Perioperative assessment of regional ventilation during changing body positions and ventilation conditions by electrical impedance tomography

Transcription

1 British Journal of Anaesthesia, 117 (2): (16) doi: 1.193/bja/aew188 Respiration and the Airway RESPIRATION AND THE AIRWAY Perioperative assessment of regional ventilation during changing body positions and ventilation conditions by electrical impedance tomography A. Ukere 1,, A. März 1,, *, K. H. Wodack 1,C.J.Trepte 1, A. Haese 2, A. D. Waldmann 3, S. H. Böhm 3 and D. A. Reuter 1 1 Department of Anesthesiology, Center of Anesthesiology and Intensive Care Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany, 2 Martini Klinik, University Medical Center Hamburg- Eppendorf, Hamburg, Germany, and 3 Swisstom AG, Landquart, Switzerland *Corresponding author. a.maerz@uke.de Abstract Background: Lung-protective ventilation is claimed to be beneficial not only in critically ill patients, but also in pulmonary healthy patients undergoing general anaesthesia. We report the use of electrical impedance tomography for assessing regional changes in ventilation, during both spontaneous breathing and mechanical ventilation, in patients undergoing robot-assisted radical prostatectomy. Methods: We performed electrical impedance tomography measurements in 39 patients before induction of anaesthesia in the sitting (M1) and supine position (M2), after the start of mechanical ventilation (M3), during capnoperitoneum and Trendelenburg positioning (M4), and finally, in the supine position after release of capnoperitoneum (M). To quantify regional changes in lung ventilation, we calculated the centre of ventilation and silent spaces in the ventral and dorsal lung regions that did not show major impedance changes. Results: Compared with the awake supine position [2.3% (2.3)], anaesthesia and mechanical ventilation induced a significant increase in silent spaces in the dorsal dependent lung [9.2% (6.3); P<.]. Capnoperitoneum and the Trendelenburg position led to a significant increase in such spaces [11.% (8.9)]. Silent space in the ventral lung remained constant throughout anaesthesia. Conclusion: Electrical impedance tomography was able to identify and quantify on a breath-by-breath basis circumscribed areas, so-called silent spaces, within healthy lungs that received little or no ventilation during general anaesthesia, capnoperitoneum, and different body positions. As these silent spaces are suggestive of atelectasis on the one hand and overdistension on the other, they might become useful to guide individualized protective ventilation strategies to mitigate the side-effects of anaesthesia and surgery on the lungs. Key words: atelectasis; electrical impedance; electrodes; perioperative period; respiration, artificial During general anaesthesia, gas exchange is impaired because of a mismatch of the regional distribution of ventilation and perfusion. 1 The main pathogenic mechanism is the development of atelectasis in dorsal dependent lung areas and overdistension A. Ukere and A. März contributed equally to this work. Accepted: May 31, 16 The Author 16. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved. For Permissions, please journals.permissions@oup.com 228

2 Perioperative study of regional ventilation by EIT 229 Editor s key points Electrical impedance tomography was used to assess regional changes of lung ventilation in patients undergoing robot-assisted radical prostatectomy. The silent spaces (which receive little or no ventilation) of the lungs were increased by mechanical ventilation, by capnoperitoneum, and by Trendelenburg s positioning. Electrical impedance tomography may be useful in identifying the silent spaces, which may suggestive of atelectasis of the lungs. in ventral non-dependent lung areas. 2 Given that artificial ventilation can cause harm to healthy lungs, protective ventilation strategies have been advocated. 3 As the complexity and duration of laparoscopic procedures have increased over the years and are expected to increase even more (robot-assisted prostatectomy in the deep Trendelenburg position is only one example), bedside tools capable of identifying the consequences of therapeutic interventions with sufficient sensitivity and specificity could allow optimization of ventilation strategies for such settings. The working principle of electrical impedance tomography (EIT) is the measurement of changes in resistance of the lung against the flow of alternating electrical currents applied to the thorax during breathing. This resistance of a medium to alternating current is called impedance. Electrical impedance tomography reconstructs functional images with high temporal resolution based on the assessment of impedance changes during the respiratory cycle. In this way, regional ventilation can be seen continuously at the bedside. Stretched lung tissue poses higher impedance to electrical current than non-stretched lung tissue. The variations of impedance during the breathing cycle seen in EIT are therefore the result of changes in tissue stretch. 