Protective PEEP and Lung Capacity May Be Determined by a Rapid PEEP-step Procedure

Protective and lung capacity determined rapid Abstract words Background: A protective ventilation strategy should be based on assessment of lung mechanics and transpulmonary pressure, as this is the pressure that directly “hits” the lung. Esophageal pressure has been used for this purpose but has not gained widespread clinical acceptance. Instead, respiratory system mechanics and airway driving pressure have been used as surrogate measures. We have shown that the lung P/V curve coincides with the line connecting the end-expiratory airway P/V points of a PEEP trial. Consequently, transpulmonary pressure increases as much as PEEP is increased. If the change in end-expiratory lung volume ( Δ EELV) is determined, lung compliance (CL) can be determined as Δ EELV/ Δ PEEP and Δ PTP as tidal volume times ΔPEEP/Δ EELV. Methods: In ten patients with acute respiratory failure, ΔEELV was measured during each 4 cmH 2 O PEEP- step from 0 to 16 cmH 2 O and CL for each PEEP interval calculated as ΔEELV/ΔPEEP giving a lung P/V curve for the whole PEEP trial. Results: Lung P/V curves showed a marked individual variation with an overall lung compliance of 43 – 143 ml/ cmH 2 O (total inspiratory volume divided by end-inspiratory transpulmonary plateau pressure at PEEP 16 cmH 2 O). The two patients with lowest lung compliance were non-responders to PEEP with decreasing lung compliance at high PEEP levels, indicating over-distension. Patients with higher lung compliance had a positive response to PEEP with successively higher lung compliance when increasing PEEP. A two- step PEEP procedure starting from a clinical PEEP level of 8 cmH 2 O gave almost identical lung P/V curves as the four PEEP-step procedure. The ratio of airway driving pressure ( Δ PAW) to transpulmonary driving pressure (ΔPTP/ΔPAW) varied between patients and changed with PEEP, reducing the value of Δ PAW as surrogate for Δ PTP in individual patients. Conclusion: Separation of lung and chest wall mechanics can be achieved without esophageal pressure if when PEEP .


Introduction
Inadequately applied mechanical ventilation may cause injury to lung tissue and multi-organ failure, ventilator induced lung injury (VILI) [1][2][3][4][5]. Decreasing tidal volume was shown to be effective in decreasing mortality in ARDS [6,7] and it was concluded that a protective ventilation strategy included tidal volumes of 6 ml per kg ideal body weight [8]. In spite of the fact that experimental studies showed that increasing PEEP led to a decrease of VILI [9-11], large multi-centre studies failed to show significant positive effect of high PEEP on mortality [12][13][14]. A large multi-centre study with a recruitment manoeuvre and PEEP titration, where PEEP was set two cmH2O above the PEEP level with best respiratory system compliance, showed even increase in mortality in the study group compared to the control group, which was managed with low tidal volume/low PEEP according to ARDSnet [6,15]. This was a setback as the PEEP level with the highest respiratory system compliance also has the lowest airway driving pressure, which is strongly correlated to reduced mortality [16,17]. But, as VILI is a lung problem, it would be rational to individualize according to best lung compliance and transpulmonary driving pressure instead [18,19], especially so, as there is no correlation in individual patients between airway and transpulmonary driving pressure [20].
However, adding esophageal pressure measurements to a standard multi-step PEEP titration makes the procedure even more time consuming and cumbersome and is still not a standard procedure in every patient.
We have found that the size of the end-expiratory lung volume change after a PEEP change is determined by the size of the PEEP change and the elastic properties of the lung, ΔEELV = ΔPEEP x CL [20][21][22][23][24][25][26], which indicates that the chest wall does not affect lung mechanics during end-expiratory zero-flow conditions. The data from a study by Yoshida using esophageal pressure measurement during a PEEP-trial 0-24 cmH2O confirms our findings. In that study, ΔEELV is not determined [27], but lung compliance (CL) can be calculated at each PEEP level by data retrieved from figures of the study.
Consequently, ΔEELV can be determined for each PEEP step and tidal airway-and lung P/V curves can be displayed at respective end-expiratory lung volume level. This reveals that end-expiratory-and end-inspiratory transpulmonary P/V points are positioned along a single, common lung P/V curve coinciding with the end-expiratory airway P/V curve (Fig. 1). [27]. This makes it possible to determine transpulmonary and chest wall driving pressure and lung and chest wall elastance (EL, ECW) and respiratory system elastance (ERS) as the sum of tidal lung and chest wall elastance at each PEEP level. Also, ΔEELV between PEEP levels can be calculated as ΔPEEP/EL (ΔPEEP x CL).

