Measurement of pleural pressure swings with a fluid-filled esophageal catheter vs pulmonary artery occlusion pressure

    Measurement of pleural pressure swings with a fluid filled esophageal catheter versus pulmonary artery occlusion pressure S. Verscheure, P.B. Massion, S. Gottfried, P. Goldberg, L. Samy, P. Damas, S. Magder PII: DOI: Reference:

S0883-9441(16)30432-4 doi: 10.1016/j.jcrc.2016.08.024 YJCRC 52265

To appear in:

Journal of Critical Care

Please cite this article as: Verscheure S, Massion PB, Gottfried S, Goldberg P, Samy L, Damas P, Magder S, Measurement of pleural pressure swings with a fluid filled esophageal catheter versus pulmonary artery occlusion pressure, Journal of Critical Care (2016), doi: 10.1016/j.jcrc.2016.08.024

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

RI P

T

July 25

SC

Measurement of pleural pressure swings with a fluid filled esophageal catheter versus pulmonary artery occlusion pressure

1

NU

Authors: S. Verscheure1,2, P.B. Massion2., S. Gottfried1, P. Goldberg1, L. Samy1, P. Damas2, S. Magder1 Department of Critical Care and Physiology

MA

McGill University Heath Centre 1001 Decarie Av

ED

Montreal, Quebec ²Department of General Intensive Care

PT

University Hospital Center of Liege

AC

CE

Domaine du Sart Tilman, B35

Address correspondence to: S Magder

McGill University Health Centre 1001 Decarie Av Montreal, Quebec [email protected]

Liège, Belgium

ACCEPTED MANUSCRIPT 2

Abstract Purpose: Pleural pressure (Ppl) measured with esophageal balloon catheters (Peso) can guide

T

ventilator management and help interpretation of hemodynamic measurements, but these

RI P

catheters are not readily available or easy to use. We tested the utility of an inexpensive, fluid filled esophageal catheter (Peso) by comparing respiratory induced changes in pulmonary artery

SC

occlusion (Ppao), central venous(CVP) and Peso pressures.

NU

Methods: We studied 30 patients undergoing elective cardiac surgery who had pulmonary artery and esophageal catheters in place. Proper placement was confirmed by chest compression with

ED

pressure support ventilation (PS).

MA

airway occlusion. Measurements were made during pressure-regulated volume control (VC) and

Results: The fluid filled esophageal catheter provided a high quality signal. During VC and PS

PT

change in Ppao (∆Ppao) was greater than ∆Peso (bias = -2 mmHg) indicating an inspiratory

CE

increase in cardiac filling. During VC ∆CVP bias was 0 indicating no change in right heart filling but during PS CVP fell less than Peso indicating an inspiratory increase in filling. Peso

AC

measurements detected activation of expiratory muscles, development of non-West zone 3 lung conditions during inspiration, and ventilator-triggered inspiratory efforts. Conclusions: A fluid filled esophageal catheter provides a high quality, easily accessible and inexpensive measure of change in Ppl and provided insights into patient-ventilator interactions. Key words: central venous pressure, pulmonary artery occlusion pressure, pleural pressure, active expiration, heart-lung interactions, transmural pressure

ACCEPTED MANUSCRIPT 3

Pleural pressure measurements (Ppl) can provide important information for the management of critically ill patients (1, 2). For example, negative swings in Ppl during

T

spontaneous breathing give an indication of inspiratory muscle effort and work of breathing.

RI P

Measurement of Ppl in mechanically ventilated patients can identify patient-ventilator dyssynchrony and unrecognized inspiratory efforts during what is thought to be controlled

SC

ventilation (3). More recently, there has been increasing interest in the use of Ppl to obtain the

NU

best value of positive end-expiratory pressure (PEEP) that can maximize pulmonary recruitment while avoiding airway opening volutrauma (4, 5). Monitoring respiratory swings in Ppl also can

MA

help interpretate the processes behind variations in vascular pressures during the respiratory

ED

cycle.