4 For these changes, the term relative stretch is used. Given that in poorly ventilated lungs or in areas of the lungs that are not ventilated or stretched at all, the change in impedance during the ventilation cycles is reduced compared with well-ventilated areas, EIT might be used as a monitoring device for guidance of lung-protective ventilation during surgery. However, its practical use is strongly dependent on parameters allowing clear and reliable interpretation of those pathophysiological changes. Therefore, we quantified regional changes in pulmonary impedance, defining lung areas with impedance changes of <1% of the maximum as silent spaces. There is still clinical uncertainty regarding how to assess regional inhomogeneity in aeration induced by mechanical ventilation and perioperative treatment at the bedside. We hypothesized that changes in the distribution of regional lung impedance as represented by the centre of ventilation (CoV) and the amount of silent space within the dorsal and ventral lung areas quantified by EIT could represent the expected pathophysiological changes in aeration. Therefore, we aimed to document systematically the effects of induction of anaesthesia with cessation of spontaneous breathing and implementation of mechanical ventilation, of perioperative changes in body positioning (Trendelenburg position) in combination with application of capnoperitoneum, and of the reversal of these manoeuvres on EIT findings, and to compare them with global parameters of oxygenation and lung mechanics. Methods This study was performed within the framework of a larger two-centre observational trial focusing on lung function after robot-assisted prostatectomy (NCT66246). With approval by the local ethics committee of the Medical Board Hamburg (PV374) and after obtaining written informed consent, 4 patients undergoing robot-assisted prostatectomy were included. Exclusion criteria were patients with known chronic obstructive pulmonary disease or bronchial asthma. Patients with sleep apnoea, with ASA risk classification IV or more, and those with BMI >3 kg m 2 were also excluded. We used an EIT device (Swisstom BB²; Swisstom, Landquart, Switzerland) with 32 textile-embedded electrodes placed around the chest along the sixth intercostal space. The accuracy and precision of the device were evaluated previously. 6 The primary outcome measurements of the present study, namely the changes in electrical impedance, were acquired at a scan rate of 46 Hz. The individual s height and weight determined the appropriate image reconstruction matrix. 7 Electrical impedance tomography images, containing pixels, were created by a modified GREIT algorithm. 8 We selected the regions of interest (ROI) within the chest contours of the right and left lung based on three-dimensional thoracic models, which were created from computed tomography (CT) scans (see Supplementary material, Fig. S1). 9 Tidal EIT images from 1 consecutive breaths were created by determining, for each pixel, the change in impedance between the end of inspiration and the preceding end of expiration. Using changes of impedance within both lung ROIs only, we then calculated silent spaces in the dorsal and the ventral lungs. However, given that EIT is usually used in the supine, semi-recumbent position, in which gravity exerts its action on the dorsal lung tissue within the EIT plane, this gravity dependence of the dorsal silent spaces was expressed by the term dependent silent space (DSS), whereas in this context the term non-dependent silent space (NSS) was used for the ventral lungs. Silent spaces were determined based on the CoV and relative stretch, as depicted in Figs 1 and 2. In an attempt to characterize the functional state of the lungs at the regional level, the median value for each one of the EIT-derived breathwise parameters was calculated and expressed as a percentage value. The CoV is a previously described parameter to express the geometric focal point of overall ventilation as a single digit in per cent The vertical position of the CoV is then expressed as a percentage of the anterior posterior extension of the identified lung region, where % refers to ventilation occurring in the most ventral lung region and 1% in the most dorsal part. 12 Impedance changes during tidal breathing within the ROIs were measured as the difference between the end of inspiration and the end of the preceding expiration (Fig. 1), whereby the global impedance change resulting from the breathing-induced changes in the electrical properties of all lung regions (the relative deformation or stretch of the lung tissue during tidal breathing according to Nopp and colleagues) 4 within the EIT plane represented the tidal volume. For the above impedance changes to be displayed as tidal images, their pixel values needed to be categorized and colour coded. Therefore, the amplitude of the pixel within the image showing the maximal impedance change or stretch during breathing served as a 1% reference (Z total ). This amplitude was then divided by 1, which resulted in 1 different amplitude categories (c 1 c 1 ). Depending on its amplitude, each pixel within the lung ROIs was then assigned to one of these categories. In order to determine the relative contribution, as a percentage, of each one of the 1 stretch categories to the overall tidal volume (Vt total ), the amplitude of each pixel within a given category was multiplied by its relative tidal stretch. From these values, a 1- bar histogram was created, whereby the sum of all 1 categories equalled the total impedance change of the tidal breath (Fig. 1).

