical, physiologic, and psychologic mechanisms
that involve activity in most parts of the nervous system concerned with
sensory, motivational, and cognitive processes as well as psychodyna-
mic mechanisms. Just as the CNS has a range of modulatory systems that

1128

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

either facilitate pain or inhibit it, the peripheral primary aﬀerents not only
signal tissue damage and become sensitized by inﬂammatory mediators
but can be inhibited by endogenous analgesic mechanisms, such as opioid
administration.

Acute pain, frequently seen in the critical care setting, consists of a com-

plex constellation of unpleasant sensory, perceptual, and emotional experi-
ences with associated autonomic and behavioral responses. These associated
responses can be provoked by noxious stimulation produced by injury or
disease of skin, deep somatic structures, or abnormal function of muscle
or viscera. The pathophysiology of acute pain is fairly well understood;
therefore, its diagnosis should not diﬃcult. As a result of eﬀective treatment,
the self-limiting nature of the disease or injury, or both, as well as the pain
and associated responses usually disappear within days or weeks. Improper
therapy, however, can cause the acute pain to persist and the pathophysiol-
ogy to progress to a chronic condition.

Physiologic and metabolic responses to pain

Pain in the critical care patient has both direct and reﬂex-related eﬀects

on the cardiovascular, pulmonary, gastrointestinal, musculoskeletal, immu-
nologic, and renal systems. Trauma-related pain, primarily that caused by
thoracic and upper abdominal injury, causes decreased pulmonary function
via chest splinting and reﬂex-activated diaphragmatic dysfunction. The con-
sequent hypoventilation and atelectasis result in ventilation-perfusion mis-
matching and hypoxemia. Functional residual capacity and vital capacity
are decreased with the consequent retention of secretions and progressive
atelectasis, which may lead to the development of secondary pneumonia.
Pain may also contribute to pulmonary dysfunction through localized
guarding of muscles around the area of pain and a generalized muscle
rigidity or spasm that restricts movement of the chest wall and diaphragm
[4]. Increased sympathetic tone from pain may decrease gastrointestinal
tract motility, producing an ileus complicating the delivery of nutritional
support. Pain accentuates these catabolic processes (eg, corticotropin
release), slows restoration of function by inhibiting movement, and elicits
increases in sympathetic nervous system outﬂow that stress the other organ
systems [5].

Critical care unit patients commonly have pain and physical discom-

fort from obvious factors, such as trauma, preexisting diseases, or invasive
procedures. Patient pain and discomfort can also be caused by monitoring
and therapeutic devices (eg, catheters, drains, nasal oxygen tubing, endotra-
cheal tubes) and routine nursing care (eg, airway suctioning, physical
therapy, dressing changes, patient restraint) [6,7]. Unrelieved pain may
contribute to inadequate sleep and agitation. It also evokes a stress response
characterized by tachycardia, increased myocardial oxygen consumption,
hypercoagulability, immunosuppression, and persistent catabolism. The

1129

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

combined use of analgesics and sedatives may ameliorate this stress response
in critically ill patients [8].

Stress response

The stress response causes tremendous neurohumoral elevations of

plasma catecholamine, cortisol, glucose, antidiuretic hormone, and acute
phase protein levels. These elevations can result in tachycardia, hyperten-
sion, vasoconstriction, increased oxygen consumption, ventilation-perfusion
mismatch, and reduced gastrointestinal motility. Pain has consequences in
the critically ill patient that can also lead to clinically signiﬁcant physiologic
responses, such as hypercoagulability, increased protein catabolism, and
immunosuppression [8,9]. The primary goals of analgesia are to relieve pain
and anxiety, which can attenuate this stress response. Left untreated, the
stress response can result in higher morbidity and prolong the rehabilitation
of our patients.

Philosophy of pain management

Veterinarians often think in terms of our sensory experience of pain as

human beings. Yet, acute pain is a complex experience that extends well
beyond simple nociceptor stimulation. Pain alone initiates a cascade of hor-
monal changes termed the suprasegmental response. As the contemporary
deﬁnition of pain implies, neuronal input is altered by emotive elements,
such as fear and anxiety. These modiﬁers of pain are common in the veteri-
nary patient that is in an unfamiliar environment and unable to verbally
communicate. Disrupted sleep patterns and repeated invasive and painful
procedures augment the patient’s distress and may precipitate agitated or
aggressive behavior. Diﬃculties with pain management in the critically ill
veterinary patient cannot be explained by a lack of available and appropri-
ate analgesics. There are a multitude of agents that can be used alone and in
combination to completely alleviate pain. What needs to occur is the appre-
ciation that pain should always be treated. Next, the decision of which anal-
gesic, how much, and how often needs to be made. Analgesic medications
have traditionally been administered on an ‘‘as-needed’’ basis; however, this
type of regimen proves to be ineﬀective for many patients. Ideally, contin-
ued evaluation of the patient dictates if the level of analgesia provided by a
chosen agent at the chosen dose and frequency is adequate. Switching from
one agent to another in the same class of drugs, adding one from another
class, increasing the dose, or decreasing the dosing interval are all methods
to increase the eﬀectiveness of the chosen analgesia regimen.

Assessment of pain

Numerous methods for assessing pain in people have been developed, and

attempts have been made to apply these to animals. Although increases in

1130

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

heart rate and respiratory rate are commonly used as indicators of pain, they
are unreliable [10]. Just as human infants are unable to verbally communicate
their pain, we must rely on the same types of pain scoring systems that use
other methods. In veterinary medicine as directly opposed to human medi-
cine, we are unable to communicate with our patients and ask them what
hurts. Despite our limited ability to measure pain, an increased awareness
of the reasons why pain should not be left untreated should eventually lead
to more eﬀective pain control. Even in the human critical care setting, pain
control is often not adequately provided for a variety of reasons, including
fear of oversedation, concern of altering physical ﬁndings, or underestima-
tion of patient needs [42–44]. Most commonly, inadequate analgesia results
from attempting to avoid certain side eﬀects of opiate analgesics, such as res-
piratory depression in spontaneously breathing patients; hypotension, which
is most likely to occur in patients with hypovolemia; and gastric ileus, which
is common in critically ill patients and can be enhanced by opiates. Despite
these legitimate concerns, adequate analgesia should remain a primary goal
in the care of the critical care veterinary patient.

Analgesic agents

Appropriate use of these medications requires a thorough understanding

of drug indications, metabolism, side eﬀects, and monitoring techniques.
Conventional modes of delivering analgesia, such as intramuscular adminis-
tration of opioids, provide inadequate analgesia, because some periods of
subtherapeutic opioid blood levels depend on dosing intervals. When larger
doses of opioids are required, patients suﬀer more side eﬀects, including
excessive sedation, respiratory depression, vomiting, ileus, and constipation.
It is now common knowledge that preventing pain is more eﬀective than
treating established pain. When patients are administered drugs on an as-
needed basis, they may encounter signiﬁcant delays in treatment. Analgesics
should be administered on a continuous or scheduled intermittent basis, with
supplemental bolus doses as required. Constant intravenous administration
usually requires lower and more frequent doses than intramuscular adminis-
tration to titrate to patient comfort. Intramuscular administration is less
preferable than the intravenous route in hemodynamically unstable patients
because of altered perfusion and variable absorption from peripheral tissue.

Opioids

Opioid analgesics are the drugs of choice for pain relief in the critic-

ally ill patient. Opioid agonists, either exogenous or endogenous, produce
analgesia and other central eﬀects by binding to speciﬁc opioid receptors
in the brain and spinal cord. The l-and j-opioid receptors then mediate
these analgesic eﬀects. When an opioid agonist like morphine binds to the

1131

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

-receptor, it results in excellent analgesia but can also cause bradycardia

and respiratory depression. The bradycardia is caused by a vagolytic eﬀect
and is easily counteracted with an anticholinergic. The respiratory depres-
sion results from a decrease in responsiveness of the medullary respiratory
centers to carbon dioxide tension. Panting can also be seen and is attributed
to an opioid-induced resetting of the hypothalamic temperature set point
rather than to actual respiratory depression.

Opioid binding can either stimulate or depress diﬀerent neuronal popula-

tions. For instance, in most animal species, opioids stimulate the Edinger-
Westphal nucleus of the oculomotor nerve to produce miosis and stimulate
the chemoreceptor trigger zone in the area postrema to decrease the thresh-
old of vomiting. Opioids also have a generalized depressant eﬀect on gastro-
intestinal motility. They reduce longitudinal peristalsis and increase
sphincter tone, with constipation as the predictable result. Pain may cause
ileus, however, also resulting in constipation, and it should not be used as
a reason to avoid opioid analgesics [40].

The selection of which opioid to use depends on the degree of pain that

the animal is perceived to be experiencing. Morphine is the prototypic
opioid with which all others are compared and is an excellent analgesic suit-
able for even the most severe pain. Other opioids that are comparable to
morphine in the level of analgesia they provide include the agonists fentanyl,
hydromorphone, and oxymorphone. These pure agonist opioid drugs seem
to have no ceiling eﬀect for analgesia. As the dose is raised, analgesic eﬀects
increase until either analgesia is achieved or the patient loses consciousness.
In practice, it is the appearance of adverse eﬀects, including extreme seda-
tion or profound respiratory depression, that imposes a limit on the useful
dose. A balance between analgesia and the side eﬀects that occur with
increasing dosages should determine the overall eﬃcacy of any drug in a spe-
ciﬁc patient (Table 1) [41].

Another group of opioid analgesics are the mixed agonist-antagonists

and include the agents butorphanol and buprenorphine. These agents do
have a limit to the amount of analgesia that they provide, meaning that
higher doses do not provide greater analgesia. They also do not provide the
same degree of pain relief that the pure opioid agonists can confer and
should therefore be reserved for cases with a mild to moderate degree of
pain.

Morphine

Morphine (morphine sulfate) is derived from the milky exudate of the

unripe seed capsules of the opium poppy (Papaver somniferum). Although
only approximately 10% of the extract is morphine, it is still obtained using
this method. Morphine binds to both the l-and j-opioid receptors,
although it does have actions on other receptors. It can be administered
by either intermittent doses or as a constant rate intravenous infusion

1132

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

(see Table 1). Care must be exercised to avoid rapid intravenous administra-
tion, because hypotension may result from histamine release and sub-
sequent vasodilation. A constant rate infusion of morphine induces eﬀects
similar to those obtained with intramuscular administration every 4 hours
in dogs and does not cause histamine release as seen with intravenous
administration. Also an infusion can attenuate the peaks and valleys seen
with intermittent bolus administration because it provides a continuous level
of analgesia [11]. It can cause an increase in vagal tone, and this may lead to
bradyarrhythmias, which can be prevented with prior use of an anticholiner-
gic. Hypoventilation may also occur with infusions of morphine, and respi-
ratory monitoring is imperative.

Fentanyl

Fentanyl (Sublimaze) is a synthetic opiate with greater potency and lipo-

philic properties than morphine, resulting in a faster onset of action. It has a
shorter duration of action but does not cause histamine release. It lasts
approximately 30 to 60 minutes because of its rapid redistribution to periph-
eral compartments. Therefore, it should be administered by continuous
intravenous infusion when long-term pain control is required. Most patients
are adequately treated with an infusion of fentanyl at 1 to 2 lg/kg/h. As with
morphine, bradycardia and hypoventilation may result, and this under-
scores the need for constant patient monitoring.

Fentanyl patch

Fentanyl may also be administered via a transdermal patch (Duragesic) in

those patients with more chronic analgesic needs or when an infusion is not
available. The patch provides consistent drug delivery, but the extent of
absorption varies depending on the permeability, temperature, perfusion, and
thickness of the skin. There is a large interpatient variability in peak plasma
concentrations. It provides a convenient regimen for the use of a drug previ-
ously limited by a short duration of action and a noninvasive parenteral route.
Transdermal fentanyl can provide continuous controlled systemic delivery of

Table 1
Opioid analgesics

Drug

Dose (mg/kg) IV

Duration (h)

CRI (mg/kg/h)

Buprenorphine

0.01–0.02

4–6

Butorphanol

0.1–0.4

1–4

Fentanyl

0.005–0.02

0.5–1.0

0.01–0.06

Hydromorphone

0.05–0.2

2–4

Morphine

0.5–2.0

2–4

0.10–0.15

Oxymorphone

0.05–0.2

2–4

Naloxone

0.04

0.5

Abbreviations: IV, intravenous; CRI, constant rate infusion.

1133

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

fentanyl base for 72 hours in dogs and for longer than 104 hours in cats [12].
The patch itself is a rectangular transparent unit composed of a protective peel
strip and four functional layers. The amount of fentanyl released from each
system (25 lg/h per 10 cm

) is proportional to the surface area of the patch.

When the system is applied, a fentanyl depot concentrates in the upper skin
layers. Fentanyl plasma concentrations are not measurable until 2 hours after
application, and it can take up to 24 hours until full clinical fentanyl eﬀects are
observed. Fentanyl patches are therefore not a recommended modality for
acute analgesia because of their delay to peak eﬀect. When compared with epi-
dural morphine, transdermal fentanyl provided signiﬁcantly more pain relief
in dogs undergoing major orthopedic surgery [13].

It must be noted that transdermally administered fentanyl results in less

analgesia than expected in some dogs and that there can be substantial indi-
vidual variation [14]. The same variation in plasma drug concentrations with
transdermal absorption in cats has also been observed [12,15]. Because the
eﬃcacy of a transdermal system is primarily dependent on the barrier prop-
erties of the skin of the targeted species and these patches were developed for
human skin, there may be variation in drug delivery [16]. As a result of the
variation in achieving analgesia, patients with a fentanyl patch should be
closely monitored and administered additional opioid agonists if needed.

Hydromorphone

Hydromorphone (Dilaudid) is a semisynthetic morphine derivative and

is approximately six to eight times more potent than morphine. It does
not increase plasma histamine concentration as can be seen with intraven-
ous morphine administration [17]. Hydromorphone’s duration of eﬀect is
greater than that of morphine and equal to that of oxymorphone. It is recom-
mended as a substitute for oxymorphone because it is less expensive.

Oxymorphone

Oxymorphone (Numorphan) is a semisynthetic opioid that is 10 times as

potent as morphine and has a longer duration of action. It also does not
cause the same degree of vomiting and defecation that is seen with morphine
administration. The onset of action of parenterally injected oxymorphone is
rapid, and the initial eﬀects are usually perceived within 5 to 10 minutes,
with analgesia persisting for approximately 2 to 4 hours.

Buprenorphine

Buprenorphine (Buprenex) has 30 times the potency of morphine and can

last up to 6 hours when administered intramuscularly. It is a partial
l

-agonist with high receptor aﬃnity. This attribute makes buprenorphine dif-

ﬁcult to reverse with an opioid antagonist. This opioid is only a partial
l

-agonist; it is less eﬃcacious than a pure l-agonist but can produce

1134

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

moderate levels of analgesia with less respiratory depression. It can be given
intravenously with little eﬀect on the cardiovascular system and does not
result in respiratory depression at clinical doses. It has a slow onset and may
take up to 45 minutes for the complete clinical eﬀect to become evident.
Buprenorphine exhibits a bell-shaped dose response curve for analgesia,
which means that with increasing doses, antinociceptive eﬀects are reduced
to almost zero. As with butorphanol, a pure agonist is preferable compared
with buprenorphine in patients in severe pain.

Butorphanol

Butorphanol (Torbugesic) is a semisynthetic opioid approximately ﬁve

times more potent than morphine. The principal actions of butorphanol
result from its agonist activity at j-receptors and antagonist activity at
l

-receptors. Because of this, it can be used to reverse l-agonists when respira-

tory depression is severe without removing j-mediated analgesia. It should
therefore not be given concurrently to a patient with a fentanyl patch
because it diminishes the eﬀect of fentanyl. Its degree of analgesia is not
as profound as that of a pure agonist, and it should be used in only patients
in mild pain. Butorphanol exhibits a ‘‘ceiling’’ for inducing respiratory
depression; greater dosages of butorphanol do not result in a corresponding
increase in respiratory depression nor do they result in greater analgesia.
When presented with a patient in acute pain, a pure agonist is preferable,
because the level of attainable analgesia is superior.

Naloxone

Naloxone (Narcan) is an eﬀective agent for the reversal of the cardiovas-

cular and respiratory depression associated with opioid administration. It is
essentially a pure narcotic antagonist and acts by antagonizing the actions
of opioids. It is a short-acting medication (30–45 minutes), and supplemen-
tal doses may be needed in patients with respiratory depression induced by
long-acting opioids.

Ketamine

Ketamine (Ketaset) is an anesthetic/analgesic agent. Its primary eﬀect is

blockade of the phencyclidine site in the ion channel associated with the N-
methyl-

-aspartate (NMDA) receptor. Ketamine is classiﬁed as a dissociative

anesthetic, but there is a clinical belief that ketamine can provide a signiﬁcant
degree of analgesia. The current thinking on this aspect of drug action is that it
reﬂects on its action as an antagonist at the glutamate receptor of the NMDA
subtype. The NMDA site is thought to be essential in evoking a hyperalgesic
state in response to repetitive small aﬀerent (C ﬁber) input.

1135

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

Its use is gaining popularity in the critical care setting, especially when

used in conjunction with opioid agents. Speciﬁcally, it has been used in
patients with dermal wounds or burns. Burn wound pain is severe, inconsis-
tent, and commonly underestimated. Ketamine has been found to be a use-
ful agent for analgesia in dogs with burn wounds at a dose of 10 mg/kg
administered orally every 8 hours [18]. In cases where opioids alone are not
suﬃcient for pain control, the addition of ketamine either as an infusion or
intermittent administration provides excellent analgesia. It does not produce
changes in gastrointestinal motility and may allow decreases in the dosages
of opioids required for pain control [19].

Nonsteroidal anti inﬂammatory agents

Nonsteroidal anti-inﬂammatory drugs (NSAIDs) provide analgesia via

the nonselective competitive inhibition of cyclooxygenase (COX), a critical
enzyme in the inﬂammatory cascade. NSAIDs have the potential to cause
signiﬁcant adverse eﬀects, including gastrointestinal bleeding, bleeding sec-
ondary to platelet inhibition, and the development of renal insuﬃciency.
Patients with hypovolemia or hypoperfusion and those with preexisting
renal impairment may be more susceptible to NSAID-induced renal injury
[20,21]. Administration of NSAIDs may reduce opioid requirements,
although the analgesic beneﬁt of NSAIDs has not been systematically
studied in critically ill veterinary patients. It is known that the selective
COX-2–inhibiting agents cause less gastrointestinal irritation with long-term
use than traditional NSAIDs [22].

