Show Summary Details
Page of

Respiratory monitoring 

Respiratory monitoring
Respiratory monitoring

Carl Waldmann

, Andrew Rhodes

, Neil Soni

, and Jonathan Handy

Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2022. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

date: 21 May 2022

Chapter contents

Pulmonary function tests in critical illness

Monitoring mechanical ventilation

The latest generation of mechanical ventilators has obtained an unprecedented level of sophistication. Many of them incorporate multiple microprocessors, which allow the monitoring and analysis of many variables. The data acquired are particularly useful for deciding on the most appropriate ventilator strategy for each patient to minimize the appearance of ventilator-induced lung injury (VILI).

Lung volumes

Automated functional residual capacity (FRC) measurements are useful for assessing the aerated lung available for ventilation. Nitrogen washout requires accurate measurement of inspired and expired oxygen, CO2, and changes in alveolar fraction of nitrogen through a metabolic monitoring module. Nonetheless, it requires relatively stable conditions and moderate fractional inspired oxygen concentration (FiO2: 0.4–0.65). FRC is often reduced in mechanically ventilated patients for several reasons, such as supine position, use of sedatives, high abdominal pressure, or underlying lung disease.

Electrical impedance tomography (EIT) is a relatively recent technology for monitoring lung function in critically ill patients on mechanical ventilation. Unlike other techniques for measuring end-expiratory lung volume, EIT is a non-invasive approach that is easy to use in mechanically ventilated patients and provides bedside lung imaging. It has been shown that a positive end-expiratory pressure (PEEP)-induced increase in end-expiratory lung volume is accompanied by a proportional increase in end-expiratory lung impedance. However, the use of end-expiratory lung impedance results may slightly underestimate the changes in lung volume. Therefore, EIT may be useful for setting the ventilator to avoid alveolar overdistension and de-recruitment.

Vital capacity (VC) is the maximum volume exhaled after maximum inspiration and can be measured during forced exhalation (FVC), thus reflecting respiratory muscle strength. VC should be measured with the patient upright rather than supine because certain conditions, such as diaphragmatic paralysis, have demonstrated positional dependence. Therefore, as critically ill patients cannot cooperate to create a maximum inspiratory effort, VC has been shown to be an unreliable measure of strength. Fortunately, a one-way valve can be used to calculate VC by the sum of the measure of inspiratory capacity during tidal breathing efforts and expiratory reserve volume.

Analysis of ventilator waveforms

Airway pressures

The airway pressure (PAW) curve has a well-known morphology during volume-controlled ventilation and constant flow inflation, which offers useful information via visual analysis during simple respiratory manoeuvres. In paralysed patients, these manoeuvres should be performed to obtain valid measurements. Moreover, in hold manoeuvres, a 4-s duration is mandatory.

Peak pressure (Pmax) is the maximum pressure measured by the ventilator and represents the pressure in the major airways. It thus provides a strong reflection of airway resistance. When an airway occlusion manoeuvre is applied at end-inspiration, it results in a sudden drop in pressure (P1) followed by a pause/plateau pressure (P2). This phenomenon is the consequence of the equilibration between mouth pressure and alveolar pressure and it estimates the small airway and alveoli pressure.

In contrast, during an airway occlusion manoeuvre at the end of expiration, a redistribution of the air inside the lung occurs, leading to the equalization of the different alveoli pressures (close to zero at the end of a normal expiration). In the presence of high resistance to expiratory flows or short expiratory times, the expiratory process may be incomplete (represented by a persistence of flow at the end of expiration in the flow-time waveform), resulting in a positive end-expiratory alveolar pressure (PEEPi) or AutoPEEP and volume trapping. This leads to increased work of breathing failure to trigger the ventilator and a higher risk of VILI.

Compliance and resistance

The compliance (Crs) and the resistance of the respiratory system (Rrs) in paralysed patients can be assessed during constant flow inflation and volume-controlled ventilation with the end-inspiratory occlusion technique. It may provide valuable information about the evolution of the ongoing disease, thus allowing the optimization of ventilatory settings to minimize VILI. It can also help to monitor treatment response.

Crs is a measure of distensibility and reflects the ability of the lung to stretch and expand. It can be defined as the change in volume at any given applied pressure. Transpulmonary pressure is defined as airway pressure minus pleural pressure. When there is no flow of air and the patient is paralysed, the transpulmonary pressure is the P2 or alveolar pressure (Palv) minus PEEP. Therefore, static compliance (the no-flow condition) is measured as Vt/Palv – (PEEP + PEEPi). Normal values for Crs are >100 ml/cmH2O but in lung diseases such as acute respiratory distress syndrome (ARDS) they may fall below 25 ml/cmH2O.

