State of the art: Monitoring of the respiratory system during veno-venous extracorporeal membrane oxygenation

Abstract

Monitoring the patient receiving veno-venous extracorporeal membrane oxygenation (VV ECMO) is challenging due to the complex physiological interplay between native and membrane lung. Understanding these interactions is essential to understand the utility and limitations of different approaches to respiratory monitoring during ECMO. We present a summary of the underlying physiology of native and membrane lung gas exchange and describe different tools for titrating and monitoring gas exchange during ECMO. However, the most important role of VV ECMO in severe respiratory failure is as a means of avoiding further ergotrauma. Although optimal respiratory management during ECMO has not been defined, over the last decade there have been advances in multimodal respiratory assessment which have the potential to guide care. We describe a combination of imaging, ventilator-derived or invasive lung mechanic assessments as a means to individualise management during ECMO.

Keywords

extracorporeal membrane oxygenation veno-venous extracorporeal membrane oxygenation acute respiratory distress syndrome respiratory physiology respiratory monitoring critical care medicine ventilator associated lung injury patient self-inflicted lung injury

Introduction

The use of veno-venous extracorporeal membrane oxygenation (VV ECMO) has increased exponentially over the last decades¹ and is now recommended in selected patients by international guidelines.² Although landmark randomised controlled trials support a reduction in mortality with ECMO in a group of patients with acute severe and potentially reversible respiratory failure,^3–5 respiratory management, patient monitoring, and outcomes all vary between centres¹ and optimal management of the native lung is not yet defined. Monitoring enables underlying physiological principles to be used to develop strategies to individualise care.

The profound physiological changes induced by extracorporeal support affect the result and interpretation of many monitoring techniques. Furthermore, understanding the interplay between the membrane and native lungs is essential to personalising native lung management during ECMO. Over the last decade, there has been significant progress in both invasive and non-invasive tools to assess and personalise care in severe respiratory failure. In this review, we will outline current concepts in monitoring of the respiratory system during VV ECMO, including a practical, multimodal approach to individualising respiratory support to enhance recovery.

Gas exchange of the native and membrane lung

O₂ consumption and CO₂ clearance requirements

Understanding the overall gas exchange requirements of the body are essential to optimise respiratory support. Total oxygen consumption (VO₂tot) at rest in health is around 250 mL/min (∼3 mL/kg/min). To provide for this oxygen ‘delivery’ (DO₂, usually around 1000 mL/min) is 3–4 times higher, resulting in a VO₂:DO₂ of around 25–30%.^6–8 Consumption of oxygen generates CO₂ with the volume of CO₂ required to be cleared per minute (VCO₂) in a proportion – the respiratory quotient (RQ) which varies with the type of nutritional substrate. A normal RQ is generally considered to be 0.8 and results in a VCO₂ of around 200 mL/min with a VO₂ of 250 mL/min.⁶ Metabolic demands however are not static, but vary according to activity and change significantly during critical illness.^9–11 For example, neuromuscular blockade can reduce VO₂ by up to 15%¹² whilst hyperthermia increases VCO₂ approximately 5–10% per degree Celsius of body temperature change.^12,13 Critical illness can result in ketosis, reducing the RQ (even to <0.7¹⁴) whilst administration of intravenous dextrose, parenteral nutrition or propofol will all change the RQ and hence, VCO₂. Finally, the measurement of VCO₂ and RQ are affected by the extracorporeal support: the higher the CO₂clearance from ECMO support the lower the VCO₂ of the native lung, leading to a decrease in RQ, according to the alveolar gas equation.¹⁵

Although in the literature the oxygen transfer from the membrane lung is commonly described as the VO₂ML, conceptually it is more akin to O₂ delivery than O₂ consumption. In real world practice, the average O2 transfer (VO₂ML) from the membrane lung has been found to be 25–50 mL per litre of ECBF.^16,17 Therefore, when there is no native lung function (e.g. a shunt fraction of 1.0- see below) and patients are fully dependent on VV ECMO, high extracorporeal blood flows (ECBF) are required. For example. to supply a normal resting VO₂ of 250 mL/min in fully dependent patients an ECBF of ≥5 L/min may be needed. This necessitates the use of large bore cannula for VV ECMO (21-29Fr in adults), an adequate haemoglobin level¹⁸ and often measures to reduce tissue oxygen consumption early after initiation.

Gas partial pressures and their relation to the content of blood

To understand and titrate gas exchange during ECMO, it is important to distinguish between the gas tension (partial pressure), blood content (mL/L) and the overall gas transfer (mL/min) of respiratory gases. Figure 1 depicts the relationship between the partial pressure and content in blood for oxygen and CO₂:

(1) The dissolved content of gases in blood is very limited (grey line), and its linearly related to their partial pressure through their solubility constant (which is around 20x higher for CO₂ than O₂).

(2) The O₂ content of a blood sample is largely determined by the O₂ saturation and quantity of haemoglobin (equation 1 in Table 1).^6,7,19–22 Therefore, the O₂ content of blood plateaus as haemoglobin becomes fully saturated.

(3) Unlike O₂, CO₂ reacts with water to form bicarbonate which is its largest component in blood and does not plateau until extreme PCO₂ values are reached. This means that the relationship between the partial pressure and blood content is more linear for CO₂ than O₂ and changes in PCO₂ are more reflective of changes in blood content.

(4) For the same change in partial pressure there is a much greater change in blood content for CO₂ compared to O₂.

Figure 1.

Blood content and gas exchange.

Table 1.

Gas exchange, transport and delivery during ECMO.

	Variable	Calculation	Considerations during ECMO
1	O₂ content of a blood sample (ml/L)	$C_{x} O_{2} = 1.39 \cdot Hb \cdot S_{x} O_{2} + P_{x} O_{2} \cdot 0.003)$	-Can be measured from anywhere in the circuit
2	Douglas equation for CO₂ content in blood (mmol/L)	$C_{x} C O_{2} = 0.0308 \cdot P_{x} C O_{2} \cdot (1 + 10^{(pH - 6.1)}) \cdot (1 - \frac{0.0289 \cdot Hb}{(3.352 - 0.456 \cdot S_{x} O_{2}) \cdot (8.142 - pH)}$ )	-Small measurement errors can confound result
3	Full alveolar gas equation for alveolar oxygen tension (mmHg)	$PA O_{2} = Pi O_{2} - (\frac{PAC O_{2}}{{RQ}_{NL}}) + (Fi O_{2} \cdot PAC O_{2} \cdot (1 - \frac{{RQ}_{NL}}{{RQ}_{NL}})$	-The PACO₂ and PaCO₂ are influenced by ML gas exchange
3			-The RQNL can be altered by ML gas exchange
4	Oxygen transfer from native lung (ml/min)	$V O_{2} NL = CO \cdot (C_{a} O_{2} - C_{mv} O_{2})$	-CO measurement challenging in ECMO
			-Requires pulmonary artery catheter for mixed venous sample
			-Measured CmvO2 includes blood which has already passed through ML
			-Therefore, it does not allow an estimation of the oxygenation potential of the native lung when ECMO is ceased
5	Oxygen transfer from ML (ml/min)	$V O_{2} ML = (C_{Postoxy} O_{2} - C_{Preoxy} O_{2}) \cdot ECBF$
6	Total O₂ transfer (ml/min)	$V O_{2} tot = V O_{2} ML + V O_{2} NL$	-VO₂NL measurement challenging (see above)
7	CO₂ cleared by native lung (ml/min)	$V C O_{2} N L = V E C O_{2} \cdot R R$	-Volumetric capnography is the most convenient and reliable method
7	CO₂ cleared by native lung (ml/min)	OR $VC O_{2} NL = CO \cdot (CaC O_{2} - CmvC O_{2})$
8	CO₂ cleared by ML (ml/min)	$VC O_{2} ML = ECBF \cdot (C_{preoxygenator} C O_{2} - C_{postoxygenator} C O_{2}) \cdot 25$	-Blood CO₂ content calculations via Douglas equation are liable to measurement error
		OR
		$VC O_{2} ML = {C O_{2}}_{% Exhaust} \cdot SGF$
		OR
		$VC O_{2} ML = P_{E x h a u s t} {CO}_{2} \cdot P_{B a r o m e t r i c - H_{2} O} \cdot SGF \cdot 1000$
9	Total CO₂ clearance (ml/min) and percentage cleared by native lung	$VC O_{2} tot = VC O_{2} ML + VC O_{2} NL$
9		$VC O_{2} % NL = \frac{VC O_{2} NL}{VC O_{2} Tot}$
10	Respiratory quotient	$RQ = \frac{VC O_{2}}{V O_{2}}$	-Can be calculated for the membrane lung and native lung separately: The former may alter the later
11	Shunt fraction of the membrane lung	${Shunt}_{ML} = \frac{Ccapillaryoxy O_{2} - Cpostoxy O_{2}}{CcapillaryoxyO 2 - Cpreoxy O_{2}}$	-Ideal capillary O₂ content requires calculation using a modified ‘alveolar’ gas equation
		Where
		$P_{capillaryoxy} O_{2} = P_{Barometric - H_{2} O} \cdot FdO 2 - (\frac{P_{postoxy} CO 2}{{RQ}_{ML}})$
		Using
		${R Q}_{M L} = \frac{V O_{2} M L}{V C O_{2} M L}$
12	Shunt fraction of native lung	$\frac{Qs}{Qt} = \frac{Cc O_{2} - Ca O_{2}}{Cc O_{2} - Cmv O_{2}}$	-V/Q matching may be altered by ECMO
13	Dead space of membrane lung	$\frac{{VD}_{ML}}{{VT}_{ML}} = \frac{P_{Postoxy} C O_{2} - P_{exhaust} C O_{2}}{P_{Postoxy} CO 2} \cdot SGF$
14	Bohr dead space of native lung	$\frac{{VD}_{Bohr}}{VT} = \frac{PAC O_{2} - PEC O_{2}}{PAC O_{2}}$	-Requires evaluation of volumetric capnography curve to identify mean PACO₂
15	Oxygen delivery (ml/min)	$D O_{2} = (1.39 \cdot Hb \cdot Sa O_{2} + (0.003 \cdot Pa O_{2})) \cdot CO$	-CO may be unknown
16	Oxygen extraction ratio	$\frac{V O_{2}}{D O_{2}} = \frac{Sa O_{2} - Sv O_{2}}{Sa O_{2}}$	-Loses meaning during ECMO as the mixed venous oxygen saturation reflects the VO₂ML, the ratio of ECBF to CO as well as the oxygen extraction ratio of the tissues
17	Cardiac output via doppler echocardiography (L/min)	$Stroke Volume = Velocity Time Integral \cdot π r^{2}$	-Operator and tissue window dependent
17	Cardiac output via doppler echocardiography (L/min)	$Cardiac output = Stroke Volume \cdot Heart Rate$	-Operator and tissue window dependent
18	Contributions to mixed venous oxygen saturation	$SvO 2 = Pa O_{2} - \frac{V O_{2} t o t}{CO} \cdot \frac{1}{Hb \cdot 1.39}$	-During VV ECMO the significance of this equation is lost

NL: Native lung; ML: Membrane lung; Cx: Content of gas in blood compartment x (ml/L); Ca: Content of gas in arterial blood (ml/L); Cv: Content of gas in mixed venous blood (ml/L); C_{preoxygenator}: Content of gas in preoxygenator blood (ml/L); C_{postoxygenator}: Content of gas in postoxygenator blood (ml/L); Hb: Haemoglobin (g/L); SxO₂: Saturation of haemoglobin of blood compartment x; SaO₂: Saturation of haemoglobin of arterial blood; SvO₂: Saturation of haemoglobin of mixed venous blood; Px: Partial pressure of gas of blood compartment x (mmHg); RQNL: Respiratory Quotient of the native lung; PaO₂: Partial pressure of arterial O₂ (mmHg); PAO₂: Partial pressure of alveolar O₂ (mmHg); PaCO₂: Partial pressure of arterial CO₂ (mmHg); PACO₂: Partial pressure of alveolar CO₂ (mmHg); FiO₂: Fraction of inspired O₂; VO₂NL: O₂ transfer from native lung (ml/min); VO₂ML: O₂ transfer from membrane lung (ml/min); VO₂tot: Total O₂ transfer (ml/min); DO₂: total oxygen delivery (cardiac output multiplied by the content of arterial blood) (ml/min); VCO₂NL: CO₂ clearance from native lung (ml/min), VCO₂ML: CO₂ clearance from membrane lung (ml/min); VCO₂tot: Total CO₂ clearance (ml/min); P_ExhaustCO₂: Partial pressure of CO₂ measured from ML exhaust (mmHg); CO_2%Exhaust: Percentage of CO₂ measured from ML exhaust; ECBF: Extracorporeal Blood Flow (L/min); SGF: Sweep gas flow (L/min); CO: Cardiac Output (L/min, r: Radius of the outflow tract where velocity time integral is measured (cm).

