Jerome G. L. Cockings
Sarah Dawson and Andrew Gratrix
Sarah Dawson and Andrew Gratrix
Michael R. Pinsky
Aerobic respiration is the most efficient method of energy production in the mammalian cell. It utilizes oxygen to produce adenosine triphosphate (ATP). The absence of oxygen or low oxygen levels result in more inefficient anaerobic respiration. Cellular energy levels become inadequate and this can lead to loss of cellular homeostasis, which in turn can lead to cellular death and very possibly organism death. A substantial part of critical care is targeted at treating and/or preventing hypoxia.
Pathophysiology of oxygen delivery
In critical illness, the delivery (DO2) and uptake (VO2) of oxygen are often abnormal. Currently there are few therapeutic strategies for improvement of VO2. Most methods of oxygen therapy target improvement in DO2.
DO2 from the environment is necessary to provide for cellular metabolism. In single-celled organisms (e.g. amoeba), simple diffusion suffices. However, in the multicellular, multiorgan human, more sophisticated mechanisms have evolved, each with their problems in illness.
Transport of oxygen to the cells follows six stages reliant only on the laws of physics.
1. Convection from the environment (ventilation).
2. Diffusion into the blood.
3. Reversible chemical bonding with haemoglobin.
4. Convective transport to the tissues (cardiac output).
5. Diffusion into the cells and organelles.
6. The redox state of the cell.
This chain of events is DO2. Failure of DO2 to match VO2 leads to shock. This occurs when DO2 declines to below approximately 300 ml/min. Shock is defined loosely as failure of DO2 to match tissue demand. Commonly this refers to circulatory failure, but low DO2 can result from several pathological mechanisms, which can occur as a single problem or in combination (Table 1.1).
Table 1.1 Types of hypoxia
Type of hypoxia
Reduced supply of oxygen to the body leading to a low arterial oxygen tension
Normal arterial oxygen tension, but circulating haemoglobin is reduced or functionally impaired
Massive haemorrhage, severe anaemia, carbon monoxide poisoning, methaemoglobinaemia
Failure of oxygen transport owing to inadequate circulation
Left ventricular failure, pulmonary embolism, hypovolaemia, hypothermia
Impaired cellular metabolism of oxygen despite adequate delivery
Cyanide poisoning, arsenic poisoning, alcohol intoxication
The impact of low DO2 can be made worse by an increase in VO2. Metabolic rate increases with exercise, inflammation, sepsis, pyrexia, thyrotoxicosis, shivering, seizures, agitation, anxiety, and pain. This mismatch leads to the need for early detection of shock and prompt treatment.
Clinical signs such as heart rate, blood pressure, and urine output can be misleading, especially in the young. This therefore requires the concept of an effective cardiac output. This couples the clinical signs with evidence of normal DO2 and VO2 balance. The assessment includes peripheral temperature, oxygen haemoglobin saturation and arterial partial pressure, the presence of acidosis with a base excess greater than –2, lactataemia, and abnormal venous oxygen saturation (SvO2). These more technical measures of adequacy of oxygen delivery and uptake must always be taken in the clinical context. For example, in cyanide poisoning, both circulatory and ventilatory indices appear normal, yet the severe acidosis and lactataemia seen in this condition demonstrates tissue hypoxia. Manipulating DO2 by increasing the environmental oxygen fraction (FiO2) or cardiac output in this setting is unlikely to be helpful, and, even in sepsis and other more common types of shock, achieving supranormal values for DO2 is not thought to be beneficial.
Strategies for increasing DO2
By assessing the type of hypoxia and its likely cause, the correct choice of DO2-improving strategy can be chosen. In the critically ill, the commonly seen combination of mechanisms leading to hypoxia may require several techniques to be instigated in parallel. The methods for improving DO2 to the tissues are based on reversing problems seen at each of the six stages of DO2. Improving the transport of oxygen once in the body will be covered later in this book. Chapter 27 is concerned with improving DO2 from the environment to the bloodstream. DO2 at this stage should be considered a support mechanism and treatment of the underlying cause is most important to reverse hypoxia.
Oxygen therapy apparatus
In the hypoxic self-ventilating patient, delivery of oxygen to the alveoli is usually achieved by increasing the FiO2. Commonly this involves the application of one of the many varieties of oxygen masks to the face, such that it covers the mouth and/or nose. Each type of delivery system consists of broadly the same six components:
1. Oxygen supply: delivery of oxygen can be from pressurized cylinders, hospital supply from cylinder banks, or a vacuum-insulated evaporator, or an oxygen concentrator.
2. Oxygen flow control: for example, an OHE ball valve flow meter.
3. Connecting tubing: both from supply to control, and from control to patient. The bore of the tubing is important as it has effects on the oxygen flow rate. In some systems, it can also act as a reservoir.
4. Reservoir: all have reservoirs. In the simple oxygen mask, it is the mask itself. Nasal cannulae use the nasopharynx as the reservoir. An oxygen tent is a large-volume reservoir. The reservoir serves to store oxygen, but must not allow significant storage of exhaled gases, leading to rebreathing of carbon dioxide.
