Auto PEEP
Auto PEEP
Auto PEEP
– 1 cm H2O
Auto-PEEP increases
the work of breathing
To overcome the positive
pressure in the alveoli
during inspiration,
the diaphragm must +9
generate enough cm H2O
negative pressure to +8
exceed the auto-PEEP – 3 cm cm H2O
and transmit negative H2O
pressure to the central
airways, generating
airflow.
+ 6 cm H2O
– 13 cm H2O
CCF
©2005
FIGURE 1
Valve closed
pressure
Airway
+
0
- in Valve open
Valve closed
Flow
0
out
Valve open
increasing
Volume
0 Valve closed
Auto-PEEP 0 2 4 6 8 10 12 14 16
is measured Valve open
by occluding
FIGURE 2. Expiratory hold techniques to estimate auto-PEEP. The exhalation valve is
the airway at closed during an expiratory hold at the end of the set expiratory time. When the
end-expiration flow equals zero, airway pressure rises to the auto-PEEP level. With the valve open,
flow continues, and the additional exhaled volume equals the volume of trapped gas.
for several MACINTYRE NR. INTRINSIC PEEP. PROB RESPIR CARE 1991; 4:45, WITH PERMISSION.
seconds
tributes to alveolar pressure, often with nor- ■ CONSEQUENCES OF AUTO-PEEP
mal or even low lung volumes. If the flow per-
sists to the end of the expiratory cycle, there Increases the work of breathing
will be an end-expiratory gradient of alveolar Auto-PEEP causes a considerable increase in
to central airway pressure—an auto-PEEP the resistive and elastic work of breath-
effect without lung distention.13,14 This auto- ing,16,17which may interfere with attempts at
PEEP phenomenon is due to dynamic airway weaning from mechanical ventilation.18 This
collapse with exaggerated expiratory activity. can cause significant discomfort and precipi-
Zakynthinos et al15 demonstrated that in tate patient-ventilator asynchrony.
intubated patients who are spontaneously
breathing and actively exhaling, auto-PEEP Worsens gas exchange
due to expiratory muscle contraction can be Brandolese et al compared the impact of auto-
estimated by subtracting the average expirato- PEEP and external PEEP on pulmonary gas
ry rise in gastric pressure from the end-expira- exchange in mechanically ventilated patients.19
tory airway pressure during airway occlusion. Arterial oxygen tension was lower in patients
804 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 72 • NUMBER 9 SEPTEMBER 2005
with auto-PEEP than in patients with a compa- increase in intrathoracic pressure may
rable level of external PEEP, an effect the falsely increase the pulmonary capillary
authors attributed to a less homogenous distrib- wedge pressure and right atrial pressure,
ution of auto-PEEP among lung units. which can lead to mistakes in hemody-
namic management.
Can cause hemodynamic compromise • Erroneous calculations of static respirato-
Auto-PEEP also has hemodynamic conse- ry compliance: the true value of static
quences. Elevated intrathoracic pressure compliance will be underestimated in the
reduces the preload of the right and left ven- presence of auto-PEEP.19
tricles, decreases left ventricular compliance, • Inappropriate fluid administration or
and can increase right ventricular afterload by unnecessary vasopressor therapy.
increasing pulmonary vascular resistance. This
can lead to hemodynamic compromise.12,20 ■ RECOGNIZING AUTO-PEEP
In a dog model described by Marini et al,21
selective hyperinflation of the lower lobes Four practical clues may suggest the diagnosis
(particularly the right lower lobe) or any dis- of auto-PEEP:
tention of lung tissue adjacent to the right side • Exhalation that continues until the next
of the heart was associated with decreased breath starts, as determined on physical exam-
stroke volume. The decrease in stroke volume ination23 or on graphic display of expiratory
was more closely related to an increase in right flow vs time in a patient on a ventilator that
atrial pressure than in left atrial pressure, is set to deliver a certain number of breaths
implying that impaired venous return was the per minute
dominant cause of reduced cardiac output. • A delay between the start of inspiratory
This mechanism is likely the cause of hypoten- effort and the drop in airway pressure or the
sion in patients with inadvertent PEEP. start of machine-delivered flow in a patient on
Hemodynamic effects of auto-PEEP a ventilator that is set to deliver breaths on
should be considered as a possible reversible demand
cause of pulseless electrical activity. In one • Failure of peak airway pressure to change Suspect
report,22 auto-PEEP may have played a part in when external PEEP is applied auto-PEEP
up to 13 (38%) of 34 patients with electro- • In paralyzed or heavily sedated patients,
mechanical dissociation. reduction of plateau pressure after prolonged if exhalation
During cardiopulmonary resuscitation, exhalation. continues
dynamic hyperinflation can develop in
patients with obstructive airway disease, ■ HOW TO MEASURE AUTO-PEEP until the next
owing to rapid manual ventilation with inad- breath starts
equate time for exhalation. This elevated Static auto-PEEP. Auto-PEEP can be
end-expiratory pressure (auto-PEEP) decreas- accurately measured only in patients without
es venous return and may depress cardiac out- active respiratory effort. It is routinely deter-
put even after a cardiac rhythm has been mined under static conditions by occluding
established. Transient withdrawal of ventila- the airway at end-exhalation. During con-
tion allows the dynamic hyperinflation to trolled mechanical ventilation, reliable quan-
diminish, reducing intrathoracic pressure and tification of auto-PEEP requires an end-expi-
permitting the return of spontaneous circula- ratory hold maneuver, terminating expiratory
tion. flow and allowing equilibration of alveolar
pressure and the airway pressure (FIGURE 2). The
Can lead to inappropriate treatment resulting airway pressure represents the aver-
Failure to recognize auto-PEEP and adjust for age total PEEP present within a nonhomoge-
it can lead to inappropriate treatment in sev- neous lung, and auto-PEEP is calculated by
eral ways: subtracting external PEEP from total PEEP.
