You are on page 1of 31

2000 Lippincott Williams & Wilkins, Inc.

Volume 38(1), Winter 2000, pp 25-53


Physiology of the Lateral Decubitus Position and One-Lung Ventilation
[Thoracic Anesthesia]

Dunn, Peter F. MD

One-lung ventilation (OLV) remains one of the more challenging techniques of daily
anesthetic practice. Twenty-eight years ago, John H. Kerr from Oxford reviewed the
physiological aspects of one-lung anesthesia in this journal. 1 Subsequently, Kerr and
associates 2,3 published their scientific observations of ventilation and oxygenation
during OLV. Their work, and the work of many others before them, established the
foundations upon which our knowledge of the physiology of OLV has been building over
the last 3 decades. Incorporating the findings from the studies over this period of time
into clinical practice has made OLV a safe technique in the majority of patients.
However, in some patients, severe hypoxemia may occur. In this review, the physiology
of the lateral decubitus position as it pertains to OLV, the method of performing OLV, and
the therapeutic options and their rationale for treating hypoxemia during OLV will be
discussed.

Physiology of Lateral Decubitus Position

When the anesthetized patient is moved from the supine to the lateral decubitus
position, significant alterations in the matching of ventilation and perfusion occur. A brief
review of the changes in ventilation and lung perfusion follows and is essential for the
understanding of the perturbations seen with OLV.

Ventilation
Closed Chest

In the anesthetized-paralyzed patient in the lateral decubitus position, approximately


55% of each tidal volume is delivered to the nondependent lung. 4 General anesthesia
with paralysis results in both lungs moving down on the pressure-volume curve
compared with the awake state (Fig 1). 5 The dependent lung moves from an initial,
favorable position on the steep portion of the curve to a position on the lower, flat
portion of the curve, implying a reduction in functional residual capacity (FRC) and
compliance. The change is due to the dependent lung being compressed (with loss of
FRC) and its excursion limited (decrease in compliance) by (1) the mediastinum, (2) the
cephalic movement of the abdominal organs via the flaccid diaphragm, and (3) the
exaggerated flexed (jackknife) position, with or without chest rolls to free the axillary
contents. The nondependent lung moves from an initially high, flat position to the lower,
steep part of the curve. This results in an improved compliance for the nondependent
lung. Although the anesthetized state results in atelectasis 6 and a net loss of total FRC,
7 the nondependent lung's FRC may actually increase in the lateral position 8 to
approximately 1.5 that of the dependent lung. 4 These are the major alterations
determining the distribution of ventilation in the anesthetized patient in the lateral
position. Although gravity may have an effect on the distribution of ventilation during
spontaneous ventilation in the lateral decubitus position, 9 it has no significant effect
during positive pressure ventilation. 10
Figure 1. Schematic diagram shows the distribution of ventilation and the changes that
occur in the lateral position in the (A) awake and (B) anesthetized patient. With the
induction of anesthesia, the nondependent lung moves from an initial, flat,
noncompliant portion of the pressure-volume curve to a steep, more compliant part.
The dependent lung moves from the steep, compliant part of the curve to a lower, flat,
less compliant part. This results in the majority of the tidal ventilation occurring in the
nondependent lung in the anesthetized patient in the lateral position. (From
Benumof.5With permission.)

Open Chest

Upon opening the nondependent hemithorax, there is an increase in the compliance and
FRC in the nondependent lung and a decrease in compliance and FRC of the dependent
lung with two-lung ventilation. 5 This results in a further increase in the amount of each
tidal volume delivered to the nondependent lung and a worsening of ventilation-
perfusion (V/Q) matching.

Pulmonary Blood Flow

Maldistribution of pulmonary blood flow is the most common cause of impaired


oxygenation of arterial blood. 9 The major determinants of the distribution of blood flow
include hypoxic pulmonary vasoconstriction (HPV), gravity, and nongravitational factors.

Hypoxic Pulmonary Vasoconstriction

The matching of perfusion to ventilated areas of the lung is paramount to the


maintenance of adequate oxygenation and the removal of carbon dioxide (CO 2). The
primary, intrinsic mechanism of the lung to maintain this balance was first described in
1946 and has been termed hypoxic pulmonary vasoconstriction (HPV). 11 HPV is the
constriction of the small arterioles (less than 500 m), and to a lesser extent the
capillaries and venules, in response to alveolar hypoxia, a response that is unique to the
pulmonary system. The stimulus for HPV is a combination of both the alveolar partial
pressure of oxygen (alveolar PO2, PaO2) and the mixed venous PO2 (PvO2). The stimulus
PO2 (PsO2) may be estimated using the following equation 12 :EQUATION
Equation U1

The alveolar PO2 is the predominant stimulus, but under conditions such as atelectasis,
the mixed venous PO2 may have a greater role. When mixed venous PO2 is high (e.g.,
sepsis), the HPV response is decreased. HPV results in the diversion of blood flow away
from hypoxic or atelectatic alveoli to those that are better ventilated. The HPV response
is most effective when the size of the atelectatic or hypoxic segment is intermediate
(between 20% and 80% of the lung) (Fig 2). 13 The percentage of lung that may be
hypoxic during one-lung ventilation (30% to 70%) falls within this range of effective
HPV. 13 During one-lung anesthesia, the HPV response may result in a 50% reduction in
the expected shunt. 14 The HPV response can be modified by many factors, including
the presence of pulmonary disease, 15 pharmacologic agents (see below), acid-base
status (acute respiratory or metabolic alkalosis blunts HPV; metabolic acidosis enhances
HPV), surgical manipulation 16 and hemodynamic changes.
Figure 2. Effect of hypoxic pulmonary vasoconstriction (HPV) on alveolar partial
pressure of oxygen (PaO2). The decrease in PaO2 when the amount of lung that is
hypoxic is intermediate (30%70%), as occurs with one-lung ventilation, is reduced in
the presence of HPV compared to values obtained when there is no HPV. FiO2 inspired
oxygen concentration. (From Benumof.13With permission.)

Benumof and Wahrenbrock 17 demonstrated in a canine, open-chest model that the HPV
response was abolished when left atrial pressure reached 25 mm Hg. Likewise, increases
in cardiac output and pulmonary arterial pressure attenuate HPV. The attenuation of HPV
with an increased cardiac output may be due to an increase in pulmonary arterial
pressure and an increase in mixed venous PO2. In the presence of a low cardiac output
(e.g., hemorrhage), the HPV response may be decreased due to alveolar vessel
compression and high pulmonary vascular resistance in the ventilated lung. 18 In
addition to lowering PaO2 by attenuating HPV, low cardiac output states will result in a
low PaO2 in the presence of physiological shunts greater than 5% of the cardiac output
(Fig 3). 19 The lung volume can have deleterious effects on HPV. If the delivered tidal
volume and the total positive end-expiratory pressure (PEEP; intrinsic and applied, see
below) result in overdistension of the ventilated alveoli, blood flow may be diverted to
the atelectatic or hypoxic alveoli, attenuating the HPV response and worsening V/Q
mismatch.

Figure 3. Relationship between cardiac output and arterial partial pressure of oxygen
(PaO2) for different shunt values (5%, 10%, 20%, and 30% of cardiac output). Hb =
hemoglobin; Q = perfusion. (From Kelman.19With permission.)

Gravity

The classic teaching is that the principal determinant of blood flow is gravity dependent
and results in the creation of distinct zones of pulmonary blood flow ( Fig 4). 20 Zone 1
occurs when the airway pressure exceeds the pulmonary arterial pressure, and no
perfusion through the compressed capillary bed of those alveoli occurs. However, a small
amount of blood flow does occur in zone 1 through the extra-alveolar vessels located in
the septae between alveoli. 21,22 It has also been shown that these vessels participate
in gas exchange. 22 Below this level, blood flow increases progressively as pulmonary
arterial pressure increases. In zone 2, the alveolar pressure exceeds pulmonary venous
pressure, and thus flow at this level is determined by the arterial-alveolar pressure
difference, a phenomenon termed the waterfall effect. 23
Figure 4. Schematic diagram of the distribution of blood flow in the lung based on
pressures affecting the capillaries. The external pressure is determined by the alveolar
pressure (PA), and the pressure within the vessel is determined by the hydrostatic
pressure gradient established due to the effect of gravity. (From West.20With
permission.)