Left panel: Tidal airway P/V curves at PEEP 4 to 24 cmH2O with tidal volume of 200 ml from Yoshida
et al [27] where ΔEELV was determined for each 2 cmH2O PEEP step as ΔPEEP/EL, and EL calculated as ΔPTP/VT.

Right panel:
Tidal airway and lung P/V curves are plotted in the same diagram, showing that the lung P/V curve coincides with the end-expiratory airway P/V curve, which confirms that lung elastance can be determined as ΔPEEP/ΔEELV (lung compliance, CL = ΔEELV/ΔPEEP).
Even if transpulmonary driving pressure is the dominating factor, also the PEEP level, minute ventilation, flow and respiratory rate are important factors of VILI, as described by the "Power" concept [28][29][30]. However, transpulmonary pressure is difficult to determine by esophageal pressure. To make the concept work for a majority of ventilator patients, airway driving pressure has been introduced in the concept as a surrogate of transpulmonary driving pressure [31,32], but there is no correlation between the two in individual patients [20].
The aim of the study is to demonstrate that the lung P/V curve can be measured without esophageal pressure measurements with a two PEEP-step procedure and to determine the PEEP level, where the transpulmonary driving pressure of any chosen tidal volume is lowest.

Patients and Methods
The study was performed in a mixed ICU of a university hospital and approved by the Local Ethics Committee of Gothenburg, Sweden, Box 100, S-405 30 Gothenburg, protocol number: 112-08. Informed consent was obtained from next of kin. Twelve mechanically ventilated patients with acute respiratory failure were included in the original study, where conventionally measured transpulmonary pressure by esophageal pressure was compared to transpulmonary pressure calculated on the basis of lung compliance determined as ∆EELV/∆PEEP [33]. In the present study, ten of the patients could be used for analysis, while two of the patients had to be excluded from the analysis because of incomplete EIT data and cuff leakage (Table 1).

Study protocol
Patients were ventilated in a volume-control mode with a breathing rate of 20 breaths/min, with tidal volumes of 6-7 ml/kg ideal body weight in supine position. Before the start of protocol, sedation was deepened and muscle relaxant (rocuronium 40-50 mg) given.

Pressure and volume measurements
Ventilatory flow, volume and pressure were measured at the Y-piece with a D-lite side-stream spirometer connected to an AS/3 multi-module monitor (GE Healthcare, Helsinki, Finland).

End-expiratory lung volume change, ΔEELV
End-expiratory lung volume increase was determined as the cumulative expiratory tidal volume difference between PEEP levels and traced for the whole PEEP trial by inspiratory tidal volume calibrated EIT (Dräger Medical, Lübeck, Germany) using a 16-electrode belt placed around the chest wall at the 5th intercostal space (Fig. 2)

