Interpretation and use of intraoperative protective ventilation parameters: a scoping review

Tidal volume

Because high V_T ventilation can prevent or minimize pulmonary atelectasis, in the past it was common practice to use a V_T of 800–1000 mL, which translated into a V_T of up to 10–15 mL kg^-1 [6]. Volumes that have been used in patients without ARDS under mechanical ventilation have been progressively decreasing [7]. In a 39-year period from 1975 to 2014, V_T decreased significantly in ICUs (annual decrease of 0.16 mL kg^-1) and in the operating theatre (annual decrease of 0.09 mL kg^-1) [8].

However, there is usually a misconception that the benefits of ventilating critically ill patients with low V_T of 6–8 mL kg^-1 are irrelevant in surgical patients, due to the relatively healthy lungs and shorter duration of ventilation in the latter [6]. Intraoperative ventilation at low V_T has been shown not only to decrease airway pressures and generation of pro- inflammatory cytokines, but also to improve patient-centred outcomes, such as need for reintubation, length of hospital and ICU stay, and postoperative pulmonary complications (PPCs) [9–12]. PPCs are increased by V_T in a “dose-dependent” manner and generate higher mortality in patients undergoing mechanical ventilation for general anaesthesia [13].

Volume selected in a protective way (6–8 mL kg^-1) can be distributed differently according to the aerated lung volume that a certain patient has to receive. This means that even the same set V_T can produce different lung stress (distending pressure applied to lung parenchyma) between patients with similar PBW [14]. The size of the aerated lung can vary considerably in patients with consolidation, fluid, pulmonary exudate, or atelectasis [15]. It is precisely these patients that are at risk of overdistention injury even when ventilating with low V_T, because they have reduced aerated lung volumes [15]. In patients with low lung compliance, such as ARDS patients, or in the operating theatre those with pulmonary atelectasis, even V_T below 6 mL kg^-1 of PBW can result in high strain (lung deformation from its resting position) [16].

In this sense, 3 randomized clinical trials have demonstrated a decrease in postoperative pulmonary complications in protectively ventilated surgical patients with V_T of 6–8 mL kg^-1 together with positive end-expiratory pressure (PEEP) levels between 6 and 12 cm H₂O, a ventilatory strategy that was related to lower airway distending pressure (airway driving pressure or ΔP_aw) and greater lung compliance [12, 17, 18]. On the other hand, the use of strategies with low V_T is associated with hypercapnia and acidosis, probably due to less efficient alveolar ventilation [7]. These patients could benefit from strategies such as reducing instrumental dead space, increasing respiratory rate, lung recruitment manoeuvres associated with selection of an optimal PEEP, and prone position [15].

Therefore, titration of V_T according to PBW could be a starting point, but it does not ensure the absence of damage. A possible solution is titrating V_T according to ΔP_aw [5, 14, 16] or to transpulmonary pressure (P_L) [15]. These parameters could be better surrogates for limiting ventilator-induced lung injury (VILI) due to overdistention [15].

PEEP

PEEP is the positive pressure at the end of expiration. Low PEEP levels (5 cm H₂O) are routinely used in patients undergoing mechanical ventilation. This practice aims to keep the lung open at the end of expiration, prevent small airway and alveoli closure, minimize the damaging effects of cyclic opening and collapsing of the alveoli, and increase the lymphatic flow, facilitating drainage of pulmonary oedema [16, 26]. Furthermore, PEEP would facilitate lung homogenization by recruiting previously non-ventilated alveolar units, thus avoiding excessive tension in the margins between aerated and non-aerated regions of the lung parenchyma (stress raisers) [1, 4, 27]. Both mean airway pressure (MAP) and the plateau pressure (P_{aw plateau}) will increase with increasing PEEP, but this increase may be accentuated to a greater or lesser extent depending on the compliance of the respiratory system (if the latter is low, the increase in pressures will be higher, and vice versa). In contrast, airway driving pressure is not affected by changes in PEEP levels unless they result in changes in lung compliance. Increases in PEEP will only be protective when they result in an increase in lung compliance, which will be reflected in a decrease in ΔP_aw [7, 28]. PEEP, therefore, should be selected individually according to the patient’s requirements [29].

PEEP = PEEP_volume × E_rs,

where PEEP is the positive pressure at the end of expiration, PEEP_volume is the volume generated by PEEP, which is calculated as the difference between EELV (end expiratory lung volume) and FRC (functional residual capacity), and E_rs is the respiratory system elastance (Figure 1 and Table 1).