Ppl most often has been estimated from pressure in an air-filled balloon in the esophagus

PT

(Peso) (6). These balloons can be combined with oral or naso-gastric tubes that are used for

CE

feeding or gastric drainage. However, current commercial esophageal balloon catheters are expensive and not readily available. The appropriate balloon volume and position also need to be

AC

regularly checked, which makes long term monitoring difficult. We have found that an inexpensive, relatively stiff tube which commonly is used for suctioning airways in children can provide an excellent measure of Ppl swings during the respiratory cycle when it is filled with fluid. Our primary objective was to confirm the reliability of this device in a series of critically ill patients. A second objective was to use measurements of Ppl to gain insight into common variations in thoracic hemodynamics that are due to heart-lung interactions in post-cardiac surgery patients. The rational for our approach is based on a previous study in which we showed that the peak fall in pulmonary artery occlusion pressure (Ppao) during a spontaneously triggered breath

ACCEPTED MANUSCRIPT 4

closely follows the peak change in Ppl measured with an esophageal balloon (7). In this study, we compared ventilation induced swings in Ppl obtained with the fluid filled catheter to changes

T

in Ppao. We chose cardiac surgery patients because pulmonary artery catheters are routinely

RI P

inserted for surgery in most of these patients at our institution. This group thus provided a readily available population for the study and, unlike the situation in most intensive care subjects,

SC

consent could be obtained when subjects were awake and competent prior to surgery. Subjects

NU

were studied during pressure-regulated volume ventilation (VC) just after surgery and then a few

MA

hours later with pressure support (PS) when they woke up. Methods

ED

Patients

PT

We performed a single center study in the medical-surgical intensive care unit (ICU) of the McGill University Heath Center (MUHC). The protocol was approved by the research ethic

CE

review board of the MUHC. All patients undergoing elective cardiac surgery (coronary artery

AC

bypass graft and/or valve replacement) were eligible. We studied thirty patients in whom preoperative consent was obtained, were sufficiently stable when they came from the operating room, and had a pulmonary artery catheter in place as part of their routine management. Exclusion criteria included gastrointestinal bleeding, anatomical esophageal abnormalities, such as esophageal varices, ulcers or tumor, presence of hemodynamic or respiratory instability, age less than 18 years, refusal by the attending physician or arrival too late in the ICU to make the measurements. Protocol and measurement

ACCEPTED MANUSCRIPT 5

After stabilization in the ICU, subjects were placed in the supine position with the head of the bed elevated to 30º. A sterile Mülly (Unomedical) suction catheter (external diameter 2.7

T

mm, length 53 cm) was inserted with the help of a guide-wire through one nostril. The tube was

RI P

filled with 0.9% saline and connected to a standard hemodynamic transducer (Transpac® IV Monitoring Kit, IcuMedical). A constant pressure of 300 mmHg was applied to the saline bag as

SC

used for maintaining invasive arterial pressure catheters. Proper positioning of the catheter in the

NU

esophagus was confirmed during controlled ventilation by doing three chest compressions in the mid-chest during an end-expiratory pause (1)(figure 1). With this procedure, the measurement of

MA

Peso was considered acceptable if change in Peso was at least 0.85 of the change of airway pressure (Paw). In all subjects the ratio was >95%. Unlike air filled catheters, pressures

ED

measured with fluid filled catheters are dependent upon a reference level. However, this was not

PT

necessary in this study because we were interested in comparing the change (delta, ∆) in vascular and esophageal pressures. Hemodynamic pressures were based on a level 5 cm below the sternal

CE

angle (8 ).

AC

The first measurement was performed just after the patient arrived in the ICU and still recovering from the anaesthesia and ventilated with VC. Maximum inspiratory increase in Ppao and CVP were compared to the maximum increase in Peso. When subjects awoke and began to trigger breaths the mode of ventilation was switched to PS. During this mode maximum negative deflections in Ppao and CVP were compared to the maximum negative deflections in Peso. Measurements of Ppao and CVP were taken at the base of the ‘c’ wave when visible or the base of the ‘a’ wave when the ‘c’ could not be identified. A second reviewer independently reviewed recordings from 20 subjects to test reproducibility of the measurements.