Calculation of the lung P/V curve
The term compliance (C) is normally used in the clinic, but for calculations, inverted compliance, 1/C = elastance (E), is more suitable, as respiratory system elastance is the sum of lung and chest wall elastance and consequently airway driving pressure the sum of transpulmonary and chest wall (pleural) driving pressure, ΔPAW = ΔPTP + ∆PPL. The relationship between compliance and elastance is shown in table 2.  Tidal pleural pressure variation of a tidal volume equal to ΔEELV (VT=ΔEELV) is the difference between airway and transpumonary driving pressure, ΔPAW -ΔPTP, and as transpulmonary driving pressure is equal to ΔPEEP, tidal pleural pressure variation is On the basis of this derivation, the lung P/V curve was obtained by plotting the end-expiratory transpulmonary pressure (PEEP) versus the end-expiratory lung volume at each PEEP level. Lung compliance is determined as the difference in end-expiratory lung volume between two PEEP levels divided by the change in PEEP.
The transpulmonary pressure at each PEEP level is equal to the end-expiratory airway pressure (PEEP), and, the lung P/V curve between ZEEP and PEEP 16 cmH2O can be obtained by plotting the end-expiratory transpulmonary pressure (PEEP) versus the end-expiratory lung volume at each PEEP level. By solving the equation of the best-fit lung P/V curve for the end-inspiratory lung volume level at each PEEP level, end-inspiratory transpulmonary pressures are obtained. It is an inherent feature of the PEEP step method that end-inspiratory transpulmonary pressure at the highest PEEP level (16 cmH2O) cannot be determined, but has to be estimated by calculating end-inspiratory transpulmonary pressure as end-inspiratory airway pressure minus tidal pleural pressure variation (ΔPPL) extrapolated from the lower PEEP levels. Chest wall compliance (CCW) and ΔPPL are mainly constant when PEEP is increased [21, 23, 24, 37]. However, analysis of individual patient´s behavior of the present study and in Stahl and coworkers studies [38,39], shows that CCW increases slightly when PEEP is increased in extrapulmonary ARDS patients, but remains mainly unchanged in pulmonary ARDS patients, which makes it necessary to perform the extrapolation procedure (Fig. 3).
Finally, a best-fit lung P/V curve was calculated from end-expiration at the lowest PEEP level to endinspiration at the highest PEEP level and the third degree equation for the curve used for determining the PEEP level where transpulmonary driving pressure was lowest for the clinically used tidal volume.
Total lung compliance (CLTOT) was calculated as the lung volume at end-inspiration (ΔEELV 0-16 + VT) divided by the end-inspiratory transpulmonary pressure at PEEP 16 cmH2O.

Results
The static steady state end-expiratory and end-inspiratory transpulmonary pressure and volume points were obtained for the whole PEEP trial from end-expiration at the lowest PEEP level to endinspiration of the highest, and used to compose lung P/V curve. The lung P/V curves of the ten patients are ranked according to total lung compliance calculated as end-inspiratory lung volume divided by end-inspiratory transpulmonary pressure at the highest PEEP level ( fig. 4). There are significant differences in the shape of the P/V curve of individuals with almost the same total lung compliance, but in spite of this, in all cases, the end-expiratory and end-inspiratory transpulmonary P/V points are aligned on a common single lung P/V curve. In all patients, the lung P/V curves show a lower inflection point (LIP), or rather a lower inflection zone, when lung compliance is increasing with PEEP increase from 0 to 8 -12 cmH2O (Fig. 4).

Fig. 4. Lung P/V measured data points and best fit curves (according to a third degree equation) in ten
patients with lung compliance ranging from 43 to 143 ml/cmH2O. The correlation coefficient (r 2 ) of the end-expiratory and end-inspiratory transpulmonary P/V points to the best-fit curve is above 0.99.
Note that in patients with almost equal compliance, patient 2 and 3, and patient 8 and 9, lung compliance increases at the highest PEEP level in patient 3 and 8, and decreases in patient 2 and 9.
Thus, in patient 2 and 9 there is an unfavorable response to high PEEP and in patient 3 and 8 there is a positive response to high PEEP.
As the calculation of the lung P/V curve requires measurement of ΔEELV, total lung compliance (CLTOT) can be regarded as an indicator of Inspiratory Capacity. In figure 5, the lung P/V curves of the ten patients are plotted to illustrate that patients treated for acute respiratory failure has very diverse lung mechanical properties, ranging from signs of emphysema with supernormal total lung compliance to a patient showing lung elastic properties comparable of a baby lung situation. The ratio of transpulmonary to airway driving press, ∆PTP/ΔPAW changes with changes in PEEP.
When changing from PEEP 12 cmH2O to 16 cmH2O, ∆PTP and ΔPAW may even change in opposite direction ( Fig. 6). To reduce the time required for the measurement procedure, the lung P/V curve was calculated from PEEP 8 cmH2O (regarded as a possible clinical PEEP level) with two steps up to 12 and 16 cmH2O of PEEP. The transpulmonary plateau pressure at PEEP 16 cmH2O was extrapolated by principally the same method as described for the original four-PEEP-step procedure, but that reduces the number of PEEP levels, which can be used as basis for extrapolation procedure from four to two. Estimated endinspiratory transpulmonary plateau pressure with the two-step procedure was 20.3 ± 1.2 cmH2O and the corresponding pressure after the original four-step procedure was 20.3 ± 1.8 cmH2O. (Fig. 7).
Lung compliance calculated as ΔEELV from 8 to 16 cmH2O PEEP divided by a ΔPEEP of 8 cmH2O correlated closely to lung compliance calculated as ΔEELV from 0 to 16 cmH2O PEEP divided by a ΔPEEP of 16 cmH2O, Y = 1.28X, r 2 = 0.97. The red arrow shows the tidal lung P/V curve with the lowest transpulmonary driving pressure for the tidal volume used clinically in the patients (6-7 ml/kg ideal body weight).