In clinically normal dogs, NSAIDs like carprofen (Rimadyl) have little

eﬀect on renal function and integrity. Bleeding times can be signiﬁcantly
prolonged, however, and NSAIDs are not recommended in patients that are
anemic or have a coagulopathy [23,24]. Because critically ill veterinary
patients can present with renal disease, coagulopathies, or gastrointestinal
disorders, the NSAIDs are best reserved for use in healthy patients with
chronic pain conditions rather than in acutely ill critical care patients.

Adjunct agents (sedatives/anxiolytics)

Sedatives are common adjuncts to analgesics used to decrease the anxiety

and agitation that may be seen in the critical care patient. The causes of
anxiety in critically ill patients are multifactorial and may be the result of
an inability to communicate their pain, continuous ambient lighting, and
excessive stimulation by caregivers. Recent studies in human patients have
conﬁrmed that agitation may have a deleterious eﬀect on patients by contri-
buting to an increase in tissue oxygen consumption and inadvertent removal
of devices and catheters [25]. Sedatives can reduce the stress response and

1136

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

improve the tolerance of routine ICU procedures [26]. Acepromazine (Pro-
mace) is one of the most commonly used tranquilizers in veterinary medi-
cine. It is a long-acting phenothiazine compound that can exert its eﬀects
for up to 8 hours even at standard dosages (Table 2). Its mode of action in-
volves blockage of dopamine receptors in the CNS. Because of a

-adrenergic

blockade, it may cause profound vasodilation with resulting hypotension
and should be avoided in hypotensive, shocky, or volume-depleted patients.

The benzodiazepines, which also have anticonvulsant eﬀects, can be used

for sedation in critical care patients. They undergo hepatic metabolism, some
to active metabolites that may accumulate in patients with renal and hepatic
insuﬃciency. The most reliable and eﬀective route of administration is intra-
venous, because the absorption of intramuscular doses may vary, especially
with diazepam. Diazepam (Valium) rapidly penetrates the CNS, allowing
for the sedative eﬀect to be seen within 2 to 3 minutes and the peak eﬀect to
be seen within 3 to 5 minutes. Midazolam (Versed) is a short-acting water-
soluble benzodiazepine that becomes a lipophilic compound in the blood and
rapidly penetrates the CNS to produce an onset of sedation (2–2.5 minutes)
similar to that of diazepam. Midazolam is similar to diazepam in all respects,
except that it is well absorbed when administered intramuscularly and can be
combined in the same syringe with other agents without causing precipitation
[27]. Flumazenil (Romazicon) is a competitive antagonist of benzodiazepine
eﬀects and is thus used to reverse excessive sedation in patients after the
administration of a benzodiazepine. Its duration of antagonism is less than
that of either diazepam or midazolam; thus, resedation may result. Flumaze-
nil is best administered intravenously and may need to be repeated at
20-minute intervals until the degree of arousal is achieved. It is ultimately im-
portant to note that these sedatives lack analgesic activity and should be used
in conjunction with an analgesic regimen and not used as a replacement.

Local analgesia

Local anesthesia is useful in a wide variety of clinical situations. It can

increase patient comfort and facilitate patient cooperation during painful
procedures. Local anesthetics act by reversibly blocking nerve impulses by
disrupting permeability to sodium during an action potential [28]. Potency

Table 2
Sedatives/anxiolytics

Drug

Dose (mg/kg) IV

Duration (h)

Acepromazine

0.05–0.1

4–6

Diazepam

0.1–0.4

0.5–1.0

Midazolam

0.1–0.2

0.5–1.0

Flumazenil

0.1

0.5

Abbreviations: IV, intravenous.

1137

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

and duration of action diﬀer among the various agents; the more hydro-
phobic an agent, the greater is the potency and the longer is the duration of
action [29]. Using epinephrine (1:200,000 ratio) mixed with local anesthetics
results in vasoconstriction, which decreases clearance of the agent, increases
the duration of action, and decreases the total required dosage. Vasocon-
striction also aids in hemostasis during wound care. The addition of epi-
nephrine to agents already possessing a long duration of action, such as
bupivacaine (Marcaine), does little to prolong anesthesia. Although many
local anesthetics are available in veterinary medicine, the most commonly
used agents are lidocaine and bupivacaine.

Techniques for the administration of local anesthetics include topical

application, local inﬁltration, and peripheral nerve block. The type of local
anesthetic used determines the duration of analgesia resulting from the
blockade. The duration of action ranges from lidocaine with a duration of
30 minutes to 1 hour to bupivacaine with duration of 4 to 6 hours. There-
fore, bupivacaine may be more appropriate for complex wounds requiring
long repair times. The addition of epinephrine to the local anesthetic solu-
tion also prolongs the duration of anesthesia; however, it should not be used
in nerve blocks of distal appendages, because ischemia and tissue death may
result.

Topical anesthesia

Several agents have been developed to replace or augment local inﬁltra-

tion anesthesia. Topical anesthesia is used to reduce the pain of local inﬁl-
tration, the pain of superﬁcial laceration repair, and the pain associated
with vascular cannulation.

Eutectic mixture of local anesthetics cream

Eutectic mixture of local anesthetics (EMLA) cream is a 1:1 mixture of

two local anesthetics (lidocaine and prilocaine). It is a eutectic mixture,
which is a substance that has a melting point lower than that of either sub-
stance by itself. This combination has excellent dermal penetration, and
because the total amount of anesthetic used is decreased, local irritation and
toxic eﬀects are reduced [30]. Application of EMLA cream allows insertion
of catheters into the cephalic vein in dogs and cats and into the marginal ear
vein in rabbits without causing any detectable pain or discomfort [31]. The
dose of EMLA is between 0.5 and 1 g, and the cream should be applied 30
minutes to 1 hour before the procedure [30]. EMLA cream, once applied,
should be covered with an occlusive adhesive dressing (eg, Tegaderm) and
kept in place for at least 1 hour. If an adequate amount was applied, the
cream is still visible when the dressing is removed; if the cream is no longer
visible at removal, the amount applied was insuﬃcient. The eﬃcacy of
EMLA varies by site. In highly vascular sites, such damaged skin, onset is

1138

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

prompt (15 minutes). For all other sites, an anesthetic depth of 3 mm is
achieved after 60 minutes. The depth increases 1 mm per 30 minutes up
to 5 mm at 120 minutes [30]. Although there have not been any veterinary
studies to date to determine if the same occurs in animals, it is the author’s
clinical experience that indeed the same is true for veterinary patients. Local
side eﬀects are mild, and the only systemic side eﬀect of importance is the
risk of methemoglobinemia in cats. Because no data presently exist in this
species, it should be used with caution.

Intercostal nerve block

This block is used to diminish sensation from a thoracotomy, placement of

a chest tube, or rib fractures. Intercostal nerve blocks have been used for many
years to treat rib fracture–related pain [32]. The technique has repeatedly been
shown to provide increased maximal inspiratory ﬂow rates, superior analge-
sia, and better ability to cough and take deep breaths compared with other
methods of pain control [33,34]. The chief limitation of intercostal nerve
blocks is that the relief of pain is temporary, lasting 6 to 8 hours or less.

The intercostal nerve runs along the posterior border of each rib along

with the intercostal artery and vein. The injections are made posterior to the
ribs near the intervertebral foramen three ribs anterior and two ribs poste-
rior to the injury. Several ribs need to be blocked because of the overlapping
of nerves. Injection of bupivacaine not to exceed a total dose of 1.5 mg/kg is
divided over ﬁve sites. The action of bupivacaine can last up to 6 hours.
Complications include pneumothorax, intravascular injection, and pulmo-
nary damage. Although the technique is not complicated, it may be diﬃcult
to perform if positioning poses a challenge in trauma patients.

Interpleural bupivacaine

This local anesthetic technique is used to provide pain relief in patients

with thoracic trauma or after thoracotomies. After placement of a chest tube
and attainment of negative intrathoracic pressure, 1.5 mg/kg of bupivacaine
is instilled through the chest tube. The tube can be ﬂushed with 2 to 5 mL of
sterile saline after instillation of the bupivacaine. The bupivacaine is distrib-
uted by gravity, so the patient should be in lateral recumbency with the
injured side down for 5 to 10 minutes after instillation. The analgesia pro-
vided can last up to 8 hours. In the absence of a chest tube, an over-the-
needle catheter is introduced into the pleural space for administration of the
bupivacaine. If the instillation of the bupivacaine is painful, several milli-
liters of lidocaine may be injected prior. Interpleural administration of bupi-
vacaine can provide analgesia equal to that produced by systemic
administration of morphine or an intercostal nerve block, and the incidence
of respiratory depression is decreased. After intercostal thoracotomy, inter-
pleural bupivacaine provided prolonged analgesia with fewer blood gas
alterations than morphine [35–37].

1139

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

The major concerns with interpleural catheter placement are that the

peak plasma levels of local anesthetics are relatively high. In addition, in
patients with chest tubes, the risk of suctioning the injected local anesthetics
must be guarded against. This is achieved by delaying the suction of the
thoracostomy tube for 15 to 30 minutes after injection of the local anesthetic
through the interpleural catheter.

Epidural

Epidural analgesia is an eﬃcacious method of pain control that has min-

imal systemic eﬀects and is a useful technique for relieving pain in critical
care patients. Before administration, patients must be thoroughly assessed
to identify any preexisting conditions that preclude the safe use of this
technique. Coagulation defects, cardiovascular instability, sepsis, untreated
bacteremia, or infection at the proposed site of skin puncture is a contra-
indication to epidural placement. Analgesia of the thorax, abdomen, and
lower extremities can be accomplished with less respiratory depression than
with parenteral opioids.

The spinal cord in the dog ends at about L6, and there is little chance of

damaging the spinal cord during a lumbosacral puncture. It is also unlikely
for cerebrospinal ﬂuid (CSF) to be encountered in this space; however, aspi-
ration through the needle should be attempted before injection to ensure
that the subarachnoid (intrathecal space) has not been breached. In the cat,
the spinal cord ends more caudally (S1), so care is needed during needle
placement, because it is possible to damage the cord.

For lumbosacral epidural placement, ideally, the patient is placed in ster-

nal recumbency. The wings of the ilium are the starting point for ﬁnding the
lumbosacral space when placing the needle for epidural injection. After
povidone iodine (Betadine) preparation and sterile draping, the bony land-
marks are palpated and the last lumbar vertebra is identiﬁed at the level of
the iliac crest. An intradermal wheal is raised with the local anesthetic (1 mL
of lidocaine) over the lumbosacral space, and deep inﬁltration of local anes-
thetic is made using a 25-gauge needle to facilitate placement in nonanesthe-
tized patients. A spinal or epidural needle should then be used for epidural
anesthetic administration. The needle is inserted into the middle of the inter-
space at right angles to the skin between the dorsal spinal processes of L7
and S1. It is advanced until it penetrates the interspinous ligament and then
is advanced to penetrate the ligamentum ﬂavum. A syringe containing air or
saline is attached to the needle, and light continuous pressure is maintained
on the syringe plunger to identify the epidural space by loss of resistance.
Once the position of the spinal needle is veriﬁed, the syringe containing the
agent is connected and the agent is injected slowly.

The opioids are often used in preference to local anesthetic agents,

because the side eﬀects of opioids are less severe, duration of analgesia is
much longer, motor function is preserved, and sympathetic blockade does

1140

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

not occur. Morphine is administered in preference to other opioid agents
because it has low lipid solubility and is slowly removed from the spinal
space, thus providing a long duration of analgesia. The epidural morphine
dosage is 0.1 mL/kg of preservative-free morphine (Duramorph) and can
provide up to 24 hours of analgesia [38].

An epidural catheter can be placed for chronic epidural drug administra-

tion. Standard orders address the need to monitor the patient, treat side
eﬀects, and adjust the medication to provide adequate analgesia. The com-
plication rate associated with temporary epidural catheterization of dogs
seems to be low, and complications generally are not serious. Catheters can
be maintained for up to 7 days safely. In a study of 80 dogs that received
epidural catheters, dislodgement was the most common complication and
was seen in 16% of the animals. A thorough discussion of epidural catheter
placement is beyond the scope of this article, and readers are encouraged to
seek additional sources [38,39].

Monitoring

Monitoring of analgesia in the critically ill patient is diﬃcult. Reliable

objective measures of pain are unavailable, and underlying disease or med-
ications may alter blood pressure and heart rate, which are commonly used
indicators of pain. Monitoring the patient’s condition involves visual obser-
vation and assessment of the level of consciousness and physiologic changes.
The components of monitoring may include level of consciousness, respira-
tory rate, blood pressure, oxygen saturation, percentage of exhaled carbon
dioxide, heart rate, blood pressure, and electrocardiographic (ECG) rhythm.
Patients who are not arousable are potentially at risk for airway compro-
mise, need higher levels of monitoring, and may be candidates for drug
reversal agents.

Pulse oximetry provides continuous noninvasive estimates of arterial oxy-

gen saturation and is a reliable tool in detecting early decreases in oxygen sat-
uration and changes in the patient’s heart rate. Under most circumstances,
there is excellent correlation between the pulse oximeter saturation and
arterial hemoglobin oxygen saturation. The limitations of oximetry include
its inability to detect early decreases in the adequacy of ventilation, and thus
the onset of hypercarbia that may occur before the development of apnea. It
has been clearly demonstrated that many of the drugs used in procedural
sedation and analgesia predispose the patient to the development of hypo-
xemia and that drug combinations, especially benzodiazepines and opioids,
have a potentiating eﬀect in suppressing respirations. Ventilatory function
should be continually monitored by observation and/or auscultation. When
possible, blood pressure should be determined before sedation/analgesia is
initiated. Once sedation/analgesia is established, blood pressure should be
measured at regular intervals. ECG monitoring should be used in patients

1141

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

with signiﬁcant cardiovascular disease as well as during opioid constant rate
infusion when dysrhythmias (bradycardia) are anticipated.

Summary

All medical interventions, including the provision of analgesia, are asso-

ciated with risks and beneﬁts, which, when considered together, comprise
that intervention’s risk/beneﬁt ratio. All interventions have alternatives
(including no intervention), and each alternative possesses its own risk/ben-
eﬁt ratio. Clinical decision making involves comparing and contrasting the
risk/beneﬁt ratios of alternative interventions (relative risk/beneﬁt ratio).
The most formidable limitations of drug treatment relate to their potential
to produce pharmacologic side eﬀects or complications. Careful monitoring
and the use of strategies for preventing and managing drug side eﬀects are
often all that is required to maintain eﬃcacy.

References

[1] Merskey M. Pain terms: a list with deﬁnitions and notes on usage. Recommended by the

IASP Subcommittee on Taxonomy. Pain 1979;6:249–52.

[2] Melzack R, Wall PD. Pain mechanisms: a new theory. Science 1965;150:971–9.
[3] Abram SE. Advances in chronic pain management since gate control. Reg Anesth 1993;

18:66–81.

[4] Desai PM. Pain management and pulmonary dysfunction. Crit Care Clin 1999;15:151–66.
[5] Bereiter D, Plotsky P, Gann D. Tooth pulp stimulation potentiates the ACTH response to

hemorrhage in cats. Endocrinology 1982;111:1127–32.

[6] Epstein J, Breslow MJ. The stress response of critical illness. Crit Care Clin 1999;15:17–33.
[7] Novaes MA, Knobel E, Bork AM, et al. Stressors in ICU: perception of the patient,

relatives and healthcare team. Intensive Care Med 1999;25:1421–6.

[8] Lewis KS, Whipple JK, Michael KA, et al. Eﬀect of analgesic treatment on the physio-

logical consequences of acute pain. Am J Hosp Pharm 1994;51:1539–54.

[9] Kehlet H. The surgical stress response: should it be prevented? Can J Surg 1991;34:565–7.

[10] Conzemius MG, Hill CM, Sammarco JL, et al. Correlation between subjective and

objective measures used to determine severity of postoperative pain in dogs. JAVMA 1997;
210:1619–22.

[11] Lucas AN, Firth AM, Anderson GA, et al. Comparison of the eﬀects of morphine

administered by constant-rate intravenous infusion or intermittent intramuscular injection
in dogs. JAVMA 2001;218:884–91.

[12] Scherk-Nixon M. A study of the use of a transdermal fentanyl patch in cats. J Am Anim

Hosp Assoc 1996;32:19–24.

[13] Robinson TM, Kruse-Elliott KT, Markel MD, et al. A comparison of transdermal fentanyl

versus epidural morphine for analgesia in dogs undergoing major orthopedic surgery. J Am
Anim Hosp Assoc 1999;35:95–100.

[14] Kyles AE, Papich MG, Hardie EM. Disposition of transdermally administered fentanyl in

dogs. Am J Vet Res 1996;57:715–9.

[15] Lee DD, Papich MG, Hardie EM. Comparison of pharmacokinetics of fentanyl after

intravenous and transdermal administration in cats. Am J Vet Res 2000;6:672–7.

[16] Riviere JE, Papich MG. Potential and problems of developing transdermal patches for

veterinary applications. Adv Drug Deliv Rev 2001;50:175–203.

1142

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

[17] Pettifer G, Dyson D. Hydromorphone: a cost-eﬀective alternative to the use of oxymor-

phone. A comparison of transdermal fentanyl versus epidural morphine for analgesia
in dogs undergoing major orthopedic surgery. Can Vet J 2000;41:135–7.

[18] Joubert K. Ketamine hydrochloride: an adjunct for analgesia in dogs with burn wounds.

J S Afr Vet Assoc 1998;69:95–7.

[19] Fass J, Bares R, Hermsdorf V, et al. Eﬀects of intravenous ketamine on gastrointestinal

motility in the dog. Intensive Care Med 1995;21:584–9.

[20] Pearce CJ, Gonzalez FM, Wallin JD. Renal failure and hyperkalemia associated with

ketorolac tromethamine. Arch Intern Med 1993;153:1000–2.

[21] Schlondorﬀ D. Renal complications of nonsteroidal anti-inﬂammatory drugs. Kidney Int

1993;44:643–53.

[22] Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of

rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000;343:
1520–8.

[23] Grisneaux E, Pibarot P, Dupuis J, et al. Comparison of ketoprofen and carprofen admin-

istered prior to orthopedic surgery for control of postoperative pain in dogs. JAVMA 1999;
215:1105–10.

[24] Lobetti RG, Joubert KE. Eﬀect of administration of nonsteroidal anti-inﬂammatory drugs

before surgery on renal function in clinically normal dogs. Am J Vet Res 2000;61:1501–7.

[25] Fraser GL, Prato S, Berthiaume D, et al. Evaluation of agitation in ICU patients:

incidence, severity, and treatment in the young versus the elderly. Pharmacotherapy
2000;20:75–82.

[26] Cohen D, Horiuchi K, Kemper M, et al. Modulating eﬀects of propofol on metabolic and

cardiopulmonary responses to stressful intensive care unit procedures. Crit Care Med
1996;24:612–7.