5. Patient attachment: this permits delivery of oxygen to the airway. This is achieved either by directly covering the upper airway, e.g. plastic mask/head box, or by increasing the oxygen concentration in the wider environment, e.g. oxygen tent.
6. Expired gas facility: expired gas needs to dissipate to the environment. This can be achieved by having a small reservoir with holes, one-way valves as in the non-rebreather masks, or high oxygen flows as seen in some of the continuous positive airway pressure (CPAP) systems.
Additional features of oxygen breathing systems are the presence of humidification (such as a water bath) to prevent drying of the mucosal membranes. Some devices have an oxygen monitor incorporated into the apparatus to permit more accurate definition of the FiO2.
Factors that affect the performance of oxygen delivery systems
Most of the simpler oxygen delivery devices, e.g. plastic masks, nasal cannulae, etc., deliver oxygen at relatively low oxygen flow rates. The patient inspiratory flow rate varies throughout inspiration (25–100+ l/min) and exceeds the oxygen flow rate. This drains the small reservoir and causes entrainment of environmental air. The effect is to dilute the oxygen concentration to the final FiO2. The actual FiO2 that reaches the alveolus is therefore unpredictable and is dependent on the interaction of patient factors and device factors (Table 1.2). In the hypoxic patient, it is common to find significant increases in inspiratory flow rates as well as the loss of the respiratory pause. This causes significant entrainment of air, lowering the alveolar FiO2. This is particularly true of the variable performance masks but is also seen in Venturi-type masks, particularly when higher FiO2 inserts are used. The presence of a valve-controlled reservoir bag on a non-rebreather mask should compensate for high inspiratory flows, hence the belief that such devices can deliver an FiO2 of 1.0, which does not actually happen. This is not seen in models of human ventilation (Figure 1.1).
Table 1.2 Factors that influence the FiO2 delivered to a patient by oxygen delivery devices
Inspiratory flow rate
Oxygen flow rate
Presence of a respiratory pause
Volume of mask
Air vent size
Tightness of fit
Source data from Leigh J. Variation in performance of oxygen therapy devices. Anaesthesia 1970; 25, pp. 210–222.
Classification of oxygen delivery devices
Methods of delivering oxygen to the conscious patient with no airway instrumentation can be broadly divided into the following categories.
• Variable performance systems.
• Fixed performance systems.
• High flow systems.
Variable performance systems are so called because their FiO2 can vary. Fixed performance systems cannot. High-flow systems use high oxygen flows to maintain a fixed performance. The common types and their properties are summarized in Table 1.3.
Table 1.3 Classification of oxygen delivery systems
Oxygen delivery system
Nasal cannulae, semi-rigid masks (Hudson, MC), non-rebreathing masks, tracheostomy mask, T-piece systems
Non-sealed masks or nasal cannulae. Oxygen at low flow (2–15 l/min). Small reservoir. Significant entrainment of environmental air. Accurate FiO2 not possible. Comfortable and simple to use.
Venturi-type masks, anaesthetic breathing circuits (Waters circuit, Ambu-bag)
Venturi-type masks rely on the Venturi principle to dilute oxygen predictably to FiO2. Need to change valve to alter FiO2. Higher FiO2 valves have larger orifices, so behave more like a variable performance system. Comfortable. Simple to use, but needs attention to detail.
Anaesthetic breathing systems require sealing mask to prevent entrainment. Valves prevent rebreathing. Large reservoir. Accurate FiO2. Sealed mask can be uncomfortable. Knowledge of breathing systems required.
T-piece systems, Optiflow® (humidified high-flow nasal cannulae)
Rely on high oxygen flows to match patient’s inspiratory flow rate. T-piece systems require small reservoirs and sealed masks. These provide an accurate FiO2, but are uncomfortable with risk of mucosal dryness and are more complicated to set up. Newer high-frequency oscillation (HFO) systems can deliver a high-flow air/oxygen blend through a nasal cannula.
Intravascular oxygenation (cardiopulmonary bypass, interventional lung assist devices [Novolung®], extracorporeal membrane oxygenation)
Unusual in the self-ventilating patient. Oxygenation achieved across synthetic membrane. CO2 removal can be an issue. FiO2 can be difficult to measure. Complicated and limited to specialist centres.
Hazards of oxygen therapy
Oxygen is a drug and, like most drugs, its use is not without risk. It is also a gas and commonly delivered from compressed sources.
Medical oxygen is supplied at 137 bar from a cylinder, and 4 bar from hospital pipelines. Direct administration at delivery pressures is highly dangerous and requires properly functioning pressure-limiting valves. Oxygen supports combustion. Patients must not smoke cigarettes when receiving oxygen therapy, and oxygen should be removed from the environment when sparking may occur, e.g. during defibrillation.
Central nervous system toxicity (Paul Bert effect)
Seen in diving, oxygen delivered at high pressures (>3 atm) can lead to acute central nervous system (CNS) signs and seizures.
Lung toxicity (Lorrain Smith effect)
Prolonged exposure to a high FiO2 results in pulmonary injury. Possibly mediated by free oxygen radicals, there is a progressive reduction in lung compliance, associated with interstitial oedema and fibrosis. Avoidance of long periods of high oxygen concentrations reduces this effect. Clinically it can be difficult to prevent long exposure times; however, in general, patients should remain below an FiO2 of 0.5 where possible and not remain above this value for much longer than 30 h.
