Clinical Commentary
Occult Positive End-Expiratory Pressure in Mechanically
Ventilated Patients with Airflow Obstruction1-2
The Auto-PEEP Effect
IVlechanically-assisted ventilation with
intermittent positive pressure (IPPV)
increases intrathoracic pressure and
can depress cardiac output (QT) by di-
minishing preload (1, 2). Reduction in
QT occurs most predictably in patients
depleted of intravascular volume and in
those ventilated with positive end-
expiratory pressure (PEEP) (1, 3, 4).
However, patients with chronic ob-
structive pulmonary disease (COPD)
are also susceptible to QT depression
during mechanical ventilation, even
when a PEEP valve is not used (5, 6).
In these patients, elevated mean intra-
thoracic pressure and reduced QT have
previously been attributed to retarded
exhalation and enhanced transmission
of airway pressure to the pleural space
through compliant lungs (6, 7). Our
purpose is to document that another
important mechanism can cause mean
intrathoracic pressure to increase dur-
ing IPPV—the spontaneous develop-
ment of positive end-expiratory pres-
sure at the alveolar level.
Although "air-trapping" is well
known to occur in patients with airflow
obstruction, it is less commonly recog-
nized that alveolar pressure may not
decrease to zero at the end of exhala-
tion during IPPV. This "auto-PEEP"
phenomenon occurs with normal venti-
lator operation but is not readily ap-
parent. However, unless considered in
evaluating hemodynamic status, the
auto-PEEP effect may lead to serious
overestimation of effective ventricular
filling pressure, prompting inappropri-
ate fluid restriction and vasopressor
therapy.
In this report, we describe the patho-
physiologic aspects of this phenome-
non and propose a simple bedside
maneuver to detect and quantify alveo-
lar pressure (PL) at end-exhalation.
Several representative cases are pre-
166
SUMMARY Alveolar pressure can remain positive throughout the ventilatory cycle of mechani-
cally-ventilated patients with airflow obstruction, even when positive end-expiratory pressure
(PEEP) is not applied intentionally. The increase of intrathoracic pressure associated with this
"auto-PEEP" phenomenon can severely depress cardiac output as well as elevate the end-
expiratory pulmonary artery wedge pressure. Such effects may be exaggerated in patients with
chronic obstructive pulmonary disease because abnormally compliant lungs transmit a high
fraction of alveolar pressure to intrathoracic vessels. Failure to recognize the hemodynamic con-
sequences of auto-PEEP may lead to inappropriate fluid restriction or unnecessary vasopressor
therapy. Although not apparent during normal ventilator operation, the auto-PEEP effect can be
detected and quantified by a simple bedside maneuver: expiratory port occlusion at the end of the
set exhalation period. AM REV RESPIR DIS 1982; 126:166-170
sented to demonstrate the use of our sure (BP) in the Trendelenberg position
method as well as to illustrate the im- increased to 80 mmHg, systolic. Arteri-
portance of auto-PEEP in the clinical al blood gas analysis indicated appro-
management of patients with ventila- priate alveolar ventilation: pH, 7.44;
tory failure. The first presentation Paco 2 , 53 mmHg; arterial oxygen ten-
details the methodology used in all of sion (Pao2), 71 mmHg. A chest roent-
these cases. genogram demonstrated proper endo-
Patient 1. A 64-yr-old woman with- tracheal tube position without pneu-
out known cardiac disease was hospi- mothorax or other notable change
talized for an acute exacerbation of from previous radiographs. Intraven-
chronic airflow obstruction. When ously administered dopamine HC1 (18
clinically stable, her FEVi had been jug/kg/min) was titrated to maintain
20% of predicted (370 ml) and she had systolic BP at 90 mmHg during place-
radiographic evidence of bullous lung ment of a Swan-Ganz pulmonary
disease. Respiratory acidosis developed artery catheter. An initial end-
despite aggressive medical therapy with expiratory pulmonary artery wedge
oxygen, bronchodilators, and cortico- pressure (Pw) (8, 9) of 20 mmHg was
steroids (arterial pH, 7.32; carbon di- interpreted by her physician to reflect
oxide tension, (Paco2) 65 mmHg; cal- adequate left ventricular filling pres-
culated bicarbonate, 33 mEq/L); IPPV sure, and no further fluid challenges
was initiated using a volume-cycled were given. The Pw tracing had an ap-
ventilator (MA-1; Bennett Respiration propriate wave form ("a" and "v"
Products, Los Angeles, CA) set to deli- waves present), reflecting the Zone III
ver a 400-ml (11 ml/kg body weight) condition of West (10).