The Poiseuille law defines resistance as the ratio between a gradient of pressure and flow. According to the values used, different subcategories of airway resistances can be obtained:

Total resistances (Rmax):(PmaxP2)

Airway flowdependent resistance (RAW): (PmaxP1)

Pressure gradient or additional lung resistance value: (P1 – P2); reflects the pendelluft phenomenon and viscoelastic lung behaviour.

Normal RAW values in spontaneously breathing patients are <2.5 cmH2O/l/s and may rise to ~4 in ventilated patients with healthy lungs. Resistance measures should be considered in pathologies with high airway resistance such as obstructive airway disease or asthma, where resistances are >20 cmH2O/l/s, to establish the severity of illness and assessment treatment response. Patients with cardiogenic pulmonary oedema and ARDS may also present a moderate increase in RAW (>0 and >15 cmH2O/l/s, respectively).

Ventilator loops

The interpretation of ventilatory curves facilitates the optimization of ventilator settings and early detection of pathological changes. For example, pressure–volume loops may be helpful in detecting changes in compliance or inadequate trigger sensitivity. The presence of excessive secretions, air leak, or air trapping can be identified by flow–volume loops.

Monitoring the weaning process

Despite the utility of the classical clinical criteria such as adequate level of consciousness, correct oxygenation, or haemodynamic stability, many tests are also available for clinicians taking weaning-related decisions at the bedside. Successful weaning predictors are measurable objective variables that can be used to estimate whether a patient is likely to tolerate the withdrawal of mechanical ventilation. Some of the most accurate predictors are listed here:

  • The rapid shallow breathing index is the ratio of respiratory rate to tidal volume (f/Vt in litres). Patients who cannot tolerate spontaneous breathing tend to breathe rapidly (high frequency) and shallowly (low Vt). Thus, they generally have a high rapid shallow breathing index (>105). In contrast, a value <80 predicts a high likelihood of weaning success.

  • Maximal inspiratory pressure (Pimax) is a measure of the strength of the inspiratory muscles. Values of –30 cmH2O or less may be associated with successful weaning.

  • P0.1 occlusion pressure (airway pressure generated in the first 0.1 s of the inspiratory effort against an occluded airway) estimates the respiratory drive. In normal subjects, P0.1 is less than 2 cmH2O. Values greater than 4 cmH2O predict weaning failure. The predictive accuracy of the P0.1 appears to increase significantly when the airway occlusion pressure is normalized for maximal inspiratory pressure (P0.1/maximal inspiratory pressure).

Blood gas analysis

Analysis of blood gases provides valuable data for diagnosing respiratory and metabolic disturbances and for assessing the effect of therapeutic interventions. The following data can be obtained via blood analysis:

  • Partial arterial pressure of oxygen and carbon dioxide (PaO2, PaCO2) measurement can identify arterial hypoxaemia and hyperoxia, hypercapnia, and hypocapnia, thus enabling monitoring of disease progression and efficacy of treatment. In addition, precise ventilator and FiO2 adjustments can be made.

  • pH, PaCO2, and base deficit values can be useful for diagnosis of acidosis or alkalosis of either respiratory or metabolic origin, and whether or not compensation has occurred.

  • Haemoglobin (Hb) oxygen saturation and total Hb level can be measured using a co-oximeter. Moreover, some co-oximeters provide extra measurements as a total fraction of methaemoglobin (metHb), carboxyhaemoglobin (COHb), deoxyHb, and foetal Hb.

  • Measurement of mixed venous oxygen saturation (SvO2) for calculating oxygen consumption and monitoring the balance between oxygen supply and demand.

  • Measurement of oxygen exchange. The PaO2/FiO2 ratio is a convenient and widely used bedside index of oxygen exchange. However, although simple to calculate, it is affected by changes in SvO2, does not remain equally sensitive across the entire range of FiO2 (especially when shunt is the major cause of admixture), and does not take into account the functional status of the lung as a result of PEEP or the patient’s position. By contrast, the oxygenation index (PaO2/(FiO2 × mean PAW) considers the effects of PEEP and inspiratory time fraction and has gained widespread popularity, especially in paediatric practice.


Pulmonary function tests provide important and useful clinical information that may help in many clinical situations, particularly during mechanical ventilation. The analysis of ventilator waveforms and their derived measurements may facilitate mechanical ventilation set-up, thus decreasing VILI incidence. Pulmonary function tests can help in the assessment of treatment response and weaning process.

End-tidal CO2 monitoring

There are two types of commonly used gaseous CO2 monitoring used in clinical practice.

  • Capnography: this is the instantaneous graphical record of CO2 concentration in the respired gases during respiration.

  • Capnometry: this is the measurement and display of CO2 concentrations on a digital or analogue indicator.

Both methods can be employed in the critical care environment. Analysis of the CO2 concentration and waveform can help in the management of both the respiratory and cardiovascular systems.