The haemoglobin saturation and haemoglobin level of a sample can be directly measured, hence calculating blood O₂ content is relatively straightforward (equation 1). Conversely, the CO₂ content of blood requires calculation using complex equations, for example the Douglas equation (equation 2, Table 1).^7,19

Gas transfer across the native lung (without ECMO)

Gases cross the alveolar-capillary membrane according to the partial pressure gradients between alveolar gas and capillary blood. The partial pressure of alveolar oxygen (PAO₂) depends upon the fraction of inspired oxygen, the pressure in the alveolus, the pressure of alveolar CO₂ and the RQ, mathematically expressed as the classical alveolar gas equation (equation 3, Table 1). The alveolar gas equation can therefore explain hypoxaemia due to a low partial pressure of inspired O₂ (e.g., at altitude) or due to hypoventilation. Beyond changes in alveolar gases, respiratory failure can result from diffusion abnormalities or, most commonly, an altered ratio of ventilation to perfusion (V/Q).

In Riley’s classic 3 compartment model of the lung, shunt refers to alveoli where there is perfusion but no ventilation (V/Q = 0).²³ Shunt blood returns to the left atrium with the same oxygen content as the mixed venous blood. Shunt predominantly leads to hypoxaemia, as the oxygen content of blood plateaus, and increased ventilation or FiO₂ of the healthy alveoli cannot compensate for the shunt fraction (Figure 1 Panel B). However, if ventilation of other lung units can be sufficiently increased, the PaCO₂ can still be maintained, until the shunt fraction becomes extremely high or in conditions of very low cardiac output when the mixed venous CO₂ tension is much higher than the arterial or capillary pCO₂. In practice there is a spectrum of V/Q relationships within the lung from 0 to 1 giving rise to ‘venous admixture’ which refers to the amount of mixed venous blood which would be necessary to explain an observed difference between capillary and arterial O₂. Dead space refers to the ventilated but not perfused portion of the tidal volume (true dead space has a V/Q = ∞). Physiological dead space includes both anatomical dead space (oro-nasopharyngeal and airways that don’t participate in gas exchange) and alveolar dead space. Dead space reduces the alveolar ventilation and therefore increases the minute ventilation required to achieve a given PaCO₂ (Panel B Figure 1). The spectrum of abnormal ventilation perfusion relationships (V/Q ≠ 1) is referred to as V/Q mismatch. The indices used to evaluate the magnitude of respiratory failure and quantify shunt, venous admixture or dead space are described in Table 2.

Table 2.

Gas exchange indexes evaluating the native lung.

Gas exchange index	Calculation/Measurement	Considerations in conventionally supported patients	Considerations during ECMO
PaO₂:FiO₂ ratio	$\frac{P a O_{2}}{F i O_{2}}$	-In widespread use including for making a diagnosis of ARDS or the indication of therapies at different severity thresholds	-Any metric relying on PaO₂ loses meaning in evaluating native lung function during ECMO
Alveolar-arterial gradient (mmHg)	$P A O_{2} - P a O_{2}$	-Limited utility in the intubated patient	-Both the content of alveolar gas and PaO₂ are influenced by ECMO
Oxygenation index	$\frac{F i O_{2} ∙ M e a n A i r w a y P r e s s u r e}{P a O_{2}}$	-Only moderately prognostic in adult ARDS	-Any metric relying on PaO₂ loses meaning in evaluating native lung function during ECMO
Bohr dead space fraction	$\frac{VDBohr}{VT} = \frac{PAC O_{2} - PEC O_{2}}{PAC O_{2}}$	-Requires volumetric capnography to measure PACO₂	-Valid and quantifies true anatomic and alveolar dead space
Enghoff dead space fraction	$\frac{VDEnghoff}{VT} = \frac{PaC O_{2} - PEC O_{2}}{PaC O_{2}}$	-Requires volumetric capnography to measure mixed expired CO₂	-Not validated during ECMO, PaCO₂ influenced by VCO2ML
		-Sensitive to venous admixture
		-Can be used to aid PEEP selection
		-Established evidence base in ARDS
Modified Enghoff dead space fraction	$\frac{VDModifiedEnghoff}{VT} = \frac{PaC O_{2} - PETC O_{2}}{PaC O_{2}}$	-More straightforward to calculate from end tidal CO₂	-Not validated during ECMO, PaCO₂ influenced by VCO₂ML
Modified Enghoff dead space fraction		-Sensitive to venous admixture	-Not validated during ECMO, PaCO₂ influenced by VCO₂ML
Ratio of end tidal to arterial partial pressure of CO₂	$\frac{PET {CO}_{2}}{Pa {CO}_{2}}$	-Ideal gas exchanger has value of 1-Sensitive to venous admixture and intended as a ‘holistic’ index-Evolving evidence base in ARDS	-Similar to Enghoff, PaCO₂ is influenced by VCO₂ML
Ratio of end tidal to arterial partial pressure of CO₂	$\frac{PET {CO}_{2}}{Pa {CO}_{2}}$		-However, some data suggests predictive of weaning outcomes during ECMO
Ventilatory ratio	$\frac{(V E \cdot P a {CO}_{2})}{(P B W \cdot 100 \cdot 37.5)}$	-The ratio of the actual minute ventilation and PaCO2 and that expected by population estimates	-During VV ECMO the usual relationship between minute ventilation and PaCO₂ is lost
		-Sensitive to venous admixture
		-Established evidence base in ARDS

PaO₂: Partial pressure of arterial O₂ (mmHg); PAO₂: Partial pressure of alveolar O₂ (mmHg); PaCO₂: Partial pressure of arterial CO₂ (mmHg); PACO₂: Partial pressure of alveolar CO₂ (mmHg); PECO₂: Partial pressure of mixed expired CO₂ (mmHg); PETCO₂: Partial pressure of end tidal CO₂ (mmHg); FiO₂: Fraction of inspired O₂; VT: Tidal volume (L); VE: Minute ventilation (L/min); PBW: Predicted body weight (kg); VCO₂ML: CO₂ clearance from membrane lung (ml/min).

Gas transfer across the membrane lung (ML)

During ECMO, extracorporeal blood flow (ECBF) is drained from the IVC, SVC or right atrium and passed through a ML before being returned ‘arterialised’ (O₂ added, CO₂ partially removed) to the right atrium. Within the artificial lung, hollow fibres carry fresh sweep gas flow (SGF), of titratable oxygen fraction (FdO₂) and rate. Similar to the native lung, gas exchange occurs passively across membranes down pressure gradients for each gas. However, the membrane fibres are thicker, have less total volume and are much less efficient than the alveolar membranes in the native lung.

For the membrane lung the ‘alveolar’ gas is the sweep gas and the fraction of device O₂ (FdO₂) can be directly manipulated which will affect the O₂ transfer across the membrane lung (VO₂ML). The other determinants of VO₂ML are the blood flow and pre-membrane oxygen saturation. Similar to increasing alveolar ventilation in the native lung, the sweep flow rate rather than FdO₂ predominantly alters the CO₂ clearance (VCO₂ML) (although the FdO₂ may have some influence due to the Haldane effect). Gas diffusion for a given partial pressure gradient is largely fixed by the mechanical properties of the hollow fibres within the membrane lung (most commonly made from polymethylpentene) but may worsen with pseudomembrane formation. Further, similarly to in the native lung, high ECBF flow rates may limit CO₂ transfer by creating diffusion impairments for a given SGF when the ‘transit time’ of blood is shortened. V/Q mismatch also applies to the membrane lung.^{6,20,21,24,25} As the SGF is increased for a given ECBF, the VCO₂ML plateaus due to a diminishing gradient for CO₂ removal. Oxygenator function tends to worsen over time due to the accumulation of thrombus, pseudomembrane formation and gas condensation. This may lead to increased dead space (occluded blood channels) or shunt (occluded gas channels).

Monitoring gas exchange during VV ECMO

During VV ECMO gas exchange in the ML and native lung occurs ‘in series’. A proportion of the venous return passes through the extracorporeal circuit before it enters the pulmonary circulation. Figure 2 displays the interactions between the ML and native lung during VV ECMO.

Figure 2.

The interdependence of membrane lung and native lung gas exchange during V-V ECMO.

Monitoring the membrane lung

Evaluating the VO₂ML

The volume of oxygen added to blood extracorporeally (VO₂ML) can be calculated using equation 5 (Table 1). This can be understood as the difference between the oxygen content of blood entering and exiting the membrane lung multiplied by the ECBF to give the oxygen transfer per minute.²⁶

The saturation of preoxygenator blood is a critical determinant of the VO₂ML. The preoxygenator SO₂ is influenced by both the overall VO₂:DO₂ and recirculation (Q_R) - blood which has already undergone extracorporeal gas exchange being re-aspirated into the circuit. This will ‘dilute’ the true venous blood leading to a higher preoxygenator SO₂ and reducing the VO₂ML at any given ECBF. Although Q_R varies with cannula positioning and at different ECBF rates, techniques used experimentally to directly quantify it^27,28 are currently research tools and in general it is best to qualitatively assess the degree of Q_R by monitoring the preoxygenator and peripheral SO₂ when changing the ECBF. The preoxygenator SO₂ is usually an overestimate of the true venous SO₂ because of Q_R, hence a very low preoxygenator SO₂ (<50%) should prompt concern about the overall VO₂:DO₂. Finally, if all other factors are held constant, a change in the preoxygenator SO₂ (e.g. due to altered VO₂:DO₂ of the patient) will alter the VO₂ML, even if oxygenator function is static.

Importantly, measuring the postoxygenator PO₂ alone is not a reliable indicator of oxygenator failure.¹⁷ Indeed, in a recent study a low post-oxygenator PO₂ was the sole or contributory reason for an ECMO circuit change in around one third of cases, but a circuit change could have been avoided or delayed in some cases had formal calculation of VO₂ML occurred.¹⁷ An aging membrane can decrease the functional units of gas exchange within the membrane lung analogously to the concept of shunt in the native lung. The shunt fraction of the membrane lung can be formally calculated (equation 12), similar to that of the native lung.²⁴

Evaluating the VCO₂ML

Similar to the VO₂ML, the VCO₂ML can be calculated by the use of pre- and post-oxygenator blood gas analysis using the Douglas equation (equation 2). From this, the VCO₂ML can be calculated as the difference in CO₂ content across the membrane multiplied by blood flow^6,20,21,25 (equation 8). The complexity of this equation means that there is a potential for error by magnifying small variations in measurements between samples.²⁹

Capnometry can be used to calculate the VCO₂ML by measuring the CO₂ content of the exhaust gas from the ML (Figure 1).^{20,21,24,30–32} Specific devices have been developed which warm the exhaust gas to avoid condensation and the influence of temperature as well as directly measuring the flow rate of gas from the exhaust.^31,33 However, providing that there is no resistance to exhaust flow, substituting a standard portable capnometer provides acceptable performance (equation 8)^30,31 Some authors suggest increasing the SGF to at least 10L per minute for 30s (a ‘sigh’) prior to measuring the exhaust CO₂ after a period of equilibration in order to remove any condensation from the gas channels.³⁴

Monitoring native lung function during VV ECMO

Classical assessments of native lung gas exchange (Table 2) are challenging during VV ECMO. True venous blood is mixed with post-oxygenator blood in the right ventricle before being distributed to the alveolar capillaries (Figure 2). The mixed venous blood therefore has a higher oxygen saturation than that of true venous blood and no longer reflects the VO₂:DO₂. As hypoxic pulmonary vasoconstriction is sensitive to arteriolar as well as alveolar tensions of oxygen,^35,36 VV ECMO will alter native lung VQ matching.^6,22 Further, the VO₂ and VCO₂ of the native lung are influenced by extracorporeal gas exchange: for example, if the mixed venous oxygen saturation of O₂ is >90% due to ECMO, regardless of native lung function minimal further oxygen can be added. As the arterial content of O₂ and CO₂ is a combination of ML and native lung gas exchange occurring in series, many indices to assess native lung gas exchange are no longer truly reflective of lung function (Table 2). Understanding the limitations of monitoring native lung gas exchange is therefore crucial during VV ECMO.