effective tidal volume (VT) with a
back-up frequency of 12/min in the
{Received in original form July 31, 1981 and in
"assist/control" (AMV) mode. Set in- revised form January 4, 1982)
spiratory flow rate and inspired oxygen
fraction were 90 L/min and 0.3, re- 1
Presented in part at the Annual Meeting of
spectively. Shortly after IPPV initia- the American Thoracic Society, Detroit, Michi-
tion, severe systemic hypotension (60 gan, May 1981.
2
mmHg, systolic) was discovered. After Requests for reprints should be addressed to
John J. Marini, M.D., Respiratory Diseases
a rapid intravenous infusion of 500 ml Division, Seattle Public Health Hospital, 1131
of normal saline, systemic blood pres- 14th Avenue South, #1031, Seattle, WA 98114.
CLINICAL COMMENTARY 167

.E 50
'- rupted suddenly by onset of the in- /-expiratory port open
| 3.0- 437 * to atmosphere
* QT spiratory phase. When the ventilator
1.0- 2.62 2S1 ' rate was momentarily set to zero, the
^50- spirometer filled with 500 ml of gas in
100- addition to VT before passive expira-
50 BP tory flow ceased completely. Failure of
WEDGE the lungs to empty to their relaxed SEVERE OBSTRUCTION
expiratory port open
volume at the end of the set expiratory
ESOPH
period implied that PL remained
positive at end-expiration, even though
a PEEP valve was not in use.
We reasoned that one way to quanti-
fy PL at the end of the set exhalation
Fig. 1. Effect of temporary discontinuation of in- SEVERE OBSTRUCTION
termittent positive pressure ventilation (IPPV) on
period would be to delay a ventilator- expiratory port occluded
cardiac output (QT), systemic blood pressure (BP), delivered breath and to occlude the ex-
wedge and esophageal (ESOPH) pressures in a pa- haust port of the ventilator when the
tient with severe airflow obstruction (Patient 1). next inflation would have begun. With
expiratory gas flow blocked, pressure
in the tubing would equilibrate with
After placement of a radial artery PL, allowing PL to register on the ven-
Fig. 2. Measurement of auto-PEEP by expiratory
catheter, it was observed that systolic tilator manometer (VM) (figure 2). The port occlusion. Normally (see top panel), alveolar
BP increased consistently to 140 measured pressure (Pvm) would approx- pressure is atmospheric at the end of passive ex-
mmHg within 30 s of a brief interrup- imate PL during the ventilatory cycle so halation. With severe airflow obstruction (see mid-
tion of IPPV, whereas Pw decreased to long as the patient did not cough or dle panel) alveolar pressure remains elevated (in
10 to 13 mmHg. Simultaneously, mean make other unusual respiratory efforts. this example at 15 cmH 2 0) and slow flow con-
tinues, even at the end of the set exhalation
pulmonary artery pressure (Ppa) and Accordingly, we lowered the ventilator period. The ventilator manometer senses negli-
heart rate decreased. Whenever IPPV frequency from its set rate of 12/min gible pressure because it is open to atmosphere
was reinstituted, heart rate, BP, Pw, to 0/min, occluding the expiratory port through large bore tubing and downstream from
and Ppa returned to their previous 5 s after initiation of the last ventila- the site of flow limitation. With gas flow stopped
tordelivered breath, and Pvm increased by occlusion of the expiratory port at the end of
values (figure 1 and table 1). the set exhalation period (see lower panel),
Ventilator operation appeared nor- suddenly from 0 to 14 cm H 2 0. When pressure equilibrates throughout the lung-
mal despite the clear association of the occlusion was released, Pvm de- ventilator system and is displayed on the ven-
hemodynamic compromise with IPPV. creased immediately to 0 cmH 2 0. The tilator manometer.