Measurement techniques


Four main methods are used to measure CO2 concentration in respiratory gases: mass spectrography, Raman spectrography, photoacoustic spectrography, and, most commonly, infrared spectrography. This final method uses the physical principle that polyatomic gases absorb infrared light. CO2 selectively absorbs light at a specific wavelength of 4.3 μ‎m. The absorption of this wavelength of light is proportional to the concentration of CO2 molecules and, by comparing them with a known standard, the concentration can be displayed—usually as a partial pressure (mmHg or kPa).


A pH-sensitive chemical indicator in a plastic housing is placed in the system between the endotracheal tube (ETT) and the source of ventilation. Exposure to CO2 causes a change in the paper’s colour: purple to yellow. The degree of colour change permits approximation of the CO2 concentration; however, it cannot be used for accurate measurement.

Location of CO2 monitors

Capnometers are classified as side-stream or main-stream, depending on the position of the sensor relative to the gas delivery system. Both have advantages and disadvantages.


The sensor is located distal to the patient. The gas is sampled through capillary tubing attached to the distal end of the ventilation apparatus, as close to the subject’s airway as possible. This may be a T-piece inserted into a breathing system, or a Luer lock on a heated membrane exchange. The tubing can also be inserted into the nostril of a spontaneously breathing patient. Optimal gas flow is 50–200 ml/min, which needs to be considered if using expired gas flow to measure Vt. It also causes a delay in reading of ~150 ms.


Mounted on the breathing system, the capnometer consists of an infrared generator and sensor within a cuvette through which the respiratory gases pass. The sensors are heated to 39°C to prevent water condensation that may produce falsely elevated readings. Care is required as the heat can cause facial burns. The additional weight of the apparatus on the breathing system can cause traction on the ETT. Unlike the side-stream method, the main-stream technique provides immediate readings with respiration.

The capnogram waveform

The normal capnogram is usually displayed at 7 mm/s. It has a waveform that is divided into three phases.

  1. I. CO2-free phase. Anatomical dead space and that of the breathing system, heated membrane exchange, catheter mount.

  2. II. An S-shaped upswing which denotes mixed dead space and alveolar gases.

  3. III. A plateau phase owing to alveolar emptying. The normal plateau often has a slight upward slope; the highest point of this slope is the end-tidal CO2 partial pressure (PetCO2) and, if the slope has achieved a plateau, is the closest approximation of the PaCO2.

Volumetric capnography

Plotting the partial pressure of expired CO2 (PECO2) against time does not account for expiratory flow. While giving useful information about the presence of ventilation and an estimate of the PaCO2, the inclusion of flow measurement allows for a greater amount of pulmonary physiology to be detected and applied in management. This leads to the use of volumetric capnography based on the single breath test for CO2 (SBT-CO2). It differs from standard capnography by the integration of gas flow to produce a graph of PECO2 against volume rather than time. The SBT-CO2 gives more information on the ventilation/perfusion (V/Q) status of the lung.

Clinical applications of capnography

The use of capnography has been a standard requirement for monitoring at induction of general anaesthesia. It is now a requirement for monitoring of ventilated patients on the intensive care unit (ICU) and transport of ventilated patients both intrahospital and interhospital.

Confirmation of endotracheal intubation

The presence of a capnogram aids in the identification of an appropriately placed ETT and is mandatory for oral or nasal tubes and tracheostomy. Following intubation, six expired CO2 waves need to be observed to confirm correct placement, as oesophageal and gastric CO2 may be present from bag and mask ventilation, the presence of carbonated drinks, etc. The use of capnometry for such procedures is acceptable, but caution must be observed as tracheal secretions or gastric contents may contaminate the capnometer, interfering with its chemical reaction and causing it to read positive for CO2 in error.

Presence and adequacy of ventilation

The presence of a regular capnogram of correct morphology allows continuous monitoring of ventilation. This is in preference to the display of the PetCO2 value alone. Accidental disconnection or extubation, endotracheal obstruction, ETT cuff leak, and ventilator malfunction can all be detected due to a change in the morphology of the capnogram.

Capnography can be used to monitor ventilation in the spontaneously ventilating subject. Sampling can be undertaken at the nose, by adaptation of the nasal cannulae, or by placing the sampling line within the face mask. When oxygen is concurrently at low flow rates, the PetCO2 is a good predictor of PaCO2. At high oxygen flow rates, the mixing with the exhaled gases leads to underreading of the PetCO2. In patients at high risk of apnoea, e.g. type 2 chronic obstructive pulmonary disease (COPD) or neuromuscular disorders, CO2 monitoring can contribute to a prompt response in the event of hypopnoea or respiratory arrest.

