Mechanical performance of clinically available, neonatal, high-frequency, oscillatory-type ventilators.

OBJECTIVE

To perform a functional evaluation of five different high-frequency, oscillatory-type ventilators that are currently being marketed for neonatal high-frequency oscillation.

DESIGN

Observational animal study.

SETTING

Laboratory.

SUBJECTS

New Zealand White male rabbits.

INTERVENTIONS

Oscillator waveforms and delivered volumes were measured plethysmographically for the following ventilators: the SensorMedics 3100 A; the Dräger Baby Log 8000; the Metran Humming V; the Infant Star; and the Infant Star 950. The independent variables which were adjusted included frequency (5 to 15 Hz), amplitude (25% to 100%), mean airway pressure (5 to 25 cm H2O) and lung injury.

MEASUREMENTS AND MAIN RESULTS

At 15 Hz, the volume delivered at the 100% amplitude setting varied from 2.1 to 8.8 mL. Generally, the delivered volume decreased with increasing frequency, and with increased percentage of amplitude. Volume delivery was relatively unaffected by mean airway pressure but decreased with lung injury. Waveforms ranged from pure sinusoidal to a complex square wave. The handling of inspiration/expiration time ratios was ventilator specific. The SensorMedics inspiration/ expiration ratio is user selected over a range from 1:2.3 (30% inspiratory time) to 1:1 (50% inspiratory time) and once selected it is consistent over its entire range of operating frequencies. The Drager ratio is machine determined and varied from 1:2.5 at 5 Hz to 1:1 at 15 Hz. Inspiratory time of the Infant Star is machine set at 18 msecs such that the inspiration/expiration ratio is 1:10.1 at 5 Hz and 1:2.7 at 15 Hz. The Humming V has a fixed inspiration/expiration ratio of 1:1. The relationship of the mean airway pressure displayed on the ventilator to the alveolar occlusion pressure varied considerably among devices. The displayed mean pressure could either overestimate (SensorMedics at 33% inspiratory time or Infant Star), approximate (Humming V), or underestimate (Dräger) the mean lung distending pressure measured during a brief occlusion maneuver.

CONCLUSIONS

The findings demonstrate large variations in machine performance. The ventilators also differed profoundly in complexity of operation and versatility as neonatal ventilators.