Resting cuff volumes (the performed just inflated volume) were measured using a calibrated syringe. A mechanical lung (Vent Aid TI L, Michigan Instruments, Inc), which was connected to a lubricated (lightweight household oil, WD-40) 19 x 22-mm model trachea (Imatrach, Mallinckrodt) was used to simulate changes in Cl from 100 to 15 mL/cm HaO. A volume-cycled ventilator (Bear 2) delivered a tidal volume (Vt) of 1,000 ml; inspiratory and expiratory volumes were measured with two rotating vane-type spirometers (Wright) connected in series just proximal to the ETT.
Ten ETTs of each size (7.0- and 8.0-mm ID) and cuff type (LO, MED, HI) were inflated in the trachea at 21°C using a calibrated syringe until a “minimal” leak of 10 percent (100 ml) and then 5 percent (50 ml) of the 1,000 mL Vt occurred (American National Standards Institute defines a minimal leak as 5 percent of the delivered volume.) Airway and ETT intracuff pressures were measured at both the 10 and 5 leak volumes using pressure transducers (Gould-Statham F-50) and a multichannel polygraph recorder (Crass model 7B). Leak volume was ascertained by the difference between the inspiratory and expiratory volumes. Complete occlusion of the trachea was demonstrated by lack of bubble formation in a soap solution coating the proximal trachea-cuff interface and showed this method of leak measurement to be accurate ±2 percent.