[27] Brown SA, Jacobson JD, Hartsﬁeld SM. Pharmacokinetics of midazolam administered

concurrently with ketamine after intravenous bolus or infusion in dogs. J Vet Pharmacol
Ther 1993;16:419–25.

[28] Norris Jr RL. Local anesthetics. Emerg Med Clin North Am 1992;10:707–18.
[29] Catterall W, Mackie K. Local anesthetics. In: Gilman AG, Goodman LS, Gilman A,

editors. The pharmacological basis of therapeutics. 6th edition. New York: MacMillan;
1980. p. 331–47.

[30] Gajraj NM, Pennant JH, Watcha MF. Eutectic mixture of local anesthetics (EMLA)

cream. Anesth Analg 1994;78:574–83.

[31] Flecknell PA, Liles JH, Williamson HA. The use of lignocaine-prilocaine local anaesthetic

cream for pain-free venipuncture in laboratory animals. Lab Anim 1990;24:142–6.

[32] Bryant LR, Trinkle JK, Wood RE. A technique for intercostal nerve block after

thoracotomy. Ann Thorac Surg 1971;11:388–91.

[33] Bergh NP, Dottori O, Lof BA, et al. Eﬀect of intercostal nerve block on lung function after

thoracotomy. Acta Anaesthiol Scand Suppl 1966;18(Suppl 24):85–95.

[34] Liu M, Rock P, Grass JA, et al. Double-blind randomized evaluation of intercostal nerve

blocks as an adjuvant to subarachnoid administered morphine for post-thoracotomy
analgesia. Reg Anesth 1995;20:418–25.

[35] Flecknell PA, Kirk AJ, Liles JH, et al. Postoperative analgesia following thoracotomy in

the dog: an evaluation of the eﬀects of bupivacaine intercostal nerve block and nalbuphine
on respiratory function. Lab Anim 1991;25:319–24.

[36] Kushner LI, Trim CM, Madhusudhan S, et al. Evaluation of the hemodynamic eﬀects of

interpleural bupivacaine in dogs. Vet Surg 1995;24:180–7.

[37] Thompson SE, Johnson JM. Analgesia in dogs after intercostal thoracotomy. A compar-

ison of morphine, selective intercostal nerve block, and interpleural regional analgesia
with bupivacaine. Vet Surg 1991;20:73–7.

[38] Wetmore LA, Glowaski MM. Epidural analgesia in veterinary critical care. Clin Tech

Small Anim Pract 2000;15:177–88.

1143

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

[39] Swalander DB, Crowe DT, Jr, Hittenmiller DH, et al. Complications associated with the

use of indwelling epidural catheters in dogs: 81 cases (1996–1999). JAVMA 2000;216:
368–70.

[40] Coleman DL. Control of postoperative pain: nonnarcotic and narcotic alternatives and

their eﬀect on pulmonary function. Chest 1987;92:520–8.

[41] Kremer E, Atkinson JH, Ignelzi RJ. Measurement of pain: patient preference does not

confound pain measurement. Pain 1981;10:241–8.

[42] Miller MD, Ferris DG. Measurement of subjective phenomena in primary care research:

the Visual Analogue Scale. Fam Pract Res J 1993;13:15–24.

[43] Revill SI, Robinson JO, Rosen M, et al. The reliability of a linear analog for evaluating

pain. Anaesthesia 1976;31:1191–8.

[44] Wilson J, Pendleton J. Oligoanalgesia in the emergency department. Am J Emerg Med

1989;7:620–3.

1144

M.M. Glowaski / Vet Clin Small Anim 32 (2002) 1127–1144

Nutritional support in critical

care patients

Rebecca L. Remillard, PhD, DVM

Angell Memorial Animal Hospital, 350 South Huntington Avenue,

Boston, MA 02130, USA

Clinical importance

At any age or stage of life, patients with inadequate nutrient intake

become malnourished. By most estimates, many hospitalized people and
companion animals are not receiving adequate nutrition. In a study of
276 hospitalized dogs, a positive energy balance was achieved in only 27%
of some 821 dog-days recorded; whereas a negative energy balance was
observed on the majority (73%) of the dog-days [1]. The major consequences
of malnutrition in all patients, but more prominently in sickand (or) injured
patients, are decreased immunocompetence [2,3], decreased tissue synthesis
and repair [4], and altered intermediary drug metabolism [5]. Malnutrition
in people is associated with prolonged ventilatory dependence and increased
complication rates, with increased hospital stays and costs [6,7]. Similarly,
malnutrition is thought to increase morbidity and mortality in veterinary
patients.

Assess the animal

Malnutrition can be recognized in a patient by using a nutritional assess-

ment protocol. Nutritional assessment helps to identify which patients
require nutritional support to avoid or reduce nutrient deﬁciencies and the
associated complications. A veterinary nutritional assessment protocol
should include a history, physical examination with special attention to cer-
tain riskfactors, body condition score, and laboratory tests [8]. To date, few
clinical studies have been performed in veterinary patient populations to

Vet Clin Small Anim 32 (2002) 1145–1164

E-mail address: rremillard@mscpa.org (R.L. Remillard).

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 0 5 0 - 5

determine which parameters are applicable and how accurate they are in
determining nutritional status and predicting outcome [1,9].

It is important to remember that diseased patients have insulin resistance

and decreased glucose utilization, with increased proteolysis and lipolysis.
Neuroendocrine responses and local mediators common to stressful stimuli
mediate this metabolic change from glucose to protein/fat metabolism. The
speciﬁc hormonal and subsequent physiologic changes that occur in a
stressed patient are unique to that disease condition, the time course of the
disease, and the presence of other complicating diseases or conditions. Hor-
monal and tissue substrate changes have best been characterized in disease
conditions with a known acute onset, such as trauma or infection [10]. The
response has been described as having an acute (ebb) phase and then an
adaptive (ﬂow) phase [11]. The acute and adaptive phases can vary in dura-
tion and intensity depending on the speciﬁcs of the incidence. Nutritional
support to these patients is done well when the nutrient proﬁle (glucose-
to-fat-to-protein ratios) best ﬁts the catabolic state of the patient. Far fewer
refeeding complications arise when critically ill patients are initially fed a
high-fat low-glucose form of nutritional support.

Recommendations for avoiding refeeding complications are as follows:

1. Anticipate the potential for the problem and refeed with formulations

known to have adequate potassium, phosphate, and magnesium.

2. Initial nutritional refeeding rates should not exceed RERs and 2 to 6 g

of protein per 100 kcal. These can be increased as needed over sub-
sequent days. Consider refeeding a high-fat low-carbohydrate formula
for patients that have not eaten for longer than 5 days.

3. Monitor serum potassium, phosphate, and magnesium levels as needed;

however, once a day is suﬃcient for most cases.

4. Supply water-soluble vitamins ad libitum, particularly thiamine, to

facilitate energy metabolism.

5. Monitor patients daily for signs of ﬂuid overload and (or) congestive

failure.

Keynutritional factors

The goal of nutritional support is to slow catabolism and provide precur-

sors for optimal immune function, tissue synthesis and repair, and drug
metabolism as summarized in Table 1. The patient’s ﬂuid, electrolyte, and
acid-base abnormalities as well as blood glucose levels must be within nor-
mal or approaching normal limits before refeeding, because administering
enteral or parenteral nutrition (PN) may compromise the patient further.
It is also advisable to have severe tachycardia, hypotension, and colloid and
volume deﬁcits corrected before starting to feed [12]. A practical goal is to
begin nutritional support, or at least do a nutritional assessment, within
24 hours of the injury, illness, or presentation [13,14].

1146

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

Table 1
Summary of the key nutritional factors

Nutrient

Associated conditions

Recommendations

Fluid

Dehydration

Increase ﬂuid rate

Cardiac/pulmonary

congestion

Decrease ﬂuid rate

Energy

Decreased body weight
Depressed attitude and

lethargy

Provide at least RER using

dextrose and fat

Omega 3 fatty acids

Suppress inﬂammatory

responses

N/A with infections

Omega 6 fatty acids

Accentuate inﬂammatory

responses

N/A in immune-mediated

diseases

Protein

Decreased body weight
Muscle wasting

Provide 3–6 g per 100 kcal

of RER

Arginine

Decreased serum proteins

2% of oral intake

Glutamine

Depressed immune response

2% of oral intake

B-complex vitamins

Poor energy metabolism

Administer B-complex

vitamins (1 mL per 100
kcal of RER per day IV)

Thiamin

Thiamin deﬁciency signs

Administer thiamin (1–2

mg/kg/d IV or PO)

Electrolytes (calcium,

sodium, magnesium)

Cardiac arrhythmias
Muscle weakness

Replete electrolytes

Potassium

Hypo/hyperkalemia

30 mEq/L IV in PN

Phosphorus

Hypo/hyperphosphatemia

10 mM/L IV in PN

Trace elements

Poor energy metabolism

Administer nutritional

support with trace
elements (1 mL per 100
kcal of RER per day IV)

Fat-soluble vitamins

Long history of anorexia or

fat malabsorption

Administer nutritional

support with fat-soluble
vitamins) (1 mL of
vitamins A, D and E IM
once)

(1 mg of vitamin

per kilogram SC twice

daily)

Abbreviations: RER, resting energy requirements; N/A, not appropriate; IV, intravenous;

PO, by mouth; PN, parenteral nutrition; IM, intramuscular; SC, subcutaneous.

B-Vitamin Complex; Butler Company; Columbia, OH.

MTE-4; Abbott Laboratories, Chicago, IL.

Vital E-A+D; Schering Plough Animal Health Corporation; Kenilworth, NJ.

1147

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

Energy

Diseased patients have metabolic rates and energy requirements that

are less than those of a comparable normal healthy individual. In malnutri-
tion, without disease or injury, decreased triiodothyronine concentrations
decrease the metabolic rate in an eﬀort to conserve body functions. With
an ongoing disease process or traumatic injury, however, the neuroendo-
crine responses to stress increase the metabolic rate above that found in sim-
ple starvation. Respiration calorimetry measurements of more than 3000
human cases with a wide variety of diseases, speciﬁcally excluding hyper-
thyroidism, illustrated that 90% of the patients were within

^15% of resting

energy requirements (RER) [15]. In people with trauma, the energy ex-
penditure peaks in 3 to 4 days and then subsides by day 7 to day10 unless
complicated by sepsis [16]. A few preliminary respiration calorimetry
measurements in ill dogs with speciﬁc disease conditions, excluding trauma,
suggested that most have requirements near RER as well [17–19]. The met-
abolic rate of an anorectic sickor traumatized patient is therefore greater
than that of the animal in simple uncomplicated starvation but less than that
of the animal in a normal healthy state.

In all probability, veterinary patients are similar to ill people and have

metabolic rates near their RER. Estimating the RER of a hospitalized
patient can be relatively simple using the equations RER kcal/d

¼ 15 · BW

BW lb for dogs, 20

· BW lb for cats, and 25 · BW lb for either under 5 lb,

where BW is body weight. Most hospitalized veterinary patients should be
fed at their calculated RER, realizing that their actual energy requirement
is likely to change over the time course of the disease process and recovery.
In human surgical patients, there was relatively little additional beneﬁt to
increasing intake once half of the caloric requirement of the patient had
been achieved [20]. Therefore, initially feeding patients at their RER, or
slightly greater than 50% of their RER if 100% is not possible, is a rational
and safe recommendation that signiﬁcantly decreases the probability of met-
abolic complications associated with overfeeding.

Knowing the patient’s approximate caloric requirement is important,

because feeding more than needed of any food can cause metabolic compli-
cations. Providing maintenance or even supraphysiologic quantities of
nutrients to these patients does not reverse severe catabolism and may not
achieve nitrogen balance [21]. Overfeeding patients is possible through a
feeding tube or with PN support. In people and several diﬀerent animal
models, excessive carbohydrate intake has been associated with hyperglyce-
mia, hypercarbia, fatty liver, increased ventilatory drive, and failure to wean
from a ventilator [22–24]. Excessive fat administration has been associated
with hyperlipidemia, hypoxia, increased rate of infection, and a higher post-
operative mortality rate [25–27]. The proportion of fat, carbohydrates, and
protein in diets fed to hospitalized patients should be similar to those that
the liver is estimated to be using from body stores. In general, foods and
solutions with a 1-kcal/mL caloric density meet the patient’s RER in a

1148

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

tolerable oral or one half of maintenance (15 mL/lb) intravenous (IV) vol-
ume. The goal in providing nutrition to these catabolic patients is to fuel the
catabolism with exogenous sources of protein and fat, sparing endogenous
sources. Feeding spares endogenous protein, because the loss of 25% to
30% of body protein stores has been associated with reduced heart muscle
mass and function, decreased pulmonary function and diminished respira-
tory drive, compromised immune function, and therefore increased mortal-
ity. Regular subjective and objective assessments of the patient are strongly
recommended so as to make adjustments to these initial feeding rates.

Protein

Providing a protein source to catabolic animals spares endogenous skel-

etal muscle proteins and supplies essential amino acids and amino groups
for acute-phase proteins and immune response. Suﬃcient calories must be
available from fat and (or) glucose before amino acids are used for tissue
synthesis and repair [28–30]. Prolonged insuﬃcient protein intake has been
linked to low albumin, poor immune response, poor healing, and increased
risks for dehiscence and muscle wasting. Conversely, excessive protein feed-
ing requires energy expenditure to rid the body of excess nitrogen, which
may or may not be handled well by the liver and kidneys in some cases.

Most eﬃcient use of protein in people occurs when 2 to 6 g of protein per

100 kcal is administered [31]. A starting average of 5 to 6 g of protein per
100 kcal enterally [32] and 3 to 4 g of protein per 100 kcal parenterally
[33,34] can be used for most canine patients when their ability to handle pro-
tein waste products is not in question and there are no known extraordinary
protein losses. Higher ratios (6–8 g/100 kcal enterally [32] and 4–5 g/100 kcal
parenterally) are more reasonable estimates for cats given their constant
state of gluconeogenesis and higher requirement for protein. Protein intake
may then be adjusted based on the patient’s changing need and ability to
handle these initial protein recommendations.

Arginine supplementation (350–850 mg/kg of BW) has been shown to

improve nitrogen balance and immune function during stress and enhances
formation of new collagen in traumatized animals [35–39]. All veterinary
enteral diets contain arginine (0.5%–1.25% dry matter [DM]) because
arginine is an essential amino acid to dogs and cats. Given the recently
reported positive eﬀects of arginine on the immune system and wound heal-
ing, diets with increased arginine (>2% DM) concentrations are now avail-
able. The optimal arginine intake is not known, and selection of enteral diets
based solely on arginine content is not recommended.

Glutamine has been considered a nonessential amino acid in dogs, cats,

and human beings. The intestinal uptake of glutamine is known to increase
with surgery and (or) trauma, most likely because glutamine is the preferred
fuel for enterocytes [40]. At least 80% of the published studies using gluta-
mine-enriched feedings in animals demonstrate a positive eﬀect. Feeding a
glutamine-enriched enteral diet has positive eﬀects on protein metabolism,

1149

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

intestinal and pancreatic repair and regeneration, nutrient absorption, gut
barrier function, systemic and intestinal immune function, and animal sur-
vival [41]. Glutamine is therefore considered a conditionally essential amino
acid but probably only needs to be administered during early periods of
physiologic stress to stimulate DNA synthesis and increase mucosal mass
early in recovery. Glutamine added to nutritional IV solutions has been
shown to reduce some aspects of intestinal atrophy and to enhance intestinal
immune function in short-term rat models. The use of IV glutamine is cost-
prohibitive and should probably be limited to short-term use (1 week) just
before oral refeeding if at all.

B vitamins

B-complex vitamins (folic acid, thiamin, riboﬂavin, niacin, pantothenic

acid, pyridoxine, and B

) are essential for hepatic metabolism of glucose,

fat, and protein. They are coenzymes for the tricarboxylic acid cycle, energy
production, and red blood cell metabolism. They are needed in small
amounts relative to other nutrients but are required for energy metabolism
to operate eﬃciently. All patients not eating but receiving ﬂuid therapy
should have B vitamins added to their ﬂuids. These vitamins are easily and
inexpensively replaced and should be included in all forms of assisted feed-
ing. Most pet foods contain adequate amounts of these nutrients and there-
fore should not be of concern if the patient is eating suﬃcient amounts of a
complete and balanced pet food to meet RER.

Trace minerals

Zinc, copper, manganese, chromium, and selenium are vital cofactors to

optimal hepatic and peripheral metabolism of energy substrates. These min-
erals should be included in all forms of nutritional support. Most pet foods
contain adequate amounts of these nutrients and should not be of concern if
the patient is eating suﬃcient amounts of a complete and balanced pet food
to meet RER. Parenteral trace element products should be added to PN
mixtures.

Fat-soluble vitamins

Hospitalized patients rarely need fat-soluble vitamins. Most acute trauma

patients have fat and hepatic stores of the fat-soluble vitamins suﬃcient to
meet metabolic needs for weeks to months. Administering fat-soluble vita-
mins should be considered in patients with prolonged malnutrition that are
severely underweight with little to no fat stores, however.

Keynutritional factors for the respiratorypatient

Inadequate caloric intake promotes protein catabolism in the critically ill

patient, and the skeletal muscle pool is consumed for gluconeogenesis [42].
Given that respiratory muscles (diaphragm and intercostals) are skeletal
muscles; they are catabolized for energy. Malnourished patients without

1150

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

lung disease have a 34% decrease in respiratory muscle strength, a 41%
decrease in maximum voluntary ventilation, and 63% less vital capacity
compared with nourished individuals [43]. There are no such data available
for canine or feline patients; however, it is only logical to conclude that our
patients also suﬀer diminished respiratory function as a result of inadequate
calorie and protein intake. Energy measurements in malnourished spontane-
ously breathing people with chronic obstructive pulmonary disease indicate
their need to be approximately 15% greater than RER, which can be
explained by increased ventilatory muscle work[44]. Given the inaccuracies
of calculating our patients’ RER and the iatrogenic induced metabolic
derangements of overfeeding, providing a balanced formulation at the esti-
mated RER is the current recommendation.

The substrate (protein, fat, carbohydrate) mix is an issue for human

patients in the intensive care unit (ICU) with respiratory disease, and it has
gone generally unrecognized in veterinary patients. Nutritional hypercapnia,
an increase in CO

production, may be an issue for critically ill veterinary

patients and is associated with the administration of glucose for two rea-
sons. First, when glucose administration matches actual caloric need, the
combustion of glucose raises CO

consumption per unit time by 22% com-

pared with isocaloric administration of lipid. Second, when glucose intake
exceeds actual caloric need, lipogenesis results, and there is marked increase
in the respiratory quotient (RQ

¼ CO

). The RQ of glucose metabolism

is 1.0, and the RQ of fat metabolism is 0.7; however, the RQ of lipid
synthesis is 8.0 because of the far greater production of CO

relative to

consumption during lipogenesis. The administration of glucose in excess

of caloric need has several other potential disadvantages. Chronic hyper-
glycemia is known to be immunosuppressive and, in diabetes, increases the
insulin requirement, which may result in sodium and ﬂuid retention. Pres-
ently, providing 80% to 100% of the animal’s RER in the form of a 20%
lipid has been clinically successful in meeting the respiratory patient’s caloric
need without negative metabolic consequences (R.L. Remillard, unpub-
lished data, 2001).