FIGURE 1

Determinants of protective ventilation parameters. Where P_{aw peak} is airway peak pressure, delta peak plateau is the difference between the peak and the plateau pressure, P_{aw plateau} is airway plateau pressure, V_T is tidal volume, PEEP_volume is volume generated by PEEP, E_rs is respiratory system elastance, E_L is lung elastance, E_cw is chest wall elastance, P_{aw PEEP} is positive end expiratory pressure or PEEP, ΔP_aw is airway driving pressure or respiratory system driving pressure (ΔPrs), P_{L PEEP} is transpulmonary pressure at end of expiration, ΔP_L is transpulmonary driving pressure or lung driving pressure, ΔP_es is oesophageal driving pressure or driving pressure across the chest wall (ΔP_cw) or the change in pleural pressure between the end of inspiration and the end of expiration (ΔP_pl), P_{L plateau} is transpulmonary pressure at the end of inspiration, P_{es plateau} is oesophageal pressure at the end of inspiration, and P_{es PEEP} is oesophageal pressure at the end of expiration

P_{aw PEEP} – airway PEEP pressure, PEEP_volume – volume generated by PEEP, EELV – end expiratory lung volume, FRC – functional residual capacity, E_rs – respiratory system elastance, P_{aw peak} – airway peak pressure, P_{aw plateau} – airway plateau pressure, V_T – tidal volume, E_rs – respiratory system elastance, ΔP_aw – airway driving pressure or respiratory system driving pressure (ΔP_rs), P_{L plateau} – transpulmonary pressure at the end of inspiration, E_L – lung elastance, ΔP_L – transpulmonary driving pressure or lung driving pressure

Airway peak pressure

Peak airway pressure (P_{aw peak}) is the maximum pressure measured in the airway at the end of inspiration. Unlike P_{aw plateau}, P_{aw peak} is influenced by both the elastic and resistive properties of the respiratory system. At the bedside, the difference between peak and plateau pressures can be easily individualized during an inspiratory pause during volume-controlled ventilation. Immediately after the inspiratory pause, a rapid pressure drop is observed, which represents the pressure dissipated to overcome airway resistance. To calculate the airway resistance, this pressure difference is divided by the inspiratory flow. In normal subjects, airway resistance values do not exceed 15 cm H₂O L^-1 s^-1 under controlled mechanical ventilation [16].

During mechanical ventilation, P_{aw peak} depends on P_{aw plateau}, and on the pressure difference between these 2 pressures (P_{aw peak} – P_{aw plateau}), which is determined by both the inspiratory flow and airway resistance.

P_{aw peak} = (P_{aw peak} – P_{aw plateau}) + P_{aw plateau}

P_{aw peak} = ΔV × R_aw + (V_T + PEEP_volume) × E_rs,

where P_{aw peak} is airway peak pressure, P_{aw plateau} is airway plateau pressure, ΔV represents inspiratory flow, R_aw is airway resistance, PEEP_volume is volume generated by PEEP, and E_rs is respiratory system elastance (Figure 1 and Table 1).

In patients without increases in airway resistance and/or obstructions in the ventilatory circuit, the maximum inspiratory pressure (MIP) or peak inspiratory pressure (PIP) is approximately equal to the plateau pressure [7]. During pressure-controlled ventilation, the set ventilator pressure (MIP) is comparable to the P_{aw plateau} during volume-controlled ventilation if inspiratory flow seen on the flow-time curve reaches zero, a situation in which MIP will constitute a surrogate of alveolar pressure.

Airway plateau pressure

Plateau pressure, or airway plateau pressure (P_{aw plateau}), is the pressure measured in static conditions during an inspiratory pause in volume- controlled ventilation, and that results from the sum of pressures associated with tidal volume and volume generated by PEEP, if it is present. Because it is estimated during zero flow conditions, it avoids taking into account the pressures required to exceed airway resistance [3]. In other words, the P_{aw plateau} is not affected by changes in inspiratory flow or airway resistance [16] and only reflects the elastic properties of the respiratory system [32].

The P_{aw plateau} is thus determined by changes in V_T, E_rs, and PEEP levels [31].

The importance of plateau pressure lies in the fact that it is considered a surrogate, a reflection of alveolar pressure (P_alv), which is the real pressure that distends the entire respiratory system (lung plus chest wall) [33]. In this sense, it also involves the pressure required to surpass the chest wall elastance, and therefore this “alveolar pressure” does not reliably reflect the pressure load to which only the lung is exposed. Pulmonary distending pressure in isolation is evaluated with transpulmonary pressure, which represents the elastic recoil pressure of the lungs [29].

The P_{aw plateau} thus represents the best clinical way to assess airway pressure under static conditions, and therefore it can be used to estimate transpulmonary pressure [3].