Challenges in measuring cardiac output

Although the cardiac output (CO) is a key variable to assess native lung function, this is difficult to measure during VV ECMO. The ‘gold standard’ direct Fick method³⁷ is invalidated by the presence of extracorporeal gas exchange. All thermodilution techniques are unreliable during ECMO as the indicator is partially aspirated into the ECMO circuit, leading to a systematic overestimation of CO.³⁸ Echocardiographic measurement of CO from the left or right ventricular outflow tract using the doppler effect^38–41 (equation 17) is reliable during VV ECMO. However, this offers only a single time point measurement and is resource intensive, dependent upon tissue windows and suffers from inter- and intra-operator variability.

Monitoring native lung oxygenation capacity and the VO₂NL

The most common indices aiming to assess native lung capacity for oxygenation used in critical care are no longer helpful during ECMO, including the P:F ratio, oxygenation index and alveolar-arterial gradient (Table 2).

The transfer from the native lung (VO₂NL) can be estimated in VV ECMO using the arterio-mixed venous content difference if the true cardiac output is known (equation 4). The difference between the inspired and expired oxygen from the native lung can be monitored and is be considered to be a marker of O₂ uptake by the lung in steady state^42,43 and used as a means of estimating the VO₂NL over time. However, unlike CO₂, the difference between inspired and expired oxygen is generally low during mechanical ventilation, especially with high FiO₂, introducing a potential source of error.⁴⁴ The proportion of venous admixture (true shunt plus V/Q <1) can be calculated if a pulmonary artery catheter is in situ during VV ECMO, but this measurement may not be representative of conditions when extracorporeal support is stopped, due to the influence of post-oxygenator blood on the mixed venous saturation and therefore V/Q.^6,22

One useful metric for native lung oxygenation capacity is the ‘oxygen challenge test’. In this test the ventilator FiO₂ is increased to 1.0 whilst other mechanical ventilation and ECMO settings are left unchanged. Then an arterial blood gas is measured after 30 min to assess the change in PaO₂. This change equates to changes in the native lung’s capacity for oxygenation over time. The utility of the oxygen challenge test to define readiness to wean ECMO is unproven.^45,46

Monitoring VCO₂NL, assessing the dead space and ventilatory efficiency of the native lung

Metrics sensitive to dead space tend to be more prognostic than oxygenation indices in acute lung injury.^47–49 Although VV ECMO invalidates Riley’s assumptions,²⁹ waveform and volumetric capnography mean indices related to dead space and measurement of the VCO₂ML are much easier to apply and have a greater evidence base during ECMO.

Volumetric capnography enables calculation of the anatomical and instrumental dead space, the mean alveolar CO₂ concentration (the midpoint of ‘phase III’, for the Bohr dead space) and the mixed expired PECO₂ (the mean pressure of expired CO₂, for the Bohr and Enghoff dead space) as well as the ETCO₂.⁵⁰ From this, ventilators can calculate the total expired CO₂ per breath (the VECO₂) and per minute (the VCO₂NL). From this, the total VCO₂ (VCO₂tot) can be calculated (equation 9).⁶ The percentage of the VCO₂tot which is achieved by the native lung (equation 9) can be tracked over time in response to changes. In a perfect gas exchange system, the terminal alveolar gas would have an identical pressure of CO2 as arterial blood. However, due to dead space, V/Q mismatch and shunt, this is never the case. Volumetric capnography enables the calculation of several metrics related to dead space, some influenced by the venous admixture, which are useful tools during VV ECMO (Table 2)^22,47,50,51 such as the ETCO₂:PaCO₂ ratio. Although these are influenced by extracorporeal gas exchange, they can track native lung function over time with static ECMO support, as ECMO support is weaned, or used to titrate mechanical ventilation. Finally, akin to cardiopulmonary exercise testing, monitoring the ventilatory efficiency (ratio of minute ventilation to VCO₂NL) over time may be useful as a measure of improving respiratory function (including during weaning).^22,29

Additional benefits of VO₂ and VCO₂ measurement during ECMO

Indirect calorimetry can be used to individualise nutrition through measuring the VO₂, VCO₂ and RQ of the patient to infer their caloric requirements.^44,52 The VO₂ and VCO₂ of the native and ML should be summed to generate values to estimate resting energy expenditure via the Weir equation.⁵³ It is important to note that this calculation is less accurate with increasing recirculation.⁵³

Targets for gas exchange

The oxygen target for critically ill patients is debated with competing concerns of tissue hypoxia versus pulmonary and systemic oxygen toxicity. Although several recent high-quality randomised control trials have been published, these have not demonstrated a consistent benefit or harm of targeting oxygen saturation levels close to the ranges usually adopted in clinical practice (SaO₂ at least >88% and <100%).^54–57 Reports have demonstrated there is no difference in neurocognitive outcomes in VV ECMO survivors even with prolonged hypoxaemia unless cerebral blood flow has been compromised.^58,59

Technical considerations are relevant to the setting of targets for oxygenation during ECMO. Firstly, the question of whether to set a peripheral oxygen saturation (SpO₂) target, SaO₂ target, PaO₂ target or DO₂ target (Table 3). A few specific points merit highlighting:

(1) A systematic overestimation of SpO₂ in hypoxic patients of black skin pigmentation occurs with many SpO₂ monitors^60,61 (although at least one device has been demonstrated to be accurate regardless of skin tone).⁶²

(2) Haemolysis, which inevitably occurs on ECMO, can generate endogenous carbon monoxide release leading to carboxyhaemoglobinaemia. Indeed, it has been demonstrated that SpO₂ may become less accurate over time during prolonged ECMO runs due to a high prevalence of mild (>3%) carboxyhaemoglobinaemia which should be measured on serial blood gases.⁶³

(3) As certain tissues including the cerebral and pulmonary vasculature sense oxygen tensions rather than DO₂ or saturations,⁶⁴ targeting PaO₂ may be more appropriate than SaO₂ in patients with brain injury or right heart failure.

Table 3.

Comparison of targets for oxygenation during ECMO.

Monitor	Strengths	Limitations
SpO₂	Continuous, non-invasive, usually correlates with SaO₂	-Carboxyhaemoglobinaemia is common during ECMO and may cause overestimation of SaO₂
		-May overestimate SaO₂ in hypoxic patients of black ethnicity
		-Monitors often post-process waveform giving illusion that there is a good signal:noise ratio (aim for pulse index >0.4 for reliability)
SaO₂	If measured, gold standard vs SpO₂	-Some gas analysers calculate SaO₂ rather than directly measure
SaO₂	Major contributor to CaO₂	-Requires blood sampling
PaO₂	Sensed by tissues including pulmonary and cerebral vasculature	-Poor contribution to CaO₂
PaO₂		-Requires blood sampling
DO₂	Better reflects oxygen delivery to tissues	-Unclear threshold for action given unable to measure oxygen extraction ratio objectively
DO₂	Better reflects oxygen delivery to tissues	-Challenging to measure cardiac output

SpO₂: Saturation of haemoglobin in peripheries; SaO₂: Saturation of haemoglobin of arterial blood; CaO₂: Content of O₂ in arterial blood; DO₂: total O₂ delivery (cardiac output multiplied by the content of arterial blood) (ml/min).

Directly assessing the adequacy of tissue oxygen delivery and oxygen extraction is challenging on ECMO as the mixed venous oxygen saturation is no longer truly representative of VO₂:DO₂. An alternative technique which is not confounded by ECMO is nearfield infrared spectroscopy (NIRS). Similar to pulse oximetry, this uses the absorption of infrared light to infer the relative concentration of oxy- and deoxy-haemoglobin.⁶⁵ As most of the signal is derived from venous blood, desaturation can detect increasing oxygen extraction as well as arterial desaturation. NIRS is influenced by pH, blood flow, PaCO₂ and haemoglobin concentration and so may represent a non-specific but integrative measure for local tissue perfusion.⁶⁶ An important limitation is that the sensor can only detect changes 2-3cm deep. However, if there is no regional intracranial pathology this is a means of assessing the adequacy of cerebral perfusion and oxygenation. NIRS defined desaturations are related to cerebral outcomes during ECMO^67,68 (although case series largely relate to VA modes), where a relative decrease of 10–20% from baseline or acutely may be more meaningful rather than the absolute value, although values below 50% are concerning.⁶⁹

Avoiding respiratory acidosis (pH <7.3) with VCO₂ML is a reasonable target, however as ECMO is initiated it is important to correct hypercapnia slowly to avoid reactive cerebral vasoconstriction. An association between rapid changes in PaCO₂ after ECMO initiation and cerebral complications such as intracranial haemorrhage has been demonstrated.⁷⁰ Changes in PaCO₂ should be limited to <5–10 mmHg (<0.6-1.3 kPa) per hour.

Daily measurements to assess the native and membrane lungs

Although there is no guidance, it seems reasonable to perform daily measurement of the VO₂ML, VCO₂ML, VCO₂NL and an oxygen challenge test. This allows basic assessment of the gas transfer capability of both the membrane and native lungs. The assessment is supplemented using dead space indices including the ETCO₂:PaCO₂ ratio and monitoring the ventilatory efficiency and the percentage of total VCO₂ carried out by the native lung over time.

Monitoring to minimise native lung injury and respiratory distress

Importantly, we believe that the mortality benefit from VV ECMO is driven by the facilitation of lung protective ventilation rather than changes in gas exchange per se. Although optimal mechanical ventilation has not been defined for VV ECMO, it is reasonable to minimise the risk of further lung injury. Hence multimodal respiratory assessment is especially relevant during VV ECMO.

Mechanical power and ventilation

Calculating mechanical power at the bedside

Ventilator and patient self-induced lung injury (VILI and P-SILI) are important concepts that can worsen the impact of the underlying lung injury, worsen patient outcome and are well described using the concept of ‘mechanical power’ (MP). This measures the total energy transferred to the native lung during ventilation in J/min and can be obtained through integration of the pressure volume loop over time.⁷¹ This requires offline analysis, hence surrogate calculations have been suggested^72,73 (Table 4).

Table 4.

Formulas and definitions related to ergotrauma, recruitability and respiratory effort.

