Bongard Frederic , Darryl Sue. Diagnosis and Treatment Critical Care

Подождите немного. Документ загружается.



247

0012

Respiratory Failure

Darryl Y. Sue, MD

Janine R. E. Vintch, MD

PATHOPHYSIOLOGY OF RESPIRATORY FAILURE

Respiratory failure is inability of the respiratory system to

maintain a normal state of gas exchange from the atmos-

phere to the cells as required by the body. Simply, the role of

the respiratory system is to maintain normal arterial blood

, and pH. Respiratory failure can result from disor-

ders of the lungs, heart, chest wall, respiratory muscles, and

central ventilatory control mechanisms. In addition, dys-

functions of the heart, the pulmonary and systemic circula-

tions, the oxygen-carrying capacity of the blood, and

systemic capillaries have important implications for patients

with respiratory failure.

Definition

Respiratory failure is present (1) if arterial P

(Pa

) is less

than 60 mm Hg or (2) if arterial P

(Pa

) is greater than

45 mm Hg. A Pa

of less than 60 mm Hg, indicating hypox-

emic respiratory failure, is valid during room air breathing

(inspired O

fraction [F

] = 0.21), but hypoxemia while

breathing supplemental oxygen also indicates respiratory

failure. An exception to the rule that Pa

greater than

45 mm Hg defines hypercapnic respiratory failure is meta-

bolic acidosis. Patients with metabolic acidosis normally

decrease Pa

to compensate for low pH, but if Pa

abnormally elevated even though below 45 mm Hg during

metabolic acidosis, respiratory failure is present.

Effectiveness and Efficiency of the Respiratory

System

It is useful to distinguish between the effectiveness and the effi-

ciency of the respiratory system in maintaining arterial blood

gases. An arterial P

of 100 mm Hg, for example, indicates

effective oxygenation (more than adequate). Effective elimi-

nation of CO

is evidenced by an arterial P

of 40 mm Hg

if this Pa

is consistent with an acceptable acid-base status.

On the other hand, there is an obvious difference between two

patients who each have a Pa

of 100 mm Hg if the first

patient is breathing room air (F

= 0.21) and the other is

breathing 100% O

= 1.0). The first patient is exchanging

oxygen more efficiently from the atmosphere to the arterial

blood than the latter. Pa

determines the effectiveness of oxy-

genation; the relationship between inspired oxygen concentra-

tion and Pa

is a marker of efficiency.

measures the effectiveness of ventilation. Two

patients with Pa

of 40 mm Hg have equally effective ven-

tilation. If one of them needs a higher minute ventilation

(respired gas volume in 1 minute) than the other, however,

the patient requiring the higher minute ventilation is less

efficiently eliminating CO

than the one with the lower

minute ventilation. Thus the relationship between Pa

and

minute ventilation (

) reflects the efficiency of ventilation.

Measurement of the degree of inefficiency of oxygenation

and ventilation is discussed below.

Classification of Respiratory Failure

One classification of respiratory failure separates disorders

that affect the lungs (airways, alveolar spaces, interstitium,

and pulmonary circulation) from those that affect primarily

the nonlung components of the respiratory system.

Respiratory failure from diseases that directly affect the

lungs almost always has hypoxemia, but these patients may

or may not have hypercapnia depending on the type of dis-

ease and its severity. Examples include pneumonia, aspira-

tion of gastric contents, acute respiratory distress syndrome

(ARDS), pulmonary embolism, asthma, and interstitial lung

diseases.

Disorders of the nonpulmonary respiratory system usu-

ally cause hypercapnia plus hypoxemia. Examples include

diseases that cause weakness of the respiratory muscles, CNS

diseases that disrupt ventilatory control, and conditions that

affect chest wall shape or size, such as kyphoscoliosis. In

hypercapnic respiratory failure, the lungs may in fact be nor-

mal, so hypoxemia out of proportion to hypercapnia most

CHAPTER 12

likely indicates additional lung disease. An example might be

a patient with neuromuscular weakness from myasthenia

gravis who initially presents with hypercapnic respiratory

failure but who subsequently develops pneumonia from an

inability to clear tracheal secretions. At this point, the patient

may be considered to have hypoxemic respiratory failure in

addition to hypercapnia.



Hypercapnic Respiratory Failure

Patients with hypercapnic respiratory failure have an

abnormally high Pa

. Because CO

is elevated in the alve-

olar spaces, O

is displaced from the alveoli, and Pa

decreases. Thus these patients usually have both hypercap-

nia and hypoxemia unless the inspired gas is enriched with

oxygen. The lungs themselves may or may not be abnormal,

especially if the primary disease affects nonlung parts of the

respiratory system such as the chest wall, respiratory mus-

cles, or brain stem. On the other hand, chronic obstructive

lung disease not infrequently leads to hypercapnic respira-

tory failure, and some patients with severe asthma, end-

stage pulmonary fibrosis, and severe ARDS can develop

hypercapnia.



Physiologic Considerations

A. Alveolar Hypoventilation—In the steady state, a patient

produces CO

from metabolic processes each minute and

must eliminate that CO

from the lungs each minute. If the

minute output of CO

(

) exchanges into the gas-

exchanging spaces of the lungs, the fractional concentration

of CO

in the alveolar space is:

where F

ACO

is the fractional concentration of CO

in the

alveoli,

is the volume of air exchanging in the alveoli dur-

ing that minute (alveolar ventilation), and 0.826 adjusts for

temperature and water vapor. The sum of partial pressures of

individual gases equals the total pressure, so the fraction of

alveolar gas that is CO

also can be written as

where P

ACO

is the alveolar partial pressure of CO

and P

barometric pressure. Because P

ACO

cannot be measured,

is usually substituted. Substituting and rearranging

the preceding equation gives

where 863 includes factors that adjust for

at standard

temperature and pressure, dry (STPD); for

at body tem-

perature and pressure, saturated (BTPS); and for Pa

in mil-

limeters of mercury. For constant CO

output, the relationship

between Pa

and

describes a “ventilatory hyperbola” in

which Pa

and

are inversely related. Thus hypercapnia is

always means alveolar hypoventilation, and hypocapnia is syn-

onymous with alveolar hyperventilation. Because alveolar ven-

tilation cannot be measured, estimation of alveolar ventilation

can be made only by using Pa

and this formula.