There are some interesting data now appearing in the literature where

the type of fat (n3 vs. n6) administered may be beneﬁcial in managing respi-
ratory diseases with an inﬂammatory component. Enteral feeding of an
eicosapentaenoic acid (EPA)–enriched or gamma-linolenic acid (GLA)–
enriched (n3) diet within 4 days altered the alveolar macrophage phos-
pholipid fatty acid proﬁle and promoted a shift toward less inﬂammatory
eicosanoids but did not impair alveolar macrophage bactericidal function
relative to that of rats fed a linoleic acid (n6) diet [45]. The beneﬁcial eﬀects
of the EPA + GLA diet on pulmonary neutrophil recruitment, gas exchange,
lower requirement for mechanical ventilation, shorter length of ICU stay,
and fewer new organ failures suggested that an n3-enriched enteral formula
would be a useful adjuvant therapy in the clinical management of acute
respiratory distress syndrome (ARDS) patients [46].

1151

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

Feeding plan

Recording the intake of hospitalized patients is essential to determining

whether or not assisted feeding is necessary. In addition to having the com-
plete feeding prescription, the medical record should contain the time of day
and the amount actually consumed by the patient. Consumption can be sim-
ply recorded as some percentage of the food oﬀered (eg, 0%, 50%, or 100%).
If feeding orders are properly written and consumption is recorded, after 24
hours of hospitalization, it should be apparent whether or not the patient is
consuming suﬃcient food to meet RER and thus how and what type of
nutritional support are required.

Devising a feeding plan for hospitalized patients requires knowledge of the

case, and the plan often needs to be individually tailored because of the
unique circumstances surrounding each case. Nutritional plans require an
understanding of the patient’s metabolic state relative to the changes in
metabolism as food deprivation continues. Refeeding patients in the early
phase versus the later phases of food deprivation dictates the proportion of
fat and carbohydrate in the refeeding formula. Preexisting condition(s)
requiring speciﬁc nutritional (eg, renal insuﬃciency) or dietary modiﬁcations
(eg, food hypersensitivities) must be understood and incorporated into
the new feeding plan for the patient. Forehand knowledge that a patient
requires other medical and surgical procedures should be taken into account
when formulating an assisted feeding plan. Nutritional plans should also
take into consideration the whole treatment plan and owner’s expectations,
because some nutritional plans can only be implemented while the patient
is in the hospital versus those that can be implemented at home by the owner.

There are only two methods by which nutrients can be supplied to the

body: enteral and parenteral. Enteral feeding can provide adequate nutrition
simply and cost-eﬀectively whether done orally or by feeding tube. Enteral
feeding is the method of choice compared with parenteral feeding in most
clinical cases, because using the gastrointestinal tract is less expensive, stim-
ulates the immune system, and most likely avoids metabolic complications
of refeeding. When the small intestine (SI) is not functioning or food cannot
be delivered to the SI adequately to meet the patient’s nutrient requirements
by the enteral route, nutrients must be administered parenterally. The two
methods are not mutually exclusive; in many cases, supplementing calories
and protein parenterally over what the patient consumes voluntarily is pos-
sible in most veterinary practices. Therefore, overall medical assessment of
the patient with particular attention to gastrointestinal function is essential
to the decision of how nutritional support should be provided.

The goal of nutritional support is to provide adequate nutrition at the

patient’s RER, and the logistics of that support have to be determined on
a case-by-case basis. General guidelines are discussed here to help establish
a foundation; however, greater detail on enteral nutrition and PN is avail-
able elsewhere [47].

1152

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

Enteral feeding

Enteral assisted feeding is providing nutrients to the patient using some

portion of the gastrointestinal tract. Patients that cannot or will not eat
but can digest and absorb nutrients from the SI should receive enteral nutri-
tional support. Feeding via the gastrointestinal tract can be the simplest,
fastest, easiest, safest, least expensive, and most physiologic method of feed-
ing patients (Table 2).

Nasoesophageal tubes

Nasoesophageal tubes are generally used for periods of time, such as 3 to

7 days, but are occasionally used longer. Polyurethane tubes with or without
a weighted tip (Kangaroo Enteral Feeding Tube, Sherwood Medical, St.
Louis, MO; KeoFeed II Feeding Tube; IVAC Corporation, San Diego,

Table 2
Daily in-hospital enteral nutrition/tube feeding suggestions

Food product

Units per day

10-lb cat

30-lb dog

Resting energy requirements

kcal ME

218

497

Normal protein intake

Abbott’s CliniCare

Canine (5.4 g per 100 kcal)

—

500

Feline (8.2 g per 100 kcal)

220

—

Hill’s a/d (8.3 g per 100 kcal)

170

385

High-protein intake

Abbott’s CliniCare

Feline (8.2 g per 100 kcal)

—

500

Hill’s a/d

(8.3 g per 100 kcal)

—

385

Ross Laboratories

Promote

(15 g per 100 kcal)

220

500

Low-protein intake

Ross Laboratories Ensure

(4.0 g per 100 kcal)

220

500

Abbott’s CliniCare Feline RF

(6.1 g per 100 kcal)

220

—

Calorie-dense diets

Hill’s a/d (1.3 kcal/mL)

170

385

Ross Laboratories

PulmoCare (1.5 kcal/mL)

150

330

TwoCal HN (2.0 kcal/mL)

110

250

Monomeric diets

ClinTec’s Peptamen

220

500

Abbott’s Perative

220

500

Abbreviation: ME, metabolizable energy.

Abbott Laboratories, Chicago, IL; telephone: 847-935-4849.

Hill’s Pet Nutrition, Topeka, KS; telephone: 800-548-8387. Two cans (5.5 oz per can) of

a/d

þ 50 mL of water make about 300 mL of 1-kcal/mL slurry.

Ross Laboratories, Columbus, OH; telephone: 800-551-5838.

1153

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

CA) and silicone feeding tubes (Feeding Tube; CookVeterinary Products,
Bloomington, IN) may be placed in the caudal esophagus. An 8-French tube
passes through the nasal cavity of most dogs, but a 5-French tube is more
comfortable in cats. Nasoesophageal feeding tubes may be used in anorectic
patients that do not have nasal, oral, or pharyngeal disease or trauma. Gen-
eral anesthesia or tranquilization is not necessary to place a nasoesophageal
tube; therefore, these tubes provide enteral access to patients considered to
be at anesthetic risk.

Pharyngostomy/esophagostomy/gastrostomy tubes

Pharyngostomy or esophagostomy tubes (8–16 French) may be placed in

patients with disease or trauma to the nasal or oral cavity. The tip of the
tube is placed in the caudal esophagus and can be used long term (weeks
to months) for in-hospital or home feedings. Gastrostomy tubes (G-tubes;
mushroom tipped, 16–22 French; Pezzer Catheter, Mill Ross Laboratories,
Mentor, OH) placed intraoperatively or percutaneously using either an endo-
scope or a G-tube introduction device (Gastrostomy Tube Introduction Set;
CookVeterinary Products) are used in patients that require bypassing the
pharynx and esophagus [48]. G-tubes are also recommended for long-term
use if needed; they have generally replaced pharyngostomy tubes, because
G-tubes are convenient and safe for both in-hospital and at-home feedings.

Jejunostomytubes

Jejunostomy tubes (J-tubes; 5–8 French) are placed within the SI, ideally

at the time of surgery, to bypass the proximal gastrointestinal tract [49].
Another method of J-tube placement is to thread a small feeding tube
through a larger esophagostomy, pharyngostomy, or gastrostomy feeding
tube and place the tip of the smaller tube in the jejunum. A tungsten
weighted-tip feeding tube may be threaded through the pylorus into the jeju-
num using an endoscope or during a surgical procedure, with the cranial end
exiting through a larger tube in the stomach, pharynx, or esophagus [50].
There is risk, however, that even a weighted-tip tube may be returned to the
stomach by reverse peristalsis.

Selecting a food

Food selection depends on tube size and location within the gastrointes-

tinal tract, the availability and cost of products, and the experience of the
clinician. Commercial foods available for enteral use in veterinary patients
can be divided into two major types: liquid or modular products and
blended pet foods. Nasal and J-tubes usually have a small diameter (<8
French), which requires using liquid foods. Orogastric, pharyngostomy,
esophagostomy, and G-tubes have large diameters (>8 French) and are suit-
able for liquid and (>16 French) blended pet foods.

There is only one liquid veterinary polymeric product (CliniCare Canine

and Feline products; Abbott Laboratories, Chicago, IL) that meets the

1154

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

current Association of American Feed Oﬃcials (AAFCO) nutrient recom-
mendations for adult dogs and cats. This product is a homogenized liquid
containing 1 kcal/mL and usually has better acceptability than other liquid
products containing medium chain triglycerides (MCT) oil. This liquid diet
is the best option currently available when small-diameter nasogastric and
jejunostomy feeding tubes have been placed and (or) when continuous-drip
feedings are necessary.

Enteral feeding schedule

The feeding schedule is often determined by the patient’s ability to tolerate

food and the logistics of feeding. Feeding an amount equal to the patient’s
RER in the ﬁrst 24 hours of food reintroduction, if physically tolerated, is
highly recommended. If the patient is volume sensitive, initially feeding a
third of RER and then increasing the amount fed by a third every 24 hours
until reaching an adequate food dosage may be better tolerated. Diets should
be warmed to room temperature but no higher than body temperature before
feeding. Any tube that has been placed into the esophagus or stomach allows
bolus or meal type feeding schedules. Bolus infusion of foods must be slow
(

1 minute) to allow gastric expansion. The daily food dosage should be div-

ided into several meals well below the expected stomach capacity. Capacities
for cats and dogs are approximately 5 to 10 mL/kg of BW during initial food
reintroduction. Maximum capacities have been measured in cats and dogs as
high as 45 to 90 mL/kg of BW when fully realimented.

The exception to this rule is the patient that cannot tolerate bolus feeding

to the stomach without vomiting. Such patients beneﬁt from slow continu-
ous-drip administration (by pump or gravity ﬂow) of food to the stomach.
Salivating, gulping, retching, and even vomiting may occur when too much
food has been infused or when the infusion rate is too fast. Feeding should
be stopped at the ﬁrst sign of retching or salivating, with the meal size
reduced by one half to two thirds for 24 hours and then increased gradually.
Administration of the diet through J-tubes ideally should be via low contin-
uous-drip administration delivered preferably by a pump, although some
patients tolerate frequent (10–20 meals per day) small-bolus (10–20 mL per
meal) feedings.

Parenteral feeding

The term parenteral indicates administration in a manner other than

through the gastrointestinal tract and could involve IV, intramuscular, sub-
cutaneous, intraosseous, or intraperitoneal methods. PN is the administra-
tion of nutrients by some means other than through the digestive tract
and has been further characterized in human medicine as total (TPN) or
partial relative to nutrient requirements and central (CPN) or peripheral rel-
ative to venous access. Administration of PN to veterinary patients is not

1155

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

and has never been by means of TPN. This terminology should be discon-
tinued. At best, PN to our patients provides only a portion of the patient’s
nutritional needs. The term partial parenteral nutrition is not necessary and
may be confused with the term peripheral parenteral nutrition. Hence, use of
the term parenteral nutrition should be encouraged and is used throughout
this article because it correctly identiﬁes a general method of administrating
nutrients to a patient (Table 3).

Table 3
Daily parenteral nutrition formulation

Solutions

10-lb cat

30-lb dog

50% dextrose

25 mL

20% lipid

90 mL

200

8.5% amino acid with electrolytes

100 mL

175

Abbott’s Norm R

135 mL

700

KCl (2 mEq/mL)

1.4 mL

Potassium phosphate (KPO

(3 mmol of phosphorus

þ 4.4 mEq

of potassium per mL)

0 mL

MTE

2 mL

B-complex vitamins

2 mL

Total volume for 24 hours

355 mL

1158

Osmolarity

747 mOsm/L

540

Electrolyte concentrations

Sodium

67.0 mEq/L

96.1

Potassium

29.8 mEq/L

30.2

Magnesium

4.11 mEq/L

3.35

Phosphate

9.82 mM/L

9.85

Chloride

53.5 mEq/L

70.5

Calcium

0 mEq/L

Zinc

6.7 lg/mL

4.3

Copper

2.7 lg/mL

1.7

Manganese

0.7 lg/mL

0.4

Chromium

26.8 ng/mL

17.2

Protein administered

8.7 g

14.9

Calories administered

218 kcal ME

497

Calories as fat, %

Calories as glucose, %

Abbreviations: ME, metabolizable energy; MTE, multitrace elements.

This formula assumes the patient does not have a protein-losing disease, can tolerate a

moderate protein intake, and has normal hydration and serum electrolytes before starting
parenteral nutrition.

American Pharmaceutical Partners, Los Angeles, CA, providing 1 mg of zinc, 0.4 mg of

copper, 0.1 mg of magnesium and 4 lg of chromium per mL.

Butler Company, Columbus, OH.

An osmolarity greater than 550 mOsm/L should be administered into a central vein; less

than 550 mOsm/L may be administered through a peripheral vein.

1156

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

Parenteral products

Compounding a PN solution is not a procedure for most veterinary prac-

tices; however, most veterinary practices can administer PN to a patient.
Individual PN solutions of dextrose and lipid and amino acids can be com-
bined as a ‘‘three-in-one’’ solution and is then referred to as a total nutrient
admixture (TNA). TNA refers to one ﬂuid bag containing a suﬃcient mix-
ture of parenteral solutions to meet a particular patient’s ﬂuid, energy,
amino acid, electrolyte, and B-vitamin needs for a 24-hour period. This is
a convenient method requiring only one bag, one infusion pump, and an
administration set, and most any infusion pump can be used. The formula-
tion is designed speciﬁcally for the patient based on the current BW, daily
ﬂuid and maintenance electrolyte requirements, approximate protein need,
and ability to handle dextrose versus lipid. The TNA solution should be cal-
culated primarily to meet the patient’s RER and protein needs; the total
ﬂuid volume is then adjusted with a standard crystalloid solution to meet the
patient’s daily ﬂuid requirements, and electrolytes are adjusted. Potassium
and phosphorous supplementation is usually needed to prevent electrolyte
shifts between the serum and intracellular ﬂuid. Finally, the water-soluble
vitamins and trace minerals are added. Alternatively, the additional crystal-
loid ﬂuids with added potassium may be administered via a separate IV line
piggybacked into the same catheter.

There are several methods by which to obtain bags of PN solutions. Some

human hospitals and independent pharmaceutical companies compound
TNAs for veterinarians. A prescription must be provided indicating the vol-
ume or ﬁnal concentration of each nutrient (fat, dextrose, amino acids, and
each electrolyte). These solutions are expensive because of the cost of indi-
vidual nutrient additions, and they are likely to be incorrectly referred to as
TPN. Some veterinary schools as well as large referral and private veterinary
hospitals maintain parenteral solution compounders and supplies not only
for their own use, but they also compound and sell TNA bags directly to
practitioners. Several bags of PN solution (up to 7 days’ worth) can be sent
by overnight mail service and delivered directly to the practice. This is often
the safest, most convenient, and most economic method of obtaining an all-
in-one PN solution for the occasional patient in most veterinary practices.

Drug additions

Although it becomes most convenient to administer IV drugs with the PN

solution, extreme caution must be taken before any medications are added
to the TNA. Drug and TNA solution compatibility studies are ongoing, and
there are published lists of drugs known to be compatible and safe [51]. The
drugs of most interest to veterinarians that can be incorporated into a three-
in-one mixture include the following [52]:

• Aminophylline
• Ampicillin

1157

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

• Cefazolin
• Chloramphenicol
• Cimetidine
• Clindamycin
• Digoxin
• Diphenhydramine
• Dopamine
• Erythromycin
• Furosemide
• Gentamicin
• Heparin
• Insulin (regular)
• Lidocaine
• Metoclopramide
• Penicillin G
• Phytonadione
• Ranitidine
• Ticarcillin

Metoclopramide added to PN bags for a continuous rate infusion to

vomiting patients is a frequent indication. It is also worth noting that once
added to the day’s PN solution, a decision to discontinue that medication
can be costly, because a new bag of PN solution must be compounded.
Therefore, use of a second peripheral catheter or a double-lumen central
catheter may be preferable to adding drugs to PN solutions.

Administration

The parenteral administration of nutrients to respiratory patients is quite

useful, particularly when the dyspneic patient refuses to eat voluntarily and
(or) a naso-oxygen tube or face maskis already in place. PN solutions can
be delivered to a patient by a central, peripheral, intraosseous, or intraperi-
toneal catheter. PN solutions with an osmolarity greater than 550 mOsm/L
should be administered in a central vein to avoid thrombophlebitis, whereas
solutions less than 550 mOsm/L may be administered via a peripheral vein.
Osmolarity should be noted for all TNA solutions.

Peripheral vein infusion. There is an increased interest in human medicine in
administering PN via a peripheral vein because of the increased riskand
additional expense associated with central venous catheters and newer cath-
eter materials that decrease the incidence of peripheral vein thrombophle-
bitis [53–55]. PN solutions of 600 to 1250 mOsm/L are administered
peripherally to people for short periods (3 days) [56–59]. Phlebitis is the pri-
mary complication, usually occurring within the ﬁrst 72 hours in 26% to
48% of human patients [60]. There are no published clinical studies or

1158

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

reports evaluating the administration of TNA solutions to dogs or cats
peripherally. Administering calories peripherally to hundreds of dogs and
cats per year has been successfully done, however, using either a TNA
(400–600 mOsm/L) or an isomolar 20% lipid solution at volumes suﬃcient
to meet RER (R.L. Remillard, unpublished data, 2001).

Peripherally inserted central lines (PIC lines) can be used in dogs and cats

for administering TNA solutions. Placing a 10- to 20-cm polyurethane
(L-Cath [16- and 18-gauge], Luther Medical Products, Santa Ana, CA; Cen-
tral Venous [16–20-gauge] catheter, CookVeterinary Products) or silicone
(CookCritical Care, Bloomington, IN) catheter

into the medial saphenous

vein at the level of the tarsus and advancing the catheter up the vein places
the tip of the catheter in the caudal vena cava of the cat. A similar but longer
(20–30 cm) polyurethane or silicone catheter placed in the lateral saphenous
vein by a similar method is more useful in dogs weighing less than 20 kg. For
long-term applications (>3 days), silicone and polyurethane remain the only
acceptable catheter materials, and they need not be removed at a predeter-
mined time.