P_{aw plateau} = PEEP + ΔP_aw

(V_T + PEEP_volume) × E_rs = (PEEP_volume × E_rs) + (V_T × E_rs)

Where P_{aw plateau} is airway plateau pressure, PEEP is positive pressure measured at the end of expiration, ΔP_aw is the airway driving pressure, and E_rs is the respiratory system elastance (Figure 1 and Table 1).

The recommendation of the ARDS network protocol to limit the P_{aw plateau} below 30 cm H₂O to improve survival derives from the evidence obtained during spontaneous ventilation, in which total lung capacity is achieved with a transpulmonary pressure of 25 cm H₂O. If a patient has normal chest wall elastance, it corresponds to a P_{aw plateau} of 30 cm H₂O, a pressure level below which was shown to generate minimal inflation and therefore the absence of VILI in an animal study [34] (Figure 1).

Limitations

During a patient’s mechanical ventilation, a high P_{aw plateau} may not be injurious per se, when instead of being applied to inflate the alveoli, it is mostly used to inflate a high elastance chest wall [29].

In 1988 Dreyfus et al. [35] demonstrated that injurious pulmonary oedema occurred in healthy paralyzed animals during pressure-controlled ventilation with high V_T and high airway pressures, but it did not occur in those ventilated with similar airway pressures and low V_T due to straps applied around their abdomen and chest. These straps were a simple way to increase chest wall elastance, and therefore, for the same airway pressure or plateau pressure, pulmonary distending pressure or transpulmonary pressure was lower and did not cause injury [29]. These experiments demonstrated that the volume that caused lung stretching, and not the airway pressure, was the most important factor in determining injury, a finding that led them to employ the term ‘volutrauma’. We currently interpret these findings as an indirect demonstration of the importance of transpulmonary pressure in determining ‘lung trauma’ and injury, which in fact does not occur if this pressure is kept within certain limits, no matter how high the plateau pressure is [29].

In this way, ventilating patients with high chest wall elastance and limiting the plateau pressure to 30 cm H₂O, far from inducing damage by baro- or volutrauma, this can lead to atelectasis with shunt and hypoxaemia, in addition to lung damage due to increased stress raisers (e.g. in obese patients, pregnant women, patients with pleural effusion, neuromuscular diseases, or capnoperitoneum) [3]. On the one hand, high chest wall elastance is effectively protective for the lungs, but it entails an increase in pleural pressure, which will have haemodynamic repercussions through an increase in pressure in the right atrium, which will then decrease venous return and cardiac output [3]. On the other hand, ventilating patients with the same P_{aw plateau} of 30 cm H₂O but with a low chest wall elastance, will generate a high transpulmonary pressure, which is the true pulmonary distending pressure and the one that generates lung damage at high values [3] (Figure 2).

FIGURE 2

Representation of airway and pleural (oesophageal) pressures in 2 patients with different chest wall and lung elastance during end inspiration and end expiration. Two patients with different mechanical properties of the lung and chest wall are presented; both ventilated with the same high plateau pressure (P_{aw plateau}) and high airway driving pressure (ΔP_aw) (represented in yellow). In A, the patient has high chest wall elastance, and therefore presents a low transpulmonary pressure (P_{L plateau}) and a low transpulmonary driving pressure (ΔP_L) (represented in green), thus with a low risk of lung damage. At the same time, it presents high pleural pressures (e.g. ΔP_es) (represented in red), which probably implies some haemodynamic repercussion. Patient B with high pulmonary elastance has high transpulmonary pressure (P_{L plateau}) and high transpulmonary driving pressure (ΔP_L) with the consequent risk of lung damage (represented in red) but with low pleural pressures therefore (e.g. ΔP_es) (represented in green), without haemodynamic repercussions. Actually, the risks of mechanical ventilation in patients ventilated with similarly high airway pressures range from the risk of lung damage if the lung elastance is high and the chest wall elastance is low, and haemodynamic instability in the opposite situation [3, 31]. Where P_{aw plateau} is airway plateau pressure, PEEP is positive end expiratory pressure, ΔP_aw is airway driving pressure, P_{L PEEP} is transpulmonary pressure at end of expiration, ΔP_L is transpulmonary driving pressure o lung driving pressure, ΔP_es is oesophageal driving pressure or driving pressure across the chest wall (ΔP_cw) or the change in pleural pressure between the end of inspiration and the end of expiration (ΔP_pl), PL plateau is transpulmonary pressure at the end of inspiration, P_{es plateau} is oesophageal pressure at the end of inspiration, and P_es PEEP is oesophageal pressure at the end of expiration. All pressures are expressed in cm H₂O

Therefore, we consider that using a single P_{aw plateau} level to define potential harm could be overly simplistic, and it could lead to the administration of insufficient pressure to some patients and excessive pressure to others [36]. Given that what generates VILI is not the pressure applied to the airway but rather the pressure applied to the lungs (transpulmonary pressure), it follows that a better limit or safe pressure threshold could be a transpulmonary pressure level rather than a certain P_{aw plateau}level [36].