PEEP response
Lung compliance increases with increasing PEEP in PEEP responders and is associated with higher over-all lung compliance, than in non-responders, where over-all lung compliance is low, as seen in the two most non-responsive patients (pat. 1 and 2) with lung compliance around 50 ml/cmH2O, which is half the lung compliance seen in lung healthy [23,37]. The PEEP non-responder is characterized by decreasing lung compliance with an upper deflection point on the lung P/V curve at higher PEEP levels (Fig. 8). and a transpulmonary plateau pressure, which is well below VILI levels. Note, that inflation by PEEP occurs from ventral to dorsal region (grey-black, thick arrow), which causes tidal ventilation to be moved in dependent direction when PEEP is increased.

Discussion
In this observational study, we could show that it was possible to obtain precise lung P/V curves without esophageal pressure measurements during a PEEP trial. Lung mechanics showed marked individual variation with over-all lung compliance between 143 and 43 ml/cmH2O. Eight of the ten patients showed varying degree of positive PEEP response lung compliance increasing with PEEP, while the two patients with the lowest lung compliance were PEEP non-responders lung compliance decreasing with increasing PEEP. In the individual patient, there was no correlation between transpulmonary-and airway driving pressure. We also showed that a lung P/V curve covering the transpulmonary pressure/volume range within minimal risk of VILI could be obtained with only two PEEP steps, making the measurement procedure fast and easy in a clinical situation. This best-fit lung P/V curve could be used to determine the PEEP level where transpulmonary driving pressure was lowest for any given tidal volume.

