A comparison of endotracheal tube compensation techniques for the measurement of respiratory mechanical impedance at low frequencies

Abstract

Measurement of respiratory impedance ( Z rs ) in intubated patients requires accurate compensation for pressure losses across the endotracheal tube (ETT). In this study, we compared time-domain (TD), frequency-domain (FD) and combined time-/ frequency-domain (FT) methods for ETT compensation. We measured total impedance ( Z tot ) of a test lung in series with three different ETT sizes, as well as in three intubated porcine subjects. Pressure measurement at the distal end of the ETT was used to determine the true Z rs . For TD compensation, pressure distal to the ETT was obtained based on its resistive and inertial properties, and the corresponding Z rs was estimated. For FD compensation, impedance of the isolated ETT was obtained from oscillatory flow and pressure waveforms, and then subtracted from Z tot . For TF compensation, the nonlinear resistive properties of the ETT were subtracted from the proximal pressure measurement, from which the linear resistive and inertial ETT properties were removed in the frequency-domain to obtain Z rs . The relative root mean square error between the actual and estimated Z rs ( rRMSE, Ẑrs ) showed that TD compensation yielded the least accurate estimates of Z rs for the in vitro experiments, with small deviations observed at higher frequencies. The FD and TF compensations yielded estimates of Z rs with similar accuracies. For the porcine subjects, no significant differences were observed in rRMSE, Ẑrs across compensation methods. FD and TF compensation of the ETT may allow for accurate oscillometric estimates of Z rs in intubated subjects, while avoiding the difficulties associated with direct tracheal pressure measurement.