Description	Calculation	Derivation, considerations
Mechanical power for pressure control modes	$MP = 0.098 \cdot RR \cdot VT \cdot (PEEP + Δ P a w)$	-Assumes normal airway resistance
Mechanical power for volume control modes	$MP = 0.098 \cdot RR \cdot VT \cdot (Peak inspiratory pressure - 1 / 2 (Δ P a w))$	-Square wave flow necessary
Approximate mechanical power in volume control	$MP = \frac{VE \cdot (Peak pressure + PEEP + \frac{Inspiratory Flow}{6})}{20}$	-Square wave flow necessary
		-Can be used without performing hold manoevers
		-However, assumes 60 L/min flow and normal airway resistance
Estimated transpulmonary mechanical power	$E s t {M P}_{L} = 0.098 \cdot 0.5 \cdot (P S - 0.93 \cdot P_{o c c}) \cdot V T \cdot R R$	-Requires external validation study
Recruitment to inflation ratio	$\frac{R e c r u i t m e n t}{I n f l a t i o n} = \frac{C_{R e c}}{C_{L o w P E E P}}$	-See Figure 3
Transpulmonary pressure (direct method)	$P_{L} = P_{a w} - P_{e s}$	-Preferred method to calculate end expiratory P_L
Transpulmonary pressure (elastance derived)	$P_{L} = P a w - (P a w ∙ \frac{E c w}{E_{R S}})$	-Preferred method to calculate end inspiratory P_L
	Where
	$E_{RS} = \frac{Δ P a w}{Δ V}; E_{CW} = \frac{Δ P e s}{Δ V}$
Estimated transpulmonary pressure	${E s t P}_{L} = P_{i n s p} - 0.66 \cdot P_{occ}$	-As P_occ is always negative with inspiratory effort, the estimated transpulmonary pressure will be higher than the peak pressure
Estimated total muscle pressure	$E s t P_{m u s} = - 0.75 \cdot P_{occ}$	-P_occ correlates well with ΔP_es but the airway must be occluded to measure it. A correction factor (empirically derived) adjusts for the ensuing increased effort
Total muscle pressure	$P_{m u s} = Δ P_{e s} + E_{C W} \cdot VT$	-The difference between P_es and the estimated recoil pressure of the chest wall

MP: Mechanical Power (J/min); RR: Respiratory Rate per minute; VT: Tidal Volume (L); PEEP: Positive End Expiratory Pressure (cmH₂O); ΔP_aw: Plateau pressure minus PEEP (the driving pressure) (cmH₂O), EstMP_L: Estimated transpulmonary MP (J/min); C_Rec: Compliance of recruited lung (L/cmH₂O); C_LowPEEP: Compliance at low PEEP (L/cmH₂O); P_L: Transpulmonary pressure (cmH₂O); PS: Pressure support (cmH₂O); EstΔP_L: Estimated driving transpulmonary pressure (cmH₂O); P_aw: Pressure measured at the airway (cmH₂O); ΔP_aw: Change in pressure measured at the airway (cmH₂O); P_es: Pressure measured in the oesophagus (cmH₂O); ΔP_es: Change in pressure in the oesophagus (cmH₂O); ΔP_L: Driving transpulmonary pressure (cmH₂O); ΔV: Change in volume (L); E_RS: Elastance of the respiratory system (cmH₂O/L); E_CW: Elastance of the chest wall (cmH₂O/L); P_mus: Pressure exerted by the respiratory muscles (cmH₂O); EstP_mus: Estimated pressure exerted by the respiratory muscles (cmH₂O).

The importance of the MP concept is that it integrates all the contributors to lung injury. In animal experiments where the delivered MP remained constant but applied in different manners (RR, VT, PEEP), the resulting degree of lung injury was equivalent,⁷⁴ with profound VILI experienced in piglets with healthy lungs ventilated to a MP of >12 J/min.⁷⁵ The most important implications are that:

(1) Low tidal volume ventilation (<6 mL/kg IBW) may still be injurious if it is associated with high driving pressures (e.g., >15 cmH₂O).⁷⁶

(2) Respiratory rate is an important modifier of the risk of MP and mortality in ARDS.

(3) PEEP contributes to MP and therefore may also contribute to lung injury, however appropriately set PEEP may reduce the driving pressure and distribute the applied MP more homogenously thereby reducing its impact.⁷⁶

A ‘safe’ threshold for mechanical power is unclear,⁷⁷ but higher MP correlates with mortality during invasive ventilation,⁷⁸ ARDS⁷⁹ and during ECMO.⁸⁰ It is likely that the same mechanical power is more injurious when applied to a smaller functional lung volume and for a longer duration.^81–84 Conversely, an excessively cautious approach to mechanical ventilation during ECMO (for example, a MP of 0) will quickly result in total lung collapse. Although unproven, given that severe lung injury is the primary indication for VV ECMO, our practice is to target a MP 6–10 J/min (around 2–3 times that of normal breathing). For example, our usual baseline setting on the ventilator is 10 cmH₂O PEEP, a driving pressure of 10 cmH₂O and a respiratory rate of 10 prior to assessment. Even if there is minimal minute ventilation with such an approach, the priority in the early phase is in protecting the lung by managing gas exchange extracorporeally.

Ventilator assessments of recruitability and overdistension

During tidal ventilation, lung units may open in inspiration and then close in expiration. This is referred to as intratidal recruitment and may be harmful (‘atelectotrauma’).^85,86 Higher PEEP or mean airway pressure may provide sustained recruitment of previously closed lung units. Optimal PEEP may be assessed by finding the point at which compliance is optimal, however as compliance is normally measured at end inspiration, it may increase with both intratidal and sustained recruitment. Assessing optimal PEEP using oxygenation may also be undertaken, however this can also be misleading as PaO₂ may rise due to the haemodynamic effects of PEEP as well as its effects on recruitment.

Given the potential harm from aggressive recruitment⁸⁷ or excessive PEEP,⁷⁴ alternative methods to assess lung ‘recruitability’ are required. Multiple parameters from a PV loop can be used to evaluate recruitability (panel C, Figure 3). The difference in volume between the inspiratory and expiratory limbs of a PV loop is referred to as hysteresis and reflects the static properties of the lung at low flow. Hysteresis represents the energy transfer during inspiration which is not recovered from elastic recoil-energy and is tidally dissipated into the lung tissue.⁸⁸ In general, a large hysteresis level is predictive of recruitability with higher PEEP.⁸⁶ This can also be quantitatively assessed at the bedside-for example the normalised maximal distance (NMD). This quantifies the maximal hysteresis volume between inspiration and expiration divided by the total volume insufflated during the loop (e.g. from 5 to 45 cmH₂O) (Figure 2). An NMD >41% is predictive for recruitment^86,89 and suggests consideration of a higher PEEP or mean airway pressure strategy. Up to ∼25% of patients with ARDS have complete airway closure below a certain airway pressure.⁹⁰ This can be assessed during a low flow PV Loop (Figure 3) and setting PEEP above this level will minimise intratidal recruitment.

Figure 3.

Ventilator assessments to personalise ventilation during ECMO.

The ‘recruitment to inflation ratio’ (R:I) has been validated as a marker of recruitability by CT and the low flow PV loop.^89,91 The principle is that if a change in PEEP leads to recruitment of previously closed lung units the compliance of the recruited volume will be higher than if the added ‘PEEP volume’ is distributed across the same lung units which are already open at low PEEP (potentially causing overdistension). In a passively ventilated patient, the R:I is measured by dividing the compliance of the volume recruited by PEEP (measured as the expired volume when PEEP is reduced) by the compliance at low PEEP (Table 4 and Panel D Figure 3). An R:I >0.5 is suggestive of recruitability. This simple manoeuvre can be applied with any ICU ventilator.⁹¹

When using volume control with constant inspiratory flow, visualisation of the pressure curve can identify a positive concavity at end inspiration suggestive of overdistension (an elevated ‘stress index’ (panel B Figure 3).⁹² Conversely, a negative concavity suggests the presence of intratidal recruitment (an increase in PEEP may prevent this). The stress index is only measurable with constant inspiratory flow and experimental studies suggest it is sensitive to changing chest wall elastance as well as mechanical ventilation settings.^93,94 Frank overdistension can also be seen with ‘beaking’ on a PV loop but its absence does not exclude regional overdistension as overdistension can co-exist with recruitment at a given pressure level.

These tools are extremely useful in optimising PEEP at the bedside and can be carried out over time as lung mechanics change.

Invasive assessment of transpulmonary pressure

Lung stress is generated by the ‘transpulmonary pressure’, not the airway pressure.⁹⁵ The transpulmonary pressure (P_L) is defined as the pressure applied from the airway, across the alveoli to the pleural space (e.g. P_L=P_aw – P_pleura). This is particularly pertinent at extremes of chest wall elastance (e.g. obesity). The gold standard to assess the transpulmonary pressure at the bedside is the use of an oesophageal balloon catheter to measure oesophageal pressure (P_es). P_es is a useful surrogate for the P_pleura which can quantify the transpulmonary pressure and guide ventilation.

The transpulmonary pressure at end inspiration can be measured in two complementary ways^95,96 (Table 4):

a. The direct method

b. The elastance derived method

The direct P_L best approximates the P_pleura in basal lung segments close to the horizontal plane of the catheter and is helpful for determining optimal PEEP. The elastance derived P_L better approximates the P_pleura in non-dependent regions and is helpful to assess the risk of overdistension.⁹⁷ A maximal inspiratory P_L of 20–25 cmH₂O, ideally measured using the elastance derived method, has been suggested and limiting transpulmonary driving pressure (ΔP_L) to maximum 10–12 cmH₂O in patients with lung injury is prudent.^95–99 Similar to the targeting of MP, a more conservative approach may be appropriate during ECMO. The transpulmonary pressure can also be used to develop a transpulmonary low-flow PV loop in a similar manner to that described above.

Imaging to assess recruitability, overdistension and heterogeneity

The most important consideration when using MP or P_L to guide ventilation is to place the measurements in context. The applied MP is distributed only to the smaller functional lung volume-the ‘baby lung’- in ARDS and it is in these regions where intense inflammatory activity can be detected with CT positron emission tomography.¹⁰⁰ Knowing the functional residual capacity for individual patients would be extremely useful as a means of quantifying the ‘baby lung’. Unfortunately, the classical tools to assess FRC (or most commonly, End Expiratory Lung Volume (EELV) when patients are receiving PEEP) such as nitrogen wash-out/wash-in (EELV) or helium dilution (true FRC) have their assumptions violated- or are impractical-during extracorporeal gas exchange.¹⁰¹ Furthermore, regional collapse distorts the normal extracellular architecture and acts as a local ‘stress raiser’ which magnifies the forces to surrounding alveoli. Therefore, although the P_L is reflective of the functional lung volume as both are dependent on the elastance of the lung and chest wall, understanding the heterogeneity of lung injury is crucial to judge the regional risks of using a ‘global’ P_L.

CT represents the clinical gold standard lung anatomical assessment and for measuring EELV in mechanically ventilated patients.¹⁰² Hounsfield unit filters can be used to identify voxels of complete collapse, partial aeration, normal aeration and hyperinflation in a quantitative manner.^103,104 However, this requires post-processing and filtering to achieve accurate results. Quantitative CT can be acquired at different airway pressure levels to assess the recruitability of lung and guide early ventilation strategies during ECMO.¹⁰⁵ There is emerging literature to suggest that AI-derived automated quantitative CT analysis is feasible.¹⁰⁶ CT scanning requires patient transfer, radiation exposure and, as lung mechanics and pathology evolve over time, can only provide an impression at the time of image acquisition. Nevertheless, our practice is to acquire thoracic CT at 5 and 45 cmH₂O following cannulation for ECMO to evaluate the underlying parenchyma, assess the functional lung volume and estimate recruitability.

Electrical impedance tomography (EIT) is a non-ionising, continuous, semi-quantitative approach to imaging the lung which can be applied at the bedside.¹⁰⁷ During EIT changes in impedance provide a real time assessment of regional ventilation and heterogeneity. Hyperinflated and collapsed regions can be identified and PEEP titrated to reduce heterogeneity, which is not necessarily the PEEP providing the optimal compliance.^108–111 EIT derived optimal PEEP varies significantly across patients supported with ECMO.¹¹⁰ Changes in EELV can be estimated by evaluating changes in end-expiratory lung impedance in response to interventions. Although changes in impedance correlate with changes in EELV, other interventions including the administration of a fluid bolus can also be also change impedance.^112,113 Limitations of EIT are its spatial resolution (restricted to one 10–18 cm horizontal plane band) and for many more advanced techniques, offline post-processing of imaging is required.