Methodological considerations
At FRC/ZEEP there is a balance between the recoil of the lung and the expanding thoracic wall, which has caused inflation of the lung from residual volume to FRC with a positive transpulmonary pressure, and an equally negative pleural pressure as result. The thoracic cage strives outwards until 70 -80 % of TLC, but cannot reach higher end-expiratory volume than FRC unless the lung is inflated by PEEP. As the chest wall complex does not lean on the lung at end-expiration, even when PEEP is increased, until lung volume reaches 70-80 % of TLC, transpulmonary pressure increases as much as PEEP is increased. Consequently, lung and chest wall mechanics can be separated by a PEEP trial as lung compliance is equal to ΔEELV divided by ΔPEEP. A lung P/V curve derived from a PEEP trial is a true static lung P/V curve, where end-expiratory and end-inspiratory transpulmonary P/V points are aligned on a single common curve, as transpulmonary pressure at a certain lung volume is independent of whether this volume is reached by tidal or PEEP inflation (Fig. 1, 3, 4). An additional section shows the complete mathematical derivation of PEEP inflation [see Additional file 1].
Our aim was to develop a method for separation of lung and chest wall mechanics that was fast and easy that it could be used under any circumstances, at any time of ventilator treatment from start until extubation. We investigated whether a two-PEEP-step, three PEEP level procedure was as precise as the four-step five PEEP level procedure and delivered similar information on lung elastic properties, which it did. The two PEEP-step-up procedure takes around three minutes. It is probable that the lung P/V curve would be more precise if yet another PEEP step was used, for example 8 -12-16 -20 cmH2O, for improved estimation (extrapolation) of end-inspiratory transpulmonary pressure. but this will lead to a prolongation of the measurement procedure from 3 to ≈ 4,5 minutes. If the PEEP step method is implemented in a ventilator, it can be programmed to make a PEEP-down procedure from a high initial PEEP level as prolonged expirations step-wise down to lower PEEP levels, which will make the procedure very rapid. However, such a procedure needs to be evaluated as hysteresis phenomenon during fast procedures may cause less adequate results [40][41][42].
It is an inherent feature of the PEEP step method that the end-inspiratory transpulmonary pressure (PTPplat) at the highest PEEP level used during the measurement procedure has to be estimated instead of measured (Fig. 3). In the present analysis, 8 out of ten patients had a PTPplat below 20 cmH2O and the estimation procedure could have been avoided by adding a PEEP step up to 20 cmH2O. In cases with severely compromised oxygenation, ECMO treatment may be avoided if PEEP can be increased safely, i.e. without risk of exceeding the upper transpulmonary pressure limit ( ≈ 24 cmH2O) [43][44][45][46]. In such cases, it seems reasonable to extend the measurement procedure to a PEEP of 24 cmH2O to obtain a correct measured lung P/V curve for the whole PEEP/EELV range within the safety limit.
In the present study ΔEELV was determined as the cumulative difference in expiratory tidal volume between PEEP levels, but as a long duration, 5 PEEP-level trial was performed, the changes in endexpiratory lung volume was enhanced by EIT tracking (Fig.2). An additional paragraph provides detailed information on the tidal volume calibration by EIT process [see Additional file 1]. During a two or three PEEP-step procedure, ΔEELV can be determined by cumulative expiratory tidal volume difference determined by the ventilator spirometry alone with high precision (± 3.5 %)[37].
To assess cumulative DEELV from ZEEP to the highest PEEP level, the two-step procedure should be preceded by a PEEP decrease down to ZEEP and back again to baseline PEEP. This also makes the correct calculation of total lung compliance (CLTOT) as (DEELV0-16 + VT)/PTPplatPEEP16 possible. This is an advantage as total lung compliance reflects the size of the lung at ZEEP, low CLTOT -baby lung, high CLTOT -normal FRC.
The two-step procedure can be regarded as a diagnostic procedure, but to allow for continual monitoring of transpulmonary pressure only a simple PEEP step up and down procedure where PEEP is changed as much as the existing transpulmonary driving pressure, is required. Such a PEEP updown procedure should be performed on a regular basis and when patient position is changed or after interventions, such as suctioning has been performed.

Transpulmonary plateau and driving pressure and lung capacity
The most striking feature of the PEEP-step method is that not only transpulmonary pressure is determined, but also the increase in end-expiratory lung volume caused by increasing PEEP. The cumulative increase in EELV, the PEEP volume [28,47], is an indicator of the volume capacity of the lung and directly related to lung compliance. In a lung healthy individual, the over-all lung compliance is 80 -125 ml/cmH2O and a PEEP increase of 10 cmH2O should accordingly result in a ∆EELV of 800 -1250 ml. In table 3, we have identified some key data for different lung conditions and calculated the transpulmonary driving pressure of a tidal volume of ≈ 7 ml/kg IBW in a person with an ideal body weight (IBW) of 70 kg, 500 ml. This shows that such a "protective" tidal volume causes almost twice the ΔPTP in a pulmonary ARDS patient, as in a lung healthy. In the ARDSnet study in 2000, a tidal volume of 12 and 6 ml IBW was compared in ARDS patients and mortality was highly significantly lower when using the lower tidal volume. Since then, a tidal volume of 6 ml/kg IBW is regarded as "protective" and commonly used. In our study, all patients were ventilated with a tidal volume of 6 -7 ml, but as lung mechanics were highly varying, with lung compliance from above normal to levels corresponding to pulmonary ARDS, such a protective tidal volume is not protective in all individual patients. In fact, these tidal volumes did not correlate with respective individual total lung compliance, which reflects the size of the lung at FRC [48,49]. Thus, from a lung mechanical point of view, it would be an advantage to adapt the tidal volume to lung capacity as a surrogate for functional residual capacity when calculating the stress/strain ratio [48].