For this reason, there are authors who propose monitoring oesophageal pressure in patients with risk factors for chest wall stiffness. In these cases of high chest wall elastance, a safe threshold of 30 cm H₂O for plateau pressure may actually be higher and still be safe [1].

Application in clinical practice

It is suggested that a P_{aw plateau} below 25–28 cm H₂O should be used in mechanically ventilated patients without ARDS during both volume-controlled and pressure-controlled ventilation [7].

Its use as the only safety parameter without taking into account other features of patient mechanics could lead to an insufficient ventilation (in patients with high chest wall elastance such as in obesity) or an excessive one (in patients with high lung elastance such as in ARDS). Therefore, it should be used considering other patient factors, such as the ones that increase or decrease chest wall elastance. We can mention 3 scenarios:

If the P_{aw plateau} is less than 25 cm H₂O and airway driving pressure is less than 13 cm H₂O, then ventilatory parameters could be maintained because we would be ventilating within limits that are considered safe.

If the P_{aw plateau} is greater than 25 cm H₂O, considering its determinants, it should be ruled out that this is not generated by high PEEP levels.

If the P_{aw plateau} is greater than 25 cm H₂O and the airway driving pressure is greater than 13 cm H₂O, considering PEEP levels within certain limits, causes of increased respiratory system elastance should be assessed (Figure 3).

FIGURE 3

Algorithm for programming protective ventilation parameters: V_T = 6–8 mL kg_-1 PBW, RR targeting an EtCO₂ of 35 to 45 mmHg, PEEP ≥ 5 cm H₂O, I : E ratio (the one that allows expiratory flow to reach zero, and the longest inspiratory time e.g. 1 : 1.5), if high peak airway pressure, check high Delta peak plateau and rule out kinked orotracheal tube, secretions, mucus, etc.). Where VCV is volume control ventilation, VT is tidal volume, RR is respiratory rate, PEEP is positive end expiratory pressure, I : E ratio is the ratio between inspiratory time : expiratory time, P_{aw plateau} is airway plateau pressure, and ΔP_aw is airway driving pressure. *The decision to recruit the lung should also be based on other parameters such as the air test, the increased capnographic gap (PaCO₂ – End tidal CO₂) and/or decreased PAFI. All pressures are expressed in cm H₂O

Airway driving pressure

The airway driving pressure (ΔP_aw) or (ΔP_rs) (respiratory system driving pressure) is the pressure required during lung expansion to overcome the elastic forces generated by the lung parenchyma, pleura, and chest wall [28]. Unlike the plateau pressure it represents the pressure above PEEP required to overcome the elastic recoil of the respiratory system as a whole generated by the delivery of tidal volume [27].

Another way of interpreting the ΔP_aw is the oscillation between the end of inspiration and the end of expiration of the respiratory system elastic pressure [37]. Thus, assuming that there is no auto-PEEP, ΔP_aw estimates the increase in alveolar pressure during inspiration and its decrease during expiration representing the pressure change during tidal volume ventilation [15, 29].

It is important to recognize that ΔP_aw involves 2 components, one related to the expansion of the lung and the other related to the expansion of the chest wall [38].

ΔP_aw = ΔP_L + ΔP_pl,

where ΔP_L is the transpulmonary driving pressure and ΔP_pl is the change in pleural pressure or driving pressure across the chest wall (ΔPCW) or oesophageal driving pressure (ΔP_es) (Figure 1 and Table 1).

This implies that ΔP_aw can be elevated by both elastic components or at the expense of one of them. Therefore, increases in chest wall elastance in isolation can affect this parameter without increasing pulmonary stress, globally represented by transpulmonary driving pressure [5].

Chest wall elastance may be increased in conditions such as obesity, abdominal compartment syndrome, kyphoscoliosis, chest wall burns, or increased muscle tone due to dyssynchrony with the ventilator. Conversely, it could be decreased by muscle paralysis [39].

ΔP_aw is easily calculated at the bedside as the difference between the plateau pressure and PEEP.

ΔP_aw = P_{aw plateau} – PEEP

It is equivalent to the relationship between V_T and respiratory system compliance (C_rs) [14]. So, assuming C_rs reflects the amount of aerated lung volume, ΔP_aw is considered a surrogate of dynamic or cyclic lung strain (or stretch). The latter is defined as the ratio between V_T and FRC [32]. Therefore, the higher the FRC, the higher the C_rs (due to greater aerated lung volume) and the lower the ΔP_aw. In this way ΔP_aw, unlike P_{aw plateau} and PEEP (which are static estimates of stress in the respiratory system), constitutes a dynamic indicator because it represents the cyclic strain that lung parenchyma (and the chest wall) are exposed to during each ventilatory cycle [2, 15].