Lung ultrasound (LUS) approaches CT for diagnostic accuracy in categorising acute respiratory failure.¹¹⁴ The regional extent of peripheral consolidation can be assessed and pleural effusions identified easily. The number of regions with 2 or more B lines (vertical linear artefacts reflecting transmission of the signal from the pleural line abnormally through parenchyma with increased fluid/air ratio) correlates well with the extravascular lung water (EVLW) measurements obtained through transpulmonary thermodilution in conventionally supported patients.^115–117 A LUS score can be generated from 6 regions in each hemithorax (a score of 1 for normal findings, 2 for >2 B lines, 3 for coalesced B lines and 4 for complete loss of aeration ‘hepatisation’).¹¹⁸ This is a particularly useful technique during ECMO as measuring EVLW through thermodilution is inaccurate.¹¹⁹ LUS can be repeated over time at the bedside to obtain prognostic information and provide a therapeutic target. LUS can be used to assess regional ventilation and aeration after ventilatory changes. For example, an improvement in the LUS score in response to PEEP correlates well with changes in PaO₂ and recruitment measured by a P/V Loop.¹²⁰ At present LUS requires experience and is limited to assessing peripheral lung, however artificial intelligence is being explored and automated B line quantification at the bedside is already possible.^121–123 LUS and EIT give different, complimentary information and their use in combination may add value at the bedside.¹²⁴

Monitoring spontaneous breathing during VV ECMO

There are benefits to allowing spontaneous efforts in severe lung injury – to prevent respiratory muscle weakness and reduce the requirements for sedation and neuromuscular blockade. However, spontaneous breathing brings a new set of challenges during ECMO. Respiratory drive, the intensity of the output of the respiratory centres, adjusts the output of the respiratory muscles and hence, minute ventilation to achieve a certain PaCO₂ ‘target’, the so-called set point.^125,126 For example, during exercise, respiratory drive and minute ventilation increase, leading to a VO₂ and VCO₂ rise whilst PaCO₂ remains stable leading to isocapnic hyperpnea.¹²⁷ The set point of the respiratory system is influenced by many factors including inflammation, hypoxia, lung mechanoceptor stimulation, pain or sedation,^125,126 and not solely on pH. In the ECMO supported patient making spontaneous efforts, the pH/PaCO₂ alone is not a reliable target as changes in the VCO₂ML and VCO₂NL interact. If for example the VCO₂ML is inappropriately decreased, respiratory drive and hence spontaneous effort will increase to maintain the PaCO₂ set point, which may occur at the cost of enormous transpulmonary pressures and risk of further lung injury.¹²⁸ Conversely, if VCO₂ML is excessive, respiratory drive decreases resulting in low respiratory effort which may predispose to muscle atrophy or intermittent apnoea.^129,130 Avoiding excessive or extremely low respiratory drive or effort is therefore increasingly proposed as a goal in acute respiratory failure.¹³¹

Monitoring respiratory drive

The simplest clinical assessment of respiratory drive is to ask the patient. The 11-point modified Borg dyspnoea scale functions similarly to a visual analogue scale¹³² and is well validated in conventional ventilation^133,134 and ECMO^135–137 The Borg score is also readily repeated allowing changes in native lung and membrane function to be tracked. Most ECMO-supported patients are not able to self-report dyspnoea and so this must be inferred by other means. Although there is emerging interest in techniques such as surface EMG of the intercostal muscles,^138,139 these are not widely available and can be affected by electrical noise. Using a gastro-oesophageal catheter, the electrical activity of the diaphragm can be measured which correlates with respiratory drive.^130,140

All ICU ventilators record the P0.1 (pressure reached during a transient valve occlusion at 100ms following the onset of inspiration), which is an acceptable surrogate of respiratory drive during mechanical ventilation.¹⁴¹ At rest, in health a value of −0.5 to −1.5 cmH₂O is obtained and an increasingly negative P0.1 (especially less than -5cmH₂O) correlates with increased drive.¹³¹ However, not all ventilators derive the value in the same manner which reduces the ability to provide absolute limits.^142,143

Monitoring respiratory effort

There is increasing interest how to manage patient self-induced lung injury (P-SILI)^131,144 whereby large spontaneous transpulmonary pressures in injured lungs with regional heterogeneities and high inspiratory effort worsens lung injury, especially as high spontaneous efforts generate a strong dorsal-ventral pressure.¹⁴⁵

Measuring oesophageal pressure (P_es) is the gold standard to quantify inspiratory efforts: changes in P_es reflect changes in the P_pleura, which are proportional to the effort of the respiratory muscles (P_mus). Studies suggests that in patients receiving respiratory support, including ECMO, sustained ΔP_es of around >15 cmH₂O are associated with exhaustion and treatment failure.^29,146–149 Further, the total ΔP_L will include both the patient and ventilator contributions and means that potentially injurious pressure swings can be identified.^95,96,145 The P_mus can be calculated using the ΔP_es and calculation of the elastic recoil pressure of the chest wall-if chest wall elastance has been measured during passive conditions (or alternatively estimated from population data). An oesophageal catheter can also be coupled with gastric manometry to detect and quantify expiratory efforts. We recommend insertion of an oesophageal catheter to measure the ΔP_es,, P_L and P_mus in patients making inspiratory efforts, either routinely or targeted according to non-invasive assessments of effort.

Several non-invasive manoeuvres on any ventilator (panel A Figure 3) can be used intermittently to provide similar information to oesophageal manometry. Although these estimates of P_mus and ΔP_L are imprecise, there is increasing evidence that they are very effective in identifying patients at extremes of effort (high or low).

(1) The pressure muscle index (PMI) is the difference between plateau pressure and peak pressure obtained during an inspiratory hold during spontaneous breathing and correlates with inspiratory effort.¹⁵⁰ A rise of at least 6 cmH₂O correlates with high inspiratory efforts or active expiration. A PMI of <1 cmH₂O correlates with minimal spontaneous efforts and suggests that mechanical support, including ECMO can be weaned.

(2) The occlusion pressure (P_occ) is the maximal negative pressure generated at zero flow during inspiratory efforts against a closed valve (during an expiratory hold). This negative pressure should be inclusive of the change from total PEEP and can be used to estimate P_mus and ΔP_L associated with inspiratory efforts (Table 4).^151–153 The P0.1 can also be used to estimate P_mus and ΔP_L but it is less accurate than the P_occ.¹⁵³ Estimated ΔP_L using P_occ has been validated for a very high ‘excessive’ threshold of ΔP_L >20 in the literature.^151–153 In practice, for patients supported with ECMO we generally target a lower ΔP_L of <10–12 cmH₂O, although the accuracy of P_occ has not yet been validated for this lower target range.

(3) Careful visual assessment of flow curves during assisted ventilation can also be helpful in assessing high inspiratory effort (panel B Figure 3).¹⁵⁴

The benefit of undertaking these measurements is that high or low work can be managed by changing the ECMO and mechanical ventilation to optimally support patients and hopefully reduce the risk of P-SILI.

Patients with respiratory muscle weakness may have dissociated respiratory drive and effort if they are unable to generate the forces demanded by their respiratory centres. Measuring the maximal negative inspiratory force (NIF) the patient can generate either by asking the patient to make maximal effort during an expiratory hold, or in sedated patients maintaining an expiratory hold until the largest swing in P_es or P_aw is obtained can be useful.¹⁵⁵ A NIF of <−15 to −25 cmH₂O has been shown to correlate with successful weaning from mechanical ventilation.¹⁵⁶ However, once a maximal NIF is known, changes in effort can be indexed against their maximum achievable effort to help judge the impact of changes in ECMO or mechanical ventilatory support.

The above tools can help characterise the magnitude of forces that are being generated by spontaneous effort. High spontaneous effort can result in dorsal hyperinflation which can be detected with EIT.^107,145 Pendelluft, intra-tidal gas movement between lung units with different time constants, is much more likely with high inspiratory efforts and can be qualitatively detected with EIT as temporally distinct changes in regional ventilation.^157–160 EIT can be repeated following changes to ECMO or ventilatory support to assess the impact of any changes.

As well as generating regional lung strain, P-SILI can result in accumulation of EVLW due to a transvascular pressure gradient generated by high inspiratory efforts.¹⁴⁴ Acute changes in EVLW lead to small but detectable changes in plasma proteins and cardiac enzymes after a change in respiratory support including changes in the haematocrit and brain natriuretic peptide.^161,162 Dynamic changes in EVLW can also be monitored with serial LUS.¹¹⁶

Multimodal monitoring during VV ECMO

In Figure 4 we outline an approach to multimodal monitoring during VV ECMO. In our practice, this includes a routine recruitment thoracic CT following cannulation, use of low flow PV loops and R:I to decide upon a higher or lower PEEP/mean airway pressure. We target an MP <6–10 J/min with a lower limit where the functional lung volumes are especially low or the lung injury is very heterogenous. During ECMO we advise routinely quantifying the VO₂ML, VCO₂ML, the VCO₂NL and the oxygen challenge test daily. In selected patients, LUS and EIT can be used to optimise and monitor the response to mechanical ventilation settings. Especially at extremes of chest wall elastance, we consider invasive measurement of the P_L very helpful as an ongoing guide to monitoring. As patients improve we consider relaxing sedation and transitioning to an assisted mode of ventilation in order to prevent muscle weakness and hasten recovery. During the spontaneous phase, we routinely monitor the P0.1, P_occ and PMI in all patients and consider oesophageal manometry where persistent high efforts are suspected. The P_occ and P0.1 can be used to estimate the ΔP_L as has been recently described and validated although the P_occ is more accurate.^151,153 In awake patients, self-reported dyspnoea is a helpful adjunct to these measures. Depending on the individual assessment, we use these tools to guide interventions to correct high or low efforts, aiming for a balance of low MP (<6–10 J/min), low P_L (End inspiratory P_L <20 cmH₂O or EstΔPL <22 cmH₂O, measured ΔP_L <10–12 cmH₂O and certainly <20 cmH₂O) and moderate, sustainable inspiratory efforts. During a weaning trial which may lead to risks of P-SILI in the assisted patient, we undertake a careful assessment of these factors at each step of a standardised approach to weaning which we have previously described.^163,164

Figure 4.

An approach to monitoring the native lung during VV ECMO.

Future steps

The increasing interest in personalising ventilation is likely to lead to more complex automated assessments derived from the ventilator being integrated with non-invasive imaging at the bedside over time. New devices under development can aid the assessment of recruitability¹⁶⁵ and overdistension¹⁶⁶ at the bedside by measuring intratidal fluctuations in PaO₂. Other studies have used electroencephalography to directly measure oscillatory neural activity in response to inspiratory loading or deliberate induction of asynchrony.^167–170 Short-term changes in ventilation which increase or decrease mechanical power during ECMO are associated with dynamic variations in blood or broncho-alveolar lavage derived inflammatory cytokines and metabolites including interleukins 1, 6, tissue necrosis factor alpha and soluble receptor for advanced glycation end products.^171–173 This raises the prospect that in the future individual biological signals of harm, rather than epidemiologically based indices of risk could be monitored as respiratory support is changed.

Conclusions

Extracorporeal membrane oxygenation is lifesaving but adds to the complexity of monitoring the critically ill patient with a dual physiology-that of the native and MLs. Understanding pertinent aspects of O₂ and CO₂ carriage and transfer are essential to interpret the information which can be derived from blood gas analysis of different blood compartments and clinical monitoring of different tissues. Assessing the adequacy of overall gas exchange for these patients may be challenging and is best judged by a clinician holistically integrating a range of different inputs. The most important therapeutic effect of VV ECMO is in enabling a reduction in the MP delivered to the injured lung. It is crucial to assess the implications of a change in ventilation for the individual patient’s MP, placing this in the context of their lung mechanics, transpulmonary pressure, functional lung volumes and lung heterogeneity. Spontaneous breathing of patients as they recover has advantages, but it is important to understand that blood gases, tidal volumes and respiratory rates are not sufficient to monitor high efforts, distress and P-SILI. With an understanding of the underlying physiology and the limitations of different tools, multimodal monitoring is key to personalising respiratory support during ECMO.

Footnotes

Acknowledgements

All diagrams were prepared with Biorender

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical statement

Guarantor

NB.

Contributorship

PDC wrote the first manuscript draft and prepared all of the figures. All authors reviewed and edited the final manuscript.

ORCID iDs

Patrick Duncan Collins

Nicholas A Barrett

Appendix

References

Marhong

Telesnicki

Munshi

, et al. Mechanical ventilation during extracorporeal membrane oxygenation. An international survey. Ann Am Thorac Soc 2014; 11: 956–961.