Zone 3 is where the pulmonary blood flow is determined by the gradient between
arterial and venous pressures, as it is in other vascular beds. There is a generalized
increase in blood flow down the gravitational gradient through zones 2 and 3 to the
most dependent portion of the lung, where flow decreases. Although not depicted in
Figure 4, this zone has been called zone 4 and has been demonstrated in human
volunteers in the sitting and lateral positions. 24 The cause of the decrease flow in zone
4 has been attributed to increased interstitial pressure and increased resistance of the
extra-alveolar vessels. 25

The gravitational model of pulmonary blood flow has been supported by studies in
animals and humans. 24,26 For example, in a study performed by Landmark and
colleagues, 24 the effect of gravity on pulmonary blood flow and V/Q was evaluated on
healthy, adult volunteers in the awake and anesthetized states. At FRC, the distribution
of pulmonary blood flow was determined by Xenon (133Xe) injection and the regional
ventilation was determined by 133Xe gas inspiration. The relative regional perfusion,
perfusion per alveolus (assuming uniform size of each alveolus at total lung capacity),
and V/Q ratios were determined in sitting, supine, and lateral positions. In Figure 5, the
perfusion per alveolus is plotted against vertical distance down the lung in each position
and in both awake and anesthetized states. There was a generalized increase in
perfusion down the gravitational gradient in each position that was not affected by
anesthesia and paralysis. There was also evidence of a decrease in perfusion in the most
dependent portion of the lungs (zone 4) in the sitting and lateral positions. The study
demonstrated that progressive increases in airway pressures result in a decrease in
perfusion per alveolus in the nondependent lung. This contributed to the wider variability
of the V/Q matching seen in the anesthetized patient in the lateral position compared
with the awake state. Although gravity explained most of the findings in the study, the
authors reported a decrease in regional blood flow in the lateral position (awake and
anesthetized) in the midlung fields. Their techniques, comparing flows of lung regions
compared to other regions along the axis of gravity, precluded detection of differences
within isogravitational (horizontal) planes. Thus, nongravitational factors could not be
assessed, which may have accounted for the decrease in perfusion in the midlung area.
Figure 5. Relative perfusion/alveolus (Y-axis) is plotted against vertical distance down
the lung (X-axis) at functional residual capacity in the awake and anesthetized-
paralyzed state in different positions. (From Landmark and colleagues.24With
permission.)

Nongravitational Factors
Vascular

The gravitational model was based on older techniques that compared differences in
blood flow between relatively large areas of the lung. As seen from the study above, the
distribution of pulmonary blood flow is incompletely explained by the gravity hypothesis.
Newer techniques have been utilized to assess differences in blood flow in very small
areas of the lung. The results of these studies give support to the theory that the
distribution of pulmonary blood flow is gravity independent, a concept proposed 25
years ago with the finding of central to peripheral gradients of pulmonary blood. 27

Hakim and associates 28 utilized single-photon emission computed tomography (SPECT)


to study the central-to-peripheral gradient of pulmonary blood flow in supine, awake
human volunteers. SPECT technology allows the three-dimensional analysis of the
pulmonary vascular tree and a quantitative, reproducible comparison of blood flow
within isogravitational planes. In this study, healthy male subjects were injected with
technetium 99mlabeled human albumin macroaggregates, and the gamma emission
data were collected with the subjects holding their breath at FRC. Figure 6 shows the
tomographic image of a midcoronal (horizontal, isogravitational) slice from one subject.
The darker areas represent higher concentration of activity and thus flow. A concentric,
onion-like distribution of blood flow can be seen. Figure 7 shows a midsaggital (vertical)
image from the right lung of a subject. There is some tendency of greater blood flow in
the lower half of the lung compared to the upper portion; however, the dominant pattern
of distribution follows the concentric pattern, with the central regions having activity
levels 3 to 10 times higher than the periphery. These results raise a question as to the
importance of gravity in determining the distribution of pulmonary blood flow.
Figure 6. Midcoronal single-photon emission computed tomographic image in the
supine position. The darker areas represent greater activity (and thus flow). Distribution
of blood flow follows a central-to-peripheral pattern within an isogravitational plane.
(From Hakim and associates.28With permission.)
Figure 7. Midsagittal single-photon emission computed tomographic image in the
supine position. Distribution of blood flow appears gravity independent, following a
central-to-peripheral gradient. Darker areas represent increased activity (flow). (From
Hakim and associates.28With permission.)

In addition, the study also showed that zone 4 conditions were not limited to dependent
portions of the lung, but rather were seen at the periphery of the entire lung. Although
the mechanism for the central-to-peripheral gradient was not studied, the authors
propose that the flow to a specific site may be inversely proportional to the distance the
blood must travel to reach that site. This may not be a simple function of distance, but
would also depend on diameters and branching patterns of the vessels. The fractal
branching model of the pulmonary vascular supports their theory and the central-to-
peripheral blood flow seen in the study. 29

In isolated lungs from dogs sacrificed in supine, prone, and upright positions, the same
group used the SPECT technology to assess again the radial distribution of blood flow
and the effect changes in body position would have on this pattern. 30 In addition, the
sacrificed lungs were cut in sagittal or coronal planes and analyzed directly to compare
these findings with the SPECT results. The purpose of the direct images was to confirm
the SPECT results without the technical obstacles such as lung movement, chest wall
attenuation, regional differences in lung expansion, and interference by the great
vessels/heart that could potentially alter the interpretation of the SPECT images. As in
the previous human studies, there was a central-to-peripheral gradient of blood flow in
all body positions. There was a strong correlation between the SPECT and direct image
results, with both methods showing a central-to-peripheral flow gradient of ~10:1.
Although the radial gradient was independent of gravity, the region with the highest
activity (flow) was shifted slightly downward along the axis of gravity.

Using labeled microspheres, others have presented results that contradict the
gravitational model of pulmonary blood flow in several species in supine, prone, and
lateral positions. 3133 This technique utilizes fluorescent or radioactive labeled
microspheres that are injected intravenously and taken up by the perfused lung. The
animals are sacrificed and the lungs are prepared for analysis for radioactivity or
fluorescence. Using radioactive labeled microspheres in dogs, a lack of gravity-
dependent change in pulmonary blood flow was demonstrated when the position of the
dogs was changed from the supine to the left lateral position. 31 As seen in Figure 8,
there is a strong correlation (r = 0.85) of the blood flow for each lung segment in the
supine and left lateral position. Using fluorescent microspheres, the same group has
demonstrated gravity-independent blood flow distribution in prone sheep. 33 The
distribution was greatest in the central lung regions and the flow decreased in proportion
to the distance from the hilum in all directions. In contrast, there was no correlation of
pulmonary blood flow to the gravitational (dorsal-ventral) gradient.
Figure 8. Relationship between blood flow to each lung unit in the lateral position (Y-
axis) and the supine position (X-axis). The line represents the line of identity. Pearson
correlation coefficient for all animals in the study was 0.85. (From Mure and
colleagues.31With permission.)

It is important to appreciate the species variation when attempting to apply the findings
from these studies to other animals or humans. For instance, dogs depend less on
hypoxic vasoconstriction than do sheep or humans, as dogs have a more developed
collateral ventilation system for V/Q matching. Gravity may play a greater role in the
human than the quadraped. 34 In addition, the geometric differences of the chest
cavities between humans (greater lateral dimension) and animals (e.g., dogs, with
greater dorsal-ventral dimensions) may account for some of the differences seen at
baseline and with position changes. 31

Other Factors

Changes in cardiac output and lung volumes can have significant effects on the
distribution of pulmonary blood flow either independently or by modulating the
determinants of flow mentioned above. It was classically thought that increased cardiac
output resulted in greater uniformity of blood flow throughout all of the lung fields. 35
The mechanisms proposed were an attenuation of HPV due to an increase in pulmonary
arterial pressure and a lessening of the gravitational gradient. 35 However, Hakim and
associates, 36 utilizing SPECT techniques, have recently demonstrated that an increase
in cardiac output results in the maintenance of flow in a central-to-peripheral gradient,
without an increase in attenuation of the gradient.

Changes in lung volumes can alter the distribution of blood flow. The lung volume at
which the pulmonary vascular resistance is lowest is at FRC. When the lung volume is
lower than FRC, the resistance of the pulmonary vascular system is increased due to
decreased caliber of the extra-alveolar vessels. This may be due in part to the loss of
radial traction supporting these vessels at low lung volumes. At lung volumes above
FRC, pulmonary vascular resistance increases due to the stretch of the capillaries ( Fig 9).
Figure 9. Pulmonary vascular resistance vs lung volume. Vascular resistance is lowest
at functional residual capacity (FRC). The increase in resistance at lower lung volumes is
due to the narrowing of the extra-alveolar vessels from the loss of radial traction.
Increased interstitial pressure and edema may also contribute to the narrowing. Above
FRC, stretching of the capillaries results in an increase in resistance. (From West.20With
permission.)

