• A B Adams, P L Bliss, and J Hotchkiss.
• Regions Hospital, University of Minnesota, St Paul 55101, USA. alex.b.adams@Health Partners.com
• Resp Care. 2000 Apr 1;45(4):390-400.

BackgroundNoninvasive positive pressure ventilation (NPPV) has been studied in several settings and shown to reduce patient morbidity associated with endotracheal intubation. Intolerance to NPPV has been estimated at 25-33%, a substantial proportion of attempts to ventilate noninvasively. Bi-level pressure ventilators (BPVs) have been designed for NPPV, yet their response to changes in respiratory impedance has not been extensively evaluated. To determine responses of BPVs to changing impedance conditions, we tested 4 BPVs to evaluate the potential for intolerance. We also developed a mathematical model for BPV performance that accounted for impedance conditions, leak, pressure settings, and inspiratory flow cutoff level.MethodsFour BPV models at the same settings were challenged to ventilate a triggered test lung under a range of impedance conditions while measuring tidal volume (VT) and intrinsic positive end-expiratory pressure (auto-PEEP). The model was used to predict VT and auto-PEEP under normal, restrictive, and obstructive conditions.ResultsThe BPV models tested delivered VT in a similar manner. VT decreased with decreased compliance and increased resistance. Auto-PEEP developed with increased resistance and compliance. The model predicted a VT delivery dependent on inspiratory flow cutoff level. For the obstructive condition, the model predicts an optimal VT delivery within a specific inspiratory flow cutoff range that becomes narrower with increasing resistance.ConclusionsVT delivery and auto-PEEP generated by BPVs are highly dependent on the prevailing impedance condition. Though there are differences between BPV models, generally, performance was similar between the models tested. This report suggests that knowledge of both respiratory system impedance and the performance of the BPV in use are required to attend to inadequate VT delivery and auto-PEEP generation. Furthermore, the model predicts a relatively narrow range for inspiratory flow cutoff that provides adequate ventilatory support without causing hyperinflation in patients with obstructive conditions.

10 keywords...

hide…