Central vein infusion. Single- or multiple-lumen polyurethane or silicone
elastomer catheters may be placed by a percutaneous or, rarely, cutdown
procedure in the external jugular vein of most dogs and cats. The tip of the
catheter should be located in the cranial vena cava. Catheters made of sili-
cone elastomer and polyurethane are softer and less irritating; thus, they are
likely to have fewer mechanical and septic complications and to be less
thrombogenic but are more expensive than the polytetraﬂuoroethylene (Tef-
lon) catheters. Again, for long-term applications (>3 days), silicone and pol-
yurethane remain the only acceptable catheter materials. These central
catheters are changed as indicated by the onset of a particular problem of
infection, subcutaneous migration, or thrombosis and not at predetermined
intervals [42]. Multiple-lumen catheters allow multipurpose venous access
for administering incompatible ﬂuid/drug therapies or diﬀerent ﬂuids at dif-
ferent rates. Although use in veterinary patients has not been adequately
evaluated, their use for PN administration has been associated with an
increase in septic complications in people [61]. It has been said numerous
times that the catheter must be ‘‘dedicated’’ to PN administration and
should not be used for blood sampling, medication or blood product admin-
istration, or central venous pressure monitoring. When venous access is lim-
ited, however, the PN catheter may be used for blood sampling and
administering medications if it is adequately ﬂushed before and after the
interruption in PN administration. As with any catheter, it is imperative that
excellent aseptic handling technique is used during these line interruptions.

Silicone (20–16-gauge) catheters (50–60 cm) can be cut to appropriate lengths.

1159

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

Complications

The TNA solution should be administered at room temperature, but it is

prudent not to extend the delivery of any one bag longer than 24 to 30
hours. The most clinically signiﬁcant problem in administering TNA
involves the catheter. The most common problems seen with a PN catheter
are loss of access, thrombophlebitis, and infection. Loss of venous access
may occur as a result of catheter kinking, catheter tip migration, or block-
age. In these instances, the catheter should be removed and a second cath-
eter placed in another vein. Thrombophlebitis is recognized as a response of
the vein intimae to the unique combination of the infusate, catheter mate-
rial, and placement as well as the ratio of the catheter to vessel size. Catheter
material is now considered to be the single most important factor in the
severity of infusion thrombophlebitis [62–65].

Infectious complications with IV infusions have been recognized for more

than 40 years and are now primarily associated with substandard catheter
care. Most catheter-related septicemias are caused by microbial invasion
at the catheter wound either during or after insertion [66,67]. There can also
be hematogenous seeding of the catheter tip thrombi by other infected sites,
such as urinary tract infections, abscess, pneumonia, or bacterial gastro-
intestinal translocation. Infusion of contaminated ﬂuid is a third source of
infection but is highly unlikely when the TNA is compounded in a closed-
circuit ﬂuid system. More often misunderstood is the fact that the current
PN solutions using crystalline amino acids do not favor microbial growth
as previously stated [68–72]. Fungi can still proliferate in admixtures; how-
ever, refrigeration at 4

C suppresses all microbial growth. Conversely, lipid

emulsions alone do support gram-positive and -negative bacterial growth as
well as fungal growth if contaminated. The Centers for Disease Control and
Prevention have recommended that lipid emulsions be administered for no
longer than 12 hours, except in TNA systems, in which they can be admin-
istered over a 24-hour period [73].

Combined enteral and parenteral feeding

There has been renewed interest in human medicine regarding the use of

tube feeding in combination with parenteral administration [74,75]. There is
increasing evidence that with prolonged fasting (>3 days), there is enterocyte
deterioration and decreased gastrointestinal immunity [76]. A possible
source of infection with parenteral administration is the translocation of
enteric bacteria as the result of a compromised intestinal mucosal barrier.
A combination of enteral and parenteral administration has been suggested,
because enteral infusion of small quantities of a liquid diet has been beneﬁ-
cial in preventing intestinal mucosal deterioration during PN in piglets and
human infants and adults [77–80]. In addition, intestinal adaptations after
disease and intestinal hypertrophy after surgery require the presence of
intraluminal nutrients. The intake of food has been shown to promote intes-

1160

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

tinal hyperplasia and brush border enzyme activity [81]. Therefore, most
recent recommendations are encouraging some enteral feeding to those
patients receiving PN support if possible. Feeding both the small bowel and
the patient is important.

Reassessment

Regular reassessment is a critical step in the successful nutritional man-

agement of a hospitalized patient regardless of whether the enteral or paren-
teral route, or a combination, is used. Malnutrition in the form of
insuﬃcient nutrient intake to support tissue metabolism undermines appro-
priate medical or surgical therapeutic management of a case. Patients resting
in a cage have been mistakenly assumed to require little or no nutrition
when, in fact, the nutrient costs of tissue repair, immunocompetence, and
drug metabolism are signiﬁcant.

References

[1] Remillard RL, Darden DE, Michel KE, et al. An investigation of the relationship between

caloric intake and outcome of hospitalized dogs. Vet Ther 2001;2:301–10.

[2] Semba RD, Meydani SN, Kramer TR, et al. Micronutrient inﬂuences on immune response.

In: Proceedings of the Conference on Nutrition and Immunity as part of the International
Conference Series on Nutrition and Health Promotion. Atlanta: International Life Sciences
Institute; 1997. p. 9–18.

[3] Burkholder WJ, Swecker WS. Nutritional inﬂuences on immunity. Semin Vet Med Surg

1990;5:154–66.

[4] Crane SW. Nutritional aspects of wound healing. Semin Vet Med Surg 1989;4:263–7.
[5] Fettman MJ, Phillips RW. Dietary eﬀects on drug metabolism. In: Hand MS, Thatcher

CD, Remillard RL, et al, editors. Small animal clinical nutrition. 4th edition. Topeka, KS:
MarkMorris Institute; 2000. p. 923–39.

[6] Remillard RL, Martin RA. Nutritional support in the surgical patient. Semin Vet Med

Surg 1990;5:197–207.

[7] Meguid MM, Mughal MM, Meguid V, et al. Risk-beneﬁt analysis of malnutrition and

perioperative nutritional support: a review. Nutr Intern 1987;3:25–34.

[8] Thatcher CD, Hand MS, Remillard RL. Small animal clinical nutrition: an interactive

process. In: Hand MS, Thatcher CD, Remillard RL, et al, editors. Small animal clinical
nutrition. 4th edition. Topeka (KS): Mark Morris Institute; 2000. p. 1–19.

[9] Michel K. Prognostic value of clinical nutritional assessment in canine patients. J Vet

Emerg Crit Care 1993;3:96–104.

[10] Cuthbertson DP. The metabolic response to injury and its nutritional implications:

retrospect and prospect. J Parenter Enteral Nutr 1979;3:108–29.

[11] Popp MB, Brennan MF. Metabolic response to trauma and infection. In: Fischer JE,

editor. Surgical nutrition. Boston: Little, Brown and Company; 1983. p. 479–513.

[12] Minard G, KudskKA. Is early feeding beneﬁcial? How early is early? New Horiz

1994;2:156–63.

[13] Burkholder WJ. Metabolic rates and nutrient requirements of sick dogs and cats. JAVMA

1995;206:614–8.

[14] Devey JJ, Crowe DT, Kirby R, et al. Postsurgical nutritional support [letter]. JAVMA

1995;206:1673–5.

1161

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

[15] Boothby WM, Sandiford I. Basal metabolism. Physiol Rev 1924;4:69–161.
[16] Moore FA, Moore EE. Trauma. In: Zaloga GP, editor. Nutrition in critical care. St. Louis:

Mosby; 1994. p. 571–86.

[17] Ogilvie GK, Walters LM, Salman MD, et al. Resting energy expenditure in dogs with

nonhematopoietic malignancies before and after excision of tumors. Am J Vet Res
1996;57:1463–7.

[18] Ogilvie GK, Salman MD, Kesel ML, et al. Eﬀect of anesthesia and surgery on energy

expenditure determined by indirect calorimetry in dogs with malignant and nonmalignant
conditions. Am J Vet Res 1996;57:1321–6.

[19] Walton RS, Wingﬁeld WE, Oglivie GK, et al. Energy expenditure in 104 postoperative and

traumatically injured dogs with indirect calorimetry. J Vet Emerg Crit Care 1996;6:71–9.

[20] Elwyn DH, Kinney JM, Askanazi J. Energy expenditure in surgical patients. Surg Clin

North Am 1981;61:545–6.

[21] Streat SJ, Beddoe AH, Hill GL. Aggressive nutritional support does not prevent protein

loss despite fat gain in septic intensive care patients. J Traumatol 1987;27:262–6.

[22] Deitel M, Williams VP, Rice TW. Nutrition and the patient requiring mechanical

ventilatory support. J Am Coll Nutr 1983;2:25–32.

[23] Pilbeam SP, Head A, Grossman GD, et al. Undernutrition and the respiratory system Part

1. Respir Ther 1983;13:65–9.

[24] Pilbeam SP, Head A, Grossman GD, et al. Undernutrition and the respiratory system Part

2. Respir Ther 1983;13:72–8.

[25] Chang S, Silvis SE. Fatty liver produced by hyperalimentation of rats. Am J Gastroenterol

1974;62:410–8.

[26] Lowry SF, Brennan MF. Abnormal liver function during parenteral nutrition: relation to

infusion excess. J Surg Res 1979;26:300–7.

[27] Hirai Y, Sanada Y, Fujiwara T, et al. High calorie infusion-induced hepatic impairments in

infants. JPEN 1979;3:146–50.

[28] Mallet JO. Calculating parenteral feedings: a programmed instruction. J Am Diet Assoc

1984;84:1312–20.

[29] Porta EA, Hartroft WS. Protein deﬁciency and liver injury. Am J Nutr 1970;23:447–61.
[30] Truswell AS, Hanson JDL, Watson CE, et al. Relation of serum lipids and lipoproteins to

fatty liver in Kwashiorkor. Am J Nutr 1969;22:568–76.

[31] Stein TP. Protein metabolism and parenteral nutrition. In: Rombeau JL, Caldwell MD,

editors. Clinical nutrition; parenteral nutrition. 1st edition. Philadelphia: WB Saunders;
1986. p. 100–34.

[32] Association of American Feed Control Oﬃcials. Oﬃcial Publication. Oxford (IN):

AAFCO; 2001.

[33] DudrickSJ, Wilmore DW, Vars HM, et al. Long-term total parenteral nutrition with

growth, development and positive nitrogen balance. Surgery 1968;64:134–42.

[34] Remillard RL, Thatcher CD. Parenteral nutritional support in the small animal patient.

Vet Clin North Am Small Anim Pract 1989;19:1287–306.

[35] KirkSJ, Barbul A. Role of arginine in trauma, sepsis and immunity. JPEN 1990;14:

226S–9S.

[36] Chyun JH, Griminger P. Improvement of nitrogen retention by arginine and glycine

supplementation and its relation to collagen synthesis in traumatized mature and aged rats.
J Nutr 1984;114:1697–704.

[37] Daly JM, Reynolds J, Thom A, et al. Immune and metabolic eﬀects of arginine in the

surgical patient. Ann Surg 1988;208:512–22.

[38] Gottschlich MM, Jenkins M, Warden GD, et al. Diﬀerential eﬀects of three dietary

regimens on selected outcome variables in burn patients. J Parenter Enteral Nutr 1990;14:
225–36.

[39] Bohles H, Segerer H, Fekl W. Improved nitrogen retention during L-carnitine-

supplemented total parenteral nutrition. J Parenter Enteral Nutr 1984;8:9–13.

1162

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

[40] Lacey JM, Wilmore DW. Is glutamine a conditionally essential amino acid? Nutr Rev

1989;97:1033–42.

[41] Ziegler TR, Young LS. Therapeutic eﬀects of speciﬁc nutrients. In: Rombeau JL,

Rolandelli RH, editors. Clinical nutrition; enteral and tube feeding. 3rd edition. Phila-
delphia: WB Saunders; 1997. p. 112–37.

[42] Long CL, Birkahm RH, Geiger JW. Contribution of skeletal muscle protein in elevated

rates of whole body protein catabolism in trauma patients. Am J Clin Nutr 1981;34:1087–93.

[43] Arora NS, Rochaester DF. Respiratory muscle strength and maximal voluntary ventilation

in undernourished patients. Am Rev Respir Dis 1982;126:5–8.

[44] Donahoe M, Rogers RM, Wilson DO, et al. Oxygen consumption of the respiratory

muscles in normal and malnourished patients with chronic obstructive pulmonary disease.
Am Rev Respir Dis 1989;149:385–91.

[45] Palombo JD, DeMichele SJ, Boyce PJ, et al. Eﬀect of short-term enteral feeding with

eicosapentaenoic and gamma-linolenic acids on alveolar macrophage eicosanoid synthesis
and bactericidal function in rats. Crit Care Med 1999;27:1908–15.

[46] GadekJE, DeMichele SJ, Karlstad MD, et al. Eﬀect of enteral feeding with eico-

sapentaenoic acid, gamma-linolenic acid, and antioxidants in patients with acute respi-
ratory distress syndrome. Enteral Nutrition in ARDS Study Group. Crit Care Med 1999;
27:1409–20.

[47] Remillard RL, Armstrong PJ, Davenport DJ. Assisted feeding in hospitalized patients:

enteral and parenteral nutrition. In: Hand MS, Thatcher CD, Remillard RL, et al, edi-
tors, Small animal clinical nutrition. 4th edition. Topeka (KS): Mark Morris Institute;
2000. p. 351–99.

[48] Clary EM, Hardie EM, Fischer W, et al. Nonendoscopic antegrade percutaneous gastro-

stomy: the eﬀect of preplacement gastric insuﬄation on tube position and intra-abdominal
anatomy. J Vet Intern Med 1996;10:15–20.

[49] Orton EC. Enteral hyperalimentation administered via needle catheter—jejunostoma as an

adjunct to cranial abdominal surgery in dogs and cats. JAVMA 1986;188:1406–11.

[50] Crowe DT. Methods of enteral feeding in the seriously ill or injured patient: Part I and Part

II. J Vet Emerg Critical Care 1986;3:1–17.

[51] Dickerson RN, Brown RO, White KG. Parenteral nutrition solutions. In: Rombeau JL,

Caldwell MD, editors. Clinical nutrition; parenteral nutrition. 2nd edition. Philadelphia:
WB Saunders; 1993. p. 310–33.

[52] Trissel LA. Handbookon injectable drugs. 9th edition. Bethesda: American Society of

Hospital Pharmacists; 1996.

[53] Payne-James JJ, Khawaja HT. First choice for total parenteral nutrition: the peripheral

route. JPEN 1993;17:468–78.

[54] Everitt NJ, McMahon MJ. Peripheral intravenous nutrition. Nutrition 1994;10:49–57.
[55] Kohlhardt SR, Smith RC, Wright CR. Peripheral versus central intravenous nutrition:

comparison of two delivery systems. Br J Surg 1994;81:66–70.

[56] Matsusue S, Nishimura S, Koizumi S, et al. Preventive eﬀect of simultaneously infused lipid

emulsion against thrombophlebitis during postoperative peripheral parenteral nutrition.
Surgery Today 1995;25:667–71.

[57] Daly JM, Masser E, Hansen L, et al. Peripheral vein infusion of dextrose/amino acid

solutions +/

20% fat emulsion. J Parenter Enteral Nutr 1985;9:296–9.

[58] Isaacs JW, Millikan WJ, Stackhouse J, et al. Parenteral nutrition of adults with a 900

milliosmolar solution via peripheral veins. Am J Clin Nutr 1977;30:552–9.

[59] Maden M, Alexander DJ, McMahon MJ. Inﬂuence of catheter type on occurrence of

thrombophlebitis during peripheral intravenous nutrition. Lancet 1992;339:101–3.

[60] Bayer-Berger M, Chiolero R, Freeman J. Incidence of phlebitis in peripheral parenteral

nutrition: eﬀect of the diﬀerent nutrient solutions. Clin Nutr 1989;8:181–6.

[61] McCarthy MC, Shives JK, Robinson RJ, et al. Prospective evaluation of single and triple

lumen catheters in total parenteral nutrition. J Parenter Enteral Nutr 1987;11:259–62.

1163

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

[62] Gaukroger PB, Roberts JG, Manners TA. Infusion thrombophlebitis: a prospective

comparison of 645 Vialon and Teﬂon cannulae in anaesthetic and postoperative use.
Anaesth Intensive Care 1988;16:265–71.

[63] McKee JM, Shell JA, Warren TA, et al. Complications of intravenous therapy: a

randomized prospective study—Vialon vs. Teﬂon. J Intravenous Nurs 1989;12:288–95.

[64] Borow M, Crowley JG. Evaluation of central venous catheter thrombogenicity. Acta

Anaesthesiol Scand Suppl 1985;81:59–64.

[65] di Costanzo J, Sastre B, Choux R, et al. Mechanism of thrombogenesis during total

parenteral nutrition: role of catheter composition. J Parenter Enteral Nutr 1988;12:190–4.

[66] Bozzetti F. Central venous catheter sepsis. Drug Gynecol Obstet 1985;161:293–301.
[67] Maki DG, Weise CE, Saraﬁn HW. A semiquantitative culture method for identifying

intravenous catheter-related infection. N Engl J Med 1977;296:1305–9.

[68] Herruzo-Bareara R, Garcia-Caballero J, Vera-Cortes L, et al. Growth of microorganisms

in parenteral nutrition solutions. Am J Hosp Pharm 1984;41:1178–80.

[69] Goldmann DA, Martin WT, Worthington JW. Growth of bacteria and fungi in total

parenteral nutrition solutions. Am J Surg 1971;126:314–8.

[70] Wilkinson WR, Flores LL, Pagones JN. Growth of microorganisms in parenteral

nutritional ﬂuids. Drug Intell Clin Pharm 1973;7:226–31.

[71] Schecekelhoﬀ DJ, Mirtallo MJ, Ayers LW, et al. Growth of bacteria and fungi in total

nutrient admixtures. Am J Hosp Pharm 1986;43:73–7.

[72] Rowe CE, Fukuyama TT, Martinoﬀ JT. Growth of microorganisms in total nutrient

admixtures. Drug Intell Clin Pharm 1987;21:633–8.

[73] Simmons BP, Hotton TM, Wong ES, et al. CDC guidelines for prevention of intravascular

infections. Infect Control 1982;3:52–67.

[74] Adams S, Dellinger EP, Wurtz MJ, et al. Enteral versus parenteral nutritional support

following laparotomy for trauma: a randomized prospective trial. J Traumatol 1986;
26:882–91.