ΔP_aw = V_T/C_rs

ΔP_aw = V_T × E_rs,

where V_T is the tidal volume, C_rs is the respiratory system compliance, and E_rs is the respiratory system elastance. ΔP_aw conveys the relationship between the V_T applied above a certain PEEP and the respiratory system compliance.

Using a statistical model known as multilevel mediation analysis, information from 3562 patients with ARDS was analysed from 9 randomized controlled trials, which demonstrated that ΔP_aw is the ventilatory parameter that best predicts survival at 60 days in ARDS [37, 40] and not V_T or PEEP [2].

Amato et al. [40] suggested that the ventilatory impact of V_T on lung injury could be better predicted if it was normalized to C_rs rather than PBW. ΔP_aw was shown to be the final mediator on the effects of lowering V_T and P_{aw plateau} on mortality in patients with ARDS [29].

A recent meta-analysis involving 17 studies and more than 2250 patients showed that changes in PEEP levels that resulted in an increase in ΔP_aw (i.e. without modifying or even decreasing C_rs) were associated with greater postoperative pulmonary complications [2, 32].

In patients with low lung compliance (e.g. with atelectasis, consolidation, oedema), even setting low V_T (< 6 mL kg^-1 PBW) could result in high ΔP_aw, i.e. in high strain (V_T/CRF) [16].

For this reason, it is suggested that V_T should be adjusted considering ΔP_aw (or V_T/C_rs). Because C_rs is directly related to the aerated lung volume, ΔP_aw reflects the level of V_T related to the aerated lung volume [1]. Using ΔP_aw as a safety limit, it could be a better way to adjust V_T to decrease dynamic or cyclic strain during mechanical ventilation [2, 5].

Most surgical patients (without ARDS) will have ΔP_aw < 10 cm H₂O, reflecting a normal or close to normal C_rs. In contrast, in patients with moderate to severe ARDS or other restrictive diseases (e.g. major pulmonary effusions, interstitial diseases, etc.), it will be more common to find an ΔP_aw > 10 cm H₂O, reflecting either a decreased C_rs or an inappropriate selection of V_T or PEEP [2].

Even more, V_T could be higher than 6 mL kg^-1 and ΔP_aw could remain below 14 cm H₂O [15]. Limiting ΔP_aw, possibly keeping it below 14 cm H₂O, can be achieved either by decreasing V_T or by increasing C_rs [37]. If the former remains constant and the latter increases, ΔP_aw will decrease proportionally. Furthermore, under steady conditions in C_rs, ΔP_aw will increase if V_T increases [7].

One way to increase compliance is through lung recruitment with recruitment manoeuvres and a higher PEEP level or through prone positioning [1, 21, 23].

In this context, ΔP_aw could also be a valuable tool for setting PEEP. Regardless of the strategy used to titrate PEEP, changes on its levels should be considered in the impact on ΔP_aw, in addition to other variables such as gas exchange and haemodynamics [2]. A decrease in ΔP_aw after increasing PEEP ne-cessarily reflects lung recruitment and a decrease to their cyclic strain. Conversely, an increase in ΔP_aw suggests a non-recruitable lung in which overdistention prevails over recruitment [7].

If, after optimizing PEEP, ΔP_aw remains > 15 cm H₂O, it is suggested that V_T should be lowered and causes of increased chest wall elastance be considered. In these cases, the placement of an oesophageal catheter to measure transpulmonary driving pressure (ΔP_L) could be a useful strategy [2], although this strategy is probably not a practical recommendation and is not applicable in the field of anaesthesiology.

It is suggested that a target of ΔP_aw < 13–15 cm H₂O be used [41].

Limitations

A problem with this parameter to help in the guidance of V_T selection is that one of its components is the plateau pressure, which, as we mentioned above, is influenced by the chest wall elastance. Thus, 2 patients with the same ΔP_aw may have different risk of VILI. A high ΔP_aw in the context of high chest wall elastance probably implies that this is due to the chest wall component and that the lung is not being subjected to high and injurious pressures [2, 15]. In this scenario, ΔP_aw may be high at the expense of ΔP_pl, with transpulmonary driving pressure being at normal or low values.

↑ ΔP_aw = ΔP_L + ↑ ΔP_pl

↑ ΔP_aw = V_T x E_L + V_T × ↑ E_CW,

where ΔP_aw is airway driving pressure, ΔP_L is transpulmonary driving pressure, ΔP_pl is oesophageal driving pressure, V_T is tidal volume, E_L is lung elastance, and E_CW is chest wall elastance. Oeso-phageal driving pressure is elevated due to increased chest wall elastance (Figures 1 and 4).