If neuromuscular blockade is being used, then CO2 monitoring can aid assessment of its adequacy. Breathing by the patient during mandatory ventilation leads to a characteristic dip within phase III, the so-called ‘curare cleft’. Titration of paralysing therapy to prevent this phenomenon can be achieved as a goal and can lead to lower dose requirements.

Estimation of PaCO2

In normal individuals, the difference between PaCO2 and PetCO2, the (a-et)PCO2 is between 2 and 5 mmHg. It increases with age and disorders that may increase anatomical or alveolar dead space, e.g. COPD, pulmonary embolism, low cardiac output, and mechanical ventilation itself. It decreases with large Vt low-frequency ventilation, as well as in pregnancy and in young children.

Because of this potential variability, it is best to have a concurrent assessment of the PaCO2. Ongoing monitoring can then use the PetCO2 as a surrogate. Dynamic changes in the patient’s cardiovascular or pulmonary condition may result in a change in the (a-et)PCO2, so the PetCO2 cannot be ubiquitously relied on as a guide to PaCO2. Although a raised PetCO2 usually signifies a raised PaCO2, the contrary is not always true.

Cardiovascular status

The delivery of CO2 to the lungs and the regional perfusion of the alveoli are reflected in the partial pressure of CO2 in the alveoli. If ventilation remains constant (V), then changes in perfusion (Q) will result in a change in PECO2. This has been demonstrated clinically in both dynamic changes in cardiac output and the detection of pulmonary embolism. Changes that occur in hypermetabolic states such as thyrotoxicosis or severe sepsis can increase PECO2 as a reflection of the increase in tissue CO2 production. A novel method of cardiac output monitoring, the partial rebreathing technique, uses a variation of the indirect Fick method. Studies suggest that there is good correlation between this method and other standards. This may have a role as a non-invasive alternative in the mechanically ventilated patient.

Capnography also has a role at cardiac arrest. It reflects well the adequacy of resuscitation. PetCO2 values of <10 mmHg at 20 min postresuscitation are associated with significant mortality. Return of a spontaneous circulation can be signified by a sudden increase in PetCO2.

Alveolar recruitment and PEEP

The application of PEEP as a method to recruit and maintain the patency of collapsed alveoli is a strategy that can be used in respiratory failure, particularly acute lung injury (ALI). Dead space is initially large in ALI, as reflected by a high (a-et)PCO2. The application of PEEP has been shown to reduce this gradient. Other studies have demonstrated that titration of PEEP does not improve (a-et)PCO2 in all cases of ALI/ARDS. However, the gradient reduces in those that have a demonstrable inflection point on the pressure–volume curve and those patients that improve oxygenation with the application of PEEP.

The shape of the volumetric capnogram

Alterations in the shape of the capnogram can give information regarding pulmonary function. If all the alveoli had the same PCO2, the phase II time would be short and the plateau of phase III would be horizontal. The lung is not homogenous owing to wide ranges of V/Q ratios. Those areas of the lung that have a low V/Q ratio (underventilated, high PCO2) empty after those with a high V/Q ratio (well-ventilated, low PCO2). This causes an upward slope of phase III. The respiratory units with the lower V/Q tend to be located distally to those with a higher V/Q. Consequently, they empty later in the respiratory cycle. The slope of phase III is thus dependent on the emptying patterns of alveoli with differing time constants and PCO2 concentrations. A lengthening of phase II and an increase in the gradient of phase III owing to an increase in the heterogeneity of alveolar ventilation and perfusion is seen in many causes of respiratory failure, notably acute asthma, COPD, ALI/ARDS, pulmonary fibrosis, and pneumonia. In pulmonary embolism, the reduction in perfusion leads to a marked increase in the affected V/Q. The absence of CO2 in the affected units does not alter the morphology of the SBT-CO2 trace; it just reduces its height.

Pulse oximetry


Pulse oximetry is a non-invasive, continuous monitoring technique for determining the oxygen saturation of arterial blood (SaO2). It has become one of the most widely used monitoring systems in hospital medicine. In the critical care setting it is considered the standard of care in many situations. By warning clinicians of the presence of hypoxaemia, it allows rapid treatment and prevents possible complications.


Pulse oximetry determines oxygen saturation by measuring light absorption of arterial blood at two spectral wavelengths: 660 nm (red) and 940 nm (infrared). Pulse oximeter probes comprise two light-emitting diodes and a photodetector. The diodes transmit red and infrared light in sequence, several hundred times per second. The light travels through the tissue to the photodetector where it is electronically processed and the absorption at the two wavelengths is compared.