• Misinterpretation of central venous and Dynamic auto-PEEP. There is no accept-
pulmonary artery catheter pressure mea- ed, reliable method to measure auto-PEEP in
surements12: the auto-PEEP-induced spontaneously breathing patients. However,
TA B L E 2
Auto-positive end-expiratory
pressure: measured value can Treatment of auto-positive
underestimate true value end-expiratory pressure
Change ventilator settings
End-expiratory airway occlusion pressure Increase expiratory time
(cm H2O) Decrease respiratory rate
Decrease tidal volume
20
PEEP reflects the lowest regional auto-PEEP
and therefore underestimates static auto-PEEP
FIGURE 3. Hypothetical model showing in the presence of heterogenous mechanical
low measured auto-positive end- properties,20 ie, if some airways are blocked
expiratory pressure despite high average and some not, or some parts of the lung are
end-expiratory alveolar pressure as a stiff and others are compliant. Maltias et
consequence of widespread airway al24demonstrated that dynamic auto-PEEP
closure.
considerably underestimates static auto-PEEP
FROM LEATHERMAN JW, RAVENSCRAFT SA: LOW MEASURED INTRINSIC
POSITIVE END-EXPIRATORY PRESSURE IN MECHANICALLY VENTILATED in patients with significant airway obstruction.
PATIENTS WITH SEVERE ASTHMA: HIDDEN AUTO-PEEP.
CRIT CARE MED 1996; 24:541–546, WITH PERMISSION.
In such patients two major problems must be
solved, therefore, to obtain a correct measure
of auto-PEEP: airway occlusion must be syn-
chronized to the end of the expiratory cycle,
an esophageal balloon catheter can be used to and respiratory muscle activity must be sup-
measure the auto-PEEP during unoccluded pressed. On the other hand, tensing of abdom-
breathing in such patients, as the esophageal inal expiratory muscles at end-expiration may
pressure is assumed to be about the same as the cause the measured auto-PEEP to greatly over-
pleural pressure. This is achieved by calculat- estimate the end-expiratory elastic recoil pres-
ing the negative deflection in esophageal pres- sure.14
sure from the start of inspiratory effort to the It is also crucial that the airway occlusion
onset of inspiratory flow. be maintained for several seconds to avoid
This method is based on the assumption gross underestimation of average end-expira-
that the change in esophageal pressure reflects tory alveolar pressure. Some lung units may
the inspiratory muscle pressure required to not communicate with the proximal airway, as
counterbalance the end-expiratory elastic the peripheral airways may be blocked by
recoil of the respiratory system (ie, auto- mucous hypersecretion or increased wall
PEEP). To obtain valid measurements, the thickness, and the alveolar pressure in these
inspiratory and expiratory muscles need to be noncommunicating lung units will not direct-
relaxed at end-expiration.13,14 ly contribute to the pressure measured during
It has been suggested that dynamic auto- airway occlusion (FIGURE 3).25
BRIEF ANSWERS
TO SPECIFIC
Q:
CLINICAL QUESTIONS
NAME
ADDRESS
CITY
STATE ZIP
PHONE EMAIL
Cleveland Clinic Journal of Medicine, 9500 Euclid Ave., NA32, Cleveland, OH 44195
PHONE 216 • 444 • 2661 FAX 216 • 444 • 9385 E-MAIL ccjm@ccf.org