Teleologically it makes sense that if there is no or only a small air
leak present that the endotracheal tube may be stenting the airway
open, and if the tube is removed, loss of the airway may occur.
However, other factors may affect leak size. It may depend not only on
the leak area but also on the resistance in the endotracheal tube, the
exhaled limb of the breathing circuit, and the mechanical ventilator.
Tubes that are partially occluded with secretions will have a higher
resistance to airflow and hence, an increase in the cuff-leak.
Patients in this study typically had endotracheal tubes with minimal
secretions. All three ventilators (Bird 8400 STi8and
Adult Star9used here and the Puritan Bennett 7200[
Carlsbad, CA]10 used by Miller and Cole3)
have very little expiratory airflow resistance and hence, would not
contribute to the leak. Additionally, when in assist-control mode, the
three ventilators deliver the preset Vt. They make no
correction for any leak.,8–10 Because the cuff-leak test
was conducted immediately before extubation, all patients in this study
had spontaneous respiration. Even though the ventilator was in
assist-control mode, patients may have had sufficient respiratory drive
to inhale around the endotracheal tube. The effect of this on the
cuff-leak test has not been quantified and needs further study. Whereas
several studies1–2,6 performed the cuff-leak test
immediately before extubation, Miller and Cole3 performed
the test within 24 h of extubation. Their patients may have been
more sedated and may not have breathed around the endotracheal tube.