Regional pulmonary effects of bronchoalveolar lavage procedure determined by electrical impedance tomography

Abstract

Background: The provision of guidance in ventilator therapy by continuous monitoring of regional lung ventilation, aeration and respiratory system mechanics is the main clinical benefit of electrical impedance tomography (EIT). A new application was recently described in critically ill patients undergoing diagnostic bronchoalveolar lavage (BAL) with the intention of using EIT to identify the region where sampling was performed. Increased electrical bioimpedance was reported after fluid instillation. To verify the accuracy of these findings, contradicting the current EIT knowledge, we have systematically analysed chest EIT data acquired under controlled experimental conditions in animals undergoing a large number of BAL procedures.

Methods: One hundred thirteen BAL procedures were performed in 13 newborn piglets positioned both supine and prone. EIT data was obtained at 13 images before, during and after each BAL. The data was analysed at three time points: (1) after disconnection from the ventilator before the fluid instillation and by the ends of fluid (2) instillation and (3) recovery by suction and compared with the baseline measurements before the procedure. Functional EIT images were generated, and changes in pixel electrical bioimpedance were calculated relative to baseline. The data was examined in the whole image and in three (ventral, middle, dorsal) regions-of-interest per lung.

Results: Compared with the baseline phase, chest electrical bioimpedance fell after the disconnection from the ventilator in all animals in both postures during all procedures. The fluid instillation further decreased electrical bioimpedance. During fluid recovery, electrical bioimpedance increased, but not to baseline values. All effects were highly significant (p < 0.001). The fractional changes in individual regions-of-interest were posture-dependent. The regional fall in electrical bioimpedance was smaller in the ventral and larger in the dorsal regions after the fluid instillation than after the initial disconnection to ambient pressure in supine animals (p < 0.001) whereas these changes were of comparable amplitude in prone position.

Conclusions: The results of this study show a regionally dissimilar initial fall in electrical bioimpedance caused by non-uniform aeration loss at the beginning of the BAL procedure. They also confirm a further pronounced fall in bioimpedance during fluid instillation, incomplete recovery after suction and a posture-dependent distribution pattern of these effects.

Keywords: Alveolar collapse; BAL; EIT; Electrical bioimpedance; Functional imaging; Regional ventilation; Ventilation monitoring.

Fig. 1

Global and regional EIT waveforms with highlighted measurement phases before, during and after bronchoalveolar lavage. The global waveform shows the sum of all image pixel values of relative impedance changes (rel. ΔZ) in arbitrary units (AU) (top), the regional ones in the ventral (middle) and dorsal image sections (bottom). The recording continued through five phases: I, continuous mechanical ventilation; II, disconnection of the endotracheal tube from the ventilator; III, administration of the lavage fluid through the endotracheal tube; IV, fluid recovery by suction, V, re-connection of the endotracheal tube to the ventilator and resumption of mechanical ventilation. (The small dent in the waveforms after approximately 60% of the suction phase IV (i.e. at 52 s of the recording) was associated with the temporary evacuation of the fluid from the suction syringe. The clearly discernible change in the waveforms with higher tidal variation amplitude and increased end-expiratory values after the resumption of ventilation in phase V at time point 89 s resulted from the adjustment in ventilator settings with higher PEEP.) EIT data was analysed at three standardised time points: t₁, after disconnection from the ventilator directly prior to the fluid instillation into the lungs; t₂, at the end of fluid instillation and t₃, at the end of suction of the lavage fluid immediately before the resumption of mechanical ventilation, in each case compared with bioimpedance at baseline phase I prior to the disconnection from the ventilator at t₀

Fig. 2

Example functional EIT images. The images show the regional fall in electrical bioimpedance in lung areas in one of the examined animals in the supine (top) and the prone positions (bottom) at three time points during the bronchoalveolar lavage procedure. The impedance decrease is shown in dark tones after the disconnection of the endotracheal tube from the ventilator before the fluid instillation (t₁), after the fluid instillation (t₂) and after the fluid suction before the resumption of ventilation (t₃) in each case relative to baseline. v, ventral; d, dorsal; r, right; l, left

Fig. 3

Change in pulmonary electrical bioimpedance at distinct time points during the bronchoalveolar lavage procedure. The sums of all image pixel values of relative impedance changes (rel. ΔZ) in arbitrary units (AU) are given after the disconnection of the endotracheal tube from the ventilator before the lavage fluid instillation (t₁), after the fluid instillation (t₂) and after the fluid suction before the resumption of ventilation (t₃), in each case compared to baseline before the disconnection to ambient pressure. The data originate from lavages performed first in supine (left), then prone (middle) and final supine (right) positions (supine I, prone and supine II, respectively). The numbers of analysed bronchoalveolar lavages in each posture are given at the top of each diagram. p values given in the diagrams highlight the highly significant effect of the time point. The significance of differences between the individual time points was obtained from post analyses at **p < 0.01 and ***p < 0.001. Significantly different values from the supine I position at corresponding time points are given as ††p < 0.01 and †††p < 0.001 and from supine II position as ‡‡p < 0.01

Fig. 4

Fractional bioimpedance change in ventral, middle and dorsal regions-of-interest. The fractional changes are given at three time points during the bronchoalveolar lavage procedure: after the disconnection of the endotracheal tube from the ventilator before the lavage fluid instillation (t₁), after the fluid instillation (t₂) and after the fluid suction before the resumption of ventilation (t₃). The data was obtained first in supine (top row), then prone (middle row) and supine (bottom row) animals (supine I, prone and supine II, respectively). The significance of differences between the individual time points was derived from post analyses at *p < 0.05, **p < 0.01 and ***p < 0.001. Significantly different values from the supine I position at corresponding time points are given as †††p < 0.001 and from supine II position as ‡‡p < 0.01 and ‡‡‡p < 0.001