This technique is based on two physical principles: the different absorption spectra of oxyhaemoglobin and deoxyhaemoglobin and the Beer–Lambert law. According to the Beer–Lambert law, the intensity of transmitted light decreases exponentially as the concentration of a substance increases and also decreases exponentially as the distance travelled through a substance increases. Oxygenated Hb absorbs more infrared light and allows more red light to pass through, whereas deoxygenated Hb acts in the opposite way. The ratio of absorbance at the two wavelengths is calculated and calibrated against direct measurements of SaO2 in healthy volunteers using a co-oximeter. To eliminate the absorbance of tissues and capillary and venous blood, pulse oximeters separate the pulsatile component of the absorption signal. In addition, most devices display a plethysmographic waveform, which can help clinicians to identify artifacts.

Clinical applications

Detection of hypoxaemia and assessment of pulmonary gas exchange

Continuous pulse oximetry is now routinely used in critical care units. It allows rapid detection of hypoxaemia and helps to prevent subsequent complications. However, clinical decisions based upon pulse oximetry should be taken with care; although SaO2 values are clearly related to PaO2, changes in peripheral capillary oxygen saturation (SpO2) may not necessarily reflect a real change in PaO2 and do not assess patient ventilation. Indeed, SpO2 has been reported to be inaccurate in assessing abnormal pulmonary gas exchange.

Monitoring during invasive procedures and transport

Pulse oximetry monitoring has become mandatory during surgery or any diagnostic or therapeutic procedure that requires sedation, as well as during intrahospital or extrahospital transport.

Cardiopulmonary diseases

The technique has been shown to be useful as a screening tool for cardiopulmonary disease together with other clinical variables such as respiratory rate. It can provide early warning of hypoxaemia and may help to predict the occurrence of respiratory failure in high-risk patients.

Titration of FiO2 during mechanical ventilation

Pulse oximetry is also reliable for titrating FiO2 in mechanically ventilated patients. SpO2 target values of 92% and 95% predict a satisfactory level of oxygenation in Caucasian and black patients, respectively. In patients with severe ARDS, a SpO2 target of 88–92% is accepted to minimize oxygen toxicity.

Interestingly, the SpO2 to FiO2 ratio (S/F) can be a surrogate measure for the PaO2/FiO2 ratio when assessing oxygenation for ARDS recognition or when calculating organ failure assessment scores.

Cardiopulmonary arrest

During cardiopulmonary resuscitation the pulse rate on the oximeter may correlate with the electrocardiogram or chest compressions, giving a general assessment of the adequacy of perfusion.


Finally, SpO2 monitoring can be used in research on sleep and respiratory or cardiac disorders.


Low perfusion states

Pulse oximetry depends on satisfactory arterial perfusion of the skin. Consequently, the signal may be inaccurate or absent in low perfusion states such as low cardiac output, vasoconstriction, or hypothermia. In these situations, the sensor may not be able to distinguish the true signal from background noise (Figure 6.1).

Figure 6.1 Common pulsatile signals on a pulse oximeter.

Figure 6.1 Common pulsatile signals on a pulse oximeter.

SpO2 and SaO2 correlation—the oxyhaemoglobin dissociation curve

Pulse oximetry correlates well with SaO2 in patients with high SpO2. Because of the sigmoid shape of the dissociation curve, a large decrease in PaO2 will not produce a significant fall in SaO2 until it reaches a level of approximately 60–70 mmHg (Figure 6.2). In critically ill patients, who may present large differences between SpO2 and SaO2, pulse oximetry may not be sufficient. The two measurements should initially be performed simultaneously to validate the pulse oximeter readings. Nor can the technique detect significant hyperoxia, which may result in oxygen toxicity in neonates. Other factors, such as temperature, pH and PaCO2, can modify the relationship between SpO2 and PaO2. Furthermore, pulse oximetry results are signal-averaged over several seconds: this delay may be of particular significance during intubation.

Figure 6.2 The oxyhaemoglobin dissociation curve.

Figure 6.2 The oxyhaemoglobin dissociation curve.

Interference with other substances

Conventional pulse oximeters distinguish only two substances: reduced Hb and oxyHb. COHb has a similar absorption coefficient to oxyHb and may result in a false high reading (Table 6.1). MetHb has a similar absorption at both wavelengths. Recently, multiwavelength oximeters have been designed that are capable of estimating blood levels of COHb and MetHb.

Table 6.1 Causes of false readings

False high SpO2


False low SpO2


Intravenous dyes (methylthioninium chloride, indocyanine green, indigo carmine)


Low perfusion state

Fluorescent and xenon arc surgical lamps

Blood concentration of HbA1c

Movement artifacts

Intravenous dyes (e.g. methylene blue, indocyanine green) and bilirubin can cause false low SpO2 values. Skin pigmentation and nail polish may also affect readings.


The pulse oximeter is a convenient way of measuring arterial oxygenation. However, it is unable to assess ventilation or measure partial PaCO2. Therefore, end-tidal CO2 monitoring or arterial blood gas analysis should be considered when alveolar hypoventilation is suspected.