3 23 Ukere et al. Table 1 Data on perioperative electrical impedance tomography, ventilation, and oxygenation data compared as follows: measurement time point M1 vs M2, M2 vs M3, M3 vs M4, and M4 vs M, respectively. Values are the mean (SD). BR, breathing rate; CoV, centre of ventilation; C RS, compliance; FIO 2, fraction of inspired oxygen; MV, minute volume; NSS, ventral (non-dependent) silence space; Pa O2, arterial pressure of oxygen; Pa CO2, arterial pressure of carbon dioxide; PEEP, positive end expiratory pressure; P insp, inspiratory pressure; ΔP insp, driving pressure; Vt, tidal volume. Statistically significant differences (P<.) between time points are indicated as follows: *between time points M1 and M2; between time points M2 and M3; between time points M3 and M4; between time points M4 and M; and between time points M3 and M Parameter M1 M2 M3 M4 M Sitting position, spontaneous breathing Supine position, spontaneous breathing Supine position, mechanical ventilation 3 head-down-tilt position, mechanical ventilation Supine position, mechanical ventilation CoV (%) 62.9 (6.6) 4.8 (6.7)* 4.4 (4.) 41.7 (8.7) 46.1 (3.9) NSS (%).3 (2.9) 4.1 (1.9)* 3.1 (1.9) 2.8 (2.7) 2.7 (1.9) DSS (%).1 (.3) 2.3 (2.3)* 9.2 (6.3) 11. (8.9) 8. (.8) MV (l min 1 ).7 (.7) 7.1 (1.6) 7.9 (1.42) BR (min 1 ) 16.9 (4.87) 16.8 (3.84) 12.1 (2.63) 14.2 (3.2) 16.1 (4.) Vt (ml) 488 (31.9) 7 (31.3) (4.8) PEEP (cm H 2 O) 1 P insp mean (cm H 2 O) 13.7 (1.) 28.3 (3.3).9 (2.8) ΔP insp (cm H 2 O) 8.7 (1.) 18.3 (3.3) 1.9 (2.8) C RS [ml (cm H 2 O) 1 ] 6.3 (1.) 31. (6.6) 4.8 (1.6) Pa O2 /FIO 2 (kpa) (13.) 4.68 (84.7) (78.9) ðpa CO2 Þ (kpa).48 (.3).6 (.).76 (.) The height of each bar represents the relative contribution of this particular stretch category to the overall tidal volume. By calculating the relative stretch in the presented manner, the risk of outliers falsifying the result are eliminated. The NSS and DSS were calculated as follows: for each breath, a virtual line perpendicular to the gravity vector through the CoV of that same breath defined the ventilation horizon. All pixels lying below this ventilation horizon and belonging to the lowermost stretch category defined the DSS. The number of these pixels was counted and expressed as a percentage of all pixels within both lung ROIs. Accordingly, the NSS value describes the percentage of poorly ventilated pixels physically located above the ventilation horizon (see Fig. 1). In the sitting position, the same calculation method was applied, without using the gravity vector but instead the coordinates of the pixels within the contour to determine ventral and dorsal silent spaces. Study protocol Electrical impedance tomography measurements were performed at five consecutive time points (see Fig. 2). Measurement M1 was carried out after arrival in the operating room in the awake, spontaneously breathing, and upright sitting patient, with legs stretched out on the operating table. During M2, data were acquired with the awake patient in the supine position still breathing spontaneously, and during M3 after induction of anaesthesia under mechanical ventilation at a PEEP of cm H 2 O. Measurement M4 was made 3 min after establishing capnoperitoneum, 3 head-down tilt position, and a PEEP of 1 cm H 2 O. Measurement M was performed back in a horizontal supine position at a PEEP cmh 2 O, immediately after release of the capnoperitoneum. In addition to the EIT measurements, data on ventilation and oxygenation were also recorded at M1 M. Anaesthesia According to our institutional care, each patient received midazolam 7. mg p.o. before the procedure. All patients were monitored with non-invasive blood pressure, peripheral oxygen saturation ðsp O2 Þ, and a five-lead ECG. After induction of anaesthesia with sufentanil. µg kg 1 i.v., propofol 2 mg kg 1 i.v., and rocuronium.6 mg kg 1 i.v., all patients were tracheally intubated and ventilated mechanically using a Primus respirator (Dräger, Lübeck, Germany) in a pressure-controlled mode (tidal volume of 7 ml kg 1 of