Auto-PEEP

BEWARE OF AUTO PEEP IN PRESSURE CONTROLLED VENTILATION

Pressure controlled ventilation is frequently used in hypoxemic patients in acute respiratory failure. The principle method of controlling hypoxia is to increase mean airway pressure by prolonging inspiration. Although this is effective, often physicians concurrently attempt to control rising PaCO2 by increasing the respiratory rate.

Clinical Scenario:
You are called to a patient on pressure controlled ventilation PEEP 15, IP 20, 1.0 second inspiratory time, respiratory rate 32. His SpO2 is measuring 88%
Blood gas reveals a severe respiratory acidosis, PaO2 55mmHg, PaCO2 88mmHg.
1 hour previously on the basis of the following gas:
PaO2 60mmHg, PaCO2 60mmHg
One of the other residents instructed that the respiratory rate be increased from 28 to 32 breaths per minute (no other changes were made).
At the time of the blood gas, the patient has a tidal volume of 400ml; now the patient's tidal volume is 260ml.

What do you think?

The mistake here is to assume that the problem is a reduction in lung compliance, and to increase the driving pressure without first looking at the respiratory waveforms.

The image above (not the patient in question) demonstrates the classic flow pattern of auto-PEEP - positive flow represents inspiration and negative flow expiration - the next breath commences before the last breath finishes. In this case the cause is the rapid respiratory rate.

The waveform demonstrates a considerable amount of auto-PEEP, which measures 9cmH2O. The ventilator is set to deliver flow between the set PEEP (P1)  and the pressure limit (PL). The presence of unexpected positive pressure within the airways curtails the driving pressure (and the tidal volume). The cartoon below demonstrates this phenomenon: prior to changing this patients vent settings (let us assume) the auto-PEEP is 0, and the tidal volume delivered for the driving pressure is 400ml. Now run your mouse over the image to see what happens with the higher respiratory rate. This patient has 9cmH2O of auto-PEEP, so the breath starts from a platform of 24cmH2O rather than 15cmH2O: the auto-PEEP eats into the driving pressure (reducing it to 11cmH2O), and thus the tidal volume falls. The patient does not develop severe hypoxemia because the mean airway pressure is high, and the lower PaO2 probably reflects higher alveolar CO2.

This situation occurs only in pressure limited modes of ventilation, and should not be confused with excessive PEEP, in which the patient is ventilated along the non-compliant part of the volume pressure curve. This can occur in either pressure control or volume control. Indeed, in volume control airway pressure is variable, so the ventilator will deliver the tidal volume above whatever end expiratory volume is already in the lung - PEEP plus auto-PEEP. The result is very high airway pressures. The blood gas may, however, look very good. The uncontrolled intrathoracic pressure may cause hemodynamic upset, and alveolar stretch injury. Run your mouse over the picture below to see how tidal volumes fall from excessive PEEP.

Note that the driving pressure is the same in both cases, the difference is the position on the volume-pressure curve.

The solution to the conundrum above is to reduce either the respiratory rate or the inspiratory time. We know that the patient's PaO2 was marginal before the respiratory rate was changed: long inspiratory times facilitate oxygenation, thus it would be unwise to reduce inspiratory time in this case. Return the respiratory rate to 28, and recheck the auto-PEEP. If necessary, reduce the rate further.

Won't that cause in increase in PaCO2?

No, the alveolar minute ventilation will increase from the reduction in auto-PEEP, and there will also probably be an reduction in alveolar dead space (less zone 1) (click here for more).
The situation with excessive PEEP is the same: ventilating the patient over the more compliant part of the volume-pressure curve will increase alveolar ventilation and reduce alveolar dead space.

       
               
     

Copyright Patrick Neligan 2002