Respiratory system mechanics during noninvasive proportional assist ventilation: A model study

To assess the accuracies of airway resistance (Raw) and compliance (Crs) calculations using the expiratory time constant (RCexp) method as well as the accuracy of Pmus estimation in obstructive lung models.

A Respironics V60 ventilator was connected to an active lung simulator. The driving pressure was maintained at 5-10 cmH2O and positive end-expiration pressure (PEEP) was 5 cmH2O. Maximal Pmus, estimated based on equations of motion and respiratory mechanical properties, was calculated by the RCexp method to derive respiratory system compliance (Crs) and inspiratory (Rinsp) and expiratory (Rexp) resistance.

During PAV, the assist proportion was adjusted to 55% and 40% with Pmus of 5 and 10 cmH2O, respectively. Pmus measurement errors were <20% of the preset values in most lung conditions. In the active lung model with PAV, an overestimation of Raw was found in the normal resistance condition, and Rinsp was underestimated in the severe obstructive model (P < 0.01). Crs was overestimated significantly except in the severe obstructive model at a Pmus of 10 cmH2O (all P < 0.01). Using the RCexp method, the target of ≤20% between the calculated and preset values in airway resistance was achieved in most obstructive models at a Pmus of 5 cmH2O.

The RCexp method might provide real-time assessments of respiratory mechanics (elastance and resistance) in the PAV mode. With low inspiratory effort, the estimation error was acceptable (<20%) in most obstructive lung models.