ACCEPTED MANUSCRIPT 6

The measurements of Ppl allowed us to make some observations on heart-lung interactions and their effects on hemodynamic interpretations. These included evidence of

T

recruitment of expiratory muscles and the appropriate time in the cycle to make hemodynamic

RI P

measurements, possible development of non-West zone 3 conditions during mechanical breaths,

SC

and ventilator-triggered ventilator responses.

We have previously described two patterns of recruitment of expiratory muscles (9, 10).

NU

In type A CVP falls during expiration. In type B CVP progressively increases during expiration.

MA

We assessed the prevalence of these patterns in the current sample as indicated by the change in Peso with the change in hemodynamic pressure.

ED

Second, when alveolar pressure is higher than pulmonary venous pressure, pulmonary veins collapse and produce flow-limitation or a Starling-resistor effect in what is called West

PT

zone 2 (11, 12). When alveolar pressure is higher than pulmonary arterial pressure, flow in that

CE

region is blocked in what is called West zone 1. In both West zone 1 and 2 Ppao reflects alveolar pressure and not left atrial pressure as is the case in West zone 3 in which there is no obstruction

AC

to pulmonary flow (13) (8, 14-17). During VC evidence of non-Zone 3 conditions was identified by an increase in Ppao that was close to the increase in Paw, more than 2 mmHg greater than the increase in Peso, and an increase in Ppao that was greater than the increase in CVP. During PS presence of non-zone 3 was suggested by a rise in Ppao with a fall in Peso. We cannot be certain that these criteria truly indicate non-zone 3 conditions, but we consider the data to be suggestive and if frequent, worthy of further study because of the significance for right ventricular loading (17, 18). Statistical analysis

ACCEPTED MANUSCRIPT 7

Data are expressed as mean ± standard deviation (SD). Agreement between measurements was assessed by Bland-Altman plots, identity plots and regression analysis. The

T

bias and limits of agreement were calculated as the difference of change of esophageal pressure

RI P

(∆Peso) during inspiration from the difference of ∆Ppao or ∆CVP during inspiration (ie ∆Peso∆Ppao, or ∆Peso-∆CVP) because in this study the change in Ppao and CVP were considered to

NU

Results

SC

be the independent variables (see discussion).

MA

Baseline characteristic of the patients are given in the table 1. Figure 2a shows a typical tracing of a subject during VC and Figure 2b an example of a subject breathing with PS. As can

ED

be seen in the figures the fluid filled esophageal catheter produced tracings with a high frequency response which even showed cardiac artifacts. See figure legends for specific details. The

PT

regression coefficients between the two reviewers were all ≥0.97 and the inter-observer

CE

variability was a mean of 0.3 and standard deviation of 0.7 mmHg.

AC

Figure 3 shows the Bland-Altman and identity plots of ∆Peso and ∆Ppao on the left side and ∆Peso and ∆CVP on the right side during VC. The bias for Ppao was -2.2 mmHg with limits of agreement of +1.4 to -5.8 mmHg, which means that Peso rose on average 2.2 mmHg less than Ppao. The regression equation is ∆Peso = 0.46 ∆Ppao + 4.0. The bias for ∆CVP was close to 0 (0.3 mmHg) and the limits of agreement were +2.8 to -2.1 mmHg. The regression equation is ∆Peso= 0.44 ∆CVP + 1.58. Figure 4 shows the Bland-Altman plot of ∆Peso and ∆Ppao and identity plot on the left side, and CVP on the right side for PS. In this case the changes refer to the peak negative deflection. The bias for ∆Ppao was -1.8 mmHg and limits of agreement of +2.1 to -5.8 mmHg.

ACCEPTED MANUSCRIPT 8

This means that during inspiration Peso fell more than Ppao and the delta was still negative as it was for VC because this is the difference of two negatives. This indicates that transmural Ppao

T

rises with both VC and PS during inspiration and left heart filling is increased by lung inflation.