B. Minute Ventilation—In a patient with alveolar hypoven-

tilation,

is reduced and Pa

is increased. Although

cannot be measured directly, the total amount of gas moving

into and out of the lungs each minute can be measured eas-

ily. This is defined as the minute ventilation (

, L/min).

the sum of the

(the portion participating in gas exchange)

and any wasted or dead-space ventilation (

and

–

Substituting,

= f × V

;

= f × V

; and

= f × V

—

where f is the respiratory frequency, V

is tidal volume, and

is dead-space volume:

where V

is the dead space:tidal volume ratio.

Substituting and rearranging gives

reflects the degree of ventilatory inefficiency of

the lungs. In a normal subject, V

is about 0.30, meaning

that about 30% of the minute ventilation is not participating

in gas exchange. In most lung diseases, the wasted proportion

increases, so V

rises. From the preceding formula

for a constant V

and

, the relationship of Pa

and

is described by a hyperbola transposed upward from

the hyperbola described by the relationship between Pa

and

. For different values of V

, these relationships are

described by a family of hyperbolic curves (Figure 12–1).

These curves are useful in estimating V

from measure-

ment of Pa

and

, or they can be used to determine the

change in

needed to cause a desired change in Pa

C. Mechanisms of Hypercapnia—Hypercapnia (alveolar

hypoventilation) occurs when

is abnormally low or when

is normal or high but V

is abnormally increased. It is

emphasized that hypoventilation refers to alveolar



×−

⎛

⎝

⎜

⎞

⎠

⎟

863



VV 1

=× −

⎛

⎝

⎜

⎞

⎠

⎟



V L/min Pa mm Hg V (L/min)

CO CO A

863

() ( )=××

P(mmHg)

ACO

VL/minSTPD)

V(L/minBTPS) 0.

ACO



(

8826



248



RESPIRATORY FAILURE

249

hypoventilation, and for this reason, hypercapnia may be

present even though the patient’s

is greater than normal if

is abnormally high or CO

output is increased (exer-

cise or other increased metabolic rate).

Alveolar dead space and V

are useful physiologic

concepts that may or may not have anatomic counterparts.

The trachea and airways are conduits for gas moving into

and out of the lungs during the respiratory cycle but do not

participate in gas exchange with the pulmonary capillary

blood. These spaces make up the anatomic dead space. An

artificial airway or part of a mechanical ventilator circuit that

is common to both inspiratory and expiratory pathways also

contributes to anatomic dead space. However, in patients

with lung disease, most of the increase in total dead space

consists of “physiologic dead space,” in which regional venti-

lation exceeds regional blood flow (ventilation-perfusion

[

Q] mismatching). While

Q mismatching usually is con-

sidered as a mechanism of hypoxemia rather than hypercap-

nia, it theoretically should cause elevated Pa

as well.

However, in all but severe instances of

Q mismatching,

hypercapnia stimulates increased ventilation, returning

to normal. Thus

Q mismatching does not usually

result in hypercapnia but in normocapnia with increased

As can be seen in Figure 12–1, increased

in the face of nor-

mal Pa

indicates increased V

—in this case an

increase in physiologic dead space.

Clinical Features

Acute hypercapnia acts largely on the CNS (Table 12–1).

Increased Pa

depresses the CNS, and the mechanism is

primarily through a fall of pH in the cerebrospinal fluid

(CSF) resulting from acute elevation of Pa

. Because CO

rapidly diffuses into the CSF, pH falls rapidly and severely

with acute hypercapnia. With chronic elevation of Pa

however, the increase in Pa

is present long enough for

plasma and CSF bicarbonate to increase in compensation for

the chronic respiratory acidosis. This explains why low pH

rather than absolute level of Pa

best correlates with

altered mental status and other clinical changes.

Symptoms of hypercapnia may overlap those of hypox-

emia. In addition, while hypercapnia stimulates ventilation in

normal subjects, hypercapnic patients may have either

decreased or increased minute ventilation depending on the

primary disorder leading to respiratory failure. Dyspnea,

tachypnea, and hyperpnea may be associated with hypercapnic

respiratory failure just as often as bradypnea and hypopnea.

The major differential distinguishing point is between

hypercapnic respiratory failure owing to lung disease and that

resulting from nonlung disorders. Patients with lung disease

often will have hypoxemia out of proportion to the degree of

hypercapnia. This can be assessed using the alveolar-arterial

difference. However, patients with nonlung problems

may have secondary hypoxemia because the effects of neuro-

muscular weakness, for example, contribute to atelectasis or

aspiration pneumonia. Disorders of the lungs—in contrast to

disorders of other components of the respiratory system—are

associated with increased V

, elevated

, and respira-

tory frequency. Patients with respiratory muscle weakness also

may be tachypneic. The effects of hypercapnia and hypoxemia

may mask neurologic disorders, overmedication with seda-

tives, myxedema, or head trauma. Alteration of mental status

may make it difficult to assess muscle strength, and the

strength of muscles in the extremities may not correlate with

respiratory muscle strength.



Figure 12–1. Alveolar ventilation (

) or minute ventila-

tion (

) as a function of Pa

for a constant CO

output of

200 mL/min. Increased Pa

(hypercapnia) is due to alve-

olar hypoventilation; decreased Pa

(hypocapnia) is the

same as alveolar hyperventilation. The other two curves

show the relationship of

and Pa

for normal dead

space:tidal volume ratio (V

= 0.3) and abnormal

dead space:tidal volume ratio (V

= 0.6).

