Detection of expiratory flow limitation during experimental mechanical ventilation

Introduction

The present study examines the possibility of non-interventional detection of expiratory flow limitation (EFL) during experimental mechanical ventilation (MV), with the aid of nonlinear regression analysis.

Methods

Eight New-Zealand rabbits under MV were included in the study. EFL was induced by the application of negative expiratory pressure (-5 to -10 hPa) and recognised by the negative expiratory pressure technique. The airway pressure (Paw) and flow (V') were recorded and treated offline with the aid of the nonlinear model Pao = EEP + E1.V + E2.V2 + k1.V'+ k2.V'.|V'| + k3.V.V', which accounts for volume dependence of elastance and flow and volume dependence of resistance. The timed signal of resistive pressure (Pres) was reconstructed after subtraction of elastic nonlinear pressure from the whole pressure.

Results and conclusions

The Pres signal presents two different configurations that permit one to separate between EFL and non-EFL. More specifically, in cases without EFL Pres is similar to the flow signal, while in cases with EFL expiratory Pres presents a minimum, which corresponds to the point of EFL installation. We conclude that the present technique, which combines computational and graphical parts, offers the possibility of EFL recognition and quantification at least during mechanical ventilation.