Expiratory flow limitation and dynamic pulmonary hyperinflation during high-frequency ventilation.

Dynamic hyperinflation of the lungs occurs during high-frequency oscillatory ventilation (HFOV) and has been attributed to asymmetry of inspiratory and expiratory impedances. To identify the nature of this asymmetry, we compared changes in lung volume (VL) observed during HFOV in ventilator-dependent patients with predictions of VL changes from electrical analogs of three potential modes of impedance asymmetry. In the patients, when a fixed oscillatory tidal volume was applied at a low mean airway opening pressure (Pao), which resulted in little increase in functional residual capacity, progressively greater dynamic hyperinflation was observed as HFOV frequency, (f) was increased. When mean Pao was raised so that resting VL increased, VL remained at this level during HFOV as f was increased until a critical f was reached; above this value, VL increased further with f in a fashion nearly parallel to that observed when low mean Pao was used. Three modes of asymmetric inspiratory and expiratory impedance were modeled as electrical circuits: 1) fixed asymmetric resistance [Rexp greater than Rinsp]; 2) variable asymmetric resistance [Rexp(VL) greater than Rinsp, with Rexp(VL) decreasing as VL increased]; and 3) equal Rinsp and Rexp, but with superimposed expiratory flow limitation, the latter simulated using a bipolar transistor as a descriptive model of this phenomenon. The fixed and the variable asymmetric resistance models displayed a progressive increase of mean VL with f at either low or high mean Pao. Only the expiratory flow limitation model displayed a dependence of dynamic hyperinflation on mean Pao and f similar to that observed in our patients. We conclude that expiratory flow limitation can account for dynamic pulmonary hyperinflation during HFOV.

The effects of passive CO2 removal on breathing pattern in humans.

To determine the effects of nonventilatory CO2 transfer on breathing pattern, we monitored breathing in 7 patients with renal failure undergoing hemodialysis. A respiratory inductance plethysmograph was used to record ventilation before and during dialysis. The duration of inspiration (TI), the duration of each breath (TTot) and the duty cycle (TI/TTot) did not differ for the pre-dialysis and the dialysis periods. In contrast, for each patient the mean tidal volume (VT) fell significantly during dialysis (p less than 0.05), accounting for the reduction in minute ventilation (p less than 0.005). The mean inspiratory flow rate (VT/TI) also fell (p less than 0.01), demonstrating that nonventilatory CO2 loss via the dialysis bath is associated with reduced respiratory drive.

Effect of bias flow rate on gas transport during high-frequency oscillatory ventilation.

Ventilatory support with low tidal volume, high-frequency oscillatory ventilation (HFOV) usually uses a bias flow system to provide fresh gas. Although the bias flow rates (Vbf) used previously have varied widely among experimental configurations, the precise role of the bias flow in HFOV-mediated gas transport has not been defined. We assessed the effect of bias flow rate on gas transport during HFOV by measuring CO2 removal rate (MCO2) in anesthetized, paralyzed dogs, using a wide range of bias flow rates (0.7-28.9 L X min-1). When a fixed tidal volume of 40 ml was applied at HFOV frequencies of 2-12 Hz, MCO2 was proportional to the time-averaged alveolar-bias flow CO2 concentration difference. Thus, when Vbf was reduced below a value which resulted in a substantial increase in bias flow CO2 concentration, MCO2 was reduced. These findings are consistent with a simple framework in which the relative magnitudes of the resistances to gas transport of the airways and of the bias flow (1/Vbf) determine the contribution of the bias flow rate to overall gas transport during HFOV. This relationship may be employed to assess the intra-airway contribution to HFOV-mediated gas transport at any bias flow rate, and may therefore allow comparison of results from experiments utilizing various bias flow rates.

Lung inflation during high-frequency ventilation.

We investigated the relationship between mean airway pressure and lung volume during low-tidal-volume, high-frequency ventilation (HFV). Eight patients requiring mechanical ventilatory support for treatment of respiratory insufficiency were studied by imposing rapid (60 to 600 breaths/min) oscillations with low tidal volumes (50 to 150 ml) at a constant mean airway pressure of 5 cm H2O. Despite this constant mean airway pressure, lung volume increased substantially during the oscillation period in 7 of 8 subjects, as indicated both by an increase in thoracoabdominal dimensions and by an increase in respiratory system relaxation pressures after the oscillations were stopped. For each patient in whom these changes occurred, the degree of lung inflation rose progressively with increases in either frequency or tidal volume. Given this dissociation between lung volume and mean airway pressure, some index of lung volume or alveolar pressure should be monitored to minimize the likelihood of adverse effects during HFV.

Physiological bases for new approaches to mechanical ventilation.

High frequency ventilatory (HFV) techniques offer potential advantages over conventional forms of mechanical ventilation in patients with diverse forms of respiratory insufficiency. In some respects, HFV challenges conventional physiologic concepts regarding gas transport in the lung. We review hypotheses regarding the mechanism of gas transport and provide a brief perspective on current clinical applications of these techniques.

Influence of the endotracheal tube on CO2 transport during high-frequency ventilation.

Low-volume, high-frequency ventilation (HFV) delivered via an endotracheal tube can maintain eucapnia in both humans and animals. Because recent animal studies have suggested that a substantial fraction of the resistance to gas transport during HFV can be attributed to the presence of the endotracheal tube, we evaluated the importance of the endotracheal tube on carbon dioxide elimination (VCO2) during HFV in humans. We compared the effectiveness of delivering the fresh gas bias flow at the proximal and the distal end of an endotracheal tube. For each bias flow position, we ventilated patients using tidal volumes of 60 ml or less and frequencies from 0.5 to 12 Hz. In each case, VCO2 was approximately 50% greater when the fresh gas was introduced at the carinal end of the endotracheal tube. Thus, the endotracheal tube contributed about one third of the resistance to HFV-induced CO2 transport in these patients. These results indicate that the position of the fresh gas source strongly influences the effectiveness of HFV.