Hypercapnia Hypoxemia

Somnolence

Lethargy

Coma

Asterixis

Restlessness

Tremor

Slurred speech

Headache

Papilledema

Anxiety

Tachycardia

Tachypnea

Diaphoresis

Arrhythmias

Altered mental status

Confusion

Cyanosis

Hypertension

Hypotension

Seizures

Lactic acidosis

Table 12–1. Clinical manifestations of hypercapnia and

hypoxemia.



CHAPTER 12

250



Hypoxemic Respiratory Failure

Hypoxemic respiratory failure is much more commonly

encountered than hypercapnic respiratory failure. These

patients have abnormally low Pa

but normal or low Pa

The latter distinguishes them from hypercapnic respiratory

failure, in which the primary problem is alveolar hypoventila-

tion. Outside of unusual environments in which the atmos-

phere has a severely reduced amount of oxygen, such as high

altitude or when oxygen has been replaced with other gases,

hypoxemic respiratory failure indicates disease affecting

the lung parenchyma or pulmonary circulation. Common

situations in which hypoxemia without elevated Pa

are

seen include pneumonia, aspiration of gastric contents, pul-

monary embolism, asthma, and acute respiratory distress

syndrome (ARDS).

Physiologic Considerations

A. Hypoxemia and Hypoxia—The term hypoxemia most

often denotes low P

in arterial blood (Pa

), but it can be

used to mean low capillary, venous, or pulmonary capillary

as well. It also may be used to signify low blood O

con-

tent or reduced saturation of hemoglobin with oxygen.

Hypoxemia should be distinguished from hypoxia, which is

decreased O

delivery to the tissues or the effects of

decreased tissue O

delivery. While hypoxia will result from

severe hypoxemia, hypoxia also can be a consequence of

decreased O

delivery owing to low cardiac output, anemia,

septic shock, or carbon monoxide poisoning, in which the

may be normal or even elevated.

B. Mechanisms of Hypoxemia—The physiologic mecha-

nism of arterial hypoxemia has important implications for

identifying the type of lung disease and the response to ther-

apy with oxygen or other treatment. Five distinct mecha-

nisms of hypoxemia can be identified, but these five can be

divided conceptually into two major groups (Table 12–2):

(1) decreased alveolar P

) and (2) increased influence

of venous admixture.

Mechanisms of arterial hypoxemia can be demonstrated

by analysis of the possible extremes of oxygenation of the

arterial blood. If desaturated systemic venous blood return-

ing to the lungs gained no oxygen during transit through the

lungs, the arterial blood would have the same oxygen con-

tent and partial pressure as the systemic venous blood (obvi-

ously a situation incompatible with life). Systemic venous

blood P

–

) determines the lower limit for arterial P

On the other hand, if all the desaturated venous blood pass-

ing through the lungs reaches equilibrium with gases in the

alveolar space, then Pa

equals P

. Thus alveolar P

determines the hypothetical upper limit for arterial P

Therefore, all possible values of Pa

must be between P

–

and P

Arterial hypoxemia is always the result either of decreased

or of a greater quantity of desaturated venous blood

(venous admixture) mixing with pulmonary capillary blood

(see Table 12–2). In some patients with hypoxemic respira-

tory failure, both these mechanisms play a role.

C. Decreased P

—The total pressure in the alveolar space

is the sum of P

, and P

. Because P

and

do not change appreciably, any increase in P

ACO

must

cause a fall in P

. Thus alveolar hypoventilation causes

decreased P

, which must result in decreased Pa

.The

alveolar gas equation, in simplified form, shows the relation-

ship between alveolar P

and P

Mechanism

ACO

) P

–a)

on 100%

(mm Hg)

†

Example

Alveolar P

↓ Inspired P

Hypoventilation

↓

↑

↓

Normal

‡

Normal

>550

High altitude.

Neuromuscular disease, obesity-

hypoventilation syndrome.

Venous admixture

Right-to-left shunt

Q mismatching

Diffusion limitation

Normal or low

Normal

Increased

<550

>550

ARDS, septal defect.

Pneumonia, asthma, COPD.

Alveolar proteinosis.

∗

–a)o

on room air. Normal <15–20 mm Hg.

†

Pao

while breathing 100% O

distinguishes right-to-left shunt from other mechanisms.

‡

–a)o

calculated using P

from alveolar gas equation.

Table 12–2. Mechanisms of hypoxemia.



RESPIRATORY FAILURE

251

where F

is the oxygen fraction of the inspired gas, P

is the

barometric pressure, and R is the respiratory gas exchange

ratio, the steady-state ratio of CO

entering and O

leaving

the alveolar space. In practice, Pa

substitutes for P

ACO

Because P

decreases with increased P

ACO

, alveolar

hypoventilation is a cause of hypoxemia (reduced Pa

The alveolar gas equation also indicates that hypoxemia

occurs if total barometric pressure is reduced, such as at high

altitude or if F

is low (such as when breathing a gas in

which some of the oxygen has been replaced by another gas).

In hypoxemia owing to reduced P

alone, the decrease in

approximately equals the fall in P

, and the alveolar-

arterial P

difference is normal.

D. Venous Admixture—The other causes of hypoxemia

result from increased amounts of deoxygenated venous

blood reaching the arterial blood without becoming fully

oxygenated by exposure to the alveolar gas. The alveolar-

arterial P

difference (P[

–a]

) is increased because

of increased venous admixture. During room air breath-

ing, P(

–a)

is normally between 10 and 20 mm Hg,

increasing with age and when the subject is in the

upright position.

During room air breathing, F

= 0.21; if R = 0.8, Pa

= 40 mm Hg, and Pa

= 55 mm Hg, then

= (0.21 × 713) – (40/0.8)

= 150 – 50 = 100 mm Hg

and

–a)

= 100 – 55 = 45 mm Hg

In this example, arterial hypoxemia is present (Pa

<60

mm Hg), and P(

–a)

is increased (>20 mm Hg). It should

be concluded that hypoxemia is due to one of the causes of

increased venous admixture.