[75] Moore EE, Jones TN. Beneﬁts of immediate jejunostomy feeding after major abdominal

trauma—a prospective randomized study. J Traumatol 1986;26:874–81.

[76] Alverdy J, Chi HS, Sheldon GF. The eﬀect of parenteral nutrition on gastrointestinal

immunity. Ann Surg 1985;202:681–4.

[77] Remillard RL, Guerino F, Dudgeon DL, et al. Eﬀects of intravenous glutamine and limited

enteral feedings in preventing small intestinal atrophy [abstract]. In: Proceedings of the
Waltham International Nutrition Symposium, Orlando, 1997. J Nutr 1998;128(12S):
2723S–6S.

[78] Remillard RL, Dudgeon DL, Yardley J. Atrophied small intestinal responses to oral

feedings of milk[abstract]. In: Proceedings of Waltham International Nutrition Sym-
posium, Orlando, 1997. J Nutr 1998;128(12S):2727S–9S.

[79] Andrassy RJ, Mahour GH, Harrison MR, et al. The role of early postoperative feeding in

the pediatric surgical patient. J Pediatr Surg 1979;14:381–5.

[80] Andrassy RJ, Dubois T, Page CP, et al. Early postoperative nutritional enhancement

utilizing enteral branched-chain amino acids by way of a needle catheter jejunostomy. Am
J Surg 1985;150:730–4.

[81] Hermann-Zaidius MG. Malabsorption in adults: etiology, evaluation and management.

J Am Diet Assoc 1986;86:1171–8.

1164

R.L. Remillard / Vet Clin Small Anim 32 (2002) 1145–1164

Bacterial translocation:

clinical implications and prevention

Douglass K. Macintire, DVM, MS*,

Ted L. Bellhorn, DVM

Department of Small Animal Surgery and Medicine, College of Veterinary Medicine,

Auburn University, Auburn, AL 36849, USA

Sepsis is the most common cause of death in human trauma patients sur-

viving more than 48 hours after an initial traumatic insult [1]. In many crit-
ically ill patients dying of sepsis, however, no source of infection can be
found. Gram-negative bacteria are the most common organisms cultured
from these patients. In the late 1980s and early 1990s, these observations led
many researchers to suspect that the gut is the reservoir of pathogenic bac-
teria and endotoxins that initiate the systemic host response leading to shock
and organ failure [2].

Deﬁnition and background

Bacterial translocation (BT) is deﬁned as the passage of viable indigenous

bacteria from the gastrointestinal (GI) tract to the mesenteric lymph nodes,
liver, spleen, and bloodstream [1]. Numerous animal and human studies
have clearly documented that microorganisms and toxins normally present
in the GI tract can translocate from inside the lumen to extraintestinal sites
[3–5].

In the 1980s and 1990s, various researchers employed a rodent hemorrha-

gic shock model to demonstrate that hemorrhagic shock causes BT to mes-
enteric lymph nodes [6]. With increasing severity of shock, they showed that
bacteria could also be cultured from the liver and spleen [3] and, ultimately,
that bacteria and endotoxin could be detected in systemic blood [4,5].
Because of these ﬁndings, many researchers saw the gut as the originator
of multiple organ failure [7–10].

Vet Clin Small Anim 32 (2002) 1165–1178

* Corresponding author.

E-mail address: macindk@vetmed.auburn.edu (D.K. Macintire).

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 0 3 7 - 2

Pathogenesis

A simple hypothesis was proposed to explain the occurrence of multiple

organ dysfunction in animals or human beings without a known septic focus
[9]. It was believed that shock or trauma resulted in reduced perfusion and
impaired oxygen delivery to the gut. The ensuing mucosal damage owing to
ischemia/reperfusion injury and oxygen stress resulted in gut barrier dys-
function, allowing translocation of bacteria, endotoxins, and cytokines into
the systemic circulation. These mediators were thought to induce a massive
proinﬂammatory response, thus aﬀecting distant organs [11]. A rodent model
of nonlethal gut ischemia/reperfusion supported the role of BT in contri-
buting to distant organ failure [12]. Occlusion of the superior mesenteric
artery for 45 minutes followed by 6 hours of reperfusion resulted in acute
lung injury in rats.

Despite the considerable mass of evidence supporting the existence of BT

in experimental animals, the clinical signiﬁcance of BT was called into ques-
tion when researchers were unable to culture bacteria from the portal or sys-
temic blood in a series of human trauma victims [13,14]. In addition, the
results of a multicenter trial in critically ill human patients evaluating selec-
tive gut decontamination in which antimicrobial agents were used aggres-
sively to depopulate the gut of pathogenic gram-negative bacteria and
fungi were somewhat disappointing [15]. No improvement in length of sur-
vival was noted, although there was a 50% reduction in the number of infec-
tious complications in these patients.

The inability to culture bacteria from the portal vein of animals or people

with shock of recent onset led to some modiﬁcations in the original concept
of BT [16]. It is now believed that the gut-derived factors contributing to dis-
tant organ injury are found in the mesenteric lymph nodes rather than in the
portal blood and that actual bacteria are not necessary to initiate the sys-
temic inﬂammatory response [17–19].

Presumably, many bacteria that translocate to the intestinal lymphatic

tissue are killed by the host, thereby initiating a massive proinﬂammatory
response characterized by the release of cytokines, vasoactive substances,
complement, and other immunomodulators [1]. Furthermore, gut-derived
endotoxemia may be the signal that triggers, perpetuates, or exacerbates the
hypermetabolic response seen in the systemic inﬂammatory response syn-
drome (SIRS) [20]. Endotoxins are known to stimulate cytokine release and
can cause impairment of the immune system, coagulation system, and GI
mucosal barrier [6,21,22]. It is therefore not necessary to culture viable bac-
teria from the bloodstream or distant organs to implicate the gut as the most
probable cause of SIRS.

One theory that has been proposed to explain the relation between the gut

and multiple organ dysfunction has been referred to as the ‘‘two-hit’’ theory
[23]. According to this concept, an initial event, such as trauma or shock, acts
to ‘‘prime’’ the immune system by activating neutrophils and macrophages.

1166

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

When a second insult occurs, such as translocation of endotoxins or bacteria
from damaged GI mucosa, an exaggerated host response follows in which
massive quantities of cytokines are produced and released into the systemic
circulation [24]. These cytokines, including tumor necrosis factor-a (TNFa),
interleukin (IL)-6, and IL-1B, cause hemodynamic instability and tissue injury
[12,25–27]. Neutrophils are activated such that their release of mediators is
ampliﬁed into a generalized systemic inﬂammatory response [28].

The systemic inﬂammatory response consists of complement activation,

oxidant activation, and endothelial-leukocyte interaction with adhesion
molecules [29]. Neutrophils are recruited to the site of injury via complement
and cause tissue damage through the release of proteolytic enzymes, reactive
oxygen species, and vasoactive substances. The endothelial-leukocyte inter-
action can lead to disseminated intravascular coagulation, microvascular
thrombi, and organ damage.

Recent studies also indicate that there is excessive synthesis of nitric oxide

associated with shock [30]. Excess nitric oxide causes vasodilation, hypoten-
sion, and decreased cardiac contractility. Nitric oxide is converted to perox-
ynitrite in macrophages. This mediator can cause cellular damage through
lipid peroxidation, glutathione depletion, ATP depletion, and mitochondrial
dysfunction [31].

Another mechanism contributing to organ damage is selective transcrip-

tion failure in which proinﬂammatory cytokines are produced in excess and
anti-inﬂammatory mediators are decreased [11]. A transcriptional failure of
acute-phase proteins has been shown to occur in animal models of fulminant
sepsis, likely contributing to organ failure [32,33].

TNFa is an important mediator of SIRS and multiple organ dysfunction

syndrome (MODS). A rodent model of hemorrhagic shock documented
translocation of GI microorganisms to the mesenteric lymph nodes within
1 hour after the onset of shock [34]. In addition, investigators documented
the expression of the TNFa gene in the ileum, mesenteric lymph nodes, and
white pulp of the spleen but not in the liver, kidney, or lung. Translocation
of microorganisms to the mesenteric lymph nodes seems to trigger the
expression of TNFa in the gut-associated lymphatic tissue, leading to a gen-
eralized host response through altered signal transduction.

Further evidence that BT involves mesenteric lymph nodes rather than

portal blood has been provided by other studies. Acute lung injury after
hemorrhagic shock or thermal injury in rats can be prevented or ameliorated
by diverting mesenteric lymph ﬂow, because the lung is the ﬁrst organ
exposed to mesenteric lymph [19]. It has also been shown that mesenteric
lymph (but not portal blood) collected after a shock episode activates neu-
trophils, increases endothelial cell permeability, and even causes cell death in
vitro [19]. Mesenteric lymph nodes drain into the thoracic duct, which
bypasses the reticuloendothelial system and goes directly to the lungs. High
levels of endotoxins have been found in lymphatic ﬂuid after shock [18],
whereas blood cultures are often negative [35].

1167

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

The liver also plays a role in the pathogenesis of BT and gut-induced sep-

sis. BT was increased almost fourfold in human patients with severe liver
disease and cirrhosis compared with noncirrhotic patients [36]. In addition,
selective decontamination of the gut reduced the level of BT for patients
with advanced cirrhosis to that found in noncirrhotic patients [38]. In rats,
experimentally induced ligation of the common bile duct caused signiﬁcant
BT to the mesenteric lymph nodes and liver. Sucralfate and/or gentamicin
reduced the degree of BT in these rats [39].

Another factor that has recently been shown to be important in the

pathogenesis of BT is bacterial virulence [38–40]. Some types of bacteria,
such as Pseudomonas aeruginosa, or strains of speciﬁc bacteria, such as
Escherichia coli, are more virulent than others. They have the ability to sense
when the host is most vulnerable to infection by detecting changes in tem-
perature, pH, osmolality, oxygen, carbon dioxide, nitrogen compounds,
oxygen free radicals, and norepinephrine. The sensing mechanism occurs
through diﬀusable molecules and allows communication among the bacte-
ria. The signaling induces gene transduction for such processes as adhesion,
colonization, and proliferation. The PA-1 adherence gene causes injury to
the tight epithelial junctions and allows cytotoxins from bacteria to enter the
host. The type III protein secretory system injects secreted invasion protein
(SIP) into cells, allowing bacteria to enter host cells. Inside the host cell, bac-
teria are able to live within macrophage vacuoles and survive oxidative
agents, DNA damage, increased osmolality, starvation, and acid pH
through macrophage gene transduction. The bacteria can then be trans-
ported to more distant sites within the host.

Bacterial overgrowth also contributes to BT [37,41–43]. Studies have

shown that use of antacids in critical patients may lead to proximal gut col-
onization by virulent bacteria because of increased gastric pH. Colonization
of the proximal gut has been associated with an increase in BT and septic
morbidity [41].

Many researchers believe that splanchnic ischemia plays a central role in

the development of multiple organ failure [43–45], because there is a strong
correlation between decreasing intramucosal pH and morbidity and mortal-
ity [46]. It is believed that intestinal ischemia leads to loss of barrier func-
tion, which results in exposure of the gut-associated lymphoid tissue
(GALT) to bacteria and toxins and the ensuing release of massive amounts
of cytokines and endotoxins. If the reticuloendothelial system is over-
whelmed, systemic endotoxemia and/or bacteremia may result.

Gastrointestinal mucosal barrier

Under normal conditions, the gut provides an eﬀective mechanical and

functional barrier to systemic absorption of intraluminal bacteria and toxins
[47]. For BT to occur, bacteria must ﬁrst adhere to the intestinal mucosa.
Adherence is reduced by intestinal peristalsis and mucus production.

1168

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

Enhanced BT has been shown to occur in conditions associated with
decreased peristalsis, such as intestinal ileus and obstruction. Administra-
tion of vasopressors, corticosteroids, and nonsteroidal anti-inﬂammatory
drugs (NSAIDS) can result in decreased mucus production and loss of the
protective mechanical barrier. Conditions of poor perfusion, such as the
splanchnic ischemia associated with shock, also result in decreased epithelial
cell turnover, cell death, and enhanced potential for mucosal breakdown.
Stress gastritis and ulceration are common in critical patients.

The gut is the largest immunologic and endocrine organ in the body [48].

The GALT consists of Peyer’s patches, lymphoid follicles, lamina propria
lymphocytes, intraepithelial lymphocytes, and mesenteric lymph nodes.
Secretory IgA is produced by the antigen-primed lymphocytes that line the
intestinal mucosa. These immune defenses are critical in defending the host
against bacterial invasion. Immunosuppressed patients are therefore pre-
disposed to BT. Poor nutrition of enterocytes can result in decreased IgA
production and impaired GI immune defenses.

A ﬁnal factor that helps to maintain the normal GI mucosal barrier is the

protective role of the normal indigenous microﬂora [49]. Anaerobes are the
most numerous bacteria in the GI tract. They compete with potential patho-
gens for nutrients and mucosal attachment sites, thereby inhibiting bacterial
overgrowth with gram-negative bacteria. Antibiotic therapy often upsets the
delicate balance of the GI microﬂora by selecting for gram-negative and
resistant organisms while suppressing the more sensitive indigenous anae-
robes [3]. Other interventions that may disrupt the normal ﬂora in critical
patients include the use of H2 blockers, which can result in bacterial over-
growth and colonization of the stomach [50], and the use of hyperosmolar
enteral diets [51].

Importance of nutrition

For many years, the GI tract was ignored in the management of critically

ill patients. The primary function of the GI tract was seen as the absorption
of nutrients, which was considered necessary to support adequate wound
healing and host response to injury or infection. Concern about possible
aspiration, vomiting, ileus, or lack of enteral access led many clinicians to
pursue a course of ‘‘bowel rest.’’ We now know that bowel rest can lead
to mucosal atrophy, altered permeability, and loss of the trophic eﬀects of
GI hormones. It has been shown in experimental models that starvation and
malnutrition alone do not induce BT but may predispose to mucosal dam-
age and the development of a potentially lethal gut origin septic state during
periods of systemic inﬂammation [52]. Currently, there is signiﬁcant interest
and ongoing research to identify individual nutrient eﬀects and to use nutri-
tion as a modulator of metabolic and inﬂammatory processes.

Early studies in rodents showed the superiority of enteral nutrition over

parenterally administered nutrition [53]. BT and mucosal atrophy were seen

1169

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

in rats administered total parenteral nutrition (TPN) but not in those fed
enterally. In human beings, the adverse eﬀects of TPN have not been pro-
ven. Mucosal atrophy does not occur with short-term TPN, and in perioper-
ative patients, there was no diﬀerence in morbidity, mortality, preservation
of gut barrier function, or development of infectious complications in
patients receiving either parenteral or enteral nutrition [54]. Early enteral
nutrition has been advocated to improve splanchnic blood ﬂow and modu-
late the immune response, especially in trauma and burn patients. Enteral
nutrition can be associated with some complications in critically ill patients,
however, such as diarrhea, bloating, vomiting, or ileus [55]. A combination
of enteral and parental nutrition may be the best method to meet the needs
of critically ill patients.

The type of diet fed may also be important. In mice, feeding a liquid diet

results in BT, whereas feeding a solid diet of rat chow does not result in BT
[50]. Another nutrient that has received considerable study is glutamine [56].
Glutamine is the preferred metabolic fuel for cells lining the small intestine
and has been considered a ‘‘conditionally essential’’ nutrient in critically ill
patients. It is essential for lymphocyte mitogenesis and enhances gut barrier
function. Many studies in rodents have shown beneﬁcial eﬀects of adding
glutamine to enteral or parenteral solutions (reduced BT, thicker GI muco-
sa, and increased survivability) [57,58]. In cats, however, a glutamine-
enriched diet was unable to prevent BT or attenuate permeability defects
secondary to methotrexate-induced enterocolitis [59].

The preferred fuel of colonocytes is short-chain fatty acids. These are

produced through fermentation of nondigestible carbohydrates, commonly
referred to as fermentable ﬁbers (pectin, b-glycan, and lactulose). Insoluble
ﬁbers, such as cellulose, have trophic eﬀects on the GI mucosa by promoting
mucus production, stimulating epithelial cell growth, and preserving growth
of normal microﬂora. Insoluble ﬁber is thought to stimulate release of tro-
phic gut hormones, which enhance gut barrier function. Current recommen-
dations regarding optimal ﬁber type and dose are lacking, but research is
ongoing. Preliminary animal studies have shown decreased BT, prevention
of mucosal atrophy, and avoidance of cecal bacterial overgrowth after the
addition of bulk ﬁber additives to enteral diets [60]. Other dietary additives
that may reduce BT include x3 fatty acids (ﬁsh oil products), arginine,
nucleic acids, and antioxidants. Other research is focusing on hormones,
such as bombesin, which exert protective trophic eﬀects on the GI mucosa.
Deﬁnitive dietary recommendations await the results of this exciting area of
research.

Clinical signiﬁcance

Based on experimental studies in animal models, three primary mecha-

nisms leading to enhanced BT have been identiﬁed: intestinal bacterial

1170

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

overgrowth, deﬁciencies in host immune defenses, and damage to the GI
mucosal barrier. Aggressive prevention of BT must therefore address these
three concerns as well as provide nutritional support of the gut.

Animal research models and human clinical reports have shown that BT

can be promoted by thermal injury, immunosuppression, trauma, hemor-
rhagic shock, endotoxin, acute necrotizing pancreatitis, TPN, neutropenia,
intestinal obstruction, and intestinal ischemia. These same conditions would
be likely to promote BT in critically ill veterinary patients. In addition, dogs
with severe parvoviral enteritis are uniquely predisposed to developing BT,
sepsis, and endotoxemia because of the combination of neutropenia and
breakdown of the GI mucosal barrier. E. coli has been cultured from the
lungs and liver of dogs that have died from parvoviral enteritis [61], and
endotoxemia has also been documented in aﬀected dogs [62].

Prevention and treatment

Prevention of BT, sepsis, and multiple organ failure is an area of ongoing

research. The most important factor for preventing BT is preservation of an
intact GI mucosal barrier, because experimental studies have shown that BT
can largely be prevented by limiting mucosal injury [63]. For this reason,
therapeutic measures are aimed at (1) decreasing the likelihood of mucosal
disruption, (2) limiting the consequences of disruption if it occurs, and (3)
supporting the gut so that mucosal defects can be rapidly repaired. The fol-
lowing recommendations can be made.

Improving gut oxygenation

It seems that most mucosal damage in critical patients is initiated by ische-

mia and potentiated by reperfusion injury [1,12,44]. Oxygen delivery to the
gut should be maximized with eﬀective aggressive hemodynamic resuscita-
tion. Adequate ﬂuid therapy with crystalloid and/or colloid ﬂuids should
be administered to maintain adequate blood pressure and GI perfusion.