A condition concerning neonates, this is a chronic fibrotic lung disease associated with ventilation at high FiO2. Pathologically it is similar to the adult condition above, but with the additional effect of immaturity. Surfactant and maternal steroid therapies have lowered the incidence and severity.
Hyperbaric oxygen therapy
Oxygen can be delivered to patients at higher than atmospheric pressures (2–3 atm). This serves to increase the amount of oxygen dissolved in the plasma, rather than that bound to haemoglobin. At rest, the metabolic demands of an average person can be met by dissolved oxygen alone when breathing an FiO2 of 1.0 at 3 atm.
Hyperbaric oxygen is delivered in a sealed chamber. The gas is warmed and humidified. The common indications for hyperbaric oxygen therapy are listed in Table 1.4. High-pressure therapy also has important side-effects. Whilst clearly of value in these situations, the availability of a hyperbaric chamber often reduces its use, particularly in carbon monoxide poisoning.
Table 1.4 Suggested indications for hyperbaric oxygen therapy
Carbon monoxide poisoning
Radiation tissue damage
Air or gas embolism
Decompression sickness (the ‘bends’)
Acute blood loss
Compromised skin flaps or grafts
Clostridial myositis and myonecrosis
Enhancement of healing of problem wounds
Ventilatory support: indications
The requirement for ventilatory support is the most common reason that patients are admitted to an intensive care unit (ICU).
The aims of ventilatory support are to:
• Improve gas exchange by correcting hypoxaemia and respiratory acidosis.
• Relieve respiratory distress by reducing the work of breathing and reducing the oxygen cost of breathing.
• Improve the pressure–volume relationships of lungs, including increasing compliance and reversing or preventing atelectasis.
• Ensure patient comfort.
• Avoid complications and permit lung healing.
Use of ventilatory support
In addition to the treatment of acute respiratory failure (ARF), ventilatory support is also used in circulatory shock (e.g. cardiogenic shock, septic shock) and in the management of cerebral injury. In an international observational study of ventilation practice in intensive care involving over 8000 patients, the indications for ventilatory support were as follows:
• ARF (70%) (including acute respiratory distress syndrome [ARDS], sepsis, cardiac failure, pneumonia, postoperative respiratory failure, trauma, aspiration).
• Coma (19%).
• Acute exacerbation of chronic obstructive pulmonary disease (COPD) (6%).
• Neuromuscular disease (1%).
Respiratory failure: is defined as the failure to maintain normal arterial blood gases breathing room air:
• Hypoxaemic (type 1)—arbitrarily defined as a partial pressure of oxygen (PaO2) of <6.7 kPa (50 mmHg) breathing room air.
• Hypercapnic (type 2)—defined as a partial pressure of carbon dioxide (PaCO2) of >6.7 kPa (50 mmHg).
Respiratory failure may be acute, chronic, or acute on chronic. The most important mechanisms of hypoxaemia are ventilation–perfusion mismatch and shunt (cardiac and intrapulmonary). Diffusion impairment and reduced inspired oxygen tension (e.g. high altitude) are less relevant.
Patients with chronic type 2 respiratory failure develop a compensatory metabolic alkalosis and maintain a normal pH despite an elevated PaCO2. An acidaemia (pH <7.30) rather than the PaCO2 value indicates the need for ventilatory support.
Ventilation–perfusion mismatch: describes areas of the lung that have excessive perfusion compared with ventilation (shunt-like effect) and areas that have excessive ventilation compared with perfusion (dead space effect). Hypoxaemia due to ventilation–perfusion mismatch is usually easily corrected by increasing the inspired oxygen tension. As long as the patient is able to increase minute ventilation by increasing tidal volume and respiratory rate, an increase in carbon dioxide tension is prevented.
Pulmonary shunt: describes the most severe form of ventilation–perfusion mismatch, where venous blood passes through the lungs without any involvement in gas exchange. It occurs if there are areas of the lung that are not ventilated (e.g. atelectasis, air spaces full of fluid, blood, or inflammatory exudate). The hypoxaemia associated with shunt is not reversed by increasing the inspired oxygen tension, and treatment needs to be directed toward opening the non-ventilated parts of the lung by the application of positive pressure ventilatory support (lung recruitment).
Intracardiac shunt: results in profound hypoxaemia that is not reversed by 100% oxygen or positive pressure ventilation. Although usually associated with congenital heart disease, intracardiac shunt may develop through a patent foramen ovale (present in 25% of the population). The foramen ovale remains closed as long as left atrial pressure is higher than right atrial pressure. In right ventricular failure (e.g. pulmonary hypertension secondary to pulmonary embolus, ARDS, etc.), right atrial pressure is elevated above left atrial pressure, which may open the foramen ovale causing a right-to-left shunt and severe hypoxaemia. Echocardiography can be invaluable in confirming right-to-left intracardiac shunts as a cause for profound hypoxaemia that is not corrected by conventional ventilatory management.
Ventilatory support and cardiac failure
Positive pressure ventilation may have adverse and beneficial effects on cardiac function.