Motion artifacts

The most recent pulse oximeter devices have incorporated several signal processing techniques to reduce motion artifacts. However, movement is still considered an important source of error and false alarms.

Other signal processing factors

Although pulse oximeters correct for ambient light, false low readings have been reported with fluorescent and xenon arc surgical lamps. Electromagnetic radiation may also interfere with the reading as well as surgical diathermy, which is also a common cause of loss of saturation reading.

Physical risks

Pulse oximetry is a non-invasive technique that is unlikely to produce adverse events. However, digital injury may occur if it is applied for several days, especially in patients with low perfusion conditions and/or receiving vasopressor therapy.

Recent advances

In recent years, pulse oximeter technology has advanced significantly and allows measurement of other variables, such as the pleth variability index and types of haemoglobin.

Pleth variability index

The pleth variability index (PVI) is derived from the analysis of the photoplethysmography waveform, which allows calculation of respiratory variations. It has been shown to correlate with pulse pressure variation and it may help to predict fluid responsiveness in mechanically ventilated patients.


The development of multiwavelength pulse oximeters allows non-invasive measurement of total Hb, COHb, and MetHb concentrations.

New sensors

The introduction of new sensors such as the forehead reflectance sensor and the nasal sensor allows the measurement of SpO2 at sites other than fingers, and may present better perfusion.

In conclusion, pulse oximetry is one of the most important advances in respiratory monitoring. It is a simple-to-use, portable, non-invasive device that permits continuous SpO2 monitoring. Although large randomized trials have not demonstrated that routine continuous pulse oximetry monitoring improves patients’ outcomes, it is considered as a standard of care in many clinical situations, especially in critically ill patients. It may be of use in clinical management at the bedside and may help the prevention of an irreversible injury.

Monitoring transpulmonary pressure

Physiological background

The comprehension of the mechanical properties of the lung requires the understanding of what forces are involved to move the respiratory system.

During mechanical ventilation, the total pressure applied to the respiratory system (PTOTAL) is the sum of the pressure provided by the ventilator (PAW) and the pressure generated by the patient’s inspiratory muscles (PMUS).


The equation of motion describes how the total pressure applied to the respiratory system must overcome the opposing forces produced by the elastic and resistive properties of the respiratory system to initiate inspiratory flow.


P0 is the value of PAW at the beginning of the breath (zero or a positive value of end-expiratory pressure); ERS is the respiratory system elastance; RRS is the respiratory system resistance; V is the difference between the instantaneous volume and the relaxation volume of the respiratory system; and V is air flow.

In clinical practice, the airway pressure (PAW) is often assumed to reflect the forces applied to the lung. However, as described by the equation of motion, PAW represents the sum of the resistive and elastic properties of the major components of the respiratory system: the lungs and the chest wall.

The transpulmonary pressure (PL) is the real force distending the lung parenchyma. It is calculated as the difference between the airway pressure (PAW) and the pleural pressure (PPL) or its proxy, the oesophageal pressure.


The measurement of transpulmonary pressure is an essential component of the assessment of mechanically ventilated patients as it allows an understanding of the relative contribution of the chest wall and the lung parenchyma to the total elastance of the respiratory system. In other words, the transpulmonary pressure allows estimation of the proportion of the applied PAW used to overcome the resistive and elastic properties of the lung and the chest wall.

Different models have been proposed to guide mechanical ventilation based on the transpulmonary pressure. All these strategies are based on the measurement of the oesophageal pressure (POES), which is the only clinically available surrogate of pleural pressure.

Oesophageal pressure as a surrogate of pleural pressure

In 1949 it was suggested to use oesophageal pressure as a surrogate for pleural pressure. The rationale for the surrogate use of oesophageal pressure is that the lower third of the oesophagus is anatomically adjacent to the pleural space and consequently PPL is simply transmitted to the wall of the oesophagus, which can be considered—for these purposes—a passive structure, particularly in the erect position.

POES measurement: calibration and validation of POES during mechanical ventilation

The POES can be measured using air-filled or liquid-filled balloon catheters or with small solid-state transducers placed in the oesophagus. The most common technique is to use an air-filled, thin-walled latex balloon catheter inserted transorally or transnasally.

To obtain reliable POES measurements, the oesophageal balloon must be placed in the lower third of the oesophagus and the balloon must be inflated with air. The adequate volume of air depends on the mechanical properties (compliance) of the balloon, which depend on its design, dimensions, geometry, and the material used.