Although not paralyzed, the patient magnitude of the pressure increase
was obtunded and the rhythmic breath- diminished gradually during subse-
ous ventilatory efforts occurred during
ing pattern was well-coordinated with quent occlusions performed serially
this period.
machine cycling. It was noted, how- over the next 15 s. Eventually an in-
ever, that the ventilator's spirometer crease could no longer be measured, in- To confirm that this "auto-PEEP"
continued to fill slowly throughout the dicating that the lungs had finally emp- effect caused an increase of intrathor-
entire expiratory period, until inter- tied to a relaxed volume. No spontane- acic (pleural) pressure at end-exhala-
tion, an esophageal balloon was in-
serted with the patient in a 45° semiup-
TABLE 1
right position. The structure, air
volume, and placement of the balloon
AIRWAY, ESOPHAGEAL, AND HEMODYNAMIC MEASUREMENTS BEFORE AND AFTER
A 30-S INTERRUPTION OF MECHANICAL VENTILATION*
followed standard recommendations
(11). On repeated determinations,
Patient 1 Patient 2 Patient 3 esophageal pressure (Pes) at end-exha-
During Without During Without During Without lation during IPPV was 10 to 11 cm
IPPV IPPV IPPV IPPV IPPV IPPV H 2 0 and decreased to 3 to 4 cm H 2 0
HR, beats/min 125 110 128 128 142 112 within 20 s after interruption of IPPV
QT, Umint 2.6 4.4 10.8 13.5 4.8 12.3 (figure 1). Occluded transpulmonary
BP, mmHg* 90§ 140§ 115 155 11011 225H pressure (Pvm-Pes) also decreased, re-
Ppa, mmHg 34 27 42 44 45 38 flecting the associated reduction in
Pw, mmHg 20 12 18 14 _1 _1
Pes, cmH 2 0 10 3 0 -5 18 2
lung volume (table 1).
Pvm, cmH 2 0 14 0 10 0 26 0 From these observations we con-
Ptp, cmH 2 0 4 -3 10 5 8 -2 cluded that depressed cardiac output
Definition of abbreviations: IPPV = intermittent positive pressure ventilation; HR = heart rate; QT = cardiac output;
was explained on the basis of continu-
BP = blood pressure; Ppa" = mean pulmonary artery pressure; Pw = pulmonary artery wedge pressure; Pes = ously elevated intrathoracic pressure
esophageal pressure; Pvm = pressure reading on the ventilator manometer after expiratory port occlusion; Ptp = trans- and impeded venous return. Diminished
pulmonary pressure (Pvm - Pes) during expiratory port occlusion. preload was not recognized initially,
* All measurements made at end-exhalation or apnea.
t Thermodilution technique (mean of 3 serial determinations). however, because of a misleading Pw.
* Systolic pressure. Surrounded by increased intrathoracic
§ Intravenous infusion rate of dopamine HCI of 18 ug/kg/min. pressure but referenced to atmosphere,
II Intravenous infusion rate of dopamine HCI of 12 ug/kg/min.
1A satisfactory Pw tracing could not be obtained during this period of data collection. Pw seriously overestimated effective
168
left ventricular filling pressure during
IPPV, even when measured at end-
exhalation.
Intermittent mandatory ventilation
(IMV) was instituted in an attempt to
lower mean pleural pressure (12). Ad-
ditional intravascular fluids were ad-
ministered, BP stabilized at 130/75
mmHg, and mental alertness improved.
At an IMV rate of 8, end-expiratory
Pw decreased to 14 mmHg, despite
volume repletion. Arterial blood gas
measurements did not change signifi-
cantly.
Patient 2. A 61-yr-old man with
COPD underwent laparotomy for in-
traabdominal sepsis and returned to
the intensive care unit oliguric while
being ventilated mechanically. Back-up
frequency, VT, and inspiratory flow
rate settings were 10/min, 850 ml (11.5
ml/kg), and 90 L/min, respectively
(AMV mode). The patient "triggered"
an additional 5 ventilator-delivered
breaths/min. Arterial blood gas analy-
sis: pH, 7.32; Paco2> 48 mmHg; Pao2,
94 mmHg (Fio2 = 0.4). Expiratory
port occlusion performed 4 s after the
onset of a ventilator-initiated breath
caused Pvm to increase from 0 to 10
cm H 2 0 whenever the patient made no
respiratory effort during the maneu-
ver. Although apparently adequate
during IPPV, vascular pressures and
cardiac output increased markedly
after temporary discontinuation of
ventilatory support (table 1). After
intravascular volume expansion and
institution of IMV, BP and urinary
output increased, notwithstanding re-
duced end-expiratory wedge pressure
and unchanging arterial blood gas
measurements.