Finally, the pharmacological agents administered to the patient and their effects on the
distribution of pulmonary blood flow must be considered. A systematic review is beyond
the scope of this paper and has been reviewed by others. 13,37 Briefly, intravenous
anesthetic agents and regional techniques have minimal effects on pulmonary vascular
resistance and pulmonary blood flow. In vitro, inhalational agents demonstrate potent
inhibition of HPV, but the results in vivo are variable. In a dog model of one-lung
hypoxia, the median effective dose for inhibition of HPV was 2 times the minimum
alveolar concentration (MAC) for isoflurane. 38 However, in humans, 1.2 MAC of
isoflurane did not affect HPV in a model of one-lung hypoxia. 39 Nitroglycerin, sodium
nitroprusside, calcium channel blockers (especially nifedipine and nicardipine),
isoproterenol, prostacyclin (given intravenously), and dobutamine all inhibit HPV.
Hydralazine, adenosine, labetalol, and aminophylline have minimal effects on HPV and
pulmonary blood flow. Norepinephrine, epinephrine, and dopamine may have minimal
effect on HPV at low doses, but will impair HPV due to alpha effects at higher doses. In
contrast, propranolol may augment HPV by inhibiting [beta] 2-mediated vasodilation. 40

Summary
Placing our anesthetized patients in the lateral decubitus position results in significant
ventilation and perfusion mismatch. With mechanical ventilation, the major portion of
each tidal volume is delivered to the nondependent lung, which, compared to the awake
state, achieves a more favorable position on the pressure volume curve, indicating
improved compliance. In contrast, the dependent lung moves to a less favorable position
on the pressure volume curve due to a decrease in compliance. The distribution of
pulmonary blood flow in the lateral position is determined by many factors. Although the
relative importance of gravity and nongravity factors is not possible to assess, both are
probably operational in humans and may be additive in some instances and offsetting at
other times. 41 Overall, there is a greater amount of blood flow to the dependent lung in
the lateral position during two-lung mechanical ventilation. The dependent lung is
therefore well perfused and poorly ventilated and the nondependent lung is well
ventilated but poorly perfused.

To improve V/Q matching in the lateral position during procedures in which two-lung
ventilation can be maintained (e.g., renal surgery, retroperitoneal abdominal aortic
aneurysm repair), the selective application of PEEP to the dependent lung may be an
option. In a study of 15 men undergoing elective surgical procedures, the patients were
intubated with double-lumen endotracheal tubes (DLT) and received differential
ventilation of each lung. 8 The only variable different between the two lungs was applied
PEEP. There was significant mismatching of V/Q in the lateral position that improved
with the selective application of 10 cm H2O of PEEP to the dependent lung (Fig 10), as did
systemic arterial oxygenation. No hemodynamic compromise was seen with the selective
application of PEEP to the dependent lung. The generalized application of 10 cm H 2O of
PEEP to both lungs did not improve V/Q or arterial oxygenation and resulted in a
significant reduction in cardiac output. It appears that the selective application of PEEP
to the dependent lung improves its compliance (moves it to a more favorable position on
the steep part of the curve [see Fig 1]) and creates better matching of ventilation and
perfusion. Of note, the selective application of 10 cm H2O of PEEP to the dependent lung
did result in a significant increase in dead space (see Fig 10). Nevertheless, this
technique of selective PEEP may not be applicable to many of our patients in the lateral
position. When we utilize DLTs, two-lung ventilation is not routinely maintained and OLV
is often employed to facilitate the performance of surgery.
Figure 10. Ventilation and perfusion distributions in the lateral position. The upper
panel shows significant ventilation-perfusion quotient ratio (VA/Q) mismatch when
conventional mechanical ventilation (CV) is used with zero positive end-expiratory
pressure (PEEP) to the dependent lung. The lower panel shows improved VA/Q matching
when 10 cm H2O PEEP is selectively applied to the dependent lung. Qs/Qt = shunt
fraction; PaO2 = arterial oxygen tension; VD/VT = dead space. (From Klingstedt and
coworkers.8With permission.)
One-Lung Ventilation

One-lung ventilation was originally performed to prevent infectious spillage from one
lung to the other during thoracic surgery and for bronchospirometry. 1,42 Gale and
Waters 43 first reported the use of selective lung ventilation during thoracic surgery in
1931. Early on, endobronchial anesthesia was performed using modifications of single-
lumen tubes, including double cuffs 44 and bronchial blockers. The precursor to our
current, disposable polyvinylchloride DLTs originated in 1949 with Carlens' double-lumen
catheter 42; Robertshaw 45 introduced a DLT without a hook in 1962. OLV is most
commonly used today to create a quiet operating field for the performance of surgery.
For many of our patients, the performance of OLV is carried out without any
consequence upon oxygenation and ventilation. However, significant hypoxemia may
occur, and the incidence has been reported to be as high as 40%50% in some series.
46,47

It is important to realize that there are relatively few absolute indications for OLV.
Absolute indications include protection from massive contamination (infection or
hemorrhage) of one lung to the other and selective lung ventilation due to severe
unilateral lung disease (bronchopleural fistula, giant bullae, bronchial
disruption/surgery). 5 Protection from spillage is also important during lavage for
pulmonary alveolar proteinosis and cystic fibrosis. 5 Although considered relative
indications, thoracic aortic aneurysm repair, thoracoscopy, pneumonectomy, lobectomy,
esophageal surgery, thoracic spine procedures, and mitral valve repair are facilitated by
OLV. 5,48 OLV may shorten operative time and reduce the injury to the lung due to
retraction and manipulation if the lung remains inflated.

Placement of Double-lumen Tubes

In our institution, the majority of cases requiring OLV are performed with disposable
DLTs. Unlike many centers, which only use left-sided DLTs, it is our practice to use both
left-and right-sided DLTs to selectively intubate the dependent lung bronchus. Many
argue that a greater margin of safety with left-sided tubes precludes the use of right-
sided tubes. 49,50 It is written that the risk of obstruction of the right upper lobe and/or
carina is greatest with right-sided tubes. 50 However, we have not found this to be the
case in our practice. In a study by Hurford and Alfille, 51 the incidence of placement
errors and complications associated with the conventional blind placement was
evaluated using disposable DLTs. In 154 disposable DLT placements, 44% of the blind
placements had to be adjusted and changed using fiberoptic bronchoscopy (FOB). The
complication rate was not influenced by the tube selected or the side intubated. Twenty-
six percent of the DLTs were placed too far initially, and the wrong side was intubated
more commonly with the left-sided tubes compared with right-sided tubes (14% vs 3%,
P = 0.0045). The use of FOB to place or confirm placement after conventional insertion
was instrumental in obtaining proper tube positioning for both right-and left-sided tubes.
Using FOB to confirm or adjust the placement of DLTs has become standard practice in
our institution and others. 49

Being proficient at placement and management of right-sided DLTs is important, as


there are instances when right-sided tubes are recommended over left-sided tubes. The
left mainstem bronchial anatomy may be distorted by tumor or by a thoracic aortic
aneurysm, precluding the placement of a left-sided tube. Right-sided DLTs are used
almost exclusively for our descending thoracic aneurysm repairs, reducing the risk of
aneurysmal rupture. 52 The conduct of an anesthetic for a left lung sleeve resection or
pneumonectomy is easier with a right-sided DLT, since tube withdrawal into the trachea
is not required during the procedure. In addition, if the DLT is not placed into the
dependent lung bronchus (right or left), the weight of the mediastinum may compromise
ventilation by either compressing the tracheal lumen of the DLT into the tracheal wall or
by obstructing the dependent lung bronchus. This potentially hazardous situation may
be prevented by intubating the dependent bronchus.

The size of the DLT used is based primarily on patient height, and to a lesser extent
gender. Average size women are most typically intubated with 35-or 37-French tubes,
whereas average size, adult men are intubated with 39-or 41-French tubes. This is in
agreement with Benumof and colleagues, 49 who correlated height with bronchial length
and recommended 39-and 41-French tubes in patients 5`8" or taller, and 35-to 39-
French tubes in patients who are shorter.

