The effect of ventilator performance on airway pressure release ventilation: a model lung study.

BACKGROUND: Using a model lung connected to six different ventilators, with each ventilator in the airway pressure release ventilation mode, we measured differences in intrinsic positive end-expiratory pressure (PEEPi) during the expiratory phase and calculated the inspiratory and expiratory pressure time product (PTP) as an index of work of breathing during the inspiratory phase.

METHODS: We compared 6 ventilators: Puritan-Bennett 840, Evita XL, Servo i, Avea, Hamilton G5, and Engström. With a constant inspiratory pressure level of 25 cm H(2)O and expiratory pressure level of 0 cm H(2)O, PEEPi was measured as the expiratory time was decremented from 1.0 second to 0.2 second in steps of 0.1 second. The inspiratory and expiratory PTPs were measured during the ventilator's inspiratory phase by simulating spontaneous breathing with a tidal volume of 300 mL, with a respiratory rate of 30 breaths/min and with expiratory flow rates of 0.5 L/s, 1.0 L/s, and 1.5 L/s.

RESULTS: In all ventilators, the progressive diminution of the expiratory time caused a significant increase in PEEPi (P< 0.001). With a 0.2-second expiratory time, PEEPi ranged from 9.4± 0.07 cm H(2)O for the Servo i to 15.7± 0.04 cm H(2)O for the Avea. The Servo i had a significantly lower inspiratory PTP than did the other ventilators (P< 0.001). When the expiratory flow rate was 0.5 L/s and 1.0 L/s, the expiratory PTP was lower with the Servo i and Evita XL than with the other ventilators (P< 0.001).

CONCLUSIONS: PEEPi varied significantly among ventilators. Inspiratory and expiratory work of breathing varied between ventilators when spontaneous breathing occurred during the ventilator's inspiratory phase.