• Hypotension is common after commencing intermittent positive pressure ventilation (IPPV) owing to the reduction in venous return. This may precipitate myocardial ischaemia in a patient with critical coronary artery disease.
• Left ventricular function may be improved. An increase in intrathoracic pressure (ITP) results in a reduction in left ventricular transmural pressure and left ventricular afterload. This improves function of the failing ventricle as it moves to a more favourable position on the left ventricular function (Starling) curve.
• Right ventricular function may be impaired. In addition to reduced preload, right ventricular function may be impaired from an increase in pulmonary vascular resistance from alveolar vessel compression associated with alveolar overdistension.
• Correction of hypoxaemia and respiratory acidosis is associated with improved cardiac function. Mechanisms include a direct effect on cellular function and from reversal of pulmonary vasoconstriction caused by hypoxaemia and respiratory acidosis.
The application of raised ITP with CPAP typically results in a rapid clinical improvement in patients with acute left ventricular failure. In right ventricular failure (e.g. massive pulmonary embolus), positive pressure ventilation with high ITPs should be avoided so that right ventricular function is not compromised further.
Ventilatory support and septic shock
Severe sepsis may result in hypoxaemic respiratory failure and is a common indication for ventilatory support. Severe sepsis is also associated with a severe metabolic acidosis and a markedly increased work of breathing. The high work of breathing increases demand for oxygen consumption in a shocked patient with inadequate utilization of oxygen. In these circumstances, the oxygen cost of breathing may approach 50% of oxygen consumption compared to a normal value of 5%. In patients with septic shock, ventilatory support may be commenced owing to a deteriorating metabolic acidosis without respiratory failure to reduce the work of breathing and the oxygen cost of breathing.
Assessment of a patient with respiratory failure
Severe hypoxaemia (PaO2 <8 kPa) despite high-flow oxygen or a deteriorating respiratory acidosis (pH <7.30) are common indications for commencing ventilatory support. Blood gas values reflect on both the severity of the acute episode and the degree of chronic impairment of respiratory function. Clinical assessment is more important than arbitrary arterial blood gas values when considering ventilatory support. Clinical signs of severe respiratory failure indicating the need for ventilatory support may include:
• Altered conscious level (from agitation to coma).
• Increased work of breathing, including tachypnoea (respiratory rate >30), use of accessory muscles, and nasal flaring.
• Paradoxical breathing pattern reflecting diaphragmatic fatigue when the flaccid diaphragm paradoxically moves cephalad during inspiration, causing inward movement of the abdominal wall.
• Central cyanosis.
• Signs of excessive catecholamine release, including diaphoresis, tachycardia, cardiac arrhythmias, and hypertension.
Lung function tests may be of value in indicating the need for ventilatory support in selected patients at risk of respiratory failure due to neuromuscular disease. A vital capacity of <10 ml/kg is associated with a markedly impaired ability to cough and, if untreated, a progressive decline in respiratory capacity occurs due to atelectasis, ending with respiratory arrest. Serial measurements of vital capacity are helpful in predicting the need for ventilatory support in Guillain–Barré syndrome. Ventilatory support should be instituted when vital capacity falls to between 10 and 15 ml/kg and before there is deterioration in arterial blood gases. Measurements of vital capacity appear to be less predictive of the need for ventilatory support in myasthenia gravis, probably because of the unpredictable and variable course of muscle function in this condition.
Complications of ventilatory support
These relate either to the requirement for tracheal intubation or to the effects of positive pressure ventilation.
Upper airway trauma is not infrequent and increases with the duration of intubation. Complications of endotracheal intubation include laryngeal swelling, prolonged laryngeal dysfunction (dysphonia and impaired swallowing) and, rarely, tracheal stenosis.
Pulmonary oxygen toxicity describes airway and lung parenchyma damage secondary to prolonged exposure to high inspired oxygen tensions. Prolonged (>24 h) exposure of the normal lung to a PaO2 of >0.5 bar has been associated with damage in previously normal lungs (atelectasis, reduced CO transfer factor). The priority is to correct hypoxaemia in the patient with ARF, but unnecessary exposure to high inspired oxygen concentrations should be avoided. A PaO2 of >8 kPa is an acceptable target in the majority of patients.
Nosocomial pneumonia (ventilator-associated pneumonia [VAP]) is the most common complication of mechanical ventilation and is thought to arise from microaspiration of colonized upper airway secretions. Mechanical ventilation should be discontinued and the patient extubated at the earliest opportunity to reduce the risk of developing VAP (see ‘Hospital-acquired pneumonia’ in Chapter 18).
Hypotension is common after commencing positive pressure ventilation owing to the reduction in venous return. It may be severe if the patient is hypovolaemic.
Barotrauma describes pressure-related damage to the lungs resulting in extrapulmonary air, which may cause pneumothorax, subcutaneous emphysema, mediastinal emphysema, and systemic air embolism. An incidence of up to 10% has been reported in patients with ARDS receiving mechanical ventilation. Although recent studies have suggested that barotrauma is more related to the degree of damage to the underlying lung than the use of high airway pressures, it is prudent to avoid high airway pressures (e.g. peak pressure <45 cmH2O and plateau pressure <30 cmH2O).