The technique used to place DLTs is that described by Hurford. 48 The appropriate size
right-or left-sided tube is prepared by lubricating the distal end of the DLT, connecting
the right-angle elbow tube connectors (OPTI-PORT, Mallinckrodt, St. Louis, MO) to the Y-
Carlens adapter, and checking the bronchial cuff (35 ml of air) and tracheal cuff (510
ml of air) for leaks. A Macintosh blade is preferable for laryngoscopy, as it may provide
more room for manipulation of the tube within the pharynx. With the stylet in the
bronchial lumen (Mallinckrodt Broncho-Cath DLTs come packaged with the stylet in the
bronchial lumen), the tube is placed into the pharynx with the concavity of the distal tip
facing anteriorly. Once the distal tip is through the cords, the stylet is removed, and the
tube is gently rotated so that the bronchial lumen and distal concavity are facing toward
the side to be intubated. During this rotation, laryngoscopy with the Macintosh blade
should be maintained. This will prevent oropharyngeal structures from interfering with
the free rotation of the tube. The tube is then gently advanced to a depth of
approximately 2728 cm in women and 29 cm in men. The disposable tubes are not
advanced until resistance is encountered, as this most certainly will mean that the tube
is in too far. 51

In certain instances, such as when a left-sided DLT is placed for a sleeve resection of a
friable tumor in the right mainstem, the placement of the DLT should be performed
under direct FOB guidance to prevent blind misplacement into the tumor. Initial tracheal
intubation may be performed with direct laryngoscopy or FOB. The tube is then guided
into the appropriate bronchus under direct vision with the FOB protruding through the
bronchial lumen.

If not utilized for initial placement, the FOB should be used to confirm placement of the
DLT. Confirming proper placement of the DLT (appropriate side and depth of insertion) is
most commonly performed using a combination of clinical assessments and FOB. As with
all endotracheal intubations, the presence of breath sounds, absence of gastric sounds
with ventilation, rise and fall of the chest, and continuous end-tidal CO 2 detection should
be confirmed. To clinically confirm placement, the following steps are recommended: (1)
Prior to cuff inflation, breath sounds are confirmed and a moderate leak around the tube
should be heard, which should disappear with 510 ml of air in the tracheal cuff. (2) The
tracheal side of the tube connector is clamped, the tracheal cap opened, and the
bronchial lumen is ventilated. The bronchial cuff is then inflated until the leak
disappears, typically with < 3 ml of air. Auscultation and observation of chest movement
will confirm if the appropriate side has been intubated. (3) The tracheal lumen cap is
replaced, the clamp removed and placed on the bronchial lumen connector, and the
bronchial cap opened. Ventilation, through the tracheal side should result in breath
sounds in and movement of the nonintubated lung. (4) Ventilation through both lungs is
again confirmed by unclamping the bronchial lumen connector and closing the bronchial
cap. The bronchial cuff should remain inflated once position is confirmed until OLV is no
longer necessary to reduce the likelihood of tube movement. During these maneuvers,
hand ventilation is performed to assess changes in compliance, which should be minimal
if the tube is in proper position.

Fiberoptic bronchoscopy is then performed with the patient in the supine position to
confirm initial tube position, since the incidence of malpositioned tubes is high with the
conventional technique. The FOB is inserted through a self-sealing diaphragm in the
elbow connector or via a bronchoscope swivel adapter that allows continued positive
pressure ventilation during bronchoscopy. First, the FOB is inserted into the tracheal
lumen until the carina and the nonintubated bronchus are visualized. The proximal
portion of the blue bronchial cuff should be barely visible. If the cuff is not seen at all,
the tube is in too far and it should be withdrawn until the cuff comes into view. If too
much of the cuff is visualized, the tube is advanced. For a left-sided DLT, passing the
FOB through the tracheal lumen will confirm the placement of the tube in the majority of
patients. However, there is some variability in the distance between the tracheal carina
and the origin of the left upper lobe. 49 We routinely pass the FOB through the bronchial
lumen to confirm that the position of left-sided DLT is proximal to the secondary carina.

For a right-sided DLT, the FOB must be passed into the bronchial lumen to confirm that
the endobronchial lumen terminates proximal to the secondary carina of the right middle
and lower lobes. In addition, the right upper lobe bronchus must be visualized through
the slot on the DLT. In addition to the tube being in too far, other problems encountered
with placement include intubation of the wrong bronchus, kinking of the tube, and
wrong size tube, which may preclude adequate lung isolation. 48 It is recommended that
the position of the tube be rechecked with the FOB after final positioning in the lateral
position, at the initiation of one-lung ventilation, and during any maneuver that may
dislodge the tube. Surveillance bronchoscopies are recommended at regular intervals to
confirm the maintenance of proper tube position during OLV.

As mentioned above, the most frequent error with initial placement of the disposable
DLT is advancing it too far distally. The corresponding lower lobe will then be the only
segment of lung ventilated. The right and left lower lobes constitute only 30%35% and
20%25% of the total lung, respectively. 49 Thus, commonly seen clinical signs of tube
malpositioning include a decrease in effective compliance, an increase in peak
inspiratory pressure (PIP), and oxygen desaturation. When the tube moves proximally,
loss of lung isolation may result. In addition to these clinical signs, the use of continuous
spirometric monitoring may aid in the diagnosis of tube malposition and reduce the need
for frequent bronchoscopies. 53 Other complications encountered when using a DLT
include tracheal and bronchial trauma, 54,55 cardiac arrest, 56 pneumothorax, and
aneurysm rupture. 52

Most procedures requiring OLV in our institution are performed with a disposable DLT.
Bronchial blockers, either as separate catheters or incorporated within a single-lumen
tube (Univent Tube, Vitaid, Lewiston, NY), may be used to obtain lung isolation in small
children, when the anatomy prevents safe placement of a DLT (e.g., tracheal stenosis),
or for cases in which changing back to a single-lumen tube at the end of the procedure
may be hazardous due to airway edema. The concept that the use of the Univent tube,
especially when placed on the right side, results in a higher incidence of malpositioning
and more frequent repositioning was recently challenged. 57 Endobronchial intubation
may be performed with long, single-lumen, single-cuff, armored tubes (Wilson) for
procedures in which isolation is required, but reinflation or suctioning of the other lung is
unnecessary or unwise (i.e., postpneumonectomy patient, patient with bronchopleural
fistula or massive intrapulmonary hemorrhage). 48 The placement of these tubes has
been described in detail elsewhere. 48

Optimizing Oxygenation

Significant and at times dangerous hypoxemia may occur in some patients during OLV. It
has been noted for many years that the most significant decreases in arterial oxygen
tension (PaO2) occur in those patients for whom OLV is used for procedures other than
pulmonary resection. 1,3 It is postulated that in those patients undergoing pulmonary
resection, the blood flow through the diseased lung (which will be the lung collapsed
during OLV) has been chronically reduced. 1,3

The hypothesis that intraoperative hypoxemia is more likely to occur from persistent
pulmonary blood flow during OLV when there is greater preoperative pulmonary blood
flow to the operative lung was tested in a study by Hurford and associates. 46 In their
study, 30 consecutive patients who received preoperative ventilation/perfusion scans
and underwent thoracic procedures in the lateral position with OLV were examined. The
percent of blood flow to the operative lung seen on the preoperative perfusion scan
correlated inversely with the PaO2 after 10 minutes of OLV (r = -.72). When the percent
of blood flow on the preoperative scan to the operative lung was greater than 45%, the
likelihood of hypoxemia (PaO2 < 75 mm Hg) was increased (Fig 11). Since preoperative
regional ventilation was well matched to perfusion in these patients, the percent of
preoperative ventilation also correlated inversely with the PaO 2 after 10 minutes of OLV
(r = -.73). Neither preoperative blood gases, pulmonary function tests, nor lung
volumes correlated with oxygenation during OLV.
Figure 11. The proportion of patients with hypoxemia (arterial oxygen tension [PaO2]
< 75 mm Hg) during one-lung ventilation (OLV) grouped according to the relative
preoperative perfusion to the operative lung. The proportion of patients with hypoxemia
during OLV increased with increasing perfusion on preoperative scan to the operative
lung. (From Hurford and associates.46With permission.)

This is in contrast to the results of Slinger and coworkers. 58 In their study, they found
that an equation using three variables, PaO 2 during intraoperative two-lung ventilation in
the lateral position, side of operation, and preoperative ratio of 1-second forced
expiratory volume to vital capacity, could be used to predict (r2 = .73) the PaO2 during
OLV with continuous positive airway pressure (CPAP) to the nonventilated lung.
Nevertheless, Katz and colleagues 59 have agreed with the findings of Hurford and
associates 46 that routine preoperative blood gases and pulmonary function tests do not
reliably predict which patients are at risk for hypoxemia during OLV.

