The ventilation cycle during high frequency oscillatory ventilation (HFOV) is too short for pressure equilibration. The theoretical advantage of HFOV thus bears the risk of uncertainty about the pressure swings effective at the tracheal level.
Objective To study the 2 mathematical models validated for calculation of intratracheal pressure during conventional ventilation (CV) (Rohrer model, Blasius-Ito model) in a laboratory model of HFOV.
Methods Through an anatomically shaped endotracheal tube of 3.0 and 4.0 mm inner diameter (ETT3, ETT4) a passive lung model (compliance 0.75 ml/cm H2O; resistance 50 cm H2O/(L/s)) was ventilated with the Sensormedics 3100 HFOV ventilator using different combinations of mean airway pressure (MAP), frequency and amplitude. Flow was measured with an ultrasonic device (Spiroson, Isler) and pressure with 2 piezoelectric transducers. For each setting 150 cycles with a probe frequency of 250 Hz were analysed.
Results Pressure drop across the ETT was a function of amplitude, frequency and ETT size, but not of MAP. Using an I:E ratio of 1:2 throughout the study the pressure drop during inspiration/expiration was 28.1–51.8%/23.8–40.7% for ETT3 and 9.5–24.7%/3.3–16.4% for ETT4. Calculation of pressure drop with each of both mathematical models yielded an underestimation of about 22 or 38 % (significant for each amplitude with p<0.001; Kruskal-Wallis test). Tabular nomograms for the ventilator settings studied were established.
Conclusions Other than in CV, pressure drop during HFOV in very small ETTs cannot be calculated by known mathematical models. Nomograms must be used if with small tubes a precise estimation of tracheal pressure is mandatory.