1. Right-to-left shunt—If some systemic venous blood

bypasses the alveoli and then mixes with blood that did go

through the lungs, the resulting arterial mixture must have a

between P

and P

–

. The exact P

depends on the

proportion of blood that bypassed the lungs and the values of

and P

–

. This mechanism of hypoxemia is known as

right-to-left shunt, one of the mechanisms of hypoxemia

owing to increased venous admixture. Right-to-left shunt may

occur because of complete atelectasis of a lung or lobe in

which blood flow is maintained, or may be seen in patients

with congenital heart disease in which there is a right-to-left

shunt through a cardiac septal defect. Patients with ARDS may

have such severe pulmonary edema, localized atelectasis, or

alveolar collapse that right-to-left shunt occurs. Clues to the

presence of right-to-left shunt include severe hypoxemia while

breathing room air, only very small increases in Pa

when

supplemental oxygen is administered, a need to give an F

greater than 0.6 to achieve an acceptable Pa

, and a Pa

less than 550 mm Hg while breathing 100% O

.By conven-

tion, when Pa

is less than 550 mm Hg while breathing 100%

, right-to-left shunt is confirmed.

2. Ventilation-perfusion mismatching—A second cause

of hypoxemia owing to venous admixture is ventilation-

perfusion (

Q) mismatching. This mechanism, sometimes

termed

Q inequality, is the most frequent cause of hypox-

emia. In contrast to right-to-left shunt, hypoxemia in

mismatching does not result from venous blood completely

bypassing ventilated areas of lung. Rather, some regions of

the lungs have insufficient ventilation for the amount of

blood flow, whereas others have excessive ventilation for the

amount of regional blood flow. Pulmonary capillary blood

draining those parts of the lungs that have relative “hypoven-

tilation” is less well oxygenated and contributes to hypox-

emia. The effects of

Q mismatching on gas exchange are

often quite complicated, but practically any lung disease that

alters the distribution of ventilation or blood flow can result

Q mismatching. Thus hypoxemia owing to

Q mis-

matching is seen in asthma and other chronic obstructive

lung diseases in which variations in airway resistance distrib-

ute ventilation unevenly.

Q mismatching also contributes

to hypoxemia in pulmonary vascular diseases such as pul-

monary thromboembolism, in which the distribution of per-

fusion is altered. In contrast to right-to-left shunt, most

patients with

Q mismatching respond dramatically to sup-

plemental oxygen therapy. A clue to

Q mismatching, there-

fore, is that the Pa

can be relatively easily brought up to

acceptable values with the administration of supplemental

oxygen.

3. Diffusion limitation—The third mechanism of hypox-

emia owing to venous admixture is diffusion limitation of O

transfer. This is an unusual cause of hypoxemia, and the basis

for this mechanism is often misunderstood. Normally, there

is more than sufficient time for blood passing through the

lungs to become fully equilibrated with the gases in the alve-

oli. Rarely, however, pulmonary capillary blood passes so

quickly through the lungs that there is insufficient time for

pulmonary capillary P

to equilibrate with P

, resulting

in hypoxemia. This is a form of venous admixture because

the hypoxemia results from the influence of deoxygenated

venous blood. Diffusion limitation resulting in hypoxemia

theoretically can occur if P

is very low—so that the diffu-

sion of oxygen is slowed—or if the transit time for pul-

monary capillary blood is very short. There are few diseases

in which diffusion limitation for oxygen transfer is thought

to be a major cause of hypoxemia. For example, hypoxemia

in patients with pulmonary vascular disease may be due to

diffusion limitation because of decreased mean transit time,

especially if cardiac output is increased. Pulmonary alveolar

proteinosis—a disease in which the alveolar spaces are filled

with a homogeneous protein and lipid fluid—may slow dif-

fusion of oxygen sufficiently to cause hypoxemia. In

PFP

AO IO B

ACO

=×−



CHAPTER 12

252

most other lung diseases presenting with hypoxemia,

Q mismatching and right-to-left shunt explain hypoxemia

more often than diffusion limitation.

Clinical Features

Manifestations of hypoxemic respiratory failure are the result

of a combination of features of arterial hypoxemia and tissue

hypoxia (see Table 12–1). Arterial hypoxemia increases ven-

tilation by stimulation of carotid body chemoreceptors, lead-

ing to dyspnea, tachypnea, hyperpnea, and usually,

hyperventilation. The degree of ventilatory response depends

on the ability to sense hypoxemia and the capacity of the res-

piratory system to respond. In hypoxemic patients with

severe lung disease or ventilatory limitation, there may be lit-

tle or no increase in ventilation and absence of hyperventila-

tion. In patients who lack carotid body function, there will be

no ventilatory response to hypoxemia. There may be

cyanosis, especially marked in the distal extremities but also

centrally prominent around the mucous membranes and

lips. The degree of cyanosis depends on the hemoglobin con-

centration and the patient’s state of perfusion.

Other effects attributable to hypoxemia are due to inade-

quate supply of oxygen to the tissues, or hypoxia. Hypoxia

causes a shift to anaerobic metabolism, which is accompa-

nied by generation of lactic acid. Increased blood lactic acid

may further stimulate ventilation. Mild early hypoxia may

cause impaired mental performance, especially for complex

tasks or abstract thinking. More severe hypoxia can cause

much more severe alteration of mental status, including

somnolence, coma, seizures, and permanent hypoxic brain

damage. Sympathetic nervous system activity is increased,

and this contributes to tachycardia, diaphoresis, and systemic

vasoconstriction, leading to hypertension. More severe

hypoxia, however, can lead to bradycardia, vasodilation, and

hypotension as well as myocardial ischemia, infarction,

arrhythmias, and cardiac failure.

Manifestations of hypoxemic respiratory failure are mag-

nified in the presence of impaired tissue oxygen delivery.

Patients with reduced cardiac output, anemia, or circulatory

abnormalities can be expected to have global and regional

tissue hypoxia at less severe degrees of hypoxemia. Examples

include the increased risk of myocardial ischemia from

hypoxemia in a patient with coexisting coronary atheroscle-

rosis or a patient with hypovolemic shock who shows evi-

dence of lactic acidosis in the presence of mild arterial

hypoxemia.