Early resuscitation is important in curbing ischemia and reperfusion

injury. Most cells deprived of oxygen for 5 to 10 minutes can become tem-
porarily or permanently damaged [16]. Simply supplying oxygen to the ani-
mal via a mask, ‘‘ﬂow by,’’ or nasal tubes can double the fractional inspired
oxygen concentration. Red blood cells or oxygen-carrying hemoglobin solu-
tions can also improve oxygen delivery (D

) to the tissues and are indi-

cated if the hemoglobin concentration drops below 10 to 12 g/dL. Fresh
whole blood or blood stored less than 14days has a higher concentration
of 2,3-diphosphoglycerate (DPG) compared with stored blood. Red blood
cells not only carry oxygen but act as antioxidants through endogenous sub-
stances, such as catalase and glutathione.

Another advantage of fresh whole blood given within 8 hours of collec-

tion is that it provides coagulation factors as well as having oxygen-carrying

1171

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

capacity. Hemoglobin substitutes have enhanced uptake of oxygen in the
lungs and may even facilitate oﬄoading of oxygen in tissue [64]. They are
also nitric oxide scavengers, a property that may improve blood pressure
in animals with hypotension secondary to hemorrhagic shock. Supplemental
oxygen should deﬁnitely be administered if the pulse oximetry reading drops
below 90% to 95%, but there is also evidence to support hyperoxygenation
of shock patients even when oxygen saturation is normal [65]. In one study
of experimentally induced hemorrhagic shock in rats, administration of
100% oxygen prevented BT and TNFa gene expression compared with that
in rats breathing room air [66]. If available, gastric tonometry is an eﬀective
method for monitoring intramucosal pH and determining whether GI per-
fusion is adequate [46].

Positive inotropes, such as dobutamine or dopamine, may be necessary to

maintain blood pressure and restore perfusion in septic patients [46]. b-Ago-
nists also have anti-inﬂammatory eﬀects that may reduce ischemia/reperfu-
sion injury. By increasing cyclic AMP, they are thought to stabilize white
blood cells and reduce oxidant release.

Recent studies performed in rodents with hemorrhagic shock have shown

a beneﬁcial eﬀect of hypertonic saline (HTS) [67,68]. Rats treated with HTS
showed attenuation of lung injury and reduced BT compared with rats
resuscitated with lactated Ringer’s solution [68]. HTS was shown to attenu-
ate platelet activating factor and postinjury priming of neutrophils, thereby
reducing the cytotoxic response [69]. Timing may be important, because
early use may suppress neutrophil activation, thereby preventing or reduc-
ing lung damage [70]. Although the eﬀects of HTS are relatively short-lived,
they can be prolonged by adding a colloid, such as hetastarch, which has
also been shown to decrease inﬂammation by attenuating neutrophils [71].

Broad-spectrum bactericidal antibiotics should be administered to any

animal exhibiting clinical signs of sepsis. Early diagnosis and surgical correc-
tion of areas of devitalized gut or abscess drainage are paramount to suc-
cessful case management.

Reperfusion injury may be prevented with allopurinol [72] or superoxide

dismutase [73]. Vitamins C, E, and A; selenium; and b-carotene as well as
the amino acids cystine, glycine, and glutamine are all involved in the body’s
antioxidant defense network. Dietary supplementation with antioxidants
may be beneﬁcial. Deferoxamine is an iron chelator shown to reduce oxi-
dant damage in experimental studies, but its use has been limited because
of side eﬀects related to hypotension [74]. Recent eﬀorts to combine deferox-
amine with hetastarch have shown improved hemodynamics and reduced
oxidant damage while avoiding toxic side eﬀects [75].

Pentoxifylline and lisofylline have been shown to attenuate lung injury and

reduce BT in rats with hemorrhagic shock [76,77]. These agents decrease tissue
injury by inhibiting neutrophil adherence and decreasing cytokine release.

Research is currently underway to identify agents that selectively improve

GI perfusion. To date, none has been found. Catecholamines, such as

1172

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

norepinephrine and epinephrine, which induce splanchnic vasoconstriction,
should be avoided.

Limiting the consequences of mucosal injury

The use of antacids and H2 receptor blockers to reduce stress ulcers and

gastritis in critical patients may result in bacterial overgrowth and an
increased incidence of hospital-acquired pneumonia in ventilated patients
[78]. The use of sucralfate and nasogastric suctioning has been recom-
mended to decrease gastric injury without increasing the gastric pH
[79,80]. Sucralfate has also been shown to decrease BT in a rat model.

The use of selective gut decontamination seems to reduce the incidence

of hospital-acquired infections but has not been shown to increase length
of survival in critically ill human patients [81]. Although not often used in
veterinary patients, a combination of amikacin, amphotericin B, and poly-
myxin B is commonly used in human patients to depopulate the gut of
pathogenic organisms. Selective gut decontamination was shown to decrease
BT in human patients with cirrhosis [36]. In rats, orally administered neo-
mycin or gentamicin alone prevented mortality and reduced BT after ther-
mal injury [37]. A combination of oral polymyxin B, charcoal, and kaolin
has been used to bind lipopolysaccharide (LPS) endotoxin. In addition, the
use of dilute chlorhexidine or povidone-iodine enemas in puppies with par-
voviral enteritis has been anecdotally reported. The use of broad-spectrum
oral antibiotics in animals that are not septic is controversial, as it may
destroy normal ﬂora and allow overgrowth of pathogenic bacteria.

Prokinetic drugs, such as metoclopramide and cisapride [82], are helpful

in decreasing bacterial overgrowth by improving gastric motility. Other
drugs that may be helpful in reducing BT are dopexamine, dobutamine, pen-
toxifylline, and angiotensin converting enzyme inhibitors.

A polyvalent equine origin antiserum against LPS endotoxin is available

for use in small animals (SEPTI-serum; Immvac, Inc., Columbia, MO). The
dosage is 4.4 mL/kg diluted 1:1 with intravenous crystalloid ﬂuids and admin-
istered slowly over 30 to 60 minutes. Although clinical trials are lacking with
this product, it should be most eﬀective if administered before antibiotic ther-
apy, because circulating endotoxin concentrations increase dramatically when
bacteria are killed [83]. Patients receiving equine origin antiserum must be
observed closely during administration for signs of anaphylaxis.

Supporting the gut with enteral nutrition

The importance of providing nutrition to critically ill patients is well

documented. In recent years, however, the importance of ‘‘feeding the gut’’
through early enteral nutrition has become more apparent.

Studies in rodents have shown that compared with enteral nutrition, TPN

was associated with increased mortality and infection rates [53]. It resulted

1173

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

in mucosal atrophy, and the lipid emulsions were shown to promote immu-
nosuppression through depressed lymphocyte blastogenesis. In addition,
x

6 fatty acids are precursors of prostaglandin and leukotrienes and can

promote inﬂammation. More recent studies in human beings have not
documented the proposed adverse eﬀects of TPN [54]. The current recom-
mendation is to provide nutrition to critical patients by whatever method
is best tolerated by the patient. Even suboptimal caloric replacement is more
beneﬁcial than none at all.

Enteral nutrition exerts its beneﬁcial eﬀects on gut function by strength-

ening the immune system (lymphocytes and macrophages), increasing IgA
and mucin secretion, and maintaining gut mass through its trophic eﬀects
[84]. Even though the beneﬁcial eﬀects of glutamine have not been proven
in species other than rodents, it is safe and may prove beneﬁcial in patients
with extensive mucosal injury. It is unstable and must be added to solutions
immediately before feeding. It is available as a powder (Cambridge Neutra-
ceuticals Baxter Health Care, Boston, MA [1-800-265-2202]) and can be
given at a dosage of 10 mg/kg/d. Glutamine can be added to the drinking
water of recovering animals or added to the enteral diet and administered
through a nasoesophageal, gastrostomy, or jejunostomy tube. Vitamins, x3
fatty acids, and antioxidants have also been used to promote a healthy gut.

Summary

The occurrence of BT has been well documented in experimental animal

models of hemorrhagic shock, trauma, severe burns, cirrhosis, pancreatitis,
and bacterial overgrowth. Translocation of viable bacteria and endotoxins
into mesenteric lymph nodes and other gut-associated lymphatic tissue is
thought to activate a complex interplay of mediators that initiates the SIRS.
Multiple humoral and cellular systems cause synthesis, expression, and
release of inﬂammatory mediators, such as toxic oxygen radicals, proteolytic
enzymes, adherence molecules, and various cytokines. A massive sustained
proinﬂammatory response can ultimately result in irreversible multiple
organ dysfunction.

Because BT is associated with splanchnic hypoperfusion, the cornerstone

of therapy involves rapid resuscitation and restoration of tissue perfusion. If
a septic focus can be identiﬁed, it should be removed. Gut protectants, pro-
motility agents, antioxidants, and immune-enhancing diets have shown
promise in improving length of survival in these critically ill patients.

References

[1] Deitch EA, Rutan R, Waymack JP. Trauma, shock and gut translocation. New Horiz

1996;4:289–99.

[2] Beal AL, Cerra F. Multiple organ failure syndrome in the 1990’s—systemic inﬂammatory

response and organ dysfunction. JAMA 1994;271:226–36.

1174

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

[3] Maejima K, Deitch EA, Berg R. Promotion by burn stress of the translocation of bacteria

from the gastrointestinal tracts of mice. Arch Surg 1984;119:166–72.

[4] Baker JW, Deitch EA, Berg RD, et al. Hemorrhagic shock induces bacterial translocation

from the gut. J Trauma 1988;28:896–906.

[5] Deitch EA, Berg R, Specan R. Endotoxin promotes the translocation of bacteria from the

gut. Arch Surg 1987;122:185–92.

[6] Deitch EA, Maejima K, Berg R. Eﬀects of oral antibiotics and bacterial overgrowth on the

translocation of the GI-tract microﬂora in burned rats. J Trauma 1985;25:385–92.

[7] Lee CC, Marill KA, Carter WA, Crupi RS. A current concept of trauma-induced multiple

organ failure. Ann Emerg Med 2001;38:170–6.

[8] Moore FA. The role of the gastrointestinal tract in post injury multiple organ failure. Am

J Surg 1999;178:449–53.

[9] Deitch EA. Multiple organ failure. Ann Surg 1992;216:117–34.

[10] Marshall JC, Christou NV, Meakins JL. The gastrointestinal tract: the ‘‘undrained

abscess’’ of multiple organ failure. Ann Surg 1993;218:111–9.

[11] Kim PK, Deutschman CS. Inﬂammatory responses and mediators. Surg Clin North Am

2000;80:885–94.

[12] Grotz MRW, Deitch EA, Ding J, et al. Intestinal cytokine response after gut ischemia: role

of gut barrier failure. Ann Surg 1999;229:478–86.

[13] Moore FA, Moore EE, Poggetti RS, et al. Gut bacterial translocation via the portal vein:

a clinical perspective with major torso trauma. J Trauma 1991;31:629–38.

[14] Sedman PC, Macﬁe J, Sagar P, et al. The prevalence of gut translocation in humans.

Gastroenterology 1994;107:643–9.

[15] Reidy JJ, Ramsay G. Clinical trials of selective decontamination of the digestive tract.

Review. Crit Care Med 1990;18:1449–56.

[16] Lemaire LCJM, van Lanschot JJB, Stoutenbeck CP, et al. Bacterial translocation in

multiple organ failure: cause or epiphenomenon still unproven. Br J Surg 1997;84:
1340–50.

[17] Deitch EA, Xu D, Franko L, et al. Evidence favoring the gut as a cytokine-generating

organ in rats subjected to hemorrhagic shock. Shock 1994;1:141–5.

[18] Magnotti LJ, Upperman JS, Xu DZ, et al. Gut-derived mesenteric lymph but not portal

blood increases endothelial cell permeability and promotes lung injury after hemorrhagic
shock. Ann Surg 1998;228:518–27.

[19] Deitch EA. Role of the gut lymphatic system in multiple organ failure. Current Opinion in

Critical Care 2001;7:92–8.

[20] Rush BF, Jr, Sori AJ, Murphy TF, et al. Endotoxemia and bacteremia during hemorrhagic

shock: the link between trauma and sepsis? Ann Surg 1988;207:549–54.

[21] O’Dwyer ST, Mitchie HR, Ziegler TR, et al. A single dose of endotoxin increases intestinal

permeability in healthy humans. Arch Surg 1988;123:1459–64.

[22] Endo S, Katsuya I, Yamada Y, et al. Plasma endotoxin and cytokine concentrations in

patients with hemorrhagic shock. Crit Care Med 1994;22:949–55.

[23] Saadia R, Schein M. Multiple organ failure: how valid is the ‘‘two hit’’ model? J Accid

Emerg Med 1999;16:163–7.

[24] Baue AE. Predicting outcome in injured patients and its relationship to circulating

cytokines. Shock 1995;4:39–40.

[25] Goris RJ. MODS/SIRS: result of an overwhelming inﬂammatory response? World J Surg

1996;20:418–21.

[26] Dinarello CA. Cytokines as mediators in the pathogenesis of septic shock. Curr Top

Microbiol Immunol 1996;216:133–65.

[27] Beutler B, Grau GE. Tumor necrosis factor in the pathogenesis of infectious diseases. Crit

Care Med 1993;21:5423–35.

[28] Korthuis RJ, Anderson DC, Granger DN. Role of neutrophil-endothelial cell adhesion in

inﬂammatory disorders. J Crit Care 1994;9:47–71.

1175

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

[29] Napolitano LM, Faist E, Wichmann MW, et al. Immune dysfunction in trauma. Surg Clin

North Am 1999;79:1385–416.

[30] Hua TC, Moochhala SM. Role of nitric oxide in hemorrhagic shock-induced bacterial

translocation. J Surg Res 2000;93:247–56.

[31] Nadler EP, Ford HR. Regulation of bacterial translocation by nitric oxide. Pediatr Surg

Int 2000;16:165–8.

[32] Andrejko KM, Deutchman CS. Acute-phase gene expression correlates with intrahepatic

tumor necrosis factor-a (abundance but not plasma TNF levels during sepsis/SIRS in the
rat). Crit Care Med 1996;24:1947–52.

[33] Andrejko KM, Deutchman CS. Altered hepatic gene expression in fecal peritonitis:

changes in transcription of gluconeogenic, B-oxidative, and ureagenic genes. Shock 1997;
7:164–7.

[34] Tani T, Mitschiro F, Kazuyoshi H, et al. Bacterial translocation and tumor necrosis

factor-a gene expression in hemorrhagic shock. Crit Care Med 2000;28:3705–9.

[35] Lichtman SM. Bacterial translocation in humans. J Pediatr Gastroenterol Nutr 2001;33:

1–10.

[36] Cirera I, Bauer TM, Navasa M, et al. Bacterial translocation of enteric organisms in

patients with cirrhosis. J Hepatol 2001;34:32–7.

[37] Akman M, Akbal H, Emir H, et al. The eﬀects of sucralfate and selective intestinal

decontamination on bacterial translocation. Pediatr Surg Int 2000;16:91–3.

[38] Ljungdahl M, Lundholm M, Katouli M, et al. Bacterial translocation in experimental

shock is dependent on the strains in the intestinal ﬂora. Scand J Gastroenterol 2000;35:
389–97.

[39] Guiney DG. Regulation of bacterial virulence gene expression by the host environment.

J Clin Invest 1997;99:565–9.

[40] Alverdy J, Holbrook C, Rocha F, et al. Gut-derived sepsis occurs when the right pathogen

with the right virulence genes meets the right host: evidence for in vivo virulence expression
in Pseudomonas aeruginosa. Ann Surg 2000;232:480–9.

[41] MacFie J, O’Boyle C, Mitchell CJ, et al. Gut origin of sepsis: a prospective study

investigating associations between bacterial translocation, gastric microﬂora, and septic
morbidity. Gut 1999;45:223–8.

[42] Yang CY, Chang CS, Chen GH. Small-intestinal bacterial overgrowth in patients with liver

cirrhosis, diagnosed with glucose H

or CH

breath tests. Scand J Gastroenterol 1998;

33:867–71.

[43] Chang CS, Chen GH, Lien HC, Yeh HZ. Small intestinal dysmotility and bacterial

overgrowth in cirrhotic patients with spontaneous bacterial peritonitis. Hepatology 1998;
28:1187–90.

[44] Bathe OF, Chow AW, Phang PT. Splanchnic origin of cytokines in a porcine model of

mesenteric ischemia-reperfusion. Surgery 1998;123:79–88.

[45] Tamion F, Richard V, Lyoumi S, et al. Gut ischemia and mesenteric synthesis of

inﬂammatory cytokines after hemorrhagic or endotoxic shock. Am J Physiol 1997;
273:G314–21.

[46] Silverman HJ, Tuma P. Gastric tonometry in patients with sepsis: eﬀects of dobutamine

and packed red blood cell transfusions. Chest 1992;102:184–91.

[47] Gebbers SO, Laissue JA. Functional morphology of the mucosal barrier. Microecology

and Therapy 1984;14:137–42.

[48] Tomasi TB. Mechanisms of immune regulation at mucosal surfaces. Rev Infect Dis

1983;5:784–90.

[49] Savage DC, Dubos R, Schaedler RW. The gastrointestinal epithelium and its auto-

chthonous bacterial ﬂora. J Exp Med 1968;127:67–72.

[50] Haskel Y, Udassin R, Freund HR, et al. Liquid enteral diets induce bacterial translocation

by increasing cecal ﬂora without changing intestinal motility. J Parenter Ent Nutr 2001;
25:60–4.

1176

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

[51] Speath G, Berg CP, Specian RD, et al. Food without ﬁber promotes bacterial translocation

from the gut. Surgery 1990;108:240–7.

[52] Deitch EA, Winterton J, Berg R. Eﬀect of starvation, malnutrition, and trauma on the GI

tract ﬂora and bacterial translocation. Arch Surg 1987;122:1019–24.

[53] Alverdy JC, Aoys E, Moss GS. Total parenteral nutrition promotes bacterial translocation

from the gut. Surgery 1988;104:185–90.

[54] Sedman PC, MacFie J, Palmer MD, et al. Preoperative total parenteral nutrition is not

associated with mucosal atrophy or bacterial translocation in humans. Br J Surg 1995;
82:1663–7.

[55] Keith ME, Jeejeebhoy KN. Enteral nutrition. Curr Opin Gastroenterol 1998;14:151–6.
[56] Foitzik T, Kruschewski M, Kroesen AJ, et al. Does glutamine reduce bacterial trans-

location? A study in two animal models with impaired gut barrier. Int J Colorectal Dis
1999;14:143–9.