FIGURE 4

Determinants of protective ventilation parameters. Where P_{aw plateau} is airway plateau pressure, V_T is tidal volume, PEEP_volume is volume generated by PEEP, E_rs is respiratory system elastance, E_L is lung elastance, E_cw is chest wall elastance, P_{L plateau} is transpulmonary pressure at the end of inspiration, P_{es plateau} is oesophageal pressure at the end of inspiration, P_{L PEEP} is transpulmonary pressure at end of expiration, ΔP_L is transpulmonary driving pressure or lung driving pressure, ΔP_es is oesophageal driving pressure or driving pressure across the chest wall (ΔP_cw) or the change in pleural pressure between the end of inspiration and the end of expiration (ΔP_pl), P_{es PEEP} is oesophageal pressure at the end of expiration, and ΔP_aw is airway driving pressure or respiratory system driving pressure (ΔP_rs) [31]

Thus, a high value of airway driving pressure may overestimate the risk of lung injury [42].

Therefore, in these cases of increased chest wall elastance, a better indicator of the dynamic stress lungs are being subjected to in isolation is transpulmonary pressure or transpulmonary driving pressure. These could be measured, limiting V_T to keep it within a safe range [14–16]. However, as we mentioned, it is a difficult measurement to implement in anaesthesiology due to the cumbersome and impractical nature of placing an oesophageal catheter and measuring pleural pressures intraoperatively.

Finally, the effects of PEEP and its continuous strain are not considered in the ΔP_aw approach. For example, a theoretically safe level of 12 cm H₂O of ΔP_aw could become harmful if PEEP is 20 instead of 0 cm H₂O [36]. This implies that an ΔP_aw still in the safe range could be injurious if PEEP levels are high, generating overdistention per se.

This is because VILI related to excessive strain can be linked to its static and dynamic components. Static strain results from application of PEEP and its resultant deformation in the lungs above their functional residual capacity. Dynamic strain results from cyclic insufflation of the aerated lung with each ventilation and is therefore linked to ΔP_aw [15] (Table 1).

Transpulmonary pressure

Transpulmonary pressure (P_L) is the pressure that distends the lungs in isolation. It constitutes the pressure generated in the lungs regardless of the effects on the chest wall and abdomen [29]. It is represented by the pressure difference between the inside of the lung (airway pressure) and the pressure on the surface of the lung (pleural pressure or its surrogate oesophageal pressure) [3]. P_L is the relevant pressure when we consider the stress applied to lung tissue in each insufflation, and because a certain stress is associated with a certain strain, P_L is strictly related to VILI generation [3, 33, 38].

Measurement of P_L can be done at the end of inspiration or at the end of expiration [32].

When measured at the end of inspiration or at plateau pressure (P_{L plateau}), it is relevant when it comes to preventing damage from lung hyperinflation [3, 38]:

P_{L plateau} = P_{aw plateau} – P_{es plateau},

where P_{L plateau} is transpulmonary pressure at the end of inspiration, P_{aw plateau} is airway plateau pressure, and P_{es plateau} is oesophageal pressure at the end of inspiration (Figures 4 and 5).

FIGURE 5

Representation of transpulmonary pressures. Airway pressure waveforms and oesophageal pressure waveforms. The double arrow represents the difference between the airway driving pressure (ΔP_aw) and the oesophageal driving pressure (ΔP_es), which is the transpulmonary driving pressure (ΔP_L). Paw plateau is airway plateau pressure, P_{aw PEEP} is positive end expiratory pressure or PEEP, ΔP_aw is airway driving pressure, ΔP_L is transpulmonary driving pressure or lung driving pressure, ΔP_es is oesophageal driving pressure or driving pressure across the chest wall (ΔP_cw) or the change in pleural pressure between the end of inspiration and the end of expiration (ΔP_pl), P_es plateau is oesophageal pressure at the end of inspiration, and P_{es PEEP} is oesophageal pressure at the end of expiration

P_L measured at the end of expiration, or transpulmonary pressure associated with PEEP (P_{L PEEP}) is relevant to preventing lung collapse [3, 38]:

P_{L PEEP} = PEEP – P_{es PEEP},

where P_{L PEEP} is transpulmonary pressure at the end of expiration, PEEP is positive pressure at the end of expiration, and P_{es PEEP} is oesophageal pressure at the end of expiration (Figure 4).