RI P

The regression equation is ∆Peso = 0.91 ∆Ppao + 1.26. The bias for ∆CVP was negative, -2.2 mmHg with limits of agreement +1.4 to -5.8 mmHg indicating that transmural CVP, too,

SC

increased in many subjects. The regression equation is ∆Peso = 0.77 ∆CVP + 0.80.

NU

. Although Type A could not be definitively identified during VC because the pressure

MA

starts elevated in all subjects from positive pressure during inflation 2 subjects had definite and one possible type B pattern of active expiration during VC (example in figure 5). During PS, 9

ED

(30%) of subjects had type B (example in figure 2b) and 5 (16%) had type A on at least some breaths. These patterns were easier to identify on CVP tracings than on the Ppao tracing. As was

PT

the case during VC, it is hard to be certain of the presence of the type A pattern because the

CE

increase in Ppl with PS could contribute to an early elevation of expiratory pressure. Development of non-West zone 3 conditions based on Ppao measurement during VC was

AC

likely in 11(42%) subjects (example figure 2a) and possible in 2 others. Although it is difficult to determine this during PS because alveolar pressure is not measured, there was a suggestion of this being the case in at least 1 patient in whom Peso decreased by 2.5mmHg and Ppao rose during inspiration by ~ 6 mmHg (figure 5). Figure 6 shows an example of a subject who had an inspiratory effort triggered by a mechanical inspiration during volume control ventilation. On the first breath in figure 6 the inspiratory effort is not evident on Paw or Ppao tracing but is evident on Peso. On the second breath Peso again shows a small inspiratory effort with a slight distortion of Ppao. On the third breath there is moderate distortion of Paw, a marked distortion of Peso and Ppao, and even some

ACCEPTED MANUSCRIPT 9

distortion of the CVP. Ventilator triggered inspirations were identified in 10 of 26 subjects during VC.

RI P

T

Discussion

Our data confirm that a fluid filled catheter placed in the esophagus reliably tracks both

SC

positive and negative changes in Ppl as indicated by changes in Ppao and CVP. We used respiratory changes in Ppao as the gold standard because we previously validated this approach

NU

by comparing peak inspiratory changes in Ppao to peak changes in esophageal balloon pressure

MA

measurements (7). In this study we used respiratory induced changes in Ppao in cardiac surgery patients to provide a convenient sample to test the fluid filled device. In reality the “gold-

ED

standard” is the fluid filled esophageal catheter. The catheter gave a high quality signal with an excellent frequency response as indicated by high resolution of ‘a’ and ‘v’ waves from the

PT

atrium. It also gave important insights into changes in filling of the atrium during the ventilation

CE

cycle whereas the pulmonary artery occlusion catheter could only be used to indicate peak

AC

changes. This is because filling of the right and left atrium varies during the ventilator cycle. A limitation to the use of a fluid filled catheter is that the fluid in the measuring device adds an offset due to the effect of gravity on the fluid column. To obtain an absolute pressure the transducer needs to be referenced to an anatomical level by using a lateral x-ray to identify the position of the catheter and correcting the values to a standard level relative to the heart, but this was not necessary in this study in which we only measured changes in pressure. The force that distends elastic structures is the difference in pressure between the inside and outside of the structure. This is called transmural pressure. For structures outside the chest the surrounding pressure is atmospheric pressure and transducers are accordingly zeroed to

ACCEPTED MANUSCRIPT 10

atmospheric pressure and the value obtained with an indwelling catheter is the transmural pressure. However, for vascular structures in the chest, the outside pressure is Ppl and Ppl must

RI P

T

be considered when interpreting the observed pressure. During lung inflation with VC, the pressure surrounding cardiac chambers increases so

SC

that at a constant cardiac volume the pressure in the heart rises relative to atmosphere. Unless pulmonary venous pressure is very low, ie less than approximately 3 mmHg, lung inflation

NU

squeezes volume out of pulmonary vessels and increases left atrial filling. This raises left atrial

MA

transmural pressure (19-21) and accordingly the increase in Ppao was greater than the increase in Peso, and the bias on the Bland-Altman plot was -2 mmHg because ∆Ppao was subtracted from

ED

∆Peso. On the right side of the heart, the average rise in CVP was similar to the average rise in Ppl and transmural CVP did not change, which can occur if CVP is close to the plateau pressure

PT

of the cardiac function curve (22).

