Objectives: To determine the influence of the dynamic properties of th
e oscillator on the oscillatory volume delivered through the endotrach
eal tube to the lung or lung surrogate (delivered volume) under condit
ions of high-frequency ventilation. In particular, the relation betwee
n the tidal volume of the pump (oscillator) and the delivered volume w
as analyzed. Paco(2) was measured further as a function of the deliver
ed volume in a number of experiments performed with healthy dogs. Desi
gn: Laboratory study. Setting: Engineering and animal laboratory. Subj
ects: Lung surrogates and healthy dogs. Interventions: An experimental
oscillatory system was connected to various lung surrogates. In addit
ion, six beagle dogs received high-frequency ventilation with differen
t delivered volumes during the study. Control of the mean airway press
ure was achieved by a peripheral pressure chamber located at the exhau
st port of the bias now tube. Results: The delivered volume, which is
the quantity of interest from a physiologic point of view, can deviate
considerably from the tidal volume of the pump due to dynamic (partic
ularly resonance) effects. Because the delivered volume and the mean a
irway pressure have to be controlled independently, two independent qu
antities are necessary for control purposes (e.g., the tidal volume of
the pump and the mean pressure at the exhaust port). Furthermore, it
was found that a minimal condition for adequate gas exchange is a deli
vered volume that exceeds the machine-related deadspace. For this reas
on, and in order to maximize the CO2 gradient, the exhaust tube must b
e as short as possible. Conclusions: a) The delivered volume has to be
monitored under clinical conditions; b) however, because the impedanc
e of the endotracheal tube in general considerably exceeds the impedan
ce of the lung, the influence of the impedance of the lung on the deli
vered volume is generally small, and thus an in vitro calibration may
serve as a useful approximation; c) at least two independent quantitie
s are needed for an adequate oscillatory control; d) a necessary (not
necessarily sufficient) condition for adequate CO2 removal is that the
delivered volume must exceed the machine-related deadspace; e) in a c
linical environment involving extremely pathologic lung conditions, e.
g., adult respiratory distress syndrome, mechanical lung characteristi
cs may deviate substantially from those characteristics used in this s
tudy (i.e., the results obtained may not necessarily be applicable und
er all clinical situations).