Ventilator-associated lung injury
Laboratory studies and clinical trials have demonstrated that exposing lungs to high tidal volumes and pressures results in an insidious and progressive worsening of the underlying lung injury. Overdistension (‘volutrauma’) and tidal lung recruitment—derecruitment (‘atelectrauma’) appear to be the important mechanisms. These mechanical stressors induce a local and systemic inflammatory response, which also contributes to the lung damage (‘biotrauma’). Selecting an inappropriately high tidal volume increases mortality in patients with ARDS and accumulating evidence suggests that high tidal volumes will damage normal lungs and increase the incidence of pulmonary complications.
Outcome from mechanical ventilation
An improvement in outcome of mechanical ventilation over time was reported in an international observational study of over 900 ICUs involving over 18 000 patients between 1998 and 2010. The mean duration of ventilation was 5 days in 2010 with hospital mortality for all patients of 35%. Compared to 1998, there was a significant reduction in the size of tidal volumes used, which was associated with a reduction in mortality over time (odds ratio 0.78 compared to 1998).
Ventilators generate a pressure gradient between the upper airway and the alveoli that results in a controlled flow of gas (air and/or oxygen) into the lungs. This can be achieved either by creating a negative pressure around the chest wall while the upper airway remains at atmospheric pressure (e.g. tank ventilator or a cuirass) or, more commonly, by creating a positive pressure in the airway. Negative pressure ventilators are rarely used in modern intensive care practice. Positive pressure ventilation may be applied via an artificial airway (invasive ventilation) or by a mask (non-invasive ventilation).
Positive pressure ventilators used in the ICU are complex microprocessor-controlled pieces of equipment. They require a high pressure (4 bar) source of oxygen and air and a power source from mains electricity. There may be a battery as a short-term (e.g. 10–30 min) back-up power source in case of failure of the mains supply.
Inspiration is the active phase of mechanical ventilation that is controlled by the ventilator. Expiration is passive and occurs when the expiratory valve is opened, with gas flow depending on the elastic recoil forces of the lungs and chest wall combined with the expiratory airway resistance.
The ventilator circuit (Figure 1.2) comprises inspiratory and expiratory limbs, which connect to the relevant ventilator port. The two limbs are joined close to the patient by the ‘Y’ piece. The Y piece may be attached directly to the airway (endotracheal tube [ETT] or tracheostomy) or with the aid of a catheter mount. A humidifier must be included in the circuit. An active humidifier (heated water bath) or a passive heat and moisture exchanger (HME) may be used. Any additional tubing (e.g. catheter mount, HME/bacterial filter) between the airway and the Y piece is termed equipment dead space as it results in rebreathing of exhaled CO2, reducing the effective tidal volume.
Classification of positive pressure ventilation
The wide and often confusing modes of ventilation offered by an ICU ventilator can be simplified by considering the following:
• Are the breaths volume or pressure targeted (or both)?
• Are the breaths initiated by the ventilator or by the patient’s efforts (or both)?
• How is the duration of inspiration controlled?
Volume- or pressure-targeted breaths
In volume-controlled ventilation the tidal volume is set and the airway pressure generated depends upon the compliance and resistance of the respiratory system (lungs and respiratory circuit). The inspiratory flow waveform is constant during volume ventilation while the airway pressure gradually increases to a peak.
In pressure-controlled ventilation the ventilator delivers a set inspiratory airway pressure and the tidal volume that is delivered will depend upon the compliance and resistance of the respiratory system. The inspiratory flow waveform has a decelerating envelope whilst airway pressure remains constant throughout inspiration.
ICU ventilators now offer dual control of the inspiratory phase where the breath is both volume-targeted and pressure-limited. This results in a breath with the characteristics of a pressure breath (constant pressure, decelerating flow) together with a predictable tidal volume.
Triggering: time-, pressure-, or flow-triggered
Breaths may be delivered according to a set frequency (ventilator or mandatory breaths) or in response to the patient making an attempt to inhale (spontaneous or supported breaths). Some modes allow a mixture of mandatory and spontaneous breaths.
Triggering describes what parameter the ventilator uses to initiate a breath and cycle to inspiration.
In time triggering, breaths are delivered according to a preset frequency. This is also termed controlled ventilation (e.g. controlled mechanical ventilation) as there is no interaction between ventilator and spontaneous respiratory efforts.
Pressure or flow triggering is used to detect spontaneous respiratory efforts to allow supported breaths. In pressure triggering, the ventilator detects the drop-in airway pressure that occurs when the patient makes a spontaneous inspiratory effort with the inspiratory valve closed. As soon as the pressure drop exceeds the trigger limit, the ventilator will cycle to inspiration. The sensitivity of the trigger may be adjustable by setting the pressure drop that will initiate inspiration (e.g. 0.5–2 cmH2O).
In flow triggering, the patient’s inspiratory effort is detected from a change in flow in the ventilator circuit. This is usually achieved by maintaining a flow rate in the ventilator circuit during the expiratory phase and monitoring the flow returning to the expiratory valve. If the patient makes an inspiratory effort, the flow returning to the expiratory valve falls below the background flow rate and the ventilator cycles to inspiration (Figure 1.3).