CE

During PS, we only studied the initial negative inspiratory component of ∆Ppl. During this phase, the environment of the heart is more negative relative to the rest of the body. This can

AC

increase right heart filling when the heart is not on the flat part of the cardiac function curve. This seemed to occur based on the identity plot (figure 4) which shows that CVP did not fall as much as Peso in many subjects indicating that CVP transmural pressure increased. Ppao, too, fell less than Peso indicating increased left heart filling during inspiration as seen with VC. These result are similar to our previous observations (7). Fluid filled catheters previously have been used in neonates (23, 24) and small animals because their small size and high frequency of breathing requires a high frequency response (25, 26) . However, fluid filled catheters only have been reported once in healthy adults (27). It has been argued that fluid filled catheters do not have a very good frequency response (28) but this was

ACCEPTED MANUSCRIPT 11

not the case in our study. This likely is because we maintained a high fluid pressure in the catheters. A recent publication also showed reliable signals with a thin esophageal catheter in

T

pigs even without fluid (29). We, however, found that the signal was very damped when

RI P

catheters were not fluid filled (data not shown).

SC

The high quality measurements of Ppl allowed us to analyze some common components of heartlung interaction. It is recommended that hemodynamic measurements be made at end-expiration

NU

for that is when chest muscles are relaxed and Ppl is closest to atmospheric pressure. However,

MA

patients often activate respiratory muscles during expiration, which can increase end-expiratory values of CVP and Ppao. We have described two basic patterns. In type A CVP increases at the

ED

start of expiration and progressively decreases during expiration. In type B, CVP progressively increases during expiration (9). These patterns were very common in our population with 30% of

PT

subjects have the B-type pattern during PS

CE

An unexpected observation that became evident with the use of esophageal pressure measurements was that non-West zone 3 (11) seemed to frequently develop in the region of the

AC

Ppao measurement at peak inspiration with VC, even though tidal volumes were not large. We identified the potential presence of this phenomenon by observing an increase in Ppao that was greater than the usual 2 mmHg more than the change in Peso and closer to the change in Paw but

these criteria do not indicate with certainty that non-West zone 3 conditions occurred. Of

importance when non-West zone 3 conditions developalveolar pressure becomes the load for right ventricular ejection (18) (12) and this right ventricular load increases one-to-one with increases in alveolar pressure (30). This might have been more frequent in our population because cardiac surgery patients often are limited by right heart filling so that pulmonary venous pressures tend to be lower and the critical closing pressure is more easily reached. The

ACCEPTED MANUSCRIPT 12

consequent increase in right heart load could have significant consequences when right ventricular function is decreased. We suspect that this phenomenon also occurred in some cases

RI P

T

during PS breaths but this is hard to determine without a measure of alveolar pressure. As recently reported by Akoumianaki et al (3), ventilator-triggered breaths were common in our

SC

subjects during VC ventilation. We observed this phenomenon in 10 of 26 subjects (example figure 6). It seemed to be more common when patients were awakening from deep sedation after

NU

their cardiac surgery and this occasionally led to significant disruptions of gas-exchange and

MA

hemodynamics. The phenomenon is identified easily with a Peso measurement, but is often not recognizable without one (examples breath 1 and 2 in figure 6).

ED

Summary

PT

An inexpensive fluid filled catheter positioned in the esophagus at the level of the heart provides an excellent estimate of change in Ppl for both positive and the negative Ppl swings during

CE

breaths. Ppl can be used for the assessment of respiratory effort, lung mechanics, and the effects

AC

of ventilation on hemodynamic pressure wave forms. Secondary observations from use of the Peso measurement in our study were that recruitment of expiratory muscles and “forced” expiration occurred frequently; development of West zone 1 or II conditions appeared to be common during peak inspiration with VC; and ventilator triggered inspiratory efforts were common in patients awakening from anaesthesia and these were often not evident without a Peso measurement. On behalf of all authors, the corresponding author states that there is no conflict of interest. Dr Sara Verscheure had funding from Fonds Leon Fredericq, Belgium. The authors had no other funds for this study.