Patient 3. A 59-yr-old man without
previous lung disease received burns
over 70% of his body surface and ex-
tensive thermal injury to the pharynx.
Progressive wheezing, dyspnea, and
disorientation prompted mechanical
ventilation to be instituted; IPPV was
administered in the "control" mode,
because sedatives and paralytic agents
were needed to prevent agitation. Dur-
ing the next 12 h increasing ventilator
rates were required to maintain accept-
able blood pH. As ventilation demands
increased, BP and urine output de-
creased. Ventilator settings at this time
included: VT, 1,200 ml (11 ml/kg); fre-
quency, 25/min; inspiratory flow rate,
110 L/min. At first, hypotension and
diminished urine output were attrib-
uted to a cardiogenic origin, in light of
an admission carboxyhemoglobin frac-
tion of 58%, an electrocardiogram sug-
gestive of coronary ischemia, and an
initial end-expiratory Pw of 20 mmHg.
An infusion of dopamine HC1 was
begun. Subsequently, an expiratory
port occlusion maneuver performed at
end-exhalation demonstrated an auto-
PEEP level of 26 cmH 2 0. Circulatory
status improved markedly during a
30-s interruption of IPPV (table 1). Ti-
tration of blood pH with sodium bicar-
bonate was suggested to decrease
minute ventilation, and hemodynamics
improved at lower ventilator rates.
Myocardial infarction was ruled out
by serial electrocardiogram and crea-
tine phosphokinase isoenzyme deter-
minations.
Discussion
These cases demonstrate that PL can
remain positive throughout the ventila-
tory cycle of patients receiving IPPV,
even when PEEP is not applied inten-
tionally. Unrecognized auto-PEEP can
depress cardiac output, confuse inter-
pretation of central vascular pressures,
and lead to unnecessary vasopressor
therapy in patients with severe airflow
obstruction, as illustrated by Patient 1.
The case of Patient 2 emphasizes that
occult PEEP may impair circulatory
function despite apparently adequate
blood pressure and cardiac output.
Hemodynamic compromise by this me-
chanism is not limited to patients with
extraordinary lung compliance; the
third case demonstrates that auto-
PEEP can develop whenever the minute
ventilation is great enough that the
lung cannot empty to its usual relaxed
volume between inflations.
Immediately after the initiation of
IPPV in patients with airflow obstruc-
tion, passive expiratory flow is often
too slow to evacuate all of the selected
tidal volume before the ventilator de-
livers the next breath. As a result, lung
volume and PL at the end of the exha-
lation period tend to build with each
subsequent cycle. At higher volumes,
increased tissue recoil enhances the
driving force for expiratory flow and
tethers open obstructed airways, lower-
ing airflow resistance (13). Within sev-
eral cycles, lung volume stabilizes at a
level at which the entire selected tidal
volume can be evacuated; the cost,
however, is increased PL at end-
exhalation.
As a first approximation, the normal
thorax can be considered as a single
compartment that empties passively at
a changing rate, proportional to
CLINICAL COMMENTARY
volume (14). The volume of this com-
partment during passive exhalation ap-
proximates an exponential function of
time, whose rate of decay can be char-
acterized by a time constant: T =
resistance x compliance (15, 16). In
COPD, a more complex model involv-
ing multiple compartments with differ-
ent time constants may be required to
describe passive exhalation (15, 17-19).
Significant gas trapping has been
demonstrated to occur during mechan-
ical ventilation of anesthetized normal
subjects if three to four time constants
do not elapse between ventilatory cy-
cles (16). Hence, an auto-PEEP effect
is likely to develop if exhalation is pro-
longed by increases in thoracic compli-
ance or resistance, or if the time allowed
for exhalation is shortened by a high
ventilatory requirement. In clinical
practice, patients with COPD who
need ventilatory support experience
auto-PEEP commonly; the slowly
emptying compartments of obstructed
lungs have lengthy time constants
(17-19), and the ventilatory require-
ment is increased by high dead-space
fractions.