As mentioned above, there is significant V/Q mismatching in the lateral decubitus


position during two-lung ventilation. The dependent lung is relatively poorly ventilated
and well perfused in contrast to the well ventilated and relatively poorly perfused
nondependent lung. Why then does OLV not result in better V/Q matching? In some
patients it does. It has been shown that during OLV the elimination of carbon dioxide is
improved. This is due to a decrease in the physiological dead space that may exist in the
nondependent lung during two-lung ventilation. Dead space has been shown to decrease
from 51% during two-lung ventilation in the lateral decubitus position to 42% during
OLV in patients undergoing pulmonary resection. 2 There was no change in dead space
seen during OLV in patients who were not undergoing lung resection. Overall,
elimination of carbon dioxide is usually not a problem during OLV.

The cause of hypoxemia during OLV is due to an increase in the physiological shunt
through both the ventilated and nonventilated lungs. Although HPV is effective at
maintaining PaO2 when the volume of lung that is atelectatic is intermediate (30%
70%), as occurs during OLV, many factors attenuate the HPV response during surgery
(see above). It has been shown that the onset of hypoxemia begins approximately 5 to
10 minutes after initiating OLV and reaches its maximal level by 15 minutes. 3,67 This
corresponds to the time it takes for the absorbable gases (oxygen and nitrous oxide) to
be absorbed completely from closed cavities when blood flow is sustained. 3 Many
different therapeutic options to reduce the shunt through the dependent or
nondependent lung and to treat the hypoxemia during OLV have been described.

Prior to discussing three of the therapies, large tidal volumes, CPAP to the nonventilated
lung, and PEEP to the dependent lung, a few points need to be emphasized. First, during
OLV, the patients should be receiving an inspired oxygen concentration (FiO 2) of 0.95
1.0 (using 0.95 may decrease the amount of absorption atelectasis in areas of low V/Q).
Once adequate oxygenation has been sustained beyond the expected time for
hypoxemia to develop (at least 15 minutes), lower levels of oxygen may be utilized,
guided by pulse oximetry and/or arterial blood gases. Second, the most common cause
of hypoxemia seen in clinical practice is a malpositioned DLT. When the saturation begins
to fall, the patient should be removed from the ventilator and hand ventilation should be
instituted. This allows a better assessment of lung compliance. A decrease in compliance
may indicate that the tube has slipped distally and occluded the upper lobe bronchus
leading to an increased shunt and desaturation. The position of the tube should be
assessed and repositioned by FOB. Using the fiberoptic bronchoscope, the anesthetist
can also remove any secretions, another common cause of desaturation. Other causes of
hypoxemia during OLV that need to be ruled out include a light plane of anesthesia and
bronchospasm. Deepening the anesthetic and administering bronchodilator therapy
should help improve gas exchange. If any of the above are not identified as the cause of
the hypoxemia, then therapies directed at reducing the increased physiologic shunt, due
to atelectasis in the dependent lung and persistent perfusion through the nonventilated
lung, should be initiated.

Large Tidal Volumes

Traditionally, the ventilatory parameters used during OLV were the same as those used
during two-lung ventilation. It has become a common practice to set the ventilator to
deliver 1012 ml/kg during OLV and maintain minute ventilation by adjusting the
respiratory rate. Early studies demonstrated that during anesthesia, the use of large
tidal volumes significantly improved pulmonary gas exchange during two-lung and one-
lung ventilation. 60,61 Although the elimination of carbon dioxide is uniformly improved
with large tidal volumes, subsequent studies have shown that the improvement in
oxygenation is variable. For instance, in a study by Katz and colleagues, 59 the use of
large tidal volumes (14 ml/kg) was effective at improving oxygenation in 16 out of 17
patients. In the other patient (undergoing an esophagectomy), however, significant
hypoxemia (PaO2) occurred, requiring a return to two-lung ventilation. Although
oxygenation was less with smaller tidal volumes (7 ml/kg), the mean difference was
small (26 mm Hg), demonstrating a significant variability in the improvement in
oxygenation with large tidal volumes. The mean difference in arterial carbon dioxide
tension (PaCO2) during ventilation with large tidal volumes compared with small tidal
volumes was also small (2.2 mm Hg).

In another study, Khanam and Branthwaite 62 showed that optimal ventilation and
oxygenation was obtained when the tidal volume for OLV was set at 7 ml/kg and the
respiratory rate was set at 20 breaths per minute. Increasing the tidal volume to 9
ml/kg caused a marked decline in arterial oxygen tension and a significant increase in
the alveolar/arterial oxygen gradient compared to the values seen with 7 ml/kg tidal
volumes. The decline in arterial oxygenation was much less when smaller (5 ml/kg) tidal
volumes were used (minute ventilation maintained by increasing respiratory rate), and
the alveolar/arterial gradient was not significantly different from the 7 ml/kg values. In
addition to the variability of response, the use of large tidal volumes to treat hypoxemia
during OLV may be questioned for other reasons.

It has been shown that lung injury associated with mechanical ventilation is due to
volutrauma. In numerous animal studies, high transpulmonary pressures result in
overdistension of the lung and subsequent injury. 63 Recent human studies have
demonstrated improved survival in patients with ARDS in whom smaller lung volumes
were used during mechanical ventilation (National Heart Lung and Blood Institute;
Prospective, randomized, multi-center trial of 12 ml/kg vs. 6 ml/kg tidal volume positive
pressure ventilation for treatment of acute lung injury and acute respiratory distress
syndrome (ARDS); March 15, 1999). In patients undergoing major pulmonary resection
or pneumonectomy, postoperative respiratory failure (postpneumonectomy pulmonary
edema [PPPE]) may occur. Although the etiology of PPPE is unknown, risk factors known
to be associated with ARDS such as barotrauma/volutrauma, complement activation,
and cytokine activation and release have been postulated as causes of PPPE. 64
Determining the actual role that intraoperative volutrauma may play in this often fatal
syndrome warrants further investigation. In addition, the use of large tidal volumes (10
ml/kg) has been associated with the development of intrinsic PEEP that may not be
recognized. This could lead to overdistension and lung injury if the level of intrinsic PEEP
is high or if excessive external PEEP is added 65 (see below).

CPAP

CPAP to the nonventilated, upper lung has been shown to be an effective method of
treating hypoxemia during OLV and is considered by many to be the first-line therapy
after the causes mentioned above have been addressed. Capan and coworkers 66 found
that CPAP of 10 cm H2O to the nonventilated lung, with zero end-expiratory pressure to
the dependent lung, was the most effective way of improving oxygenation and
decreasing shunt flow through the collapsed lung during OLV. Combining 10 cm H 2O of
PEEP to the dependent lung with CPAP to the nondependent lung was also effective at
improving oxygenation and reducing shunt, but cardiac output was reduced by up to
17%. Insufflation of oxygen to the nonventilated lung without pressure failed to improve
oxygenation.

The method in which CPAP is applied to the nondependent lung was systematically
studied by Slinger and associates. 67 They described two different methods of applying
CPAP. Method one was designed to mimic the common scenario of hypoxemia and the
application of CPAP. This method allowed the nondependent lung to deflate from a tidal
volume inflation (proximal airway pressure 20 cm H2O), equilibrating with atmospheric
pressure for 5 minutes (0 cm H2O CPAP), and then CPAP 5 cm H2O was applied for 20
minutes. In method two, the lung was deflated from a tidal volume inflation using CPAP
2 cm H2O for 5 minutes followed by an increase of CPAP to 5 cm H2O. Previous studies
had shown that compliant lungs would not deflate adequately at CPAP 5 cm H 2O but
using CPAP 2 cm H2O allowed for controlled deflation. Twenty patients having elective
thoracotomies (19 pulmonary resections, 1 hiatal hernia repair) were divided into two
groups. Group A patients were ventilated with method one followed by two, and group B
patients were ventilated with method two followed by method one. Three of the twenty
patients developed hypoxemia (PaO2 < 70 mm Hg), all of which occurred during method
one strategy. Slinger and associates 67 recommended instituting CPAP 2 cm H2O during
initial lung deflation followed by an increase to 5 cm H2O, rather than letting the lung
deflate completely followed by CPAP. Benumof 13 reported that in his experience, a low
level of CPAP is effective at improving hypoxemia without impairing surgical exposure
more than 90% of the time when proper tube position is confirmed. CPAP to the
nonventilated lung is effective at treating hypoxemia during OLV by oxygenating the
blood passing through that lung. CPAP may also improve hypoxemia by diverting more
blood flow to the ventilated lung (improving V/Q) due to an increase in vascular
resistance in the nonventilated lung. 13

PEEP

The application of PEEP to the dependent lung is intended to restore FRC, recruit alveoli,
and improve gas exchange. As logical as this seems, the studies in which dependent
lung PEEP has been studied have reported mixed results. In the study of Capan and
coworkers, 66 the selective application of PEEP to the dependent lung worsened
oxygenation and increased shunt. In patients who have low dependent lung volumes,
the application of PEEP has been shown to be beneficial. 68 However, if PEEP is applied
during OLV with large tidal volumes to the dependent lung, alveoli may become
overdistended causing intra-alveolar vascular compression and worsening V/Q
mismatch. 13 In addition, extrinsic PEEP may interact with an undetected phenomenon,
intrinsic PEEP.