After placing the patient in a semi-recumbent position, a deflated balloon catheter is inserted through a nostril and advanced into the stomach. At this point, the catheter is inflated and its intragastric position is confirmed by the presence of a positive pressure deflection during a spontaneous inspiration. Subsequently, the catheter is slowly withdrawn until the inspiratory positive pressure deflections are replaced by negative inspiratory deflections, indicating that the catheter is in the lower third of the oesophagus. This method is not valid in passive mechanically ventilated, but spontaneously breathing, patients. During mandatory ventilation, in a relaxed patient, the positioning of the oesophageal balloon can be guided by the presence and amplitude of cardiac oscillations, generally 35–45 cm from the nostril.

The conventional method to validate the POES measurement is the dynamic occlusion test (Baydur test). It consists of comparing the simultaneous negative deflections of PAW and POES following an end-expiratory occlusion manoeuvre. Because there is no change in volume and no flow, the transpulmonary pressure should be constant and, thus, POES and PAW should change equally and simultaneously. A value of ∆POES/∆PAW ratio within 0.8–1.2 indicates that the POES measurement is reliable. Otherwise the position and the volume of the balloon should be rechecked. In sedated and paralysed patients, the occlusion test as described by Baydur cannot be executed but a variation of the test can be performed by applying manual compression on the chest during an end-expiratory hold and calculating the same ratio.

Elastance-derived measurement

The elastance-derived transpulmonary pressure method assumes that the ratio between the elastance of the lung (EL) and the elastance of the chest wall (ECW) determines how the pressure applied to the entire respiratory system (i.e. PAW) is partitioned between the lung (i.e. PL) and the chest wall (i.e. PPL). For example, if EL and ECW contribute for 80% and 20%, respectively, to the ERS in a passively ventilated patient, a pressure of 30 cmH2O applied at the airway opening will generate an elastance-derived PL of 24 cmH2O (80% of PAW) and a elastance-derived PPL of 6 cmH2O (20% of PAW), respectively.

Based on this background, this method allows the calculation of the elastance-derived PL in static conditions in a passive patient receiving positive pressure ventilation as follows:


EL = Δ‎PL/Vt and ERS = Δ‎PAW/Vt under static conditions (zero flow) between the beginning and the end of a breathing cycle.

The elastance-derived method has been used to set a transpulmonary ‘open lung’ approach at 25 cmH2O in patients with ARDS. The transpulmonary inspiratory pressure was calculated as:


(PPLAT is plateau pressure.)

It has been observed that targeting PEEP to reach the upper physiological limit of transpulmonary pressure, instead of the ‘safe’ limit of plateau pressure, improved oxygenation and prevented the need for extracorporeal membrane oxygenation in those patients with abnormal chest wall mechanics.

Direct measurement (absolute value)

Tamor et al. validated the directly measured end-expiratory PL.

PL(at end-inspiration) = PAW(at PPLAT)POES(at PPLAT)

PL(at end-expiration)=PAW(at PEEP)POES(at PEEP)

End expiratory PL was corrected for the gravitational effects caused by a semi-recumbent position so that:

PL (at end-expiration) = PAW(at PEEP)[POES(at PEEP)5]

In the study, PEEP was adjusted according to measurements of POES (the POES-guided group) or according to the ARDS Network recommendations (the control group). It was concluded that a ventilator strategy using POES to estimate the transpulmonary pressure (PEEP was set to maintain positive transpulmonary pressure) significantly improved oxygenation and compliance.

However, a debate exists about whether the absolute values of POES can be interpreted as reliable estimates of PPL. Moreover, PPL varies within the pleural space because of both gravitational gradients and regional inhomogeneities. Some data suggest, however, that, even under these conditions, POES remains an acceptable effective average estimate of PPL.

Clinical use of POES

POES monitoring has been shown to be essential in improving the management of critically ill and ventilator-dependent patients. The advantages related to its measurement are:

  • A quantitative description of the mechanical properties of the lung. It helps to determine what pressures are distending the lung.

  • A guide for the management of mechanical ventilation in ARDS patients. POES measurement allows the titration of PEEP to determine a well-tolerated level of airway pressure in ARDS patients to avoid tidal overexpansion of the lung and repeated airspace collapse and de-recruitment.

  • The assessment of respiratory muscle activity during assisted ventilation.

  • The evaluation of patient–ventilator synchrony or asynchrony.

  • The calculation of respiratory muscle effort or patient’s work of breathing that can optimize the distribution of oxygen delivery, prevent diaphragm injuries, and accelerate weaning from mechanical ventilation.

POES to guide therapy in ARDS

When a transpulmonary pressure is applied to the lung, a counterforce of equal intensity (and opposite sign) develops. This counterforce is called lung stress. The associated lung deformation (change in volume) is called strain and may be expressed as the ratio of the change in lung volume to the resting lung volume during respiration (Δ‎V/V0). Because global lung stress and strain are inadequately estimated by plateau pressure and tidal volume, transpulmonary pressure has been advocated as a better guide for safe mechanical ventilation. Indeed, the elastance of the chest wall can be increased in conditions such as obesity, abdominal hypertension, etc., which are highly prevalent in critically ill patients (nearly 30% of ICU patients). Monitoring of transpulmonary pressure is the only way to understand which part of applied PAW is dissipated in distending the chest wall. The same value of PAW can generate dramatically different transpulmonary pressures and could result in alveolar hyperinflation and/or cyclical alveolar opening and collapse.