VILI assessment
A low PEEP volume, ΔEELV between ZEEP and the highest PEEP level of the measurement procedure, is a result of a small lung at FRC, in the worst cases a "baby lung" [49]. PEEP inflation, like slow and tidal inflation [50][51][52], occurs successively from non-dependent to dependent lung regions [53].
When the PEEP volume is low, it indicates that more dependent lung regions are consolidated and cannot be PEEP recruited and inflated unless very high pressures are used. In these patients, PEEP will mainly cause further inflation of the small ventral lung, like the ensuing tidal volume, with risk of overdistension and VILI (Fig. 9). This risk can be minimized by adapting ventilator settings to the mechanical condition of the lung, and this is best done by identifying when the mechanical condition diverge from "normality" or where the mechanical insult is lowest. PEEP titration is commonly used to find the PEEP level where respiratory system compliance is highest [15, [54][55][56][57] and consequently airway driving pressure lowest [16], which supposedly is the PEEP level where the risk of VILI is lowest. However, the transpulmonary driving pressure to airway driving pressure ratio, ∆PTP/ΔPAW changes significantly when changing PEEP ( fig. 6) making the airway driving pressure less suitable as surrogate for transpulmonary driving pressure [18, 48,49] . Thus, instead of seeking the PEEP level with the lowest airway driving, the PEEP level with the lowest transpulmonary driving pressure (PEEPmin) is preferable. As the tidal transpulmonary P/V curve is positioned along the lung P/V curve between end-expiration at baseline PEEP and end-inspiration at the highest PEEP level (Fig. 1, 3, 4), the transpulmonary plateau and driving pressure and can be determined for any PEEP -VT combination [58, 59]. The two PEEP-step lung P/V curve starts at PEEP 8 cmH2O, which is above the lower inflection zone, seen in the 0 -16 cmH2O P/V curve and PEEPmIn is in all patients positioned at the steepest part of the curve. In the PEEP responders, PEEPmIn may be quite high and obviously, also the end-inspiratory transpulmonary plateau pressure level must be considered when finally selecting best (mechanical) PEEP. When assessing the risk of VILI (by for example the power concept), transpulmonary driving pressure and plateau pressure are preferable to airway (respiratory system) driving and plateau pressure [19,28,29,31].

Identifying "under-treatment"
As important it is to minimize VILI, it is to identify, where it is possible to increase PEEP to improve oxygenation without risk of VILI. Grasso and coworkers managed to avoid ECMO treatment in half of the patients with influenza A (H1N1)[43], when they determined transpulmonary pressure using esophageal pressure and found that in half the patients, PEEP could be further increased without reaching the upper limit for transpulmonary plateau pressure, 25 cmH2O [44][45][46]. In the present study, two patients had the same respiratory system compliance and airway driving pressure at 8 cmH2O of PEEP. After a two-step PEEP procedure, we could show that increasing PEEP to 15 cmH2O in one of the patients, resulted in around 30 % decrease in transpulmonary driving pressure, while in the other, transpulmonary driving pressure increased with 30 % and plateau pressure reached a dangerous level, 23.2 cmH2O (Fig. 8).

Very early diagnostics
Preventive measures to avoid worsening of lung condition with consolidation of atelectasis should be taken early in the course of ventilator treatment [60-62], but little is known about very early lung mechanics conditions, as esophageal pressure measurements seldom are instigated directly after start of ventilator treatment. Most studies of PEEP strategies even include patients up to 72 hours after start of ventilator treatment. In contrast to esophageal pressure measurements, the PEEP step method provides easy assessment of lung mechanics and a rational basis for applying a protective PEEP and tidal volume strategy directly after the patient is connected to the ventilator.