ACCEPTED MANUSCRIPT 13

Figures

RI P

T

Figure 1

Example of change in Peso, CVP and Paw during the chest compression test. Change in CVP

SC

was identical to change in esophageal pressure (Peso) during the compression. CVP = central venous pressure, Paw= airway pressure, PAP = pulmonary artery pressure. All pressures are in

NU

mmHg.

MA

Figure 2a

Example of change in Paw (purple), Peso (blue), Ppao (black), CVP (red), during two pressure-

ED

regulated volume control (VC) breaths. (numbers indicate change in mmHg). In this example,

PT

the rise in Ppao was much larger than the rise in Peso and of similar magnitude to the change in Paw suggesting the development of non-West zone 3 condition during peak inspiration.

Figure 2b

AC

inspiration

CE

Abbreviations are the same as figure 1. Ppao = pulmonary artery occlusion pressure. Insp =

Example of change in Paw (purple), Peso (light blue), Ppao (red), and CVP (dark blue) during three pressure support (PS) breaths. Respiratory variations in Ppao and CVP were almost identical and are coloured on separate breaths for clarity. The double arrows indicate the inspiratory changes in Peso, Ppao and CVP. All three pressures decrease during expiration indicating an example of type A activation of expiratory muscles. Numbers in boxes are the change in pressures in mmHg. During inspiration Peso fell by 9 mmHg whereas Ppao fell by 7 mmHg and CVP by 7 mmHg indicating inspiratory increases in transmural Ppao and CVP.

ACCEPTED MANUSCRIPT 14

Abbreviations are the same as figure 2a. Figure 3. Bland-Altman and identity plots of change in Peso (∆Peso) and change in Ppao

RI P

T

(∆Ppao) (left) and ∆CVP (right) during pressure-regulated volume control breaths (n=30). Figure 4. Bland-Altman and identity plots of ∆Peso and ∆Ppao) (left) and ∆CVP (right) during

SC

PS (n=26).

NU

Figure 5

Example of possible non-West zone 3 condition for Ppao during PS inspiration. Peso fell by 3

MA

mmHg whereas Ppao rose by 6 mmHg and matched the change in airway pressure which together suggest that the Ppao reflects alveolar pressure during inspiration. Numbers in boxes are

ED

the change in pressures in mmHg.

PT

Figure 6

CE

Example of ventilator-triggered inspiratory activity during pressure-regulated volume control

AC

breaths (marked by *). The esophageal tracing during inspiration is highlighted. In the first breath Peso (traced in red) fell during the volume control breath indicating ventilator activation of inspiratory muscles. This is not evident in Paw, CVP or Ppao. There is again a small activation in the second breath and a much larger one in the third breath which produced a moderate distortion of Paw and large negative deflections of Peso and Ppao and even some distortion of the CVP. Abbreviations are the same as figure 2.