Grasselli

Calfee

Camporota

, et al. ESICM guidelines on acute respiratory distress syndrome: definition, phenotyping and respiratory support strategies. Intensive Care Med. 2023; 49: 727–759. DOI: 10.1007/s00134-023-07050-7

Peek

Mugford

Tiruvoipati

, et al. Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised controlled trial. Lancet. 2009; 374: 1351–1363. DOI: 10.1016/S0140-6736(09)61069-2

Combes

Hajage

Capellier

, et al. Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. N Engl J Med 2018; 378: 1965–1975.

Combes

Peek

Hajage

, et al. ECMO for severe ARDS: systematic review and individual patient data meta-analysis. Intensive Care Med 2020; 46: 2048–2057.

Ficial

Vasques

Zhang

, et al. Physiological basis of extracorporeal membrane oxygenation and extracorporeal carbon dioxide removal in respiratory failure. Membranes 2021; 11: 225.

Arthurs

Sudhakar

. Carbon dioxide transport. Contin Educ Anaesth Crit Care Pain 2005; 5: 207–210.

Gattinoni

Pesenti

Matthay

. Understanding blood gas analysis. Intensive Care Med 2018; 44: 91–93.

Beck

Randolph

Bailey

, et al. Relationship between cardiac output and oxygen consumption during upright cycle exercise in healthy humans. J Appl Physiol Bethesda Md 1985 2006; 101: 1474–1480.

10.

Duncker

Bache

. Regulation of coronary blood flow during exercise. Physiol Rev 2008; 88: 1009–1086.

11.

Gutierrez

Wulf-Gutierrez

Reines

. Monitoring oxygen transport and tissue oxygenation. Curr Opin Anaesthesiol 2004; 17: 107–117.

12.

Harris

Seelye

Squire

. Oxygen consumption during cardiopulmonary bypass with moderate hypothermia in man. Br J Anaesth 1971; 43: 1113–1120.

13.

Manthous

Hall

Olson

, et al. Effect of cooling on oxygen consumption in febrile critically ill patients. Am J Respir Crit Care Med 1995; 151: 10–14.

14.

Y S

, ER

. Respiratory quotients lower than 0.70 in ketogenic diets. Am J Clin Nutr; 33. Epub ahead of print June 1980. DOI: 10.1093/ajcn/33.6.1317.

15.

Gattinoni

Coppola

Camporota

. Physiology of extracorporeal CO2 removal. Intensive Care Med 2022; 48: 1322–1325.

16.

Schmidt

Tachon

Devilliers

, et al. Blood oxygenation and decarboxylation determinants during venovenous ECMO for respiratory failure in adults. Intensive Care Med 2013; 39: 838–846.

17.

Vasques

Sanderson

Correa

, et al. Prevalence and indications for oxygenator circuit replacement in patients receiving venovenous extracorporeal membrane oxygenation. ASAIO J. Publish Ahead of Print. Epub ahead of print 9 May 2023. DOI: 10.1097/MAT.0000000000001977.

18.

Spinelli

Bartlett

. Relationship between hemoglobin concentration and extracorporeal blood flow as determinants of oxygen delivery during venovenous extracorporeal membrane oxygenation: a mathematical model. ASAIO J 2014; 60: 688.

19.

Douglas

Jones

Reed

. Calculation of whole blood CO2 content. J Appl Physiol. 1988; 65: 473–7. DOI: 10.1152/jappl.1988.65.1.473

20.

Barrett

Hart

Camporota

. In-vitro performance of a low flow extracorporeal carbon dioxide removal circuit. Perfusion 2020; 35: 227–235.

21.

Karagiannidis

Strassmann

Brodie

, et al. Impact of membrane lung surface area and blood flow on extracorporeal CO2 removal during severe respiratory acidosis. Intensive Care Med Exp 2017; 5: 34.

22.

Collins

Giosa

Camarda

, et al. Physiological adaptations during weaning from veno-venous extracorporeal membrane oxygenation. Intensive Care Med Exp 2023; 11: 7.

23.

Riley

Cournand

. Ideal alveolar air and the analysis of ventilation-perfusion relationships in the lungs. J Appl Physiol 1949; 1: 825–847.

24.

Epis

Belliato

. Oxygenator performance and artificial-native lung interaction. J Thorac Dis 2018; 10: S596–S605.

25.

Zakhary

Sheldrake

Pellegrino

. Extracorporeal membrane oxygenation and V/Q ratios: an ex vivo analysis of CO₂ clearance within the Maquet Quadrox-iD oxygenator. Perfusion. 2020; 35: 29–33.

26.

Riley

Scott

Schears

. Update on safety equipment for extracorporeal life support (ECLS) circuits. Semin Cardiothorac Vasc Anesth. 2009; 13: 138–145. DOI: 10.1177/1089253209347895

27.

Abrams

Bacchetta

Brodie

. Recirculation in venovenous extracorporeal membrane oxygenation. ASAIO J Am Soc Artif Intern Organs 1992 2015; 61: 115–121.

28.

Darling

Crowell

Searles

. Use of dilutional ultrasound monitoring to detect changes in recirculation during venovenous extracorporeal membrane oxygenation in swine. ASAIO J Am Soc Artif Intern Organs 1992 2006; 52: 522–524.

29.

Lazzari

Romitti

Busana

, et al. End-tidal to arterial Pco₂ ratio as guide to weaning from venovenous extracorporeal membrane oxygenation. Am J Respir Crit Care Med. 2022; 206: 973–980.

30.

Barrett

Hart

Camporota

. In vivo carbon dioxide clearance of a low-flow extracorporeal carbon dioxide removal circuit in patients with acute exacerbations of chronic obstructive pulmonary disease. Perfusion 2020; 35: 436–441.

31.

Montalti

Belliato

Gelsomino

, et al. Continuous monitoring of membrane lung carbon dioxide removal during ECMO: experimental testing of a new volumetric capnometer. Perfusion 2019; 34: 538–543.

32.

Sun

Kaesler

Fernando

, et al. CO2 clearance by membrane lungs. Perfusion 2018; 33: 249–253.

33.

Bellancini

Cercenelli

Severi

, et al. Development of a CO2 sensor for extracorporeal life support applications. Sensors 2020; 20: 3613.

34.

Castagna

Zanella

Scaravilli

, et al. Effects on membrane lung gas exchange of an intermittent high gas flow recruitment maneuver: preliminary data in veno-venous ECMO patients. J Artif Organs 2015; 18: 213–219.

35.

Marshall

. A model for hypoxic constriction of the pulmonary circulation. J Appl Physiol 1988; 64: 68–77.

36.

Tarry

Powell

. Hypoxic pulmonary vasoconstriction. BJA Educ 2017; 17: 208–213.

37.

Mathews

Singh

. Cardiac output monitoring. Ann Card Anaesth 2008; 11: 56.

38.

Loosen

Conrad

Hagman

, et al. Transpulmonary thermodilution in patients treated with veno-venous extracorporeal membrane oxygenation. Ann Intensive Care 2021; 11: 101.

39.

Douflé

Roscoe

Billia

, et al. Echocardiography for adult patients supported with extracorporeal membrane oxygenation. Crit Care 2015; 19: 326.

40.

Lazzeri

Gensini

Peris

. The assessment of cardiac function in veno-venous extracorporeal membrane oxygenation: the emerging role of bedside echocardiography. Heart Lung Vessels 2015; 7: 99–100.

41.

Zhang

Wang

Shi

, et al. Cardiac output measurements via echocardiography versus thermodilution: a systematic review and meta-analysis. PLoS One 2019; 14: e0222105.

42.

Gadhinglajkar

Sreedhar

Unnikrishnan

. Oxygraphy: an unexplored perioperative monitoring modality. J Clin Monit Comput 2009; 23: 131–135.

43.

Weingarten

. Respiratory monitoring of carbon dioxide and oxygen: a ten-year perspective. J Clin Monit 1990; 6: 217–225.

44.

Delsoglio

Achamrah

Berger

, et al. Indirect calorimetry in clinical practice. J Clin Med 2019; 8: 1387.

45.

Hartley

Sanderson

Vasques

, et al. Prediction of readiness to decannulation from venovenous extracorporeal membrane oxygenation. Perfusion 2020; 35: 57–64.

46.

Misselbrook

Kanji

Thiara

, et al. Prediction of successful veno-venous extracorporeal life support liberation using the oxygen challenge test. Artif Organs 2023; 47: 180–186.

47.

Bonifazi

Romitti

Busana

, et al. End-tidal to arterial PCO2 ratio: a bedside meter of the overall gas exchanger performance. Intensive Care Med Exp 2021; 9: 21.

48.

Sinha

Calfee

Beitler

, et al. Physiologic analysis and clinical performance of the ventilatory ratio in acute respiratory distress syndrome. Am J Respir Crit Care Med 2019; 199: 333–341.

49.

Lecompte-Osorio

Pearson

Pieroni

, et al. Bedside estimates of dead space using end-tidal CO2 are independently associated with mortality in ARDS. Crit Care Lond Engl 2021; 25: 333.

50.

Kreit

JW.

Volume capnography in the intensive care unit: physiological principles, measurements, and calculations. Ann Am Thorac Soc. 2019; 16: 291–300. DOI: 10.1513/AnnalsATS.201807-501CME

51.

Gille

Laveneziana

. Cardiopulmonary exercise testing in interstitial lung diseases and the value of ventilatory efficiency. Eur Respir Rev 2021; 30: 200355.

52.

McClave

Martindale

Kiraly

. The use of indirect calorimetry in the intensive care unit. Curr Opin Clin Nutr Metab Care 2013; 16: 202.

53.

Bear

Smith

Barrett

. Nutrition support in adult patients receiving extracorporeal membrane oxygenation. Nutr Clin Pract 2018; 33: 738–746.

54.

Schjørring

Klitgaard

Perner

, et al. Lower or higher oxygenation targets for acute hypoxemic respiratory failure. N Engl J Med 2021; 384: 1301–1311.

55.

Gelissen

de Grooth

H-J

Smulders

, et al. Effect of low-normal vs high-normal oxygenation targets on organ dysfunction in critically ill patients: a randomized clinical trial. JAMA 2021; 326: 940–948.

56.

Barrot

Asfar

Mauny

, et al. Liberal or conservative oxygen therapy for acute respiratory distress syndrome. N Engl J Med 2020; 382: 999–1008.

57.

Conservative oxygen therapy during mechanical ventilation in the ICU. N Engl J Med 2020; 382: 989–998.

58.

von Bahr

Kalzén

Hultman

, et al. Long-term cognitive outcome and brain imaging in adults after extracorporeal membrane oxygenation. Crit Care Med 2018; 46: e351–e358.

59.

Marsh

Leach

Blane

, et al. Long-term cognitive and psychiatric outcomes of acute respiratory distress syndrome managed with extracorporeal membrane oxygenation. Respir Med 2021; 183: 106419.

60.

Sjoding

Dickson

Iwashyna

, et al. Racial bias in pulse oximetry measurement. N Engl J Med 2020; 383: 2477–2478.

61.

Valbuena

VSM

Barbaro

Claar

, et al. Racial bias in pulse oximetry measurement among patients about to undergo extracorporeal membrane oxygenation in 2019-2020: a retrospective cohort study. Chest 2022; 161: 971–978.

62.

Barker

Wilson

. Racial effects on Masimo pulse oximetry: a laboratory study. J Clin Monit Comput 2023; 37: 567–574.

63.

Nisar

Gibson

Sokolovic

, et al. Pulse oximetry is unreliable in patients on veno-venous extracorporeal membrane oxygenation caused by unrecognized carboxyhemoglobinemia. ASAIO J 2020; 66: 1105.

64.

Duke

Killick

. Pulmonary vasomotor responses of isolated perfused cat lungs to anoxia. J Physiol 1952; 117: 303–316.

65.

Hansen

Hyttel-Sørensen

Jakobsen

, et al. Cerebral Near-Infrared Spectroscopy Monitoring (NIRS) in children and adults: a systematic review with meta-analysis. Pediatr Res 2022: 1–12.