The P_L measured at the end of inspiration (P_{L plateau}), represents the total stress given by the cyclic stress to which lungs are exposed during each ventilation (transpulmonary driving pressure or ΔP_L) and the addition of the static stress measured by P_L at the end of expiration (P_{L PEEP}) [43]:

P_{L plateau} = ΔP_L + P_{L PEEP}

(V_T + PEEP_volume) × E_L = (V_T × E_L) + (PEEP_volume × E_L),

where P_{L plateau} is transpulmonary pressure at end inspiration, ΔP_L is transpulmonary driving pressure, P_{L PEEP} is transpulmonary pressure at end expiration, V_T is tidal volume, E_L is lung elastance, and PEEP_volume is the volume generated by PEEP (Figure 4).

For a given P_{aw plateau}, the P_{L plateau} depends on the relationship between lung elastance (E_L) and that of the respiratory system (E_rs), the latter being the sum of E_L plus E_CW [3, 44].

P_L = P_{aw plateau} × (E_L/E_rs),

where, P_L is transpulmonary pressure at the end of inspiration, P_{aw plateau} is airway plateau pressure, E_L is lung elastance, and E_rs is respiratory system elastance.

If we consider, for example, a patient with an elastance ratio of 0.8 (E_L/E_rs = 0.8) (e.g. ARDS) and an P_{aw plateau} of 30 cm H₂O, their P_L will be 24 cm H₂O. This is considered a pulmonary distending pressure level that corresponds to a dangerous lung volume close to total lung capacity. In contrast, for another patient with an elastance ratio as low as 0.2 (e.g. obesity or pregnancy), the same P_{aw plateau} of 30 cm H₂O will correspond to a P_L of only 6 cm H₂O, which may be associated with lung collapse and hypo-xaemia [3, 36].

In this way, it follows that for the same V_T and P_{aw plateau}, depending on the relationship between E_L and E_rs, it could result in completely different P_L and, consequently, different risks of VILI [4, 45].

Measurement of pressure that distends lungs alone may be a better approach to guiding ventilatory management, especially in patients with increased chest wall elastance in whom the P_{aw plateau} and airway driving pressure are elevated, but without this necessarily implying a risk of lung damage [4]. For a given V_T, a stiffer chest wall will lead to the development of higher pleural pressures because more of the airway driving pressure will be displacing the chest wall. In these cases, for a given P_{aw plateau}, transpulmonary pressure will become lower as chest wall elastance increases (e.g. abdominal hypertension or severe obesity) and vice versa [46] (Figure 2).

Although direct measurement of pleural pressure through pleural catheters is theoretically and practically possible, the most practical way to estimate pleural pressure in clinical practice is oesophageal manometry [3].

To better characterize ventilatory mechanics, P_L must be estimated under static conditions, i.e. when flow in the system is zero. This is mandatory to prevent having to take into account the pressure necessary to exceed airway resistance. The pressure measured under static conditions, during an inspiratory pause, is the plateau pressure, already referred to in this article [3].

Assessment of P_L could be useful for several reasons. On the one hand, for differentiating patients who could benefit from higher airway pressures due to their increased chest wall elastance from those who, despite having low airway pressure levels, are still under risk of overdistention [4]. Another bene-fit of its use would be for determining the pressure required to keep lungs open. Because oesophageal pressure is a surrogate of pleural pressure, a negative transpulmonary pressure at the end of expiration would correspond to collapsed lungs [3]. It has therefore been proposed for choosing PEEP levels, aiming to obtain a positive value of transpulmonary pressure at the end of expiration, and thus avoiding atelectasis [29].

Limiting P_L to less than 15–20 cm H₂O appears to be a physiologically reasonable approach for mitigating VILI [30, 32, 38].

Transpulmonary driving pressure

Transpulmonary driving pressure (ΔP_L) or lung driving pressure, represents the change or oscillation in transpulmonary pressure generated by V_T between the end of inspiration and the end of expiration [29]. It is calculated from the transpulmonary pressure difference between the P_L plateau and P_{L PEEP}.

ΔP_L = P_{L plateau} – P_{L PEEP}

(V_T × E_L) = (V_T + PEEP_volume) × E_L – (PEEP_volume × E_L),

where ΔP_L is transpulmonary driving pressure, P_{L plateau} is transpulmonary pressure at the end of inspiration, P_{L PEEP} is transpulmonary pressure at the end of expiration, V_T is tidal volume, E_L is lung elastance, and PEEP_volume is volume generated by PEEP (Figure 4).

In other words, the transpulmonary driving pressure (ΔP_L) would be equivalent to the transpulmonary pressure at end inspiration (P_{L plateau}) without the distending pressure generated by PEEP (P_{L PEEP}).

Another method of analysing and interpreting transpulmonary driving pressure is through ΔP_aw. The airway driving pressure is composed of 2 pressures: the one applied in the lung (ΔP_L) and the one to the chest wall (ΔP_pl) [38]:

ΔP_aw = ΔP_L + ΔP_pl,

where ΔP_aw is airway driving pressure, ΔP_L is transpulmonary driving pressure, and ΔP_pl is change in pleural pressure or driving pressure across the chest wall (ΔP_CW) [5] or oesophageal driving pressure (ΔP_es) [2] (Figures 4 and 5).