Flow triggering is considered the most sensitive method of triggering that improves synchrony between patient and ventilator. Optimal setting of the trigger ensures that all patient efforts are detected by the ventilator and that autotriggering does not occur.
Autotriggering describes the ventilator incorrectly cycling to inspiration when the patient has not made an inspiratory effort. It occurs if the trigger is set at a too sensitive level with the result that small changes in pressure or flow within the ventilator circuit that are not due to patient effort (e.g. movement) erroneously initiate inspiration.
In neurally adjusted ventilatory assist (NAVA), an increase in the diaphragmatic electrical activity is used to initiate inspiratory support. This method of triggering requires the use of a purpose-built nasogastric tube incorporating electrodes to record the diaphragmatic activity continuously; this minimizes the time from patient effort to ventilator response and results in improved patient ventilator synchrony.
Cycling: volume, flow, or time
The ventilator can use volume, time, or flow to dictate when the inspiratory phase is complete and when cycling to expiration should occur.
Volume cycling: expiration occurs as soon as the set tidal volume is delivered. There is no end-inspiratory pause. If a pause is added, the tidal volume is held within the lungs for a short period at the end of inspiration before expiration, then cycling can be by both volume and time.
Time cycling is most common method of cycling used in ICU ventilators. The inspiratory time is either set directly or derived from the set frequency and inspiratory-to-expiratory time ratio. Pressure-controlled breaths are always time-cycled.
Flow cycling is used in pressure support ventilation to terminate the inspiratory phase. The inspiratory flow has a decelerating flow profile and, when inspiratory flow falls to a predetermined percentage of the peak flow rate, the ventilator cycles to expiration. In most ventilators, this flow rate is preset at 25% of the peak flow rate, although some ventilators now have the facility for the user to choose the flow rate at which cycling occurs (e.g. between 10% and 90% of peak flow rate).
Most ICU ventilators display continuous graphical displays of airway pressure, gas flow, and volume plotted against time.
Observation of these displays can provide useful information regarding the mode of ventilation and ventilator settings, adequacy of patient ventilator synchrony, evidence of gas trapping, and an indication of the mechanical properties of the respiratory system.
In volume-controlled ventilation with a constant inspiratory flow rate, the airway pressure gradually increases to a peak during inspiration. If an end-inspiratory pause has been set, the plateau pressure can be observed (Figure 1.4).
The difference in the peak and plateau pressure reflects the pressure required to overcome resistive forces and increases as inspiratory flow rate or airway resistance increase. If inspiratory flow rate remains constant, the peak-to-plateau gradient changes proportionately with changes in airway resistance. The plateau pressure reflects the pressure required to overcome elastic forces during inspiration and depends upon tidal volume and respiratory system compliance (lung and chest wall). Inspection of the airway pressure trace during volume-controlled ventilation can therefore provide useful information about the elastic and resistive properties of the respiratory system (Figure 1.5).
In pressure-controlled ventilation (and in dual modes) the inspiratory pressure is constant throughout the inspiratory phase and it is not possible to differentiate the elastic and resistive properties of the lung from observation of the airway pressure trace. However, most ICU ventilators will calculate and display continuously values of respiratory system resistance and compliance.
The modern ICU ventilator offers a complex and wide range of ventilation modes. Unfortunately, there is little standardization of terminology between the different ventilator manufacturers, and modes that are essentially the same may have different names depending upon the particular ventilator.
Although there are a large number of modes available, there is little evidence that one mode improves clinical outcomes compared with another mode. The majority of patients can be managed with two modes: one controlled mode and one spontaneous mode.
Controlled mandatory ventilation (other terms conventional mechanical ventilation, IPPV, volume control)
This describes a fully controlled mode where the respiratory rate is set by the ventilator and the breaths are usually volume-targeted. Cycling from the inspiratory to the expiratory phase is usually by time or by volume. There is no interaction or synchronization with the patient’s efforts, and additional breaths cannot be triggered. Conventional mechanical ventilation (CMV) is typically used when patients are anaesthetized and paralysed (e.g. in the operating theatre). CMV may also be applied with a pressure-targeted breath (pressure control) (Figure 1.6).
Assist volume control (other terms CMV assist, IPPV assist)
This is similar to CMV but the patient is able to trigger ventilator breaths. All breaths are volume-targeted. If the patient does not make any respiratory efforts the respiratory rate remains at the set frequency and the mode is effectively identical to CMV. Any spontaneous effort will trigger a ventilator breath of the preset tidal volume. If the patient’s respiratory rate is above the preset frequency, the mandatory breaths are inhibited and the mode functions as a spontaneous mode with all breaths being patient-triggered. Cycling may be by volume or time (Figure 1.7).
Assist pressure control is analogous to assist volume control but with a pressure-targeted breath. An inspiratory pressure is set and the delivered tidal volume will depend upon the respiratory system compliance and resistance. Cycling is always by time, which differentiates this mode from pressure support where cycling is by flow.