Dynamic pulmonary hyperinflation occurs when expiratory flow does not end and the
lung does not relax down to FRC before the next inspiration begins. This results in the
development of positive pressure at the end of expiration, termed auto-PEEP or intrinsic
PEEP. Intrinsic PEEP is commonly seen during mechanical ventilation of chronic
obstructive pulmonary disease (COPD) patients and is due to many factors. Expiratory
flow limitation may be due to increased resistance within the patient's airways
(destruction of lung parenchyma/elastic recoil, bronchospasm, secretions) and from the
endotracheal tubes, ventilator tubing, and valves. 69 The ventilatory pattern (tidal
volume, respiratory rate, and inspiratory:expiratory (I:E) time ratio) is a very important
factor in the development of intrinsic PEEP.

Recent studies have demonstrated that intrinsic PEEP may be present in the majority of
patients undergoing OLV. 65,70,71 However, since intrinsic PEEP cannot be detected by
the manometer within the circle system, other techniques must be employed to monitor
for its presence. Typically, an end-expiratory occlusion maneuver (EEO) 65 or end-
expiratory hold 70 is used to measure the presence and magnitude of intrinsic PEEP.
Although EEO allows a quantitative measure of intrinsic PEEP, it is a static measurement
during which ventilation must be interrupted. Using a flow sensor attached to the
common connector of the DLT, Bardoczky and associates 72 found that intrinsic PEEP
was detected with a sensitivity of 78% and a specificity of 95% when there was
interruption of expiratory flow on the constructed flow-volume curves (see Fig 4, chapter
1). Although the magnitude of the intrinsic PEEP may not be determined using this
technique, the measurements are made in dynamic conditions and ventilation is not
interrupted.

Bardoczky and colleagues 73 reported that the presence of intrinsic PEEP was found
primarily in COPD patients with pre-existing pulmonary hyperinflation and airflow
obstruction. Ducros and coworkers 71 detected intrinsic PEEP in patients with normal
preoperative pulmonary function during OLV for thoracic procedures. Although the level
of intrinsic PEEP during OLV was highest for severe COPD patients undergoing lung
transplants, even patients with severe restrictive lung disease had measurable levels of
intrinsic PEEP during OLV for lung transplant. 71

The presence of intrinsic PEEP must therefore be suspected in all patients during OLV. In
severe COPD patients, using EEO or flow volume loops should be considered to detect
intrinsic PEEP. When extrinsic PEEP is applied in the presence of intrinsic PEEP, the
interaction between the two is complex. Most often the total PEEP only increases 1 to 2
cm H2O when 5 cm H2O of external PEEP is applied, and the greatest increase in total
PEEP occurs when extrinsic PEEP is added to a low level of intrinsic PEEP. When intrinsic
PEEP is high, the total PEEP does not increase with added extrinsic PEEP until a critical
value is reached. 65 Above this critical value, total PEEP increases leading to
overdistension of alveoli and further V/Q mismatch. This may be the mechanism for the
further impairment in oxygenation seen when extrinsic PEEP is used to treat hypoxemia
during OLV in some patients. Although oxygenation may improve with intrinsic PEEP,
some studies have failed to show a correlation between intrinsic PEEP and PaO 2 during
OLV. 70 If the level of intrinsic PEEP is too high, which may cause overdistension of the
lung and a reduction in cardiac output, therapies that may reduce it include adjusting
the ventilator (decrease in tidal volume, respiratory rate, and I:E time ratio), treating
bronchospasm, clearing secretions, or applying extrinsic PEEP.

Other Techniques

When these maneuvers fail to improve oxygenation, or when severe hypoxemia occurs,
two-lung ventilation should be resumed and manual ventilation started. In some
patients, intermittent reinflation of the nondependent lung may be necessary to
maintain adequate oxygenation. 3 For patients undergoing a pneumonectomy, the
surgeon may clamp the pulmonary artery to improve V/Q matching. In dogs, partial
balloon occlusion of the pulmonary artery of the nonventilated lung resulted in a
significant improvement in PaO2 by diverting blood flow to the ventilated lung. 74 This
may be an option in selected patients with severe hypoxemia in whom clamping of the
pulmonary artery is not feasible and in whom constant lung reinflation is interfering with
the progress of the surgical procedure. Combining different levels of CPAP to the
nonventilated lung and PEEP to the dependent lung may be tried to balance and
optimize V/Q matching. 13 Finally, a recent study has shown that in patients with
significant restrictive pulmonary disease (preoperative FVC < 77% predicted) pressure
controlled ventilation was superior to traditional volume controlled ventilation (10 ml/kg)
at treating hypoxemia during OLV. 75

Clinical Practice

Our practices aim to achieve adequate gas exchange using a ventilation strategy that
may also protect the lung from volutrauma. OLV is initiated using manual ventilation and
an FiO2 of 1.0. This allows the compliance of the ventilated lung to be assessed. If the
compliance is decreased, manual ventilation is continued while the tube position is
assessed and adjusted as necessary using FOB. Once adequate tube position has been
confirmed, the ventilator is set to deliver tidal volumes in the range of 5 to 7 ml/kg. The
tidal volume may then be adjusted up or down as guided by PIP or plateau pressure
(Pplat), keeping these values less than 30 cm H2O. Maintaining PIP and Pplat in this range
has been shown to reduce injury in experimental and clinical studies. 63,76,77 The
respiratory rate is then adjusted to maintain a safe level of carbon dioxide elimination
as guided by end-tidal CO2 measurements and/or blood gas analysis. At times, the
minute ventilation that is achieved using these ventilator parameters may be less than
ideal. Minute ventilation may be limited by air-trapping, not only in COPD patients, but
also in patients with normal preoperative lung function. Controlled hypoventilation has
been termed permissive hypercapnia and has been shown to be a safe technique in
ARDS, even with pH values to 7.15 and PaCO2 to 80 mm Hg. 78 The maintenance of
adequate oxygenation (PaO2 > 60 mm Hg) is imperative during this period. The safe
level of acute hypercapnia for patients under general anesthesia is not known, but
values within this range may be applicable.

Oxygenation is initiated with an FiO2 of 1.0. After 15 minutes of OLV, if oxygenation is


not a problem, the inspired oxygen concentration may then be titrated down using pulse
oximetry and/or blood gas analysis as a guide. At times, hypoxemia may occur. As
mentioned above, the first maneuver should be to increase the FiO2 to 1.0 (if lower
amounts were previously used) and the patient should be removed from the ventilator.
If overdistension of the lung was a factor, this will allow adequate exhalation from slow
lung units. The compliance of the lung should be assessed using manual ventilation and
potential causes of decreased compliance investigated (e.g., tube malposition,
secretions, light anesthesia, or bronchospasm). In the majority of cases, therapies
directed at one or more of these causes will correct the hypoxemia. Tube malposition
and secretions may be identified and corrected before hypoxemia ensues by frequent
(every 30 minutes or so) FOB use during OLV. If hypoxemia is severe, the surgeon
should be alerted and two-lung ventilation reinstituted. Rarely, reinflation of the
nondependent lung may need to be repeated every 10 to 15 minutes when other
therapies have failed. CPAP remains the first therapy instituted in less severe causes of
hypoxemia. It is accomplished by connecting a Bain circuit to the lumen ventilating the
nondependent lung. CPAP of 5 to 10 cm H2O with 100% oxygen is applied after the lung
is opened with a small tidal volume. The tidal volume used must balance the
interference that may occur in the surgical exposure (with resultant prolonged surgical
time and potential lung injury from retraction) and the degree of hypoxemia. In some,
PEEP to the dependent lung may be effective with the small tidal volumes used in this
strategy and should be titrated to effect. As PEEP is applied, both the improvement in
oxygenation and potential side effects, such as a decrease in cardiac output, should be
closely monitored.