[57] Hwang TL, O’Dwyer ST, Smith RJ, et al. Presentation of small bowel mucosa using

glutamine-enriched parenteral nutrition. Surg Forum 1986;37:56–9.

[58] Burke D, Alverdy JC, Aoys E, et al. Glutamine supplemented TPN improves gut immune

function. Arch Surg 1989;124:1396–9.

[59] Marks SL, Cook AK, Reader R, et al. Eﬀects of glutamine supplementation of an amino-

acid based puriﬁed diet on intestinal mucosal integrity in cats with methotrexate-induced
enteritis. Am J Vet Res 1999;60:755–63.

[60] Deitch EA. Nutrition and the gut mucosal barrier. In: Daly J, editor. Current opinion in gen-

eral surgery. Philadelphia: Current Science; 1993. p. 85–91.

[61] Turk J, Miller M, Brown J, et al. Coliform septicemia and pulmonary disease associated

with canine parvoviral enteritis—88 cases (1987–1988). JAVMA 1990;96:771–3.

[62] Otto CM, Drobatz KJ, Soter C. Endotoxemia and tumor necrosis factor activity in dogs

with naturally occurring parvoviral enteritis. J Vet Intern Med 1997;11:65–70.

[63] Fukatsu K, Zarzaur BL, Johnson CD. Enteral nutrition prevents remote organ injury and

death after a gut ischemic insult. Ann Surg 2001;233:660–8.

[64] Giger U, Rentko VT. Alternatives to blood transfusions. In: Proceedings of 16th American

College of Veterinary Internal Medicine Forum. Lakewood (CO): American College of
Veterinary Internal Medicine; 1998. p. 430–1.

[65] Gennari R, Alexander JW. Eﬀects of hyperoxia on bacterial translocation and mortality

during gut-derived sepsis. Arch Surg 1996;131:57–62.

[66] Tani T, Fujino M, Hanasawa K, et al. Bacterial translocation and tumor necrosis factor-a

gene expression in experimental hemorrhagic shock. Crit Care Med 2000;28:3705–9.

[67] Assalia A, Bitterman H, Hirsh TM, Krausz MM. Inﬂuence of hypertonic saline on

bacterial translocation in controlled hemorrhagic shock. Shock 2001;15:307–11.

[68] Yada-Langui MM, Coimbra R, Lancellotti C, et al. Hypertonic saline and pentoxifylline pre-

vent lung injury and bacterial translocation after hemorrhagic shock. Shock 2000;14:594–8.

[69] Ciesla DJ, Moore EE, Gonzalez RJ, et al. Hypertonic saline inhibits neutrophil (PMN)

priming via attenuation of p38 MAPK signaling. Shock 2000;14:265–70.

[70] Murao Y, Hoyt DB, Loomis W, et al. Does the timing of hypertonic saline resuscitation

aﬀect its potential to prevent lung damage? Shock 2000;14:18–23.

[71] Kaplan SS, Park TS, Gonzales ER, Gidday JM. Hydroxyethyl starch reduces leukocyte

adherence and vascular injury in the newborn pig cerebral circulation after asphyxia.
Stroke 2000;31:2218–23.

[72] Poggetti RS, Moore FA, Moore EE, et al. Simultaneous liver and lung injury following gut

ischemia is mediated by xanthine oxidase. J Trauma 1992;32:723–8.

[73] Marzi I, Buhren V, Schuttler A, et al. Value of superoxide dismutase for prevention of

multiple organ failure after multiple trauma. J Trauma 1993;35:110–9.

[74] Hallaway PE, Eaton JW, Panter SS, et al. Modulation of deferoxamine toxicity and

clearance by covalent attachment to biocompatible polymers. Pro Natl Acad Sci USA
1989;88:10108–12.

1177

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

[75] Bauer C, Walchar F, Holanda M, et al. Antioxidative resuscitation solution prevents

leukocyte adhesion in the liver after hemorrhagic shock. J Trauma 1999;46:886–93.

[76] Harada H, Ishizaka T, Yoramuru M, et al. The eﬀects of aminophylline and pentoxifylline

on multiple organ damage after E coli sepsis. Am Rev Respir Dis 1989;140:974–80.

[77] Waxman K, Daughters K, Aswani S, et al. Lisofylline decreases white cell adhesiveness

and improves survival after experimental hemorrhagic shock. Crit Care Med 1996;24:
1724–8.

[78] Heyland D, Mandell LA. Gastric colonization by gram-negative bacilli and nosocomial

pneumonia in the intensive care unit patient: evidence for causation. Chest 1992;101:
187–93.

[79] Maier KV, Mitchell D, Gentilelb L. Optimal therapy for stress gastritis. Ann Surg 1994;

220:353–63.

[80] Driks MR, Craven DE, Celli BR, et al. Nosocomial pneumonia in intubated patients given

sucralfate as compared with antacids or histamine type 2 blockers: the role of gastric
colonization. N Engl J Med 1987;317:1378–82.

[81] Selective Decontamination of the Digestive Tract Trialists’ Collaborative Group. Meta-

analysis of randomized controlled trials of selective decontamination of the digestive tract.
BMJ 1993;307:525–32.

[82] Pardo A, Bartoli R, Lorenzo-Zuniga V, et al. Eﬀect of cisapride on intestinal bacterial

overgrowth and bacterial translocation in cirrhosis. Hepatology 2000;31:858–63.

[83] Schulze C, Oesser S, Hein H, Seifert J. Risk of endotoxemia during the initial phase of gut

decontamination with antimicrobial agents. Res Exp Med 2001;200:169–74.

[84] Moore FA, Moore EE, Kudsk KA, et al. Clinical beneﬁts of an immune-enhancing diet for

early post injury enteral feeding. J Trauma 1994;37:607–15.

1178

D.K. Macintire, T.L. Bellhorn / Vet Clin Small Anim 32 (2002) 1165–1178

Index

Note:

Page numbers of article titles are in boldface type.

Airway, upper, obstruction of, medical management of, 1076–1077

Allergic airway disease, medical management of, 1084–1086

Alveolar-arterial oxygen gradient, 1052

Analgesia, in critical care, 1129–1146

monitoring of, 1143–1144

local, in pain, 1139–1143

Analgesic agents, in pain, 1133–1137

Anesthesia, topical, in pain, 1140

Anesthetic/analgesic agents, in pain, 1137–1139

Anesthetic cream, local, eutectic mixture of, 1140–1141

Antibiotics, for nosocomial infections, 1118, 1119

resistance to, nosocomial infections and, 1105–1106

Anticoagulant therapies, in acute lung injury, 1080

Anti-inﬂammatory drugs, nonsteroidal, in pain, 1138

Anxiolytics, in pain, 1139

Bacterial translocation, 1167–1180

clinical implications of, 1172–1173
deﬁnition of, 1167
gastrointestinal mucosal barrier and, 1170–1171, 1175
nutrition and, 1171–1172
pathogenesis of, 1168–1169
prevention of, 1173–1176
studies of, 1167
treatment of, 1173–1176

Barotrauma, as sequelae of mechanical ventilation, 1100

Benzodiazepines, in pain, 1139

Blood gas analysis, 1033–1050

arterial, 1037–1046, 1052

acid-base status and, 1040–1041

Vet Clin Small Anim 32 (2002) 1179–1185

0195-5616/02/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 0 5 1 - 7

Blood gas (continued)

additive eﬀect on pH and, 1046
metabolic acidosis and, 1041–1042
metabolic alkalosis and, 1043
mixed acid-base disturbances and, 1044–1045
neutralizing eﬀect on pH and, 1045
respiratory acidosis and, 1042
respiratory alkalosis and, 1043–1044
to determine oxygenation, 1038–1040
triple disorders and, 1046
ventilation and, 1037–1038

capnometry for, 1035–1036
co-oximeters for, 1035
equipment for, 1034
goal of information from, 1036–1037
pulse oximetry for, 1035
sample collection for, 1033–1034
venous, 1046–1049

acid-base status and, 1047–1048
cerebral oxygenation and, 1048
tissue perfusion and, 1047

Bloodstream, nosocomial infections of, 1106–1109

Breathing, work of (WOB), estimation of, in mechanical ventilation, 1070

Bupivacaine, interpleural, in pain, 1141–1142

Buprenex, in pain, 1136–1137

Buprenorphine, in pain, 1136–1137

Butorphanol, in pain, 1137

Capnography, 1026

clinical applications of, 1029–1030
deﬁnition of, 1023

Capnometry, for blood gas analysis, 1035–1036

Carbon dioxide monitoring, end-tidal, 1053–1054

pulse oximetry and, 1023–1031

Carbon monoxide poisoning, 1013

Catheter, intratracheal, for oxygen delivery, 1016

nasal, for oxygen delivery, 1014–1015
urinary tract nosocomial infections associated with, 1113–1116

Critical care, analgesia in, 1129–1146

monitoring of, 1143–1144

Critical care medicine, respiratory pharmacotherapy in, 1075–1088

Critical care patients, nutritional support in, 1147–1166

1180

Index / Vet Clin Small Anim 32 (2002) 1179–1185

Denitrogenation (absorption) atelectasis, 1018

Diarrhea, nosocomial, 1118–1121

Diﬀusion, impairment of, 1058

Dilaudid, in pain, 1136

Drugs, additions of, for nutritional support, 1159–1160

Dyshemoglobinemias, anemic hypoxia and, 1012–1013

Elizabethan collar canopy, for oxygen delivery, 1016

Emergency care medicine, respiratory pharmacotherapy in, 1075–1088

Energy, demands of, as factor in respiratory muscle fatigue, 1065–1067

minimizing of, to treat respiratory muscle fatigue, 1071

supply of, as factor in respiratory muscle fatigue, 1065

increasing of, to treat respiratory muscle fatigue, 1070

Enteral feeding, and parenteral feeding, for critical care patient, 1162

for critical care patient, 1155, 1157

Enteral nutrition, support of gut with, to prevent bacterial translocation,

1175–1176

Epidural anesthesia, in pain, 1142–1143

Face mask, for oxygen delivery, 1014

Fentanyl, in pain, 1135

Fentanyl patch, in pain, 1135–1136

Flow-by oxygen, 1014

Foods, selection of, for nutritional support, 1156–1157

Furosemide, in pulmonary edema, 1082

Gas exchange, and oxygen transport, 1051

Gastrointestinal mucosal barrier, bacterial translocation and, 1170–1171, 1175

Gut, oxygenation of, improvement of, in bacterial translocation, 1173–1175

support of, with enteral nutrition, to prevent bacterial translocation, 1175–1176

Hemoglobin, and oxyhemoglobin dissociation curve, 1008–1009

Hydromorphone, in pain, 1136

1181

Index / Vet Clin Small Anim 32 (2002) 1179–1185

Hypoventilation, 1054–1056

Hypoxemia, causes of, 1054

consequences of, 1058–1059

Hypoxia, and muscle fatigue, 1061

anemic, and dyshemoglobinemias, 1012–1013
anoxic, 1011–1012
categories of, 1011–1013
histiocytic, 1013
stagnant, 1013

Infections, nosocomial, 1103–1128

antibiotic resistance and, 1105–1106
catheter-associated urinary tract infections, 1113–1116
gastrointestinal, 1118–1121
hospital locations of pathogens and, 1104–1105
of bloodstream, 1106–1109
prevention of, 1121–1122
surgical wound, 1116–1118

Inspiratory pressure/maximum inspiratory pressure, conditions causing increase in,

1063, 1064

Intrapulmonary shunt, 1057

Jejunostomy tubes, for nutritional support, 1156

Ketamine, in pain, 1137–1138

Ketaset, in pain, 1137–1138

Lung, acute injury to, medical management of, 1079

Mechanical ventilation, cost of, 1089

critical care area for, requirements for, 1091–1093
for oxygen delivery, 1016–1017
indications for, 1090–1091
initiation of, 1095
modes of, 1095–1097
nursing considerations in, 1097–1099
positive-pressure, problems associated with, 1099–1100
principles of, 1089–1102
ventilator for, 1093–1095
weaning from, 1100–1101

1182

Index / Vet Clin Small Anim 32 (2002) 1179–1185

Methemoglobinemia, 1012

Morphine, in pain, 1134–1135

Muscles, respiratory. See Respiratory muscles.

Naloxone, in pain, 1137

Narcan, in pain, 1137

Nasoesophageal tubes, for nutritional support, 1155–1156

Nerve block, intercostal, 1141

Neuromuscular competence, decreased, as factor in respiratory muscle

fatigue, 1065

maximizing of, to treat respiratory muscle fatigue, 1071–1072

Nonsteroidal anti-inﬂammatory drugs, in pain, 1138

Nosocomial infections. See Infections, nosocomial.

Numorphan, in pain, 1136

Nutrition, bacterial translocation and, 1171–1172

enteral, support of gut with, to prevent bacterial translocation, 1175–1176

Nutritional factors, for respiratory patient, 1152–1154

Nutritional support, in critical care patients, 1147–1166

assessment of animal for, 1147–1153
B-complex vitamins in, 1152
clinical importance of, 1147
energy requirements and, 1150–1151
fat-soluble vitamins in, 1152
feeding plan for, 1154–1162
nutritional factors in, 1148, 1149
provision of protein in, 1151–1152
reassessment of animal for, 1163
trace minerals in, 1152

Opioids, in pain, 1133–1134, 1135

Oximetry, deﬁnition of, 1023

pulse. See Pulse oximetry.

Oxygen, concentrations of, low inspired, 1058

consumption of, 1010
delivery of, 1008

modes and techniques for, 1013–1014
total, 1009–1010

ﬂow-by, 1014
uptake and diﬀusion of, 1007–1008

1183

Index / Vet Clin Small Anim 32 (2002) 1179–1185

Oxygen cage, for oxygen delivery, 1015

Oxygen deprivation, 1010

clinical signs of, 1010–1011

Oxygen therapy, and oxygen toxicity, 1007–1022

indications for, 1010–1013
monitoring of, 1017–1018

Oxygen toxicity, diagnosis of, 1019–1020

oxygen therapy and, 1007–1022
pulmonary, pathophysiology of, 1019
treatment of, 1020

Oxygenation, physiology of, 1007–1010

Oxyhemoglobin saturation curve, 1053, 1054

Oxymorphone, in pain, 1136

Pain, analgesic agents in, 1133–1137

anesthetic/analgesic agents in, 1137–1139
assessment of, 1133
local anesthetics in, 1139–1143
management of, philosophy of, 1132
mechanisms and pathways of, 1129–1130
nociception and antinociception, 1130–1131
physiologic and metabolic responses to, 1131–1132
stress response to, 1132

Parenteral feeding, and enteral feeding, for critical care patient, 1162

for critical care patient, 1157–1159, 1160

complications of, 1161–1162

Pharmacotherapy, respiratory, in emergency and critical care medicine, 1075–1088

Pharyngostomy/esophagostomy/gastrostomy tubes, for nutritional support, 1156

Plethysmography, 1024

Pleural eﬀusion, medical management of, 1083–1084

Pneumonia, as nosocomial infection, 1109–1112

as sequelae of mechanical ventilation, 1100, 1110, 1111
medical management of, 1078–1079

Pulmonary edema, deﬁnition of, 1082

medical management of, 1082–1083

Pulse oximetry, 1053

advantages of, 1025
and end-tidal carbon dioxide monitoring, 1023–1031
deﬁnition of, 1023
for blood gas analysis, 1035

1184

Index / Vet Clin Small Anim 32 (2002) 1179–1185

Pulse oximetry (continued)

limitations of, 1026, 1143
uses of, 1024, 1143

Respiratory distress, emergency management of, 1075–1076

Respiratory distress syndrome, medical management of, 1079

Respiratory failure, causes of, 1051–1060

deﬁnition of, 1051

Respiratory muscles, fatigue of, 1058, 1061–1073

clinical conditions associated with, 1069–1070
deﬁnition of, 1061
detection of, 1067–1069
factors predisposing to, 1065
physiology of, 1062–1065
treatment of, 1070–1072

Respiratory patient, nutritional factors for, 1152–1154

Sedatives, in pain, 1138–1139

Spectrophotometry, deﬁnition of, 1023

Thromboembolism, pulmonary, medical management of, 1079–1081

pharmaceutic agents in, 1080–1081

Torbugesic, in pain, 1137

Trauma, thoracic, medical management of, 1077–1078

Urinary tract, nosocomial catheter-associated infections of, 1113–1116

Ventilation, and oxygenation, assessment of, 1052–1056

mechanical. See Mechanical ventilation.

Ventilation/perfusion mismatch, 1056–1057

Ventilator, for mechanical ventilation, 1093–1095

Vitamins, in nutritional support for critical care patients, 1152

Warfarin, in acute lung injury, 1081

1185

Index / Vet Clin Small Anim 32 (2002) 1179–1185

Document Outline

00 Preface
01 Oxygen therapy and toxici
- Oxygen therapy and toxicity Ann Marie Manning, DVM
02 Pulse oximetry and end tidal carbon dioxide monitoring
- Pulse oximetry and end tidal carbon dioxide monitoring
03 Blood gas analysis
- Blood gas analysis
04 Causes of respiratory failure
- Causes of respiratory failure
05 Respiratory muscle fatigue
- Respiratory muscle fatigue
06 Respiratory pharmacotherapy in emergency and critical care medicine
- Respiratory pharmacotherapy in emergency and critical care medicine
07 Principles of mechanical ventilation
- Principles of mechanical ventilation
08 Nosocomial infections
- Nosocomial infections
09 Analgesia in critical care
- Analgesia in critical care
10 Nutritional support in critical care patients
- Nutritional support in critical care patients
11 Bacterial translocation_clinical implications and prevention
- Bacterial translocation: clinical implications and prevention
12 Index

Wyszukiwarka

Podobne podstrony:
Blood and Blood Transfusion (Critical Care Focus 8) 0727916572
2002 02 Critical Mass
Current Clinical Strategies, Critical Care and Cardiac Medicine (2005) BM OCR 7 0 2 5
Culkin, Jennifer A Final Arc of Sky A Memoir of Critical Care (pdf)
www neroj com TIS wsmdata pl 2002 50 Focus G17386
2007 5 SEP Respiratory Physiology, Diagnostics, and Disease
www neroj com TIS wsmdata pl 2002 50 Focus G17386
Copy of Cowpark Soil respiration 2002 2003
0748616675 Edinburgh University Press After the Terror Sep 2002
0415226767 Routledge Fifty Key Figures in Twentieth Century British Politics Sep 2002
2007 5 SEP Respiratory Physiology, Diagnostics, and Disease
Ustawa z 30 10 2002 r o ubezp społ z tyt wyp przy pracy i chor zawod
ecdl 2002
ei 03 2002 s 62
2002 09 42
2002 06 15 prawdopodobie stwo i statystykaid 21643

więcej podobnych podstron

2002 5 SEP Critical care respiratory focus

Document Outline