Study limitations
In this analysis, we have only access to PEEP trials up to 16 cmH2O. As it is an inherent feature of the PEEP step method, that the transpulmonary inspiration pressure at the highest PEEP level always has to be estimated by extrapolation, it has not been possible to determine how precise estimated transpulmonry plateau pressure at 16 cmH2O of PEEP is.
Ideally, from a measurement precision point of view, PEEP steps up to 25 cmH2O should be used, as this is the upper limit for safe transpulmonary plateau pressure [44][45][46]. In most cases however, it is probably sufficient if the highest PEEP level of the measurement procedure is 16-18 cmH2O, because the plateau pressure at such PEEP levels will reach levels close to the upper safety limit.

Conclusions
There is no correlation between airway driving pressure and transpulmonary driving pressure in individual patients, indicating that airway driving pressure is less suitable for individualization of protective ventilation. However, applying PEEP causes separation of lung and chest wall mechanics, which makes it possible to identify the lung P/V curve from endexpiration at a low clinical PEEP, to end-inspiration at the highest PEEP level of a rapid PEEPstep procedure. The best-fit equation for the lung P/V curve can be used to determine the PEEP level where transpulmonary driving pressure is lowest and possibly least injurious for any given tidal volume. The measurement procedure is non-invasive and can easily be implemented from immediately after start of ventilator treatment until extubation. Thompson  Tidal airway (red arrows) and lung P/V curves (blue arrows) calculated from end-expiratory and endinspiratory esophageal and transpulmonary pressure data at each PEEP (ΔPAW/VT, (ΔPAW -ΔPES)/VT) level of the Yoshida et al study 2018 [27]. This makes it possible to determine transpulmonary and chest wall driving pressure and lung and chest wall elastance (EL, ECW) and respiratory system elastance (ERS) as the sum of tidal lung and chest wall elastance at each PEEP level. Also, ΔEELV between PEEP levels can be calculated as ΔPEEP/EL (ΔPEEP x CL). Left panel: Tidal airway P/V curves at PEEP 4 to 24 cmH2O with tidal volume of 200 ml from Yoshida et al [27] where ΔEELV was determined for each 2 cmH2O PEEP step as ΔPEEP/EL, and EL calculated as ΔPTP/VT. Right panel: Tidal airway and lung P/V curves are plotted in the same diagram, showing that the lung P/V curve coincides with the end-expiratory airway P/V curve, which con rms that lung elastance can be determined as ΔPEEP/ΔEELV (lung compliance, CL = ΔEELV/ΔPEEP).

Figure 2
Recording from an ARDS patient [33]. Upper panel: airway pressure. Lower panel: volume by electric impedance tomography (EIT). The registration is done during volume control ventilation with a tidal volume of 500 ml with an airway driving pressure (ΔPAW) of 10 cmH2O at baseline PEEP followed by an equal PEEP increase of 10 cmH2O. The red bars indicate the breath-by-breath difference in expiratory tidal volume (VTE) from baseline VTE, until VTE is again equal to baseline. Note that an airway driving pressure of 10 cmH2O is needed to in ate a tidal volume of 500 ml, while an increase in PEEP of 10 cmH2O causes an increase in EELV of 900 ml. Respiratory system compliance (VT/ΔPAW) is 500/10 = 50 ml/cmH2O. The end-expiratory respiratory system compliance, which is equal to lung compliance (ΔEELV/ΔPAWEE = ΔEELV/ΔPTPEE) is 900/10 = 90 ml/cmH2O (see further below).

Figure 3
As the end-expiratory transpulmonary pressure increases as much as PEEP is increased and the increase in end-expiratory lung volume following a PEEP increase is equal to ΔPEEP/EL, a lung P/V curve can be obtained by plotting the end-expiratory airway pressure vs cumulative end-expiratory lung volume. Red arrows: tidal airway P/V curves, Blue line: lung P/V curve. Note that the end-expiratory and inspiratory lung P/V curves and end-expiratory airway (respiratory system) P/V curve coincide. The green arrows mark the tidal pleural pressure variation, ΔPPL. The end-inspiratory transpulmonary pressure at the highest PEEP level is estimated by extrapolation of ΔPPL at PEEP 16 cmH2O from the ΔPPL at the four lower PEEP levels and estimate the end-inspiratory transpulmonary pressure at PEEP 16 cmH2O as endinspiratory airway pressure minus extrapolated ΔPPLPEEP16. The dashed blue line indicates the estimated tidal lung P/V curve at the highest PEEP level.