ideal body weight and inspiration-to-expiration ratio of 1:1.). A PEEP of cm H 2 O was established, which was increased to 1 cm H 2 O during laparoscopy in a 3 head-down tilt. Breathing frequency was adapted to maintain end-tidal carbon dioxide partial pressure between 4. and.1 kpa. Sevoflurane was delivered using minimal alveolar concentraion values of.8 1. vol%. The inspired fraction of oxygen of. was changed only if ðs p O2 Þ reduced to <96% or arterial partial pressure of oxygen ðpa O2 Þ was lower than 9.3 kpa. Each patient received a 7. Fr central venous catheter in the jugular vein and an arterial catheter in the femral artery (PiCCO, Pulsion, Munich, Germany). Surgery was performed using a DaVinci robot (Intuitive Surgical Inc., Sunnyvale, CA, USA) via a transperitoneal approach. Statistics Exposure measurements were evaluated as descriptive statistics time-point-wise. Continues variables are presented as the mean with the respective range if not normally distributed or else as mean (SD). Categorical variables are presented as absolute frequencies. We compared outcome measures of each time point with those obtained during the previous one. We also compared M3 with M to demonstrate anaesthesia- and surgery-related effects on the lung tissue. To avoid multiple testing errors and to account for data structure (repeated measures for individual patients), we used one linear mixed model for each of the variables listed above. For this post hoc pairwise comparison of time points, Bonferroni correction for multiple testing was applied if the global P-value was >.; otherwise, no correction was necessary. Values of P<.

4 Perioperative study of regional ventilation by EIT 231 Comparison M1 vs M2 M2 vs M3 M3 vs M4 M4 vs M M3 vs M Effects to be shown Comparator: knowing that anaesthesia-induced atelectasis occurs primarily in the dorsal (dependent) lungs, the dorsal silent spaces during spontaneous breathing in the sitting position served as the individual s reference value. This reference served to demonstrate that inaccuracies caused by the lung contours obtained from the device s library did not cause false-positive silent spaces exceeding an arbitrary threshold of 3%. It was further assumed that the upward lift of the abdominal content would reduce breathing in the ventral lungs, creating circumscribed silent spaces in this region. To demonstrate the aggregate effect of cessation of spontaneous breathing, anaesthesia, and mechanical ventilation at cm H 2 O of PEEP. To show the combined additional effect of the Trendelenburg position, capnoperitoneum, and mechanical ventilation with a PEEP of 1 cm H 2 O to counterbalance the impact of surgery on the lungs. To document the effects of reducing PEEP back to cm H 2 O and terminating the Trendelenburg position and the capnoperitoneum. To demonstrate potential carry-over residual effects of the aforementioned interventions of M4 on lung function during identical body position and ventilator conditions as applied immediately after the induction of anaesthesia during M3. A C B Relative tidal volume (%) D i=1 c i = Z total = VT total C 1 C 2 C 3 C 4 C C 6 C 7 C 8 C 9 C Ventral Max Relative stretch Min Dorsal silent space (%) Dorsal Ventral silent space (%) Fig 1 Schematic illustration of electrical impedance tomography-derived lung function parameters. (A) within the regions of interest (ROIs). Impedance changes during tidal breathing within the ROIs were measured as the difference between the end of inspiration and the end of the preceding expiration. For these impedance changes to be displayed as tidal images, their pixel values needed to be categorized and colour coded. (B) The amplitude of the pixel within the image showing the maximal impedance change or stretch during breathing served as a 1% reference (Ztotal). This amplitude was then divided by 1, which resulted in 1 different amplitude categories (c 1 c 1 ). Depending on its amplitude, each pixel within the lung ROIs was then assigned to one of these categories. To determine