ADJUSTMENT OF FLOW IN THE EVENT OF A LEAK BETWEEN THE TRACHEA AND ENDOTRACHEAL-TUBE

Purposes: To estimate the leak between the endotracheal tube and the t rachea in newborns in order to compensate for errors in airflow measur ement and to monitor mechanical variables from pressure and flow signa ls. Methods: Assuming that the leak resistance (Rf) is constant during a respiratory cycle, the resistive properties of the endotracheal tub e were evaluated. The method was validated in the intensive care unit with a mechanical test lung and assessed on recordings of three newbor ns during mechanical ventilation for RDS. We have used a least squares method for the estimation of positive end expiratory pressure (PEEP) on both newborns and simulated data. Results: Direct measurements of s imulated leak resistances on the mechanical lung are in agreement with our estimation of leak resistances. In newborns, the success of flow correction is evidenced on end inspiratory pauses: corrected flow drop s to zero while raw data show a constant nonzero flow. On the simulate d lung, the PEEP underestimation with uncorrected flow ranges from 10 to 20 cm H2O while the corresponding underestimation with corrected fl ow is less than 2 cm H2O. In newborns, the flow correction shifts the estimated PEEP from negative values (-0.3 +/- 1.3 cm H2O before correc tion) to positive values (3.6 +/- 0.7 H2O after correction) higher tha n the imposed PEEP (2 cm H2O). Conclusions: The efficiency of this sim ple method has been demonstrated. It could be used successfully on adu lt patients, as there will not be flow correction in the absence of le aks.