66.

Wood

Jacobson

Maslove

, et al. The physiological determinants of near-infrared spectroscopy-derived regional cerebral oxygenation in critically ill adults. Intensive Care Med Exp 2019; 7: 23.

67.

Hunt

Clark

Whitman

, et al. The use of cerebral NIRS monitoring to identify acute brain injury in patients with VA-ECMO. J Intensive Care Med 2021; 36: 1403–1409.

68.

Wong

Smith

Pitcher

, et al. Cerebral and lower limb near-infrared spectroscopy in adults on extracorporeal membrane oxygenation. Artif Organs 2012; 36: 659–667.

69.

Deschamps

Hall

Grocott

, et al. Cerebral oximetry monitoring to maintain normal cerebral oxygen saturation during high-risk cardiac surgery: a randomized controlled feasibility trial. Anesthesiology 2016; 124: 826–836.

70.

Cavayas

Munshi

Del Sorbo

, et al. The early change in PaCO2 after extracorporeal membrane oxygenation initiation is associated with neurological complications. Am J Respir Crit Care Med 2020; 201: 1525–1535.

71.

Gattinoni

Tonetti

Cressoni

, et al. Ventilator-related causes of lung injury: the mechanical power. Intensive Care Med 2016; 42: 1567–1575.

72.

Chiumello

Gotti

Guanziroli

, et al. Bedside calculation of mechanical power during volume- and pressure-controlled mechanical ventilation. Crit Care 2020; 24: 417.

73.

Giosa

Busana

Pasticci

, et al. Mechanical power at a glance: a simple surrogate for volume-controlled ventilation. Intensive Care Med Exp 2019; 7: 61.

74.

Vassalli

Pasticci

Romitti

, et al.

Does iso-mechanical power lead to iso-lung damage?

Anesthesiology 2020; 132: 1126–1137.

75.

Cressoni

Gotti

Chiurazzi

, et al. Mechanical power and development of ventilator-induced lung injury. Anesthesiology 2016; 124: 1100–1108.

76.

Amato

MBP

Meade

Slutsky

, et al. Driving pressure and survival in the acute respiratory distress syndrome. N Engl J Med 2015; 372: 747–755.

77.

Romitti

Busana

Palumbo

, et al. Mechanical power thresholds during mechanical ventilation: an experimental study. Physiol Rep. 2022; 10: e15225. DOI: 10.14814/phy2.15225

78.

PROVE Network Investigators Serpa Neto

Deliberato

, et al. Mechanical power of ventilation is associated with mortality in critically ill patients: an analysis of patients in two observational cohorts. Intensive Care Med 2018; 44: 1914–1922.

79.

Costa

ELV

Slutsky

Brochard

, et al. Ventilatory variables and mechanical power in patients with acute respiratory distress syndrome. Am J Respir Crit Care Med 2021; 204: 303–311.

80.

Chiu

L-C

Lin

S-W

Chuang

L-P

, et al. Mechanical power during extracorporeal membrane oxygenation and hospital mortality in patients with acute respiratory distress syndrome. Crit Care 2021; 25: 13.

81.

Gattarello

Coppola

Chiodaroli

, et al. Mechanical power ratio and respiratory treatment escalation in COVID-19 Pneumonia: a secondary analysis of a prospectively enrolled cohort. Anesthesiology 2023; 138: 289–298.

82.

Ghiani

Paderewska

Walcher

, et al. Mechanical power normalized to lung-thorax compliance predicts prolonged ventilation weaning failure: a prospective study. BMC Pulm Med 2021; 21: 202.

83.

Ghiani

Paderewska

Walcher

, et al. Mechanical power normalized to lung-thorax compliance indicates weaning readiness in prolonged ventilated patients. Sci Rep 2022; 12: 6.

84.

Hong

Chen

Pan

, et al. Individualized mechanical power-based ventilation strategy for acute respiratory failure formalized by finite mixture modeling and dynamic treatment regimen. EClinicalMedicine. 2021; 36: 100898. DOI: 10.1016/j.eclinm.2021.100898

85.

Maggiore

Richard

J-C

Brochard

. What has been learnt from P/V curves in patients with acute lung injury/acute respiratory distress syndrome. Eur Respir J 2003; 22: 22s–26s.

86.

Chiumello

Arnal

J-M

Umbrello

, et al. Hysteresis and lung recruitment in acute respiratory distress syndrome patients: a CT scan study. Crit Care Med 2020; 48: 1494–1502.

87.

JAMA. Effect of lung recruitment and titrated positive end-expiratory pressure (PEEP) vs low peep on mortality in patients with acute respiratory distress syndrome: a randomized clinical trial. 2017; 318: 1335–1345. DOI: 10.1001/jama.2017.14171

88.

Escolar

. Lung hysteresis: a morphological view. Histol Histopathol 2004; 19: 159–166.

89.

Nakayama

Bunya

Katayama

, et al. Correlation between the hysteresis of the pressure–volume curve and the recruitment-to-inflation ratio in patients with coronavirus disease 2019. Ann Intensive Care 2022; 12: 106.

90.

Chen

Del Sorbo

Grieco

, et al. Airway closure in acute respiratory distress syndrome: an underestimated and misinterpreted phenomenon. Am J Respir Crit Care Med 2018; 197: 132–136.

91.

Chen

Del Sorbo

Grieco

, et al. Potential for lung recruitment estimated by the recruitment-to-inflation ratio in acute respiratory distress syndrome. a clinical trial. Am J Respir Crit Care Med 2020; 201: 178–187.

92.

Grasso

Terragni

Mascia

, et al. Airway pressure-time curve profile (stress index) detects tidal recruitment/hyperinflation in experimental acute lung injury. Crit Care Med 2004; 32: 1018–1027.

93.

Chiumello

Gattinoni

. Stress index in presence of pleural effusion: does it have any meaning? Intensive Care Med 2011; 37: 561–563.

94.

Formenti

Graf

Santos

, et al. Non-pulmonary factors strongly influence the stress index. Intensive Care Med. 2011; 37: 594–600. DOI: 10.1007/s00134-011-2133-4

95.

Grieco

Chen

Brochard

. Transpulmonary pressure: importance and limits. Ann Transl Med 2017; 5: 285.

96.

Gattinoni

Giosa

Bonifazi

, et al. Targeting transpulmonary pressure to prevent ventilator-induced lung injury. Expert Rev Respir Med 2019; 13: 737–746.

97.

Yoshida

Amato

MBP

Grieco

, et al. Esophageal manometry and regional transpulmonary pressure in lung injury. Am J Respir Crit Care Med 2018; 197: 1018–1026.

98.

Kassis

Talmor

. Clinical application of esophageal manometry: how I do it. Crit Care. 2021; 25: 6. DOI: 10.1186/s13054-020-03453-w

99.

Pham

Telias

Beitler

. Esophageal Manometry. Respir Care 2020; 65: 772–792.

100.

Bellani

Amigoni

Pesenti

. Positron emission tomography in ARDS: a new look at an old syndrome. Minerva Anestesiol 2011; 77: 439–447.

101.

Piraino

. Lung volume measurement and ventilation distribution during invasive mechanical ventilation. Respir Care 2020; 65: 760–771.

102.

Gattinoni

Caironi

Pelosi

, et al. What has computed tomography taught us about the acute respiratory distress syndrome? Am J Respir Crit Care Med. 2001; 164(9): 1701–11. DOI: 10.1164/ajrccm.164.9.2103121

103.

Barbas

CSV

. Thoracic computed tomography to assess ARDS and COVID-19 lungs. Front Physiol 2022; 13: 829534.

104.

Chiumello

Busana

Coppola

, et al. Physiological and quantitative CT-scan characterization of COVID-19 and typical ARDS: a matched cohort study. Intensive Care Med 2020; 46: 2187–2196.

105.

Camporota

Caricola

Bartolomeo

, et al. Lung recruitability in severe acute respiratory distress syndrome requiring extracorporeal membrane oxygenation. Crit Care Med 2019; 47: 1177–1183.

106.

Herrmann

Busana

Cressoni

, et al. Using artificial intelligence for automatic segmentation of CT lung images in acute respiratory distress syndrome. Front Physiol 2021; 12: 676118.

107.

Tomicic

Cornejo

. Lung monitoring with electrical impedance tomography: technical considerations and clinical applications. J Thorac Dis 2019; 11: 3122–3135.

108.

Blankman

Shono

Hermans

BJM

, et al. Detection of optimal PEEP for equal distribution of tidal volume by volumetric capnography and electrical impedance tomography during decreasing levels of PEEP in post cardiac-surgery patients. Br J Anaesth 2016; 116: 862–869.

109.

Becher

Buchholz

Hassel

, et al. Individualization of PEEP and tidal volume in ARDS patients with electrical impedance tomography: a pilot feasibility study. Ann Intensive Care 2021; 11: 89.

110.

Franchineau

Bréchot

Lebreton

, et al. Bedside contribution of electrical impedance tomography to setting positive end-expiratory pressure for extracorporeal membrane oxygenation–treated patients with severe acute respiratory distress syndrome. Am J Respir Crit Care Med 2017; 196: 447–457.

111.

Soulé

Crognier

Puel

, et al. Assessment of electrical impedance tomography to set optimal positive end-expiratory pressure for venoarterial extracorporeal membrane oxygenation-treated patients. Crit Care Med 2021; 49: 923–933.

112.

Bikker

Leonhardt

Bakker

, et al. Lung volume calculated from electrical impedance tomography in ICU patients at different PEEP levels. Intensive Care Med 2009; 35: 1362–1367.

113.

Hinz

Hahn

Neumann

, et al. End-expiratory lung impedance change enables bedside monitoring of end-expiratory lung volume change. Intensive Care Med 2003; 29: 37–43.

114.

Chiumello

Umbrello

Sferrazza Papa

, et al. Global and regional diagnostic accuracy of lung ultrasound compared to ct in patients with acute respiratory distress syndrome. Crit Care Med 2019; 47: 1599–1606.

115.

Zhao

Jiang

, et al. Prognostic value of extravascular lung water assessed with lung ultrasound score by chest sonography in patients with acute respiratory distress syndrome. BMC Pulm Med 2015; 15: 98.

116.

Picano

Pellikka

. Ultrasound of extravascular lung water: a new standard for pulmonary congestion. Eur Heart J 2016; 37: 2097–2104.

117.

Neuteboom

Heldeweg

Pisani

, et al. Assessing extravascular lung water in critically Ill patients using lung ultrasound: a systematic review on methodological aspects in diagnostic accuracy studies. Ultrasound Med Biol 2020; 46: 1557–1564.

118.

Mongodi

De Luca

Colombo

, et al. Quantitative lung ultrasound: technical aspects and clinical applications. Anesthesiology 2021; 134: 949–965.

119.

Conrad

Loosen

Boesing

, et al. Effects of changes in veno-venous extracorporeal membrane oxygenation blood flow on the measurement of intrathoracic blood volume and extravascular lung water index: a prospective interventional study. J Clin Monit Comput 2023; 37: 599–607.

120.

Bouhemad

Brisson

Le-Guen

, et al. Bedside ultrasound assessment of positive end-expiratory pressure–induced lung recruitment. Am J Respir Crit Care Med 2011; 183: 341–347.

121.

Baloescu

Chen

Raju

, et al. 19 Automated Quantification Of B-lines in lung ultrasound on COVID-19 patients. Ann Emerg Med 2021; 78: S9–S10.

122.

Tan

GFL

Liu

, et al. Automated lung ultrasound image assessment using artificial intelligence to identify fluid overload in dialysis patients. BMC Nephrol 2022; 23: 410.

123.

La Salvia

Secco

Torti

, et al. Deep learning and lung ultrasound for Covid-19 pneumonia detection and severity classification. Comput Biol Med 2021; 136: 104742.

124.

Steinberg

Pasticci

Busana

, et al. Lung ultrasound and electrical impedance tomography during ventilator-induced lung injury. Crit Care Med 2022; 50: e630–e637.