the relative contribution (as a percentage) of each one of the 1 stretch categories to the overall tidal volume (Vt total ), the amplitude of each pixel within a given category was multiplied by its relative tidal stretch. From these values, a 1-bar histogram was created, whereby the sum of all 1 categories equalled the total impedance change of the tidal breath. (C) The silent spaces are visualized as grey areas within the ROIs. The blue dotted line marks the ventilation horizon. Silent spaces above the ventilation horizon are called ventral (non-dependent) silent spaces, whereas the areas below are called dorsal (dependent) silent spaces. The blue dot represents the centre of ventilation. (D) Dorsal (dependent) silent spaces and ventral (non-dependent) silent spaces expressed in a bar graph showing the percentage of the categories with the lowest impedance change in relation to the pixel with the highest change in impedance.

5 232 Ukere et al. were considered significant. Statistical analysis was carried out using SPSS version. (IBM, New York, NY, USA). Results The numerical values of CoV showed a significant and continuous decrease over the course of time points from M1 to M4 [M1, 62.9% (6.6) vs M2, 4.8% (6.7) vs M3, 4.4% (4.) vs M4, 41.7% (8.7)], which can be understood as a continuous shift of the CoV towards the ventral lungs. The lowest value was found in head-down tilt with capnoperitoneum, showing a shift by more than % compared with M1 [M1, 62.9% (6.6) vs M4, 41.7% (8.7)]. The return to the supine position and release of capnoperitoneum led to a significant reversal of this shift of CoV towards the dorsal lung region of the lung [M4, 41.7% (8.7) vs M, 46.1% (3.9)], reaching a value not different from the one measured after induction of anaesthesia immediately before the capnoperitoneum [M3, 4.4% (4.)] (see Fig. 3 and Table 1). In line with the trend of the CoV, the dorsal silent spaces increased significantly over the course of the study period time points from M1 [.1% (.3)] to M2 [2.3% (2.3)] to M3 [9.2% (6.3)] and to M4 [11.% (8.9)], whereas they decreased at M to 8.% (.8) with the release of the capnoperitoneum and the Trendelenburg position. Ventral silent spaces showed continuous and significant decreases from M1 [.3% (2.9)] to M2 [4.1% (1.9)] and to M3 [3.% (1.9)], but stayed constant thereafter (see Fig. 4 and Table 1). Discussion In spontaneously breathing patients sitting with their legs stretched out, silent spaces, reflecting poorly ventilated areas, were found in the ventral but not in the dorsal lungs. In the supine position, a small lack of ventilation occurred in the dorsal parts of the lung. With the induction of anaesthesia, silent spaces in the dependent dorsal lung regions increased significantly, and the application of capnoperitoneum and the Trendelenburg position led to even more silent spaces in the dependent but not in the non-dependent lung. The latter stayed constant during all conditions in which mechanical ventilation was applied. The effects of surgery were not fully reversible. The complete absence of silent spaces in the dorsal lung in the sitting position and the fact that they increased only marginally (<3%) as they became gravity dependent in the supine position are indicators of a reliable fit between the predicted lung contours used to calculate the EIT images and the real anatomy of the individual patients. 9 A 3 B C D M1 M2 M3 M4 M Fig 2 Representative electrical impedance tomography images and electrical impedance tomography-derived parameters of one patient at time points from M1 to M. Pictograms at the top of the figure symbolize the patient s position. (A) Picture of tidal ventilation within the region of interest. (B) within the region of interest. (C) Ten-bar histogram representing the distribution of stretch within these respective categories. (D) Silent spaces visualized as grey areas within the region of interest. The blue dotted line shows the ventilation horizon. The blue dot represents the centre of ventilation.