Oxygen Delivery & Tissue Hypoxia

Adequate O

delivery to the tissues is the most important

function of the respiratory system, and this aspect requires

normal function of the lungs, heart, and circulation.

Recognition and treatment of compromised systemic O

delivery should be primary goals in management of respira-

tory failure in addition to correcting abnormalities of arterial

blood gases.

Physiologic Considerations

A. Oxygen Delivery—Systemic O

delivery is the product of

arterial O

concentration (mL O

/L blood) and cardiac out-

put (L/min). This calculation does not help to determine

whether the blood and O

are distributed to organs in pro-

portion to their needs, so even normal or high O

delivery

may be insufficient under certain conditions such as shock,

sepsis, or end-stage liver disease.

delivery (mL/min) = arterial O

content (Ca

mL O

/L blood) × cardiac output (

Q, L/min)

where Ca

, mL O

/L blood = [O

saturation × hemoglobin

(g/dL)

× 1.34 mL O

/g hemoglobin + Pa

(mm Hg) × 0.003

mL O

/mm Hg/dL] × 10.

In normal subjects at rest, normal arterial O

concentra-

tion is about 200 mL O

/L blood (O

saturation 97%, hemo-

globin 15 g/dL of blood, Pa

100 mm Hg). Resting cardiac

output is about 5 L/min, resulting in normal O

delivery =

1000 mL O

/min.

B. Causes of Decreased Oxygen Delivery—Factors

included in the formula for O

delivery can be examined to

identify pathologic states that result in potentially decreased

delivery. First, arterial O

concentration can be reduced as

a result of decreased O

saturation of hemoglobin from arte-

rial hypoxemia (decreased Pa

) or a rightward-shifted oxy-

hemoglobin dissociation curve (eg, acidemia, hyperthermia,

or hemoglobinopathy). Anemia is an important factor

because O

concentration is largely the product of hemoglo-

bin concentration and O

saturation. A decrease in hemoglo-

bin from 12 to 8 g/dL decreases O

concentration and O

delivery by 33%—considerably more than most changes in

or O

saturation. Carbon monoxide, because of its high

affinity for hemoglobin, displaces O

and reduces arterial O

concentration. In addition, carbon monoxide shifts the oxy-

hemoglobin curve leftward, which, although it tends to

increase O

concentration at any given Pa

, causes problems

in unloading O

at the tissue level.

Cardiac output depends on multiple factors, including

adequate systemic venous return, right and left ventricular

function, pulmonary and systemic resistance, and heart rate.

Even in the absence of underlying intrinsic heart disease,

patients with respiratory failure may have impaired or

reduced cardiac output. Hypoxemia and acidosis have

adverse effects on myocardial contractility or may cause

tachycardia, bradycardia, or myocardial infarction. There is

evidence that myocardial depression can be seen in conjunc-

tion with sepsis and septic shock, mediated through products

of microorganisms, patient-produced cytokines, or other

factors. Mechanical ventilation with positive pressure inter-

acts in a number of ways with the heart and circulation.

Although much of the decrease in cardiac output during

positive-pressure ventilation is due to diminished systemic

venous return, left ventricular diastolic compliance is

impaired, pulmonary vascular resistance is increased, and



RESPIRATORY FAILURE

253

right ventricular afterload increases. The degree and signifi-

cance of interaction vary with the type and severity of respi-

ratory failure and the type of mechanical ventilation.

C. Assessment of Oxygen Delivery—The assessment of

adequacy of O

delivery remains a topic of debate. In normal

subjects, total O

consumption of the body is independent of

delivery over a wide range. Increasing or decreasing O

delivery (except at extremely low rates) by changing cardiac

output or hemoglobin does not result in a parallel increase or

decrease in O

consumption. However, in some patients with

septic shock, ARDS, or other critical illness, O

consumption

may become functionally dependent on O

delivery even

when O

delivery is in the normal range. This finding has

been taken to indicate that adequate O

delivery in a given

patient cannot be assumed even when O

delivery is normal;

an increase in O

consumption in response to increased O

delivery above normal indicates that the original O

delivery

was in fact inadequate. To explain this finding, investigators

have proposed that distribution of blood flow in the periph-

eral circulation is poorly matched to O

requirements of indi-

vidual organs—a form of “distributive shock.” In some

studies, lactic acidosis has been associated with O

delivery

dependency; this supports the concept of inadequate tissue

oxygenation. In other studies, organ dysfunction is cited as

evidence of hypoxia when O

consumption is no longer inde-

pendent of O

delivery. On the other hand, some investigators

believe that these findings are artifacts of measurements or do

not reflect the responses of most patients with these disor-

ders. There have been a few studies suggesting that increasing

delivery has resulted in improved outcome from septic

shock. Several other investigators, however, have been unable

to find any difference in patient survival in critical illness by

empirically increasing O

delivery. It is highly likely that this

dependence of O

consumption on O

delivery is patient-

specific. Therefore, attention should be paid to trying to iden-

tify evidence of inadequate O

delivery by monitoring renal,

hepatic, cardiac, and other organ system functions.



Temperature & Blood Gases

The blood gas analyzer maintains the sample of blood at

37°C while Pa

,Pa

, and pH are determined. For a given

quantity of O

and CO

in an aliquot of blood, Pa

and

will change if the temperature of the blood changes.

When the sample is cooled, Pa

and Pa

decrease; when it

is warmed, Pa

and Pa

increase.

In patients with hypothermia or hyperthermia, some lab-

oratories report “temperature-corrected” blood gases, that is,

what the Pa

and Pa

would be if measured at the

patient’s actual temperature. These corrections are deter-

mined empirically and are derived easily from tables and

nomograms or are displayed automatically by the analyzer.