The oesophageal driving pressure (ΔP_es) is the change in oesophageal pressure between inspiration and expiration and represents the pressure applied to the chest wall above PEEP due to V_T delivery.

ΔP_es = P_{es plateau} – P_{es PEEP} [3],

where ΔP_es is oesophageal driving pressure, P_{es plateau} is oesophageal pressure at the end of inspiration, and P_{es PEEP} is oesophageal pressure at the end of expiration (Figure 4).

Terminology clarification: it is possible to find in the literature the change in pleural pressure expressed as ΔP_pl or ΔP_es or ΔP_CW called oesophageal driving pressure [2], which represents the distending pressure of the chest wall in isolation, which is elevated in situations of high chest wall elastance.

Therefore, if we solve for the transpulmonary driving pressure from the previous equation:

The transpulmonary driving pressure (ΔP_L) is the same as the airway driving pressure (ΔP_aw) without the distending pressure generated by the chest wall or oesophageal driving pressure (ΔP_es) [52].

ΔP_L = ΔP_aw – ΔP_es

ΔP_L = (P_{aw plateau} – PEEP) – (P_{es plateau} – P_{es PEEP})

V_T × E_L = (V_T × E_rs) – (V_T × E_cw),

where, ΔP_L is transpulmonary driving pressure, ΔP_aw is airway driving pressure, ΔP_es is oesophageal driving pressure, P_{aw plateau} is airway plateau pressure, PEEP is positive pressure at end of expiration, P_{es plateau} is oesophageal pressure at the end of inspiration, P_{es PEEP} is oesophageal pressure at the end of expiration, V_T is tidal volume, E_L is lung elastance, E_rs is respiratory system elastance, and E_cw is chest wall elastance (Figures 4 and 5).

Therefore, ΔP_L only represents the pulmonary distending pressure generated by delivery of tidal volume without taking into account the pressure component required to mobilize the chest wall (ΔP_es) [29]. Both the properties of the chest wall and the abdomen influence ΔP_aw measurement. This influence can be misleading because an increase in chest wall elastance does not reflect an increase in the risk of lung injury and will still lead to an increase in ΔP_aw [38].

Because differences between ΔP_L and ΔP_aw are mainly due to increases in chest wall elastance, the latter can vary with respect to the former between minimal (e.g. lean patients, pneumonia) and wider (e.g. morbid obesity, abdominal hypertension) differences [2, 14]. For this reason, in these cases of high chest wall elastance, it is advisable to assess ΔP_L through an oesophageal catheter to appropriately quantify stress applied to the lungs. However, in conditions where chest wall elastance is normal and stable, changes in ΔP_aw will provide an appropriate surrogate for changes in ΔP_L and lung strain [38] (Figure 4).

The 14 cm H₂O limit of ΔP_aw suggested as a safety parameter for ventilation will depend on the elastance ratio (E_L/E_rs). ΔP_L depends, for a given ΔP_aw, on the relationship between lung elastance and respiratory system elastance:

ΔP_L = ΔP_aw × (E_L/E_rs),

where ΔP_L is the transpulmonary driving pressure and ΔP_aw is the airway driving pressure.

Taking as an example a ΔP_aw value of 14 cm H₂O, this value can range from a ΔP_L of 2.8 cm H₂O (with an elastance ratio of 0.2) to 11.2 cm H₂O (with an elastance ratio of 0.8) [3]. Therefore, the same ΔP_aw with a limit value of 14 cm H₂O but with a ΔP_L of 2.8 cm H₂O would have a lower risk of lung damage than another patient with the same ΔP_aw and a ΔP_L of 11.2 cm H₂O (value close to the limit of 10–12 cm H₂O).

ΔP_L has several benefits. First, unlike P_L but like ΔP_aw, it removes the stress caused by PEEP levels, which does not necessarily contribute to lung injury and can sometimes even mitigate it [16, 29]. Second, like P_L but unlike ΔP_aw, it removes distending pressure from the chest wall [29]. Third, it is conceived that ΔP_L would better reflect the presence of regional inhomogeneities in mechanical properties of the lungs (i.e. overdistention and atelectasis). An increase in ΔP_L would be a better parameter to estimate tissue tension than other measurements of the respiratory system. Therefore, this parameter would represent a better surrogate of pulmonary stress and even a better predictor of clinical results than P_L [16].

In the clinical setting, upper limits of transpulmonary driving pressure would be 10–12 cm H₂O [37] (Table 1).