Provided that PL is the real ‘lung-distending’ pressure, a lung-protective ventilator strategy in a patient with ARDS should take into account the measurement of POES. POES helps the clinician to titrate PEEP according to the patients’ needs, to provide appropriate and safe PL while avoiding de-recruitment and atelectrauma. One example used by Talmor in his trial could be Table 6.2.

Table 6.2 Lung-protective strategies

Controlled ventilation in ARDS patients

Tidal Δ‎PL

End inspiratory PL

End expiratory PL

< 10–12 cmH2O

< 20–25 cmH2O

> 0 cmH2O

Assisted ventilation in ARDS patients

End inspiratory PL

< 20–25 cmH2O

Transpulmonary pressure to study the effects of spontaneous breathing during assisted ventilation

Spontaneous breathing can offer protective effects to the lung, such as improved oxygenation and increased dorsal ventilation. It may also prevent the rapid respiratory muscle atrophy observed during fully controlled mechanical ventilation and the resulting ventilator-induced diaphragmatic dysfunction. Moreover, it helps to reduce the deleterious effects of prolonged sedation.

However, accumulating evidence indicates that spontaneous breathing may become deleterious in severe ARDS cases. Indeed, high spontaneous breathing efforts generate high negative pleural pressures, which can significantly increase the transpulmonary pressure and worsen lung injury. Especially in severe ARDS patients, who have high PAW and uncontrollable spontaneous demands, PL is likely to be injuriously high. In this regard, total end-inspiratory PL should be kept below 20–25 cmH2O.

Adjusting ventilator settings during assisted modes is, thus, a clinical challenge, and direct measurement of the patient’s effort is the only method available to assess a safe ventilator strategy. In this scenario, the recording of POES will allow the clinician to detect the harm of spontaneous effort in severe ARDS. Therefore, the use of the oesophageal balloon for the calculation of the chest wall elastance and the transpulmonary pressure become essential in setting a protective ventilator strategy.

Patient–ventilator asynchrony and the role of POES monitoring

Asynchrony derives from the mismatch between the patient’s respiratory drive and one or more ventilator variables controlling the breathing pattern: trigger, flow, or cycle. It is associated with increased duration of ventilation, ICU, and hospital stay, and it represents a major problem during assisted mechanical ventilation. Although asynchronies are common, their recognition may be difficult. The best way to detect asynchrony is to characterize the activity of the respiratory muscles with either an oesophageal catheter or electromyography of the diaphragm. Monitoring the POES can detect asynchrony by comparing the time occurrence of the changes in POES with those of PAW and flow–time waveforms. Hence, this will allow the clinician to adjust the ventilator settings accordingly.

As an example, ineffective efforts are characterized by a negative POES swing without ventilator pressurization. This occurs in situations combining auto PEEP with high respiratory rate and high pressure support, and can be aggravated by a poorly sensitive triggering system.

Moreover, POES is necessary to identify a phenomenon called ‘reverse triggering’ reliably. It is described as the presence of respiratory muscle contractions triggered by the ventilator and has been found in deeply sedated patients. This may have important clinical consequences, such as double inspiration, tidal volume increase, or erroneous plateau pressure measurements.

Weaning—POES monitoring during weaning from mechanical ventilation

Recording POES throughout a weaning trial may enhance the prediction of weaning success or failure. It may provide a simple method to monitor changes in patient effort or work of breathing, to determine whether worsening mechanics or haemodynamics are responsible for a weaning failure, and to institute the correct therapy, such as bronchodilators, inotropes, vasodilators, and diuretics.

Conclusions and perspectives

Despite voluminous data showing the usefulness of POES measurements in critically ill patients, in clinical practice the oesophageal catheter is still underused. As shown in prevalence studies, the oesophageal pressure catheter was used in only 0.8% of patients with ARDS.

The results of all the recent studies suggest that the measurement of transpulmonary pressure is essential to tailor mechanical ventilation to the individual patient’s needs. POES monitoring allows the partitioning of respiratory system mechanics into pulmonary and chest wall components. Its benefits are relevant for both controlled and assisted mechanical ventilation. In conclusion, POES measurements have enhanced our knowledge of the pathophysiology of ALI, patient–ventilator interaction, and weaning failure.

Multiple choice questions and further reading

Respiratory monitoring Interactive multiple choice questions to test your knowledge on this chapter and additional further reading can be found in Appendix Chapter 6 Multiple choice questions and further reading