However, temperature correction of a patient’s blood gas

results may lead to an incorrect clinical interpretation unless

they are compared with temperature-corrected normal blood

gas values. For example, in a normal animal made hypothermic,

temperature-corrected Pa

and Pa

decrease, and because

HCO

–

remains constant, pH increases. If compared with

customary normal values measured at 37°C, interpretation

of temperature-corrected Pao

,Pa

, and pH would lead to

an erroneous conclusion of hypoxemia and respiratory alka-

losis. One approach to avoiding this problem is to use refer-

ence normal values at each temperature for comparison.

A preferable method is to report all blood gas values at

the 37°C at which the blood is analyzed regardless of the

patient’s actual temperature. These results can then be com-

pared with normal values for Pa

,Pa

, and pH deter-

mined at 37°C. Interpretation will be correct for both

hypoxemia and acid-base status. Temperature correction of

blood gases is unnecessarily complex and may be misleading

if steps are not taken to provide corrected normal values.

Glenny R et al: Gas exchange in health: Rest, exercise, aging. In:

Roca J, Rodriguez-Roisin R, Wagner PD (eds), Pulmonary and

Peripheral Gas Exchange in Health and Disease. New York:

Marcel Dekker, 2000.

Levy MM: Pathophysiology of oxygen delivery in respiratory fail-

ure. Chest 2005;128:547S–53S. [PMID: 16306052]

Pierson DJ: Indications for mechanical ventilation in adults with

acute respiratory failure. Respir Care 2002;47:249–6. [PMID:

11874605]

Rice TW et al: Comparison of the Sp

ratio and the

ratio in patients with acute lung injury or acute res-

piratory distress syndrome. Chest 2007;132:410–7. [PMID:

17573487]

West JB, Wagner PD: Ventilation, blood flow, and gas exchange. In:

Murray JF, Nadel JA (eds), Textbook of Respiratory Medicine,

3rd ed. Philadelphia: Saunders, 2000.

TREATMENT OF ACUTE RESPIRATORY FAILURE

Respiratory failure is treated by a combination of specific

treatment directed at the underlying compromise of the res-

piratory system plus supportive care of oxygenation and ven-

tilation. The general principles of support are similar

regardless of the type of respiratory system disorder.



Physiologic Basis Of Treatment

Hypercapnic Respiratory Failure

Because hypercapnia is synonymous with alveolar hypoven-

tilation, supportive care restores alveolar ventilation to nor-

mal until the underlying disorder can be corrected. Alveolar

ventilation sometimes can be improved by establishing an

effective airway—suctioning to remove secretions, stimula-

tion of cough, postural drainage, or chest percussion—or by

establishing an artificial airway with an endotracheal tube or

tracheostomy.

Mechanical assistance may be necessary to maintain the

desired alveolar ventilation until the primary problem is cor-

rected. Although the mechanical ventilator theoretically can

provide any desired amount of ventilation, care should be



CHAPTER 12

254

taken to correct hypercapnia judiciously in patients with

chronic hypercapnia. This is because correction of Pa

normal in these patients can result in severe, life-threatening

alkalemia owing to their elevated plasma bicarbonate levels

as compensation.

Hypoxemia is seen often in patients with hypercapnic res-

piratory failure—especially those with lung disease—and

administration of supplemental oxygen is often necessary. In

some patients with hypercapnia, however, supplemental oxy-

gen may be hazardous if not titrated carefully. This group of

patients with chronic lung disease (either obstructive or

restrictive) or chest wall compromise (eg, kyphoscoliosis)

appears to have particular insensitivity to hypercapnia and

depend on hypoxemia to stimulate ventilation. If sufficient

oxygen is given to overcome hypoxemia, ventilatory drive

may be blunted subsequently, and the patient’s hypercapnia

may worsen.

Patients with hypercapnic respiratory failure owing to

sedative drug overdose or botulism—and most patients with

chest trauma—will improve with time, and treatment is

largely supportive. Some primary diseases associated with

hypercapnia require specific treatment, including myasthe-

nia gravis, electrolyte abnormalities, obstructive lung disease,

obstructive sleep apnea, and myxedema.

Hypoxemic Respiratory Failure

Oxygen supplementation is the most important therapy for

hypoxemic respiratory failure. In severe disorders such as

ARDS, mechanical ventilation, positive end-expiratory pres-

sure (PEEP), and other types of respiratory therapy may be

necessary. Although not a feature of most cases, hypercapnia

may develop because the high work of breathing leads to res-

piratory muscle fatigue. Attention to oxygen transport is

important, and severe anemia should be corrected and ade-

quate cardiac output maintained. The underlying disease

leading to hypoxemic respiratory failure must be addressed,

especially if pneumonia, sepsis, or other cause is identified.

Treatment may include diuretics, antibiotics, and bron-

chodilators as well as other measures.

In some patients with nonuniform lung disease,

dependent positioning of uninvolved or less involved lung

areas may improve oxygenation. Gravity and the weight of

the lungs increase perfusion and ventilation to dependent

lung regions. Patients with hemoptysis or heavy respiratory

secretions, however, should not be placed in this position

because of the likelihood of aspiration of blood or secre-

tions into uninvolved areas. In ARDS with diffuse noncar-

diogenic pulmonary edema, there has been considerable

interest in placing the patient prone. Prone ARDS patients

appear to have less tendency of dependent lung regions to

collapse as well as smaller areas of the lungs compressed by

the heart or abdominal contents. In some patients,

improvement in arterial hypoxemia is transient after turn-

ing from supine to prone, but in others the effects persist

for at least several hours.



Airway

When upper airway obstruction is the patient’s only prob-

lem, prompt restoration of an adequate airway is all that is

required to reverse respiratory failure. In all patients, estab-

lishment of the airway is essential for ventilation, oxygena-

tion, and delivery of respiratory medications.