Figure 4
Lung P/V measured data points and best t curves (according to a third degree equation) in ten patients with lung compliance ranging from 43 to 143 ml/cmH2O. The correlation coe cient (r2) of the endexpiratory and end-inspiratory transpulmonary P/V points to the best-t curve is above 0.99. Note that in patients with almost equal compliance, patient 2 and 3, and patient 8 and 9, lung compliance increases at the highest PEEP level in patient 3 and 8, and decreases in patient 2 and 9. Thus, in patient 2 and 9 there is an unfavorable response to high PEEP and in patient 3 and 8 there is a positive response to high PEEP.

Figure 5
Lung P/V curves from end-expiration at ZEEP to end-inspiration at PEEP 16 cmH2O showing the diverse lung mechanics of patients with acute respiratory failure.

Figure 6
Airway and transpulmonary driving pressure during PEEP steps 0-4-8-12-16 cmH2O (also 20 cmH2O in patient 1). Note that the relation between ΔPTP and ΔPAW changes with changing PEEP and in the individual patient, airway driving pressure does not track transpulmonary pressure during PEEP titration Lung P/V curves by a two PEEP step up-down procedure (black line). The black lled circles indicate that PEEP levels of 8, 12 and 16 cmH2O were used together with the estimated end-inspiratory transpulmonary P/V point at the highest PEEP level for determining the best-t lung P/V curve. The endinspiratory transpulmonary pressure is, except for patient 1 and 2, who are PEEP non-responders, 18-20 cmH2O. As the corresponding lung volume also is determined by the PEEP-step measurement procedure, it is possible to calculate the over-all lung compliance, which normally should be around 100 ml/cmH2O and where a low compliance indicates a low inspiratory capacity and possibly the existence of a baby lung, which is highly likely in patient 1 and 2. The red arrow shows the tidal lung P/V curve with the lowest transpulmonary driving pressure for the tidal volume used clinically in the patients (6-7 ml/kg ideal body weight).

Figure 8
Implementation of a short two-PEEP step procedure in two patients with the same respiratory system compliance at a baseline clinical PEEP of 8 cmH2O (upper panels). PEEP is increased stepwise from 8 to 12 and 16 cmH2O and the lung P/V curve (blue line) is determined from end-expiration of PEEP 8 to endinspiration at PEEP 16 cmH2O (mid panels). As the tidal lung P/V curve lies on the lung P/V curve, transpulmonary plateau and driving pressure (PTPplat, ΔPTP) can be calculated for any combination of PEEP and tidal volume. In this case PTPplat and ΔPTP is determined for a tidal volume of 400 ml at PEEP 8 and 15 cmH2O, revealing that in patient 2 (right panels) this PEEP increase leads to a signi cant increase in ΔPTP and a high and possibly risky level of PTPplat. In patient 3 (left panels), the increase in PEEP from 8 to 15 cmH2O results in a decrease in ΔPTP from 6.5 to 4.4 cmH2O and a transpulmonary plateau pressure, which is well below VILI levels. Note, that in ation by PEEP occurs from ventral to dorsal region (grey-black, thick arrow), which causes tidal ventilation to be moved in dependent direction when PEEP is increased.

Figure 9
Schematic graph showing the development from healthy lung (left upper and lower panels) to pulmonary ARDS, PEEP non-responder (upper panels) and extra-pulmonary ARDS, PEEP responder (lower panels). In pulmonary ARDS, the gradient between abdominal pressure (AP) and superimposed pressure (SP) is small. Exterior lung volume and diaphragm position are almost normal. PEEP causes in ation of the nondependent baby lung, but not of consolidated dependent lung. In extra-pulmonary ARDS, the increased edema of the abdomen pushes the diaphragm in cranial direction and decreases FRC. Dependent lung is compressed and poorly aerated. When PEEP is applied, already open lung will be further in ated and the diaphragm pushed in caudal direction, increasing space for the lung to expand.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. GrivansStenqvistProtectivePEEPforCCESUPPL26042021.pdf