ACCEPTED MANUSCRIPT 15

References

AC

CE

PT

ED

MA

NU

SC

RI P

T

1. Akoumianaki E, Maggiore SM, Valenza F, Bellani G, Jubran A, Loring SH, et al. The application of esophageal pressure measurement in patients with respiratory failure. AmJRespirCrit Care Med. 2014;189(5):520-31. 2. Hedenstierna G. Esophageal pressure: benefit and limitations. Minerva anestesiologica. 2012;78(8):959-66. 3. Akoumianaki E, Lyazidi A, Rey N, Matamis D, Perez-Martinez N, Giraud R, et al. Mechanical ventilation-induced reverse-triggered breaths: a frequently unrecognized form of neuromechanical coupling. Chest. 2013;143(4):927-38. 4. Talmor D, Sarge T, Malhotra A, O'Donnell CR, Ritz R, Lisbon A, et al. Mechanical ventilation guided by esophageal pressure in acute lung injury. NEnglJMed. 2008;359(20):2095-104. 5. Talmor D, Sarge T, O'Donnell CR, Ritz R, Malhotra A, Lisbon A, et al. Esophageal and transpulmonary pressures in acute respiratory failure. Crit Care Med. 2006;34(5):1389-94. 6. Milic-Emili J, Mead J, Turner JM, Glauser EM. Improved technique for estimating pleural pressure from oesophageal balloon. Journal of Applied Physiology. 1964;19:207-11. 7. Bellemare P, Goldberg P, Magder S. Variations in pulmonary artery occlusion pressure to estimate changes in pleural pressure. Intensive Care Medicine. 2007;33(11):2004-8. 8. Magder S. Invasive hemodynamic monitoring. Crit Care Clin. 2015;31(1):67-87. 9. Magder S. Diagnostic Information from the Respiratory Variations in Central Hemodynamics Pressures. In: Scharf SM, Pinsky MR, Magder S, editors. Respiratory-Circulatory Interactions in Health and Disease. New York: Marcel Dekker, Inc.; 2001. p. 861-82. 10. Magder S. Invasive intravascular hemodynamic monitoring: Technical issues. Critical Care Clinics. 2007;23:401-14. 11. West JB, Dollery CT, Naimark A. Distribution of blood flow in isolated lungs:relation to vascular and alveolar pressures. JApplPhysiol. 1964;19:713-24. 12. Permutt S, Bromberger-Barnea B, Bane HN. Alveolar pressure, pulmonary venous pressure, and the vascular waterfall. Med Thoracalis. 1962;19:239-60. 13. Teboul JL, besbes m, Andrivet P, Axler O, Douget D, Zelter M, et al. A bedside index assessing the reliability of pulmonary artery occlusion pressure measurements during mechanical ventilation with positive end-expiratory pressure. Journal of Critical Care. 1992;7(1):22-9. 14. Roy R, Powers SR, Feustel PJ, Dutton RE. Pulmonary wedge catheterization during positive endexpiratory pressure ventilation in the dog. Anesthesiology. 1977;46(6):385-90. 15. Tooker J, Huseby J, Butler J. The effect of Swan-Ganz catheter height on the wedge pressure-left atrial pressure relationships in edema during positive-pressure ventilation. American Review of Respiratory Diseases. 1978;117(4):721-5. 16. Jardin F, Farcot JC, Boisante L, Carien N, Margairaz A, Bourdarias JP. Influence of positive endexpiratory pressure on left ventricular performance. New England Journal of Medicine. 1981;304:38792. 17. Jardin F, Vieillard-Baron A. Right ventricular function and positive pressure ventilation in clinical practice: from hemodynamic subsets to respirator settings. Intensive Care Med. 2003;29(9):1426-34. 18. Vieillard-Baron A, Loubieres Y, Schmitt JM, Page B, Dubourg O, Jardin F. Cyclic changes in right ventricular output impedance during mechanical ventilation. J Appl Physiol. 1999;87(5):1644-50. 19. Howell JB, Permutt S, Proctor DF, Riley RL. Effect of inflation of the lung on different parts of pulmonary vascular bed. JApplPhysiol. 1961;16:71-6. 20. Magder SA, Lichtenstein S, Adelman AG. Effects of negative pleural pressure on left ventricular hemodynamics. American Journal of Cardiology. 1983;52(5):588-93.