Once the surgical procedure is complete, the return to two-lung ventilation, except in
the case of a pneumonectomy, is performed by re-expanding the nondependent lung
with peak pressures of 30 and then 40 cm H2O. This will not only determine if there are
leaks at the surgical sites, but will also recruit the atelectatic areas in the dependent
lung. Depending on the amount of lung resected, the ventilator settings may need to be
adjusted from the previous two-lung settings to reduce the risk of overdistension. For
pneumonectomy patients, the same procedure is used to check the bronchial suture line
for leaks and then the settings for OLV are continued.

Summary
OLV is most frequently utilized to provide a quiet field for the performance of many
different surgical procedures. In some patients, severe hypoxemia may result,
mandating the implementation of other therapies to provide adequate oxygenation. This
paper has reviewed the physiological consequences of the lateral position that may
contribute to the hypoxemia and the techniques we utilize at our institution for
establishing OLV, maintaining OLV, and treating hypoxemia during OLV. Our technique is
performed with the goal of maintaining adequate gas exchange and protecting the
ventilated lung from potential overdistension and injury. It remains for future study to
determine if the use of a lung protective strategy during intraoperative OLV offers any
benefit to patients at risk for postoperative lung injury, such as those undergoing major
lung resections.

References
1. Kerr JH. Physiological aspects of one-lung (endobronchial) anesthesia. Int Anesthesiol Clin
1972; 10:6178 [Context Link]

2. Kerr JH, Smith AC, Prys-Roberts C, Meloche R. Observations during endobronchial anesthesia
I: ventilation and carbon dioxide clearance. Br J Anaesth 1973; 45:159167 [Context Link]

3. Kerr JH, Smith AC, Prys-Roberts C, et al. Observations during endobronchial anaesthesia. II.
Oxygenation. Br J Anaesth 1974; 46:8492 [Context Link]

4. Larsson A, Malmkvist G, Werner O. Variations in lung volume and compliance during


pulmonary surgery. Br J Anaesth 1987; 59:585591 [Context Link]

5. Benumof JL, ed. Anesthesia for thoracic surgery. 2nd ed. Philadelphia: WB Saunders
Company, 1995 [Context Link]

6. Rothen HU, Sporre B, Engberg G, et al. Airway closure, atelectasis and gas exchange during
general anaesthesia. Br J Anaesth 1998; 81:681686 [Context Link]

7. Hedenstierna G, Strandberg A, Brismar B, et al. Functional residual capacity,


thoracoabdominal dimensions, and central blood volume during general anesthesia with muscle
paralysis and mechanical ventilation. Anesthesiology 1985; 62:247254 [Context Link]

8. Klingstedt C, Hedenstierna G, Baehrendtz S, et al. Ventilation-perfusion relationships and


atelectasis formation in the supine and lateral positions during conventional mechanical and
differential ventilation. Acta Anaesthesiol Scand 1990; 34:421429 [Context Link]

9. Nunn JF. Nunn's applied respiratory physiology. 4th ed. Oxford: Butterworth-Heinemann Ltd,
1993 [Context Link]

10. Rehder K, Sessler AD, Rodarte JR. Regional intrapulmonary gas distribution in awake and
anesthetized-paralyzed man. J Appl Physiol 1977; 42:391402 [Context Link]

11. von Euler US, Liljestrand G. Observations on the pulmonary blood pressure in cat. Acta
Physiol Scand 1946; 12:310320 [Context Link]

12. Marshall C, Marshall B. Site and sensitivity for stimulation of hypoxic pulmonary
vasoconstriction. J Appl Physiol 1983; 55:711716 [Context Link]

13. Benumof JL. One-lung ventilation and hypoxic pulmonary vasoconstriction: implications for
anesthetic management. Anesth Analg 1985; 64:821833 [Context Link]
14. Benumof JL. Isoflurane anesthesia and arterial oxygenation during one-lung ventilation.
Anesthesiology 1986; 64:419422 [Context Link]

15. Casthely PA, Lear S, Cottrell JE, Lear E. Intrapulmonary shunting during induced
hypotension. Anesth Analg 1982; 61:231235 [Context Link]

16. Ishibe Y, Marshall C, Marshall BE. Hypoxic pulmonary vasoconstriction inhibited by lung
manipulation in rabbits. Anesthesiology 1986; 69 (No. 3A):A139 Bibliographic Links [Context Link]

17. Benumof JL, Wahrenbrock EA. Blunted hypoxic pulmonary vasoconstriction by increased lung
vascular pressures. J Appl Physiol 1975; 38:846850 [Context Link]

18. Colley PS, Cheney FW, Butler J. Mechanism of change in pulmonary shunt flow with
hemorrhage. J Appl Physiol 1977; 42:196201 [Context Link]

19. Kelman GR, Nunn JF, Prys-Roberts C, Greenbaum R. The influence of cardiac output on
arterial oxygenation: a theoretical study. Br J Anaesth 1967; 39:450458 [Context Link]

20. West JB. Respiratory physiologythe essentials. 5th ed. Baltimore: Williams & Wilkins, 1995
[Context Link]

21. Rosenzweig DY, Hughes JM, Glazier JB. Effects of transpulmonary and vascular pressures on
pulmonary blood volume in isolated lung. J Appl Physiol 1970; 28:553560 [Context Link]

22. Hedenstierna G, White FC, Mazzone R, Wagner PD. Redistribution of pulmonary blood flow in
the dog with PEEP ventilation. J Appl Physiol 1979; 46:278287 [Context Link]

23. Permutt S, Riley RL. Hemodynamics of collapsible vessels with tone: the vascular waterfall. J
Appl Physiol 1963; 18:924932 [Context Link]

24. Landmark SJ, Knopp TJ, Rehder K, Sessler AD. Regional pulmonary perfusion and V/Q in
awake and anesthetized-paralyzed man. J Appl Physiol 1977; 43:9931000 [Context Link]

25. Hughes JM, Glazier JB, Maloney JE, West JB. Effect of extra-alveolar vessels on distribution
of blood flow in the dog lung. J Appl Physiol 1968; 25:701712 [Context Link]

26. Nicolaysen G, Shepard J, Onizuka M, et al. No gravity-independent gradient of blood flow


distribution in dog lung. J Appl Physiol 1987; 63:540545 [Context Link]

27. Greenleaf JF, Ritman EL, Sass DJ, Wood EH. Spatial distribution of pulmonary blood flow in
dogs in left decubitus position. Am J Physiol 1974; 227:230244 [Context Link]

28. Hakim TS, Lisbona R, Dean GW. Gravity-independent inequality in pulmonary blood flow in
humans. J Appl Physiol 1987; 63:11141121 [Context Link]

29. Glenny RW, Robertson HT. Fractal modeling of pulmonary blood flow heterogeneity. J Appl
Physiol 1991; 70:10241030 [Context Link]

30. Hakim TS, Dean GW, Lisbona R. Effect of body posture on spatial distribution of pulmonary
blood flow. J Appl Physiol 1988; 64:11601170 [Context Link]

31. Mure M, Domino KB, Robertson T, et al. Pulmonary blood flow does not redistribute in dogs
with reposition from supine to left lateral position. Anesthesiology 1998; 89:483492 [Context
Link]
32. Walther SM, Domino KB, Glenny RW, Hlastala MP. Pulmonary blood flow distribution in
sheep: effects of anesthesia, mechanical ventilation, and change in posture. Anesthesiology
1997; 87:335342 [Context Link]

33. Walther SM, Domino KB, Glenny RW, et al. Pulmonary blood flow distribution has a hilar-to-
peripheral gradient in awake, prone sheep. J Appl Physiol 1997; 82:678685 [Context Link]

34. Ross DJ, Wu P, Mohsenifar Z. Assessment of postural differences in regional pulmonary


perfusion in man by single-photon emission computerized tomography. Clin Sci (Colch) 1997;
92:8185 [Context Link]

35. Bryan AC, Beerel F, MacLeish H, Zidulka A, Bates DV. Factors affecting regional distribution
of ventilation and perfusion in the lung. J Appl Physiol 1964; 19:395402 [Context Link]