6 Perioperative study of regional ventilation by EIT 233 CoV (%) Ventral Dorsal * M1 M2 M3 M4 M The initial lack of ventilation in the ventral lungs was anticipated and can easily be explained by a reduced movement of the ventral diaphragm as a result of counterpressure exerted by the abdominal contents. If a major physical shift of abdominal contents or of the heart into the measuring plane were to have caused this increase in silent spaces, no ventilation would be noticeable at all. Moving of organs in and out of the EIT plane with breathing would have caused a massive EIT contrast and thus a large ventilation-like signal. Compared with the spontaneously breathing, awake, supine patient, the significant gain in DSS and the concomitant shift of the Gravity dependent Fig 3 Box plot of perioperative centre of ventilation (CoV) for all time points and for all patients. Boxes indicate th 7th percentile, the median is depicted by the line within the box, and the means by the cross. The whiskers mark the th and 9th confidence interval. Statistically significant differences (P<.) between time points are indicated as follows: *between time points M1 and M2; between time points M2 and M3; between time points M3 and M4; between time points M4 and M. Dorsal silent space (%) Dorsal * * M1 M2 M3 M4 Gravity dependent M Ventral Fig 4 Perioperative silent spaces with SD for all time points. Statistically significant differences (P<.) between time points are indicated as follows: *between time points M1 and M2; between time points M2 and M3; between time points M3 and M4; between time points M4 and M Ventral silent space (%) CoV towards the ventral lung together with a subnormal Pa O2 /FIO 2 ratio can be explained by the following factors: first, an anaesthesia-induced formation of atelectasis; second, its further aggravation with head-down tilt; and third, the capnoperitoneum-induced increase in abdominal pressures during the surgery, which also went along with decreases in oxygenation and compliance. As NSS remained constant throughout all anaesthetized conditions using mechanical ventilation, the presence of DSS indicates a loss of functional lung size, as recently shown by Amato and colleagues. 13 The increasing driving pressures (ΔP) at constant tidal volume (Vt) seen in the Trendelenburg position suggests a clinically significant reduction in the aerated lung volume available for tidal ventilation. The measured driving pressure normalizes tidal volume to functional lung size, which in turn is intrinsically related to respiratory-system compliance (C RS ) as shown by the following simple formula: ΔP = Vt/C RS. The combination of circumscribed significant ventilation defects and markedly reduced respiratory-system compliance indicate unequivocally that the lungs became small, not stiff. Interestingly, the loss of lung volume could not be prevented by the intentional increase of PEEP from to1cmh 2 O during capnoperitoneum, and only an active recruitment manoeuvre might have improved ventilation within the dependent lung areas (see Table 1). 14 After removing the capnoperitoneum and bringing the patient back into the supine position, the lung mechanics, oxygenation, and EIT-derived parameters remained abnormal, suggesting persistent atelectasis and a ventilation perfusion mismatch as residual effects of the surgical intervention, as indicated by both the lowest P/F ratio and the highest Pa CO2. Although compliance was higher during M than during M4, it remained substantially below its reference value immediately after induction of anaesthesia. Recent studies have shown that the concept of lung-protective ventilation might have a positive impact not only in critically ill patients, but also in patients without known lung injury who are undergoing surgery. 