Synchronized intermittent mandatory ventilation
In this mode, a mandatory frequency of breaths is delivered; this will be synchronized to any spontaneous patient effort occurring within a time window that follows the preceding breath (trigger window). If the patient breathes at a rate greater than the set rate, the additional breaths will be allowed but unsupported unless pressure support is also set. Thus, a mixture of two breaths is delivered, mandatory ventilator breaths and pressure-supported breaths for any efforts above the mandatory frequency. Cycling of the mandatory breaths is by time, while the pressure-supported breaths are flow-cycled. When first developed, the mandatory breaths in synchronized intermittent mandatory ventilation (SIMV) were volume-targeted. This mode is now available with pressure-targeted breaths (SIMV pressure control) (Figure 1.8).
Dual control modes
This term is used to describe modes that deliver breaths that are both volume- and pressure-targeted; these are becoming the default initial mode of ventilator support in the ICU. A range of names is used by the manufacturers, including AutoFlow®, Pressure Regulated Volume Control®, volume-assured pressure control, and volume control plus®.
The ventilator delivers a breath with the characteristics of a pressure-targeted breath (constant inspiratory pressure and decelerating flow profile) combined with a guaranteed tidal volume (within certain pressure limits). On commencing a dual mode, a constant flow volume controlled test breath is delivered first to assess the compliance and resistance. Subsequent breaths are pressure-controlled, with the inspiratory pressure automatically selected to ensure delivery of the set tidal volume. The ventilator monitors the tidal volume on a breath-by-breath basis and will automatically change the inspiratory pressure (usually in 2–3-cmH2O increments) to ensure that the desired tidal volume is always delivered. In dual modes, the maximum inspiratory pressure that may be delivered is a function of the high airway pressure alarm setting. This varies between manufacturers but is usually 5 or 10 cmH2O less than the high-pressure alarm limit setting. Dual control can be applied to a wide range of modes, including volume control, assist volume control, SIMV, and pressure support.
Pressure support (other term assisted spontaneous breathing)
This is a spontaneous mode where the ventilator provides an inspiratory assist by increasing airway pressure to a set level following each patient effort. The pressure support equals the difference in pressure during inspiration and expiration (positive end-expiratory pressure [PEEP]). In this mode, there is no set rate and the ventilator will only deliver inspiratory assistance in response to patient efforts. It may be combined with a back-up mandatory mode (apnoea ventilation) that will take over if the patient has a prolonged period without any inspiratory effort (e.g. >15 s) (Figure 1.9).
In pressure support, cycling is by flow. The inspiratory flow follows a decelerating profile that is monitored by the ventilator. In most ventilators, cycling to expiration occurs when inspiratory flow has fallen to 25% of the peak inspiratory flow. Some ventilators offer the facility to adjust the flow at which cycling occurs (e.g. 10–90% of peak flow). This may be useful for improving patient ventilator synchrony when the duration of the inspiratory phase is either longer or shorter than desired by the patient.
Bilevel ventilation (bilevel positive airway pressure, duo positive airway pressure)
This can be considered as a mode of ventilatory support where there is cycling between two different levels of CPAP at the set ventilator frequency. There is no synchronization with spontaneous respiratory efforts although the patient can breathe without support at any time. An adaptation of this mode may offer pressure support for spontaneous breaths during the lower pressure phase. If the patient makes no spontaneous efforts, bilevel ventilation functions in an identical manner to pressure control. Proposed advantages of bilevel ventilation are improved patient comfort, as spontaneous breathing is allowed at any point of the respiratory cycle, and better gas exchange, as maintaining spontaneous breathing enhances ventilation–perfusion matching.
Airway pressure release ventilation
During airway pressure release ventilation, a high level of CPAP (e.g. 20-30 cm H2O is maintained for periods of 5-6 seconds with brief episodes when the airway pressure is allowed to fall in order to augment CO2 clearance. The duration of the pressure releases need to be short enough (e.g. 0.5–1.0 seconds) to ensure that significant derecruitment does not occur and are set by observation of the expiratory flow waveform ensuring that full lung emptying has not occurred. It is important that spontaneous breathing is maintained throughout to ensure CO2 control. Airway pressure release ventilation (APRV) allows the application of a high mean airway pressure to improve lung recruitment and oxygenation while limiting peak airway pressure. It has been used in refractory hypoxaemia associated with ARDS, atelectasis, and morbid obesity (Figure 1.10).
Automatic tube compensation
The resistance of the endotracheal or tracheostomy tube increases the work of breathing, and this is reflected by a pressure gradient across the tracheal tube. This pressure drop, which changes during the respiratory cycle in proportion to the gas flow, can be estimated continuously by the ventilator. With automatic tube compensation (ATC), the ventilator increases airway pressure during inspiration and reduces airway pressure during expiration to offset the estimated pressure gradient across the breathing tube. ATC effectively removes the imposed work of breathing from the ETT and can be used during spontaneous breathing trials (SBTs) to assess readiness for extubation.
An increasing degree of automation is now available in ICU ventilators. For example, in adaptive support ventilation®, the target minute volume is set by the clinician and the ventilator will automatically switch between volume-targeted pressure control breaths and pressure support according to patient effort. Fully automated ventilation has been developed (Intellivent®) whereby all ventilator parameters are automatically adjusted according to the continuous measurements of oxygen saturation (SaO2) and end-tidal CO2 (ETCO2) by pulse oximetry and capnography. The inputs by the clinician are limited to the patient’s height and the target limits for SaO2 and ETCO2. Automated pressure support weaning is available where the ventilator is set to attempt to wean the patient by automatically reducing pressure support as long as the patient does not develop rapid shallow breathing (Smart Care®). Once the pressure support has been reduced to a minimal value the ventilator will display a prompt indicating that the patient may be ready for extubation.