Upper Airway Obstruction

Primary upper airway obstruction should be considered in all

patients with respiratory difficulty—but especially if they

present with any of the following: head and neck trauma, sus-

pected malignancy of the larynx or trachea, acute dyspnea

with wheezing (ie, inspiratory, expiratory, or both), dyspha-

gia, neurologic disease affecting motor or sensory function,

speech difficulty, masses in the neck owing to thyroid enlarge-

ment or lymphadenopathy, or pain, infection, or inflamma-

tion of the pharynx, larynx, or trachea. Patients with asthma

or chronic obstructive pulmonary disease (COPD) also may

have upper airway obstruction from tracheal or subglottic

stenosis if there is a history of recent or remote endotracheal

intubation or tracheostomy. Acute respiratory distress, espe-

cially in the elderly and in children, should arouse suspicion

of a foreign body in the airway. A frequent intermittent cause

of upper airway obstruction is seen in obstructive sleep apnea

syndrome (see below), in which obstruction occurs during

certain stages of sleep. It is important to consider this not

uncommon problem as a complicating factor in patients with

respiratory failure from other causes.

Natural Airway

The normal natural airway permits speech, humidifies

inspired gas, and protects against aspiration and infection;

coughing is effective, and the mucociliary function of the

trachea is maintained. When a decision is made to insert an

artificial airway, the benefits and risks of the endotracheal

tube must outweigh the benefits of the natural airway

(Table 12–3). In patients with severe upper airway obstruc-

tion, the choice is usually easy. In other patients with acute

respiratory failure, the decision rests primarily on whether

oxygen, respiratory medications, and respiratory therapy

via the natural airway will be adequate or whether an arti-

ficial airway would be preferable. A trial of aggressive treat-

ment before intubation often provides useful information.

Guidelines (Table 12–4) for selecting patients who need

endotracheal intubation may be helpful, but clinical assess-

ment of the response to therapy is usually more so.

Noninvasive positive-pressure ventilation does not require

endotracheal intubation and is an important alternative in

patients who meet criteria.

Endotracheal Tubes

Endotracheal tubes are usually made of relatively stiff plastic,

with soft, low-pressure, easily deformable inflatable tracheal



RESPIRATORY FAILURE

255

cuffs. Skilled practitioners should insert these tubes with

attention to prevention of aspiration of gastric contents; ade-

quate oxygenation during the procedure; avoidance of

trauma to the mouth, tongue, nose, epiglottis, and vocal

cords; and selection of an tube of proper size for the patient.

Experience with sedation using opioid analgesics or rapid-

acting sedatives (eg, benzodiazepines such as midazolam)

and muscle relaxants should be available for intubation of

awake patients.

Nasotracheal intubation offers greater patient comfort,

less severe positioning of the head and neck during place-

ment, and better stabilization of the tube. However, the route

of the tube and its smaller size sometimes can complicate

suctioning and weaning from a mechanical ventilator. The

smaller radius of curvature of the nasotracheal route has

been linked to higher tube resistance compared with the

same tube passed orally, but this finding has been ques-

tioned. The nasotracheal route generally is appropriate in

patients who require intubation for reasons unrelated to lung

disease, for example, sedative drug overdose. Otherwise, oro-

tracheal intubation should be used, especially if a larger tube

diameter is needed to facilitate airway suctioning or for

fiberoptic bronchoscopy. Placing an orotracheal tube

requires a laryngoscope for inspection of the vocal cords.

Laryngoscopes with straight or curved blades to hold the

tongue and other structures away are used most often, but

fiberoptic laryngoscopes have proved extremely useful for

difficult intubations or special circumstances.

Smaller-diameter endotracheal tubes impose less risk of

trauma to the vocal cords and may be more comfortable for

the patient. Larger tubes provide better access for tracheal

suctioning, delivery of medications, and fiberoptic bron-

choscopy and generally will create an adequate seal between

the tube and the trachea with lower cuff pressure. Weaning

from mechanical ventilation may be easier with a larger tube

because of lower airflow resistance. Most women can accom-

modate endotracheal tubes of 7.5–8 mm in inner diameter;

men generally will accept tubes of 8.5 or 9 mm in inner

diameter. If fiberoptic bronchoscopy is a consideration, a tube

with at least 8 mm in inner diameter is required, and 8.5 mm

is preferable.

Care of the Artificial Airway

The intubated patient must be suctioned frequently because

both cough and the mucociliary clearance mechanism are

impaired. The frequency of suctioning depends on the

amount and nature of secretions. Although the artificial air-

way becomes rapidly colonized with bacteria, suctioning

should be done using sterile technique to prevent introduc-

tion of additional organisms. A sealed system (ie, suction

catheter contained within a sheath in-line with the endotra-

cheal tube) facilitates frequent suctioning and may minimize

nosocomial contamination. For some patients, disposable

suction catheters are more effective—especially catheters with

bent tips that can be directed into the right or left main

bronchus as desired. It has become common practice to instill

small aliquots of sterile normal saline into the endotracheal

tube to facilitate suctioning of secretions, but this should be

done only if increased amounts of secretions are obtained.

Hypoxemia and tracheal trauma are the most common

complications of suctioning. Minimal negative pressure

should be used, and the suction catheter should be intro-

duced gently. During suctioning, Pa

may fall rapidly, par-

ticularly if the patient is receiving high concentrations of

inspired O

and suctioning is performed for more than

10–15 seconds. In patients with focal infiltrates or known

collections of secretions in particular parts of the tracheo-

bronchial tree, selective bronchial suctioning can be tried. By

using bent-tip catheters and by positioning the patient’s

body or head properly, successful suctioning of the left main

bronchus, for example, often can be accomplished.

Risk Benefits

Trauma of insertion

Oro- or nasopharyngeal trauma

due to chronic pressure

Tracheal damage (erosion,

tracheomalacia)

Impaired cough response

Increased aspiration risk

Impaired mucociliary function

Increased infection risk

No speech

Increased resistance and work of

breathing

Bypasses upper airway

obstruction

Route for oxygen and other

medications

Facilitates positive-pressure

ventilation and PEEP

Route for respiratory medications

Facilitates suctioning of secretions

Fiberoptic bronchoscopy route

Table 12–3. Risks and benefits of the artificial airway.