16 However, identification of regional lung dysfunction and monitoring of treatment effects at the bedside still remain difficult, because global parameters of oxygenation or lung mechanics provide only integral information, lacking resolution at the regional level. As we could show in the present study, EIT has the unique ability to display regional ventilation defects right at the bedside. Our findings are in line with those of other EIT studies demonstrating formation of gross atelectasis, shifts in the CoV, and the effects of PEEP during laparoscopic procedures We cannot state with certainty that those gravity-dependent areas declared silent (DSS) were in fact atelectatic because we did not obtain corresponding morphological data by CT or ultrasound. However, we may still dare to say that the poorly ventilated areas detected by EIT in our study were either collapsed or at particular risk for collapse. The silent spaces in the ventral non-dependent lung during mechanical ventilation are interpreted as signs of overinflation, and the fact that they did not change significantly after the patients were brought into the supine position might be explained by the use of lung-protective ventilation settings, with low tidal volumes and PEEP applied during the entire procedure. Many academic studies have investigated the influence of different postures and ventilation conditions on the lung s function. 22 Up to now, however, this knowledge on regional formation of atelectasis could not be transferred into clinical practice, because there was no parameter available to quantify these phenomena at the bedside. Based on the results of the present study, silent spaces and the CoV derived by EIT could in future serve as such non-invasive and continuously available

7 234 Ukere et al. parameters. In particular, the EIT-derived and rather sensitive parameter silent space could potentially help clinicians to optimize mechanical ventilation during general anaesthesia by individualizing PEEP, tidal volumes, and recruitment manoeuvres in the future. This interplay of personalized ventilation parameters to mitigate the side-effects of anaesthesia and surgery may achieve the best possible protective effect for the lungs. Our study has clear limitations; we did not compare different EIT-guided ventilation settings but merely documented how different ventilation conditions and body positions affected the function of the lungs. The inclusion of patients without known lung disease in this first of a series of studies was in fact intentional. Further investigations in patients with pre-existing pulmonary diseases are inevitable. Furthermore, we did not perform EIT measurements after extubation. Thus, questions related to the reversibility of the observed regional ventilation defects were not addressed. In addition, we did not perform recruitment manoeuvres to re-inflate collapsed alveoli and to redistribute regional ventilation and blood flow to improve oxygenation and lung aeration. Conclusion Electrical impedance tomography was able to identify and quantify on a breath-by-breath basis circumscribed areas, so-called silent spaces, within healthy lungs that received little or no ventilation during general anaesthesia, capnoperitoneum, and different body positions. As these silent spaces are suggestive of atelectasis on the one hand and overdistension on the other, they might become useful to guide individualized protective ventilation strategies to mitigate the side-effects of anaesthesia and surgery on the lungs. Authors contributions Conception of study design: C.J.T., A.H., A.D.W., S.H.B., D.A.R. Study design: A.U., A.M. Patient recruitment: A.U., A.M., K.H.W. Data collection: A.U., A.M., K.H.W. Data analysis: A.U., A.M., K.H.W., C.J.T., A.H., A.D.W., S.H.B., D.A.R. Drafting the manuscript: A.U., A.M. Critical revision for important intellectual content: K.H.W., C.J.T., A.H., A.D.W., S.H.B., D.A.R. Supplementary material Supplementary material is available at British Journal of Anaesthesia online. Declaration of interest A.D.W. and S.H.B. are employees of Swisstom, Switzerland. References 1. Rothen HU, Sporre B, Engberg G, Wegenius G, Hedenstierna G. Airway closure, atelectasis and gas exchange during general anaesthesia. Br J Anaesth 1998; 81: Klingstedt C, Hedenstierna G, Baehrendtz S, et al. Ventilationperfusion relationships and atelectasis formation in the supine and lateral positions during conventional mechanical and