Forceful contraction of the thoraco-
abdominal musculature can further
augment PL at end-exhalation. How-
ever, if effort is not timed precisely to
cease just as the machine-initiated in-
spiratory phase begins, the set pressure
limit of the ventilator will be exceeded,
terminating air delivery prematurely.
Conversely, an air-trapping patient
ventilated in the "assist" mode momen-
tarily diminishes PL at end-exhalation
to trigger inspiration (e.g., Patient 2).
However, such brief negative pressure
deflections cannot exert a major influ-
ence on mean intrathoracic pressure
because only a minimal change in air-
way pressure (2 to 3 cmH20) is re-
quired to initiate an "assisted" breath.
At the bedside, auto-PEEP should
be suspected if exhalatory gas flow
continues perceptibly until interrupted
by the inflation phase of the ventilator
cycle. Open to atmosphere and down-
stream from the site of flow limitation,
the ventilator manometer normally
gives no indication of alveolar pressure
during exhalation (figure 2b). However,
PL can be quantified by occluding the
expiratory port, as described in patient
1. With expiratory flow blocked, Pvm
increases to equilibrate with PL (figure
2c). If measured at the appropriate
time, expiratory port occlusion will ap-
proximate the average end-expiratory
alveolar pressure during ventilator
CLINICAL COMMENTARY
operation, provided that spontaneous
respiratory efforts do not occur during
the maneuver and that no gas flows
from a supplemental source into the
system (such as that used in some IMV
circuits or during nebulization of in-
haled medications). Measured pressure
will slightly underestimate the true
alveolar pressure at end-exhalation, to
the extent that alveolar gas decom-
presses into the compliant portions of
the lung and ventilator circuit.
In COPD, uncorrected Pw at end-
exhalation has generally been con-
sidered to be a valid index of ventricu-
lar filling (20). However, transmural,
rather than intraluminal, pressure
determines the effective distending
force applied to the heart and intra-
thoracic vessels (21). An unsuspected
increase in intrathoracic pressure caused
by auto-PEEP can elevate Pw despite
normal or low transmural pressure,
falsely assuring adequate preload and
misdirecting hemodynamic manage-
ment. Similar reasoning applies when
central venous pressure is used to
monitor volume status. Although Pes
may reflect local pleural pressure ac-
curately (22), its validity as an index of
mean intrathoracic pressure is posi-
tion-dependent. Furthermore, changes
in Pes underestimate pressure changes
in the space immediately surrounding
the heart as the lungs distend, especial-
ly in supine positions (23-25). This per-
haps explains why Pw-Pes (an estimate
of transmural left atrial pressure) failed
to increase as mechanical ventilation
was discontinued and cardiac output
improved (Patients 1 and 2).
The hemodynamic consequences of
auto-PEEP may be worse than an equi-
valent degree of PEEP that is inten-
tionally applied. At a specified amount
of end-expiratory pressure, patients
with retarded exhalation maintain
higher mean intrathoracic pressure
than nonobstructed patients because
airway (and consequently intra-
thoracic) pressure decreases more slow-
ly to end-expiration (1, 26). Further-
more, in contrast to the stiff lungs to
which PEEP is usually applied, com-
pliant lungs transmit a high fraction of
alveolar pressure to intrathoracic
vessels (27). With normal lungs and
chest wall, one-third to one-half of air-
way pressure at end-expiration is sensed
in the pleural space, as measured by an
esophageal balloon (28). Two lines of
evidence from the cases presented here
suggest that intrathoracic pressure in-
creases by a high-normal fraction of
the degree of measured auto-PEEP.
First, Pes decreased by 50% or more of
the port-occlusion pressure when IPPV
was discontinued. Second, end-expi-
ratory Pw (expressed in cm H 2 0)
decreased by approximately 50% of
the auto-PEEP when IPPV was inter-
rupted even though the simple dif-
ference between Pw measurements
with and without IPPV underestimates
the degree of transmitted airway
pressure. (Any decrease in intra-
thoracic pressure should also improve
venous return, reducing the decrease in
Pw.)
The hemodynamic effects of auto-
PEEP should be treated primarily by
maneuvers that lower mean intra-
thoracic pressure. In addition to
medical therapy aimed at decreasing
airflow obstruction, inspiratory flow
rate should be adjusted to allow max-
imal time for exhalation between cycles
(19), and minute ventilation should be
reduced to the minimal amount consis-
tent with an acceptable blood pH. Cor-
rection of fever, agitation, and meta-
bolic acidosis is indicated to diminish
the ventilatory requirement.