Table 12–4. Indications for intubation and mechanical

ventilation.∗

Physiologic

Hypoxemia persists after oxygen administration

>55 mm Hg with pH <7.25

Vital capacity <15 mL/kg with neuromuscular disease

Clinical

Altered mental status with impaired airway protection

Respiratory distress with hemodynamic instability

Upper airway obstruction

†

High volume of secretions not cleared by patient, requiring

suctioning

∗

These are guidelines that must take into account the patient’s

clinical status and other factors.

†

Consider need for tracheostomy if obstruction is above trachea.



CHAPTER 12

256

In order to provide positive-pressure ventilation, the

inflated tracheal cuff of the endotracheal tube exerts pressure

on the interior of the trachea to create an effective seal.

Almost all endotracheal and tracheostomy tubes incorporate

low-pressure, high-volume cuffs made of highly compliant

rubber or plastic. These cuffs provide a tracheal tube seal

after inflation to a relatively low pressure. If higher than nor-

mal pressures are needed to achieve a seal, damage to the tra-

chea may occur, including erosion, inflammation, softening

of the cartilage rings with tracheal dilation (tracheomalacia),

and hemorrhage. Endotracheal cuff pressure must be moni-

tored to anticipate these complications and to make certain

that the smallest effective pressure and volume are used. The

best way to do this is to slowly inflate the cuff with air, using

a small syringe, until there is minimal leak around the cuff

with inspiration and adequate tidal volume and ventilation.

The pressure read on a manometer and the amount of air put

into the cuff are recorded. The desired cuff pressure is the

smallest possible that will maintain an adequate seal between

the tube cuff and the trachea, and the pressure is ideally less

than 15 cm H

O. The incidence of tracheal complications

rises when cuff pressures exceed 20–25 cm H

O for pro-

longed periods. An automated system for maintaining opti-

mal cuff pressure is under investigation.

Complications from Endotracheal Tubes

Complications from endotracheal tubes may be classified as

early and late. Early complications are due to the trauma of

tube insertion or to malpositioning of a tube into a main

bronchus or in the esophagus. Confirmation of the tube’s

position by physical examination and chest x-ray are essen-

tial. The tip of the tube should be in the center of the trachea

and 3–5 cm above the carina, but head flexion or extension

can cause 1–5 cm of movement of the endotracheal tube tip

on chest x-ray. If the carina is difficult to see, the position of

the tip of the tube relative to the aortic arch on chest x-ray is

helpful. Analysis of expired CO

gives reliable assurance of

tracheal rather than esophageal placement; fiberoptic bron-

choscopy also can be valuable for this purpose in selected

patients.

Unplanned extubations are a complication of endotra-

cheal intubation, occurring in 3–10% of placements. The

endotracheal tube should be secured carefully, and the

patient should be educated about the need for and impor-

tance of the tube. Agitation and patient movement are asso-

ciated with extubation, and sedation and physical restraints

should be used as necessary. Several studies have shown that

unplanned extubations led to reintubation in about half of

patients, but some patients can be observed carefully for res-

piratory distress and deterioration of arterial blood gases

rather than immediate reintubation. Factors predictive of a

high likelihood for reintubation include severity of the

underlying disease, a high minute ventilation requirement

during the preceding 24 hours, a high F

requirement, and

altered mental status.

Aspiration of oral secretions or refluxed gastric secretions

is a common problem in intubated patients. Contrary to

common belief, the inflated tracheal cuff does not reliably

protect against aspiration of secretions around the tube.

Endotracheal tubes with a port above the cuff reduce the

incidence of ventilator-associated pneumonia when the port

is connected to continuous suction.

A randomized trial of orotracheal compared with nasotra-

cheal intubation showed that there was a slightly greater fre-

quency of radiographic sinusitis (by CT scan) with

nasotracheal intubation. The incidence was relatively high in

both groups (30% nasotracheal and 22% orotracheal), how-

ever, indicating that this was a frequent complication in all

patients needing endotracheal intubation for more than 7 days.

Most late complications arise from prolonged pressure of

the tube against anatomic structures. The curve of the tube

puts maximum pressure on the side of the mouth, the palate,

and the posterior pharynx (from oral intubation) or the

nasal turbinates and the posterior pharynx (nasal intuba-

tion). The greatest pressure is exerted on the vocal cords, the

narrowest part of the passage. Most late complications are

due to laryngeal trauma, followed by the development of

glottic injury and subglottic stenosis. The incidence of sub-

glottic stenosis is estimated to be less than 5% of patients

intubated for more than 10–14 days. A study of patients with

translaryngeal intubation (orotracheal or nasotracheal)

found that the extent of injury estimated by laryngoscopy

was not predictive of late complications. The duration of

translaryngeal intubation did not influence the ability of the

larynx to heal without future ill effects.

Despite the inability to correlate late complications with

the duration of intubation, early tracheostomy (within 5–7

days) is recommended if prolonged intubation is anticipated.

Translaryngeal intubation for up to 10 days does not usually

cause an appreciable increase in complications, and tra-

cheostomy is preferred if intubation for 21 days or more can

be anticipated. This conclusion is based on a similar inci-

dence of complications from translaryngeal intubation and

tracheostomy. Thus translaryngeal intubation for as long as

21 days may be acceptable unless a longer need for an artifi-

cial airway is expected or whenever there is reason to suspect

greater potential laryngeal trauma (eg, patient movement,

malnutrition, and local or systemic infection). In such cases,

earlier tracheostomy is advised, especially in view of studies

that demonstrate reduced complications when tracheostomy

is performed at 5 days.

Tracheostomy, of course, prevents or avoids laryngeal

injury but does not prevent tracheal injury from the tra-

cheostomy cuff. Attention to cuff pressure and cuff volume

is essential for tracheostomy tubes as well as endotracheal

tubes. Tracheostomy may be contraindicated if the patient

has a bleeding disorder, local infection, or neck mass.

Percutaneous tracheostomy with and without fiberoptic

bronchoscopic assistance potentially reduces complications.

Not only is the timing of tracheostomy controversial,

but so is the role of tracheostomy itself. This is based on