ACCEPTED MANUSCRIPT 16

AC

CE

PT

ED

MA

NU

SC

RI P

T

21. Permutt S, Howell JBL, Proctor DF, Riley RL. Effect of lung inflation on static pressure volume characteristics of pulmonary vessels. Journal of Applied Physiology. 1961;16:64-70. 22. Magder S, Lagonidis D, Erice F. The use of respiratory variations in right atrial pressure to predict the cardiac output response to PEEP. Journal of Critical Care. 2002;16(3):108-14. 23. Coates AL, Davis GM, Vallinis P, Outerbridge EW. Liquid-filled esophageal catheter for measuring pleural pressure in preterm neonates. JApplPhysiol (1985). 1989;67(2):889-93. 24. Asher MI, Coates AL, Collinge JM, Milic-Emili J. Measurement of pleural pressure in neonates. JApplPhysiol RespirEnvironExercPhysiol. 1982;52(2):491-4. 25. Saldiva PH, Cardoso WV, Caldeira MP, Zin WA. Mechanics in rats by end-inflation occlusion and single-breath methods. JApplPhysiol (1985). 1987;63(5):1711-8. 26. Sousa AS, Moll RJ, Pontes CF, Saldiva PH, Zin WA. Mechanical and morphometrical changes in progressive bilateral pneumothorax and pleural effusion in normal rats. The European respiratory journal. 1995;8(1):99-104. 27. Flemale A, Gillard C, Dierckx JP. Comparison of central venous, oesophageal and mouth occlusion pressure with water-filled catheters for estimating pleural pressure changes in healthy adults. EurRespirJ. 1988;1(1):51-7. 28. Talmor DS, Fessler HE. Are esophageal pressure measurements important in clinical decisionmaking in mechanically ventilated patients? RespirCare. 2010;55(2):162-72. 29. Beda A, Guldner A, Carvalho AR, Zin WA, Carvalho NC, Huhle R, et al. Liquid- and air-filled catheters without balloon as an alternative to the air-filled balloon catheter for measurement of esophageal pressure. PLoSOne. 2014;9(9):e103057. 30. Magder S, Guerard B. Heart-lung interactions and pulmonary buffering: Lessons from a computational modeling study. RespirPhysiol Neurobiol. 2012;182(2-3):60-70.

ACCEPTED MANUSCRIPT

MA

NU

SC

RI P

T

17

AC

CE

PT

ED

Figure 1

ACCEPTED MANUSCRIPT

MA

NU

SC

RI P

T

18

AC

CE

PT

ED

Figure 2a

ACCEPTED MANUSCRIPT

MA

NU

SC

RI P

T

19

AC

CE

PT

ED

Figure 2b

ACCEPTED MANUSCRIPT

MA

NU

SC

RI P

T

20

AC

CE

PT

ED

Figure 3

ACCEPTED MANUSCRIPT

MA

NU

SC

RI P

T

21

AC

CE

PT

ED

Figure 4

ACCEPTED MANUSCRIPT

MA

NU

SC

RI P

T

22

AC

CE

PT

ED

Figure 5

ACCEPTED MANUSCRIPT

MA

NU

SC

RI P

T

23

AC

CE

PT

ED

Figure 6

ACCEPTED MANUSCRIPT 24

RI P

N = 31 27 (87) 65 ± 10 67.8 ± 7.4 6 ± 1.9

AC

CE

PT

ED

SC

7.29 ± 0.06 46 ± 6 117 ± 4 22.3 ± 2.2 1.8 ± 0.8

NU

MA

Characteristic Male sex (%) Age (years) Predicted body weight (PBW) (kg) SOFA score at admission Arterial blood gas at arrival to ICU - pH - pCO2 (mmHg) - pO2 (mmHg) - Bicarbonate (mmol/L) - Lactate (mmol/L) Hemodynamic variables at arrival to ICU - Heart rate (bpm) - Systolic arterial pressure (mmHg) - Mean arterial pressure (mmHg) - Diastolic arterial pressure (mmHg) Respiratory variables at arrival to ICU - Expiratory tidal volume (ml) - Expiratory tidal volume (ml/kg ) - Respiratory rate (bpm) - PEEP cmH2O - Peak inspiratory pressure (cmH2O) - FiO2%

T

Table 1 Patient characteristics

87 ± 12 113 ± 16 76 ± 11 57 ± 10 528 ± 76 7.7 ± 0.9 15 ± 3 5.5 ± 1.5 20 ± 4 51 ± 7