Determinants of airleak flow during humming V high-frequency oscillatory ventilation in an open-compartment ex-vivo model of airleak.

High-frequency oscillatory ventilation (HFOV) using small tidal volumes and maintaining sufficient end-expiratory lung volume may be beneficial in the treatment of airleak. However, few published guidelines exist to advise clinicians on appropriate ventilator settings in this clinical scenario. The present experiment aimed to determine the effect of frequency, stroke volume (SV) and mean airway pressure (MAP) on airleak from an isolated lung model ventilated with a Humming V HFOV. We performed a crossover non-randomized experiment using the repeated measurement method to test the hypothesis that MAP is the major determinate for airleak. The lungs of 13 healthy juvenile New Zealand white rabbits were isolated and ventilated with high peak pressure to create airleak. The dataset obtained was analyzed using the generalized estimating equation method. We found that airleak flow did not change as frequency was raised from 13 to 17 Hz (P = 0.463) with MAP and SV kept constant. SV was positively correlated to the amount of change in airleak (P < 0.01, coefficients +/- SEM = 1.2 +/- 0.1 ml/min/ml). Leakage flow increased significantly from 275 +/- 168 ml/min to 1,721 +/- 552 ml/min as MAP was increased from 5 cm H(2)O to 30 cm H(2)O (P < 0.001, coefficients +/- SEM = 56.1 +/- 3.0 ml/min/cm H(2)O) while inspiratory flow increased less and amplitude pressure remained about the same. We concluded that MAP (lung volume) was the main independent factor for airleak, whilst SV (tidal volume) exerted a lesser effect. Within the operational range of the Humming V, frequency did not affect airleak.