125.

Vaporidi

Akoumianaki

Telias

, et al. Respiratory drive in critically Ill patients. pathophysiology and clinical implications. Am J Respir Crit Care Med 2020; 201: 20–32.

126.

Spinelli

Mauri

Beitler

, et al. Respiratory drive in the acute respiratory distress syndrome: pathophysiology, monitoring, and therapeutic interventions. Intensive Care Med 2020; 46: 606–618.

127.

Forster

Haouzi

Dempsey

. Control of breathing during exercise. In: Prakash

(ed) Comprehensive Physiology. Wiley, pp. 743–777.

128.

Mauri

Langer

Zanella

, et al. Extremely high transpulmonary pressure in a spontaneously breathing patient with early severe ARDS on ECMO. Intensive Care Med 2016; 42: 2101–2103.

129.

Shi

Z-H

Jonkman

de Vries

, et al. Expiratory muscle dysfunction in critically ill patients: towards improved understanding. Intensive Care Med 2019; 45: 1061–1071.

130.

Dres

Demoule

. Monitoring diaphragm function in the ICU. Curr Opin Crit Care 2020; 26: 18–25.

131.

Goligher

Jonkman

Dianti

, et al. Clinical strategies for implementing lung and diaphragm-protective ventilation: avoiding insufficient and excessive effort. Intensive Care Med 2020; 46: 2314–2326.

132.

Powers

Bennett

. Measurement of dyspnea in patients treated with mechanical ventilation. Am J Crit Care 1999; 8: 254–261.

133.

Dunn

Nelson

Hubmayr

. The control of breathing during weaning from mechanical ventilation. Chest 1991; 100: 754–761.

134.

Vitacca

Bianchi

Zanotti

, et al. Assessment of physiologic variables and subjective comfort under different levels of pressure support ventilation. Chest 2004; 126: 851–859.

135.

Cucchi

Mariani

De Piero

, et al. Awake extracorporeal life support and physiotherapy in adult patients: a systematic review of the literature. Perfusion 2023; 38: 939–958.

136.

Hayes

Holland

Pellegrino

, et al. Early rehabilitation during extracorporeal membrane oxygenation has minimal impact on physiological parameters: a pilot randomised controlled trial. Aust Crit Care 2021; 34: 217–225.

137.

Pisani

Fasano

Corcione

, et al. Effects of extracorporeal CO2 removal on inspiratory effort and respiratory pattern in patients who fail weaning from mechanical ventilation. Am J Respir Crit Care Med 2015; 192: 1392–1394.

138.

Suh

E-S

D’Cruz

Ramsay

, et al. Parasternal electromyography as a surrogate measure of neural respiratory drive: practical application and validity of surface and implanted fine wire methods. Respir Physiol Neurobiol 2021; 287: 103602.

139.

D’Cruz

Suh

E-S

Kaltsakas

, et al. Home parasternal electromyography tracks patient-reported and physiological measures of recovery from severe COPD exacerbation. ERJ Open Res 2021; 7: 00709–02020.

140.

Piquilloud

Beloncle

Richard

J-CM

, et al. Information conveyed by electrical diaphragmatic activity during unstressed, stressed and assisted spontaneous breathing: a physiological study. Ann Intensive Care 2019; 9: 89.

141.

Sato

Hasegawa

Hamahata

, et al. The predictive value of airway occlusion pressure at 100 msec (P0.1) on successful weaning from mechanical ventilation: a systematic review and meta-analysis. J Crit Care 2021; 63: 124–132.

142.

Kallet

Phillips

Summers

, et al. Expiratory pause maneuver to assess inspiratory muscle pressure during assisted mechanical ventilation: a bench study. Respir Care 2021; 66: 1649–1656.

143.

Daoud

Katigbak

Ottochian

. Accuracy of the ventilator automated displayed respiratory mechanics in passive and active breathing conditions: a bench study. Respir Care 2019; 64: 1555–1560.

144.

Sklienka

Frelich

Burša

. Patient self-inflicted lung injury-a narrative review of pathophysiology, early recognition, and management options. J Pers Med 2023; 13: 593.

145.

Yoshida

Fujino

Amato

MBP

, et al. Fifty years of research in ARDS. Spontaneous breathing during mechanical ventilation. risks, mechanisms, and management. Am J Respir Crit Care Med 2017; 195: 985–992.

146.

Al-Fares

Ferguson

, et al. Achieving safe liberation during weaning from VV-ECMO in patients with severe ARDS. Chest 2021; 160: 1704–1713.

147.

Shimatani

Kyogoku

Ito

, et al. Fundamental concepts and the latest evidence for esophageal pressure monitoring. J Intensive Care 2023; 11: 22.

148.

Tonelli

Fantini

Tabbì

, et al. Early inspiratory effort assessment by esophageal manometry predicts noninvasive ventilation outcome in de novo respiratory failure. a pilot study. Am J Respir Crit Care Med 2020; 202: 558–567.

149.

Vaporidi

Soundoulounaki

Papadakis

, et al. Esophageal and transdiaphragmatic pressure swings as indices of inspiratory effort. Respir Physiol Neurobiol 2021; 284: 103561.

150.

Foti

Cereda

Banfi

, et al. End-inspiratory airway occlusion: a method to assess the pressure developed by inspiratory muscles in patients with acute lung injury undergoing pressure support. Am J Respir Crit Care Med 1997; 156: 1210–1216.

151.

Bertoni

Telias

Urner

, et al. A novel non-invasive method to detect excessively high respiratory effort and dynamic transpulmonary driving pressure during mechanical ventilation. Crit Care 2019; 23: 346.

152.

Roesthuis

van den Berg

van der Hoeven

. Non-invasive method to detect high respiratory effort and transpulmonary driving pressures in COVID-19 patients during mechanical ventilation. Ann Intensive Care 2021; 11: 26.

153.

de Vries

Tuinman

Jonkman

, et al. Performance of noninvasive airway occlusion maneuvers to assess lung stress and diaphragm effort in mechanically ventilated critically ill patients. Anesthesiology 2023; 138: 274–288.

154.

Albani

Pisani

Ciabatti

, et al. Flow index: a novel, non-invasive, continuous, quantitative method to evaluate patient inspiratory effort during pressure support ventilation. Crit Care 2021; 25: 196.

155.

de Vries

Jonkman

Shi

Z-H

, et al. Assessing breathing effort in mechanical ventilation: physiology and clinical implications. Ann Transl Med. 2018; 6: 387. DOI: 10.21037/atm.2018.05.53

156.

Yang

Tobin

. A prospective study of indexes predicting the outcome of trials of weaning from mechanical ventilation. N Engl J Med 1991; 324: 1445–1450.

157.

Sang

Zhao

Yun

P-J

, et al. Qualitative and quantitative assessment of pendelluft: a simple method based on electrical impedance tomography. Ann Transl Med. 2020; 8: 1216. DOI: 10.21037/atm-20-4182

158.

Yoshida

Torsani

Gomes

, et al. Spontaneous effort causes occult pendelluft during mechanical ventilation. Am J Respir Crit Care Med 2013; 188: 1420–1427.

159.

Menga

Delle Cese

Rosà

, et al. Respective effects of helmet pressure support, continuous positive airway pressure, and nasal high-flow in hypoxemic respiratory failure: a randomized crossover clinical trial. Am J Respir Crit Care Med 2023; 207: 1310–1323.

160.

Moon

Huh

Hong

S-B

, et al. Dynamic inhomogeneity of aeration along the vertical axis of the lung may predict weaning failure regardless of diaphragm dysfunction. J Crit Care 2021; 65: 186–191.

161.

Anguel

Monnet

Osman

, et al. Increase in plasma protein concentration for diagnosing weaning-induced pulmonary oedema. Intensive Care Med 2008; 34: 1231–1238.

162.

Deschamps

Andersen

Webber

, et al. Brain natriuretic peptide to predict successful liberation from mechanical ventilation in critically ill patients: a systematic review and meta-analysis. Crit Care 2020; 24: 213.

163.

Collins

Giosa

Camarda

, et al. Physiological adaptations during weaning from veno-venous extracorporeal membrane oxygenation. Intensive Care Med Exp 2023; 11: 7.

164.

Vasques

Romitti

Gattinoni

, et al. How I wean patients from veno-venous extra-corporeal membrane oxygenation. Crit Care. 2019; 23: 316. DOI: 10.1186/s13054-019-2592–5

165.

Tran

Crockett

Cronin

, et al. Bedside monitoring of lung volume available for gas exchange. Intensive Care Med Exp 2021; 9: 3.

166.

Cronin

Crockett

Perchiazzi

, et al. Intra-tidal PaO2 oscillations associated with mechanical ventilation: a pilot study to identify discrete morphologies in a porcine model. Intensive Care Med Exp 2023; 11: 60.

167.

Raux

Ray

Prella

, et al. Cerebral cortex activation during experimentally induced ventilator fighting in normal humans receiving noninvasive mechanical ventilation. Anesthesiology 2007; 107: 746–755.

168.

Raux

Straus

Redolfi

, et al. Electroencephalographic evidence for pre-motor cortex activation during inspiratory loading in humans. J Physiol 2007; 578: 569–578.

169.

Dubois

Chenivesse

Raux

, et al. Neurophysiological evidence for a cortical contribution to the wakefulness-related drive to breathe explaining hypocapnia-resistant ventilation in humans. J Neurosci 2016; 36: 10673–10682.

170.

Hudson

Niérat

M-C

Raux

, et al. The relationship between respiratory-related premotor potentials and small perturbations in ventilation. Front Physiol. 2018; 9: 621. DOI: 10.3389/fphys.2018.00621

171.

Stüber

Wrigge

Schroeder

, et al. Kinetic and reversibility of mechanical ventilation-associated pulmonary and systemic inflammatory response in patients with acute lung injury. Intensive Care Med 2002; 28: 834–841.

172.

Del Sorbo

Goffi

Tomlinson

, et al. Effect of driving pressure change during extracorporeal membrane oxygenation in adults with acute respiratory distress syndrome: a randomized crossover physiologic study. Crit Care Med. 2020; 48: 1771–1778.

173.

Rozencwajg

Guihot

Franchineau

, et al. Ultra-protective ventilation reduces biotrauma in patients on venovenous extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. Crit Care Med 2019; 47: 1505–1512.

State of the art: Monitoring of the respiratory system during veno-venous extracorporeal membrane oxygenation

Abstract

Keywords

Introduction

Gas exchange of the native and membrane lung

O2 consumption and CO2 clearance requirements

Gas partial pressures and their relation to the content of blood

Gas transfer across the native lung (without ECMO)

Gas transfer across the membrane lung (ML)

Monitoring gas exchange during VV ECMO

Monitoring the membrane lung

Evaluating the VO2ML

Evaluating the VCO2ML

Monitoring native lung function during VV ECMO

Challenges in measuring cardiac output

Monitoring native lung oxygenation capacity and the VO2NL

Monitoring VCO2NL, assessing the dead space and ventilatory efficiency of the native lung

Additional benefits of VO2 and VCO2 measurement during ECMO

Targets for gas exchange

Daily measurements to assess the native and membrane lungs

Monitoring to minimise native lung injury and respiratory distress

Mechanical power and ventilation

Calculating mechanical power at the bedside

Ventilator assessments of recruitability and overdistension

Invasive assessment of transpulmonary pressure

Imaging to assess recruitability, overdistension and heterogeneity

Monitoring spontaneous breathing during VV ECMO

Monitoring respiratory drive

Monitoring respiratory effort

Multimodal monitoring during VV ECMO

Future steps

Conclusions

Footnotes

Acknowledgements

Declaration of conflicting interests

Funding

Ethical statement

Guarantor

Contributorship

ORCID iDs

Appendix

References

O₂ consumption and CO₂ clearance requirements

Evaluating the VO₂ML

Evaluating the VCO₂ML

Monitoring native lung oxygenation capacity and the VO₂NL

Monitoring VCO₂NL, assessing the dead space and ventilatory efficiency of the native lung

Additional benefits of VO₂ and VCO₂ measurement during ECMO