36. Hakim TS, Lisbona R, Dean GW. Effect of cardiac output on gravity-dependent and
nondependent inequality in pulmonary blood flow. J Appl Physiol 1989; 66:15701578 [Context
Link]

37. Tony L. Yaksh, et al, eds. Anesthesia: biologic foundations. Philadelphia: Lippincott-Raven,
1998 [Context Link]

38. Domino KB, Borowec L, Alexander CM, et al. Influence of isoflurane on hypoxic pulmonary
vasoconstriction in dogs. Anesthesiology 1986; 64:423429 [Context Link]

39. Carlsson AJ, Bindslev L, Hedenstierna G. Hypoxia-induced pulmonary vasoconstriction in the


human lung. The effect of isoflurane anesthesia. Anesthesiology 1987; 66:312316 [Context Link]

40. Shirai M, Shindo T, Ninomiya I. beta-Adrenergic mechanisms attenuated hypoxic pulmonary


vasoconstriction during systemic hypoxia in cats. Am J Physiol 1994; 266:H17771785 [Context
Link]

41. Rehder K. Postural changes in respiratory function. Acta Anaesthesiol Scand Suppl 1998;
113:1316 [Context Link]

42. Carlens E. A new flexible double-lumen catheter for bronchospirometry. J Thorac Surg 1949;
18:742746 [Context Link]

43. Gale JW, Waters RM. Closed endobronchial anesthesia in thoracic surgery. J Thorac Surg
1931; 1:432437 [Context Link]

44. Rovenstine E. Anesthesia for intrathoracic surgery: the endotracheal tube and endobronchial
techniques. Surg Gynecol Obstet 1936; 63:325330 [Context Link]

45. Robertshaw F. Low resistance double-lumen endobronchial tubes. Br J Anaesth 1962;


34:576579 [Context Link]

46. Hurford WE, Kolker AC, Strauss HW. The use of ventilation/perfusion lung scans to predict
oxygenation during one-lung anesthesia. Anesthesiology 1987; 67:841844 [Context Link]

47. Rees DI, Wansbrough SR. One-lung anesthesia: percent shunt and arterial oxygen tension
during continuous insufflation of oxygen to the nonventilated lung. Anesth Analg 1982; 61:507
512 [Context Link]

48. Hurford W. Fiberoptic endobronchial intubation. Anesthesiol Clin North Am 1991; 9:97109
[Context Link]
49. Benumof JL, Partridge BL, Salvatierra C, Keating J. Margin of safety in positioning modern
double-lumen endotracheal tubes. Anesthesiology 1987; 67:729738 [Context Link]

50. Brodsky J. Isolation of the lungs. Probl Anesth 1990; 4:264281 [Context Link]

51. Hurford WE, Alfille PH. A quality improvement study of the placement and complications of
double-lumen endobronchial tubes. J Cardiothorac Vasc Anesth 1993; 7:517520 [Context Link]

52. Cohen JA, Denisco RA, Richards TS, et al. Hazardous placement of a Robertshaw-type
endobronchial tube. Anesth Analg 1986; 65:100101 [Context Link]

53. Simon BA, Hurford WE, Alfille PH, et al. An aid in the diagnosis of malpositioned double-
lumen tubes [letter]. Anesthesiology 1992; 76:862863 [Context Link]

54. Wagner DL, Gammage GW, Wong ML. Tracheal rupture following the insertion of a disposable
double-lumen endotracheal tube. Anesthesiology 1985; 63:698700 [Context Link]

55. Brodsky JB, Shulman MS, Mark JB. Airway rupture with a disposable double-lumen tube
[letter]. Anesthesiology 1986; 64:415 Buy Now Bibliographic Links [Context Link]

56. Wells DG, Zelcer J, Podolakin W, et al. Cardiac arrest from pulmonary outflow tract
obstruction due to a double-lumen tube. Anesthesiology 1987; 66:422423 [Context Link]

57. Campos JH, Massa FC. Is there a better right-sided tube for one-lung ventilation? A
comparison of the right-sided double-lumen tube with the single-lumen tube with right-sided
enclosed bronchial blocker. Anesth Analg 1998; 86:696700 [Context Link]

58. Slinger P, Suissa S, Adam J, Triolet W. Predicting arterial oxygenation during one-lung
ventilation with continuous positive airway pressure to the nonventilated lung. J Cardiothorac
Anesth 1990; 4:436440 [Context Link]

59. Katz JA, Laverne RG, Fairley HB, Thomas AN. Pulmonary oxygen exchange during
endobronchial anesthesia: effect of tidal volume and PEEP. Anesthesiology 1982; 56:164171
[Context Link]

60. Bendixen HH, Hedley-White J, Chir B, Laver MB. Impaired oxygenation in surgical patients
during general anesthesia with controlled ventilation: a concept of atelectasis. N Engl J Med
1963; 269:991996 [Context Link]

61. Virtue RW, Permutt S, Tanaka R, Pearcy C, Bane HN, Bromberger-Barn B. Ventilation-
perfusion changes during thoracotomy. Anesthesiology 1966; 27:132146 [Context Link]

62. Khanam T, Branthwaite MA. Arterial oxygenation during one-lung anaesthesia. 1. A study in
man. Anaesthesia 1973; 28:132138 [Context Link]

63. Kacmarek RM. Ventilator-associated lung injury. Int Anesthesiol Clin 1999; 37:4764 [Context
Link]

64. Mathisen DJ, Kuo EY, Hahn C, et al. Inhaled nitric oxide for adult respiratory distress
syndrome after pulmonary resection. Ann Thorac Surg 1998; 66:18941902 [Context Link]

65. Slinger PD, Hickey DR. The interaction between applied PEEP and auto-PEEP during one-lung
ventilation. J Cardiothorac Vasc Anesth 1998; 12:133136 [Context Link]

66. Capan LM, Turndorf H, Patel C, et al. Optimization of arterial oxygenation during one-lung
anesthesia. Anesth Analg 1980; 59:847851 [Context Link]
67. Slinger P, Triolet W, Wilson J. Improving arterial oxygenation during one-lung ventilation.
Anesthesiology 1988; 68:291295 [Context Link]

68. Cohen E, Eisenkraft JB. Positive end-expiratory pressure during one-lung ventilation
improves oxygenation in patients with low arterial oxygen tensions. J Cardiothorac Vasc Anesth
1996; 10:578582 [Context Link]

69. Rossi A, Polese G, Brandi G, Conti G. Intrinsic positive end-expiratory pressure (PEEPi).
Intensive Care Med 1995; 21:522536 [Context Link]

70. Yokota K, Toriumi T, Sari A, et al. Auto-positive end-expiratory pressure during one-lung
ventilation using a double-lumen endobronchial tube [see comments]. Anesth Analg 1996;
82:10071010 [Context Link]

71. Ducros L, Moutafis M, Castelain MH, et al. Pulmonary air trapping during two-lung and one-
lung ventilation. J Cardiothorac Vasc Anesth 1999; 13:3539 [Context Link]

72. Bardoczky GI, d'Hollander AA, Cappello M, Yernault JC. Interrupted expiratory flow on
automatically constructed flow-volume curves may determine the presence of intrinsic positive
end-expiratory pressure during one-lung ventilation. Anesth Analg 1998; 86:880884 [Context
Link]

73. Bardoczky GI, Yernault JC, Engelman EE, et al. Intrinsic positive end-expiratory pressure
during one-lung ventilation for thoracic surgery. The influence of preoperative pulmonary
function [published erratum appears in Chest 1997;111(3):836]. Chest 1996; 110:180184
[Context Link]

74. Alfery DD, Zamost BG, Benumof JL. Unilateral lung lavage: blood flow manipulation by
ipsilateral pulmonary artery balloon inflation in dogs. Anesthesiology 1981; 55:376380 [Context
Link]

75. Tugrul M, Camci E, Karadeniz H, et al. Comparison of volume controlled with pressure
controlled ventilation during one-lung anaesthesia. Br J Anaesth 1997; 79:306310 [Context Link]

76. Dreyfuss D, Saumon G. Ventilator-induced lung injury: lessons from experimental studies.
Am J Respir Crit Care Med 1998; 157:294323 [Context Link]

77. Amato MB, Barbas CS, Medeiros DM, et al. Effect of a protective-ventilation strategy on
mortality in the acute respiratory distress syndrome [see comments]. N Engl J Med 1998;
338:347354 [Context Link]

78. Feihl F, Perret C. Permissive hypercapnia. How permissive should we be? Am J Respir Crit
Care Med 1994; 150:17221737 [Context Link]

You might also like