differential ventilation. Acta Anaesthesiol Scand 199; 34: Futier E, Constantin JM, Paugam-Burtz C, et al. A trial of intraoperative low-tidal-volume ventilation in abdominal surgery. N Engl J Med 13; 369: Nopp P, Harris ND, Zhao TX, Brown BH. Model for the dielectric properties of human lung tissue against frequency and air content. Med Biol Eng Comput 1997; 3: Marz A, Ukere A, Wodack K, et al. Perioperative assessment of regional ventilation during changing body positions and ventilation conditions by electrical impedance tomography with increased spatial resolution and signal quality. Critical Care ; 19: P Gaggero PO, Adler A, Waldmann AD, Mamatjan Y, Justiz J, Koch VM. Automated robust test framework for electrical impedance tomography. Physiol Meas ; 36: Ferrario D, Grychtol B, Adler A, Sola J, Bohm SH, Bodenstein M. Toward morphological thoracic EIT: major signal sources correspond to respective organ locations in CT. IEEE Trans Biomed Eng 12; 9: Adler A, Arnold JH, Bayford R, et al. GREIT: a unified approach to 2D linear EIT reconstruction of lung images. Physiol Meas 9; 3: S3 9. Woitzik C, Grychtol B, Ferrario D, et al. Ermittlung von Messdaten und Proportionen des menschlichen Thorax aus CT- Schnitten zur mathematischen Korrektur von Rohdatensätzen der Elektrischen Impedanztomografie (EIT). Fortschr Röntgenstr 14; 186: WI_PO2 1. Frerichs I, Hahn G, Golisch W, Kurpitz M, Burchardi H, Hellige G. Monitoring perioperative changes in distribution of pulmonary ventilation by functional electrical impedance tomography. Acta Anaesthesiol Scand 1998; 42: Radke OC, Schneider T, Heller AR, Koch T. Spontaneous breathing during general anesthesia prevents the ventral redistribution of ventilation as detected by electrical impedance tomography: a randomized trial. Anesthesiology 12; 116: Zick G, Elke G, Becher T, et al. Effect of PEEP and tidal volume on ventilation distribution and end-expiratory lung volume: a prospective experimental animal and pilot clinical study. PLoS ONE 13; 8: e Amato MB, Meade MO, Slutsky AS, et al. Driving pressure and survival in the acute respiratory distress syndrome. NEnglJ Med ; 372: Karsten J, Luepschen H, Grossherr M, et al. Effect of PEEP on regional ventilation during laparoscopic surgery monitored by electrical impedance tomography. Acta Anaesthesiol Scand 11; : Determann RM, Royakkers A, Wolthuis EK, et al. Ventilation with lower tidal volumes as compared with conventional tidal volumes for patients without acute lung injury: a preventive randomized controlled trial. Crit Care 1; 14: R1 16. Severgnini P, Selmo G, Lanza C, et al. Protective mechanical ventilation during general anesthesia for open abdominal surgery improves postoperative pulmonary function. Anesthesiology 13; 118: Blankman P, Hasan D, Erik G, Gommers D. Detection of best positive end-expiratory pressure derived from electrical impedance tomography parameters during a decremental positive end-expiratory pressure trial. Crit Care 14; 18: R9 18. Bikker IG, Leonhardt S, Reis Miranda D, Bakker J, Gommers D. Bedside measurement of changes in lung impedance to monitor alveolar ventilation in dependent and non-dependent parts by electrical impedance tomography during a positive end-expiratory pressure trial in mechanically ventilated intensive care unit patients. Crit Care 1; 14: R1

8 Perioperative study of regional ventilation by EIT Zhao Z, Steinmann D, Frerichs I, Guttmann J, Möller K. PEEP titration guided by ventilation homogeneity: a feasibility study using electrical impedance tomography. Crit Care 1; 14: R8. Frerichs I, Hahn G, Hellige G. Gravity-dependent phenomena in lung ventilation determined by functional EIT. Physiol Meas 1996; 17(Suppl 4A): A Riedel T, Richards T, Schibler A. The value of electrical impedance tomography in assessing the effect of body position and positive airway pressures on regional lung ventilation in spontaneously breathing subjects. Intensive Care Med ; 31: Hahn G, Just A, Hellige G, Dittmar J, Quintel M. How absolute EIT reflects the dependence of unilateral lung aeration on hyper-gravity and weightlessness? Physiol Meas 13; 34: Handling editor: T. Asai