Because of the alinear relationship
of airway resistance to lung volume,
passive exhalation may slow con-
siderably near FRC, especially in pa-
tients with severe airflow obstruction
(29). Hence, at a specified minute ven-
tilation, higher tidal volume would be
expected to minimize gas trapping at
end-exhalation (5, 29). Although man-
ipulation of ventilatory pattern may be
justified on theoretical grounds, in our
experience increasing tidal volume has
not usually been effective so long as
minute ventilation remains unchanged.
In contrast, using IMV has proved a
useful strategy; the negative pleural
pressure generated during spontaneous
inspiration lowers mean intrathoracic
pressure and enhances systemic venous
return (12).
The auto-PEEP phenomenon is not
unique to the patients presented in this
report. Although we have not yet for-
mally assessed its incidence, some de-
gree of auto-PEEP has been found in
the majority of our patients with severe
airflow obstruction who require me-
chanical assistance for ventilatory fail-
ure; serious hemodynamic conse-
quences occur in a small but significant
proportion of these. The frequency
and importance of this effect in any
specific clinical setting will depend
upon the proportion of mechanically-
ventilated patients predisposed to its
169
development. Because of the simplicity
of the bedside maneuver used to detect
auto-PEEP, we recommend perfor-
mance of port occlusion at end-expira-
tion as part of routine ventilatory care.
PAUL E. PEPE, M.D.
JOHN J. MARINI, M.D.
Division of Respiratory Diseases
Department of Medicine
University of Washington
Seattle, Washington
Acknowledgment
The writers t h a n k Robert G. Petersdorf,
M . D . and Leonard D . H u d s o n , M . D . for
their support and assistance, and Michele E .
Bundy for typing the manuscript.
References
1. Cournand A, Motley HL, Werko L,
Richards DW Jr. Physiologic studies of the ef-
fects of intermittent positive pressure breathing
on cardiac output in man. Am J Physiol 1948;
152:162-74.
2. Fewell JE, Abendschein DR, Carlson CJ,
Murray JF, Rapaport E. Continuous positive-
pressure ventilation decreases right and left ven-
tricular end-diastolic volumes in the dog. Circ
Res 1980; 46:125-32.
3. Morgan BC, Crawford EW, Guntheroth
WG. The hemodynamic effects of changes in
blood volume during intermittent positive-pres-
sure ventilation. Anesthesiology 1969; 30:
297-305.
4. Qvist J, Pontoppidan H, Wilson RS, Lowen-
stein E, Laver MB. Hemodynamic responses to
mechanical ventilation with PEEP: the effect of
hypervolemia. Anesthesiology 1975; 42:45-55.
5. Hedley-Whyte J, Pontoppidan H, Morris
MJ. The response of patients with respiratory
failure and cardiopulmonary disease to different
levels of constant volume ventilation. J Clin In-
vest 1966;45:1543-54.
6. Pontoppidan H, Geffin B, Lowenstein E.
Acute respiratory failure in the adult. N Eng J
Med 1972; 287:690-7, 743-52, 799-806.
7. Stead WW, Fry DL, Ebert RV. The elastic
properties of the lung in normal man and in pa-
tients with chronic pulmonary emphysema. J
Lab Clin Med 1952; 40:674-81.
8. Berryhill RE, Benumof JL, Rauscher LA.
Pulmonary vascular pressure reading at the end
of exhalation. Anesthesiology 1978; 49:365-8.
9. Maran AG. Variables in pulmonary capillary
wedge pressure: variation with intrathoracic
pressure, graphic and digital recorders. Crit Care
Med 1980; 8:102-5.
10. Tooker J, Huseby J, Butler J. The effect of
Swan-Ganz catheter height on the wedge pres-
sure-left atrial pressure relationship in edema
during positive-pressure ventilation. Am Rev
RespirDis 1978; 117:721-5.
11. Macklem PT. Measurement of esophageal
pressure. Measurement of lung mechanics. In:
Procedures for standard measurement of lung
mechanics. Bethesda, MD. Division of Lung Dis-
eases, National Heart and Lung Institute, 1974.
12. Downs JB, Douglas ME, Sanfelippo PM,
Stanford W, Hodges MR. Ventilatory pattern,
intrapleural pressure, and cardiac output.
170 CLINICAL COMMENTARY

Anesth Analg (Cleve) 1977; 56:88-96.
13. Murray JF. The normal lung. Philadelphia:
W.B. Saunders Co., 1976:98-104.
14. Nunn JF. Applied respiratory physiology.
2nd ed. Boston: Butterworths, 1977:145, 465.
15. Mead J, Lindgren I, Gaensler EA. The me-
chanical properties of the lungs in emphysema. J
Clin Invest 1955; 34:1005-16.
16. Bergman NA. Intrapulmonary gas trapping
during mechanical ventilation at rapid frequen-
cies. Anesthesiology 1972; 37:626-33.
17. Briscoe WA, Cournand A. Uneven ventila-
tion of normal and diseased lungs studied by an
open-circuit method. J Appl Physiol 1959; 14:
284-90.
18. Otis AB, McKerrow CB, Bartlett RA, et al.
Mechanical factors in distribution of pulmonary
ventilation. J Appl Physiol 1956; 8:427-43.
19. Connors AF Jr, McCaffree DR, Gray BA.
Effect of inspiratory flow rate on gas exchange
during mechanical ventilation. Am Rev Respir
Dis 1981; 125:537-43.
20. Rice DL, Awe RJ, Gaasch WH, Alexander
JK, Jenkins DE. Wedge pressure measurement in
obstructive pulmonary disease. Chest 1974; 66:
628-32.
21. Weber KT, Janicki JS, Shroff S, Fishman
AP. Contractile mechanics and interaction of the
right and left ventricles. Am J Cardiol 1981; 47:
686-95.
22. Gillespie DJ, Lai YL, Hyatt RE. Compari-
son of esophageal and pleural pressures in the
anesthetized dog. J Appl Physiol 1973; 35:
709-13.
23. Prewitt RM, Wood LDH. Effect of positive
end-expiratory pressure on ventricular function
in dogs. Am J Physiol 1979; 236:534-44.
24. Knowles JH, Hong SK, Rahn H. Possible
errors using esophageal balloon in determination
of pressure-volume characteristics of the lung
and thoracic cage. J Appl Physiol 1959; 14:
525-30.
25. Marini JJ, Culver BH, Butler J. Effect of
positive end-expiratory pressure on canine ven-
tricular function curves. J Appl Physiol 1981;
51(6):1367-74.
26. Colgan FJ, Barrow RE, Fanning GL. Con-
stant positive-pressure breathing and cardiore-
spiratory function. Anesthesiology 1971; 34:
145-51.
27. Chapin JC, Downs JB, Douglas ME, Mur-
phy E J, Ruiz BC. Lung expansion, airway pres-
sure transmission, and positive end-expiratory
pressure. Arch Surg 1979; 114:1193-7.
28. Cassidy SS, Eschenbacher WL, Robertson
CH Jr, Nixon JV, Blomqvist G, Johnson RL Jr.
Cardiovascular effects of positive-pressure venti-
lation in normal subjects. J Appl Physiol. 1979;
47:453-61.
29. Bergman NA. Properties of passive exhala-
tions in anesthetized subjects. Anesthesiology
1969; 30:378-87.

Necrology
NECROLOGY LISTING (OCTOBER 1981— APRIL 1982)

LELAND C. ALBERTSON, M.D. DANIEL S. LESZKIEWICZ, M.D.

Tempe, Arizona Medina, Ohio
CARMEN T. BOSCOE, M.D. KATHARINE PARDEE, M.D.
Stockton, California Cranston, Rhode Island
ABE BLAJWAS, M.D. HIKOJIRO SATO, M.D.
Washington, D.C. Nakano-ku, Tokyo, Japan
DANIEL W. DICKINSON, M.D. ALEKSANDER SKOWRON, M.D.
Wheeling, West Virginia Los Angeles, California
RENE J. DUBOS, M.D. MAURICE SONES, M.D.
New York, N.Y. Philadelphia, Pennsylvania
WILLIAM F. T. KELLOW, M.D. WILSON WEISEL, M.D.
Philadelphia, Pennsylvania Milwaukee, Wisconsin
ELMER A. LARSEN, M.D.
Denver, Colorado