Chose a mode of ventilation that is familiar and appropriate for the clinical situation. This will usually be a controlled mode (e.g. volume or pressure control) as the patient will often have received sedative agents and muscle relaxants to assist intubation.
The following parameters are set.
Regardless of mode used, the desirable tidal volume should be based on the patient’s ideal body weight (IBW). This can be calculated from the patient’s height using the following formulae:
In normal lungs, most authorities would recommend a tidal volume of 6–8 ml/kg IBW, while in patients with ARDS a tidal volume of 6 ml/kg is appropriate.
When using pressure control modes, the inspiratory pressure is adjusted once ventilation is commenced to achieve the appropriate tidal volume.
This is set with the aim of maintaining pH within normal limits (7.35–7.45). With a trend to using smaller tidal volumes to limit ventilator-induced lung injury, higher respiratory rates (e.g. 15–25) need to be applied to ensure delivery of adequate minute ventilation. If there is underlying metabolic acidosis (e.g. septic shock or cardiogenic shock), a high respiratory rate should be used to lower PaCO2 and provide some respiratory compensation and minimize the fall in pH. Conversely, in a patient with a metabolic alkalosis secondary to longstanding CO2 retention, a lower respiratory rate will need to be set to maintain a normal pH. As the underlying metabolic abnormality improves, the respiratory rate will need to be adjusted to ensure that pH remains within the normal range. Other factors that will influence the set respiratory rate are the patient’s total CO2 production and state of the lungs. Severe lung disease may increase alveolar dead space, significantly reducing the effective minute ventilation.
When adjusting the respiratory rate, the expiratory waveform should be inspected to assess adequacy of the expiration. High respiratory rates will limit time for expiration and may result in breath stacking if expiration is not completed before the onset of the next inspiration (Figure 1.10). Patients with obstructive airways disease are at particular risk of breath stacking and hyperinflation that will result in high airway pressures, high measured intrinsic PEEP, haemodynamic instability, and impaired CO2 clearance from increased alveolar dead space. In these circumstances a low respiratory rate should be set to limit this dynamic hyperinflation, which can be monitored by measuring intrinsic PEEP level following an end-expiratory pause.
In controlled modes of ventilation, the duration of the inspiratory phase is preset. How this is achieved varies according to the individual ventilator and may be by setting total inspiratory time, by selecting the inspiratory:expiratory (I:E) time ratio, or from the inspiratory flow rate (volume control modes). Appropriate initial settings are:
• Inspiratory time 1.0–1.5 s.
• I:E ratio 1:2–1:3.
• Inspiratory flow rate 30–60 l/min.
If using a ventilator where the I:E ratio is directly set, this may need to be adjusted following changes to the respiratory rate to ensure an inspiratory time >1 s. Inspiratory times of <1 s may not allow adequate time for ventilation of lung units with long time constants, resulting in impaired gas exchange.
In volume modes, inspiration is divided into an active inspiratory phase and the end-inspiratory pause. Setting an inspiratory pause during volume-controlled modes allows the plateau (end-inspiratory pause) pressure to be measured. An inspiratory pause may improve gas exchange.
Some ventilators allow the profile of the inspiratory flow waveform to be changed in volume-controlled modes. Options may include constant, decelerating, and sinusoidal flow patterns. Constant flow ventilation is the most commonly used, although decelerating flows are increasingly available and have the advantage of reduced peak airway pressure and improved patient ventilator synchrony. Sinusoidal flow mimics the normal flow pattern during spontaneous breathing although it is unknown if it is associated with any clinical benefits.
Inspired oxygen tension
In an unstable patient, ventilatory support should be commenced with an FiO2 of 1.0. This can then be adjusted according to arterial saturations recorded by pulse oximetry and from blood gas analysis. A saturation >92% and a PaO2 >8 kPa are appropriate targets in the majority of patients. To avoid oxygen toxicity, inspired oxygen tension should be adjusted to the lowest level that maintains these values.
The initial PEEP setting is 5 cmH2O in the majority of patients. This maintains the ‘physiological PEEP’ that occurs in spontaneous breathing due to exhalation through a partially closed glottis.
Patient ventilator synchrony
During spontaneous modes of ventilation, it is important to set the ventilator to ensure optimal synchrony between patient and ventilator. Dyssynchrony may manifest as patient effort not initiating inspiration (missed breaths), autotriggering where inspiration is initiated without any patient effort, and delayed cycling to expiration when the patient has to exhale actively during ventilator inspiration. Careful observation of the patient together with the ventilator graphic waveforms will allow dyssynchrony to be recognized. It can be limited by optimal setting of the inspiratory and expiratory trigger.
The mode of triggering (pressure or flow) and sensitivity may be adjustable according to the ventilator used. The trigger sensitivity should be set to ensure that all spontaneous patient efforts are detected by the ventilator. Autotriggering occurs if the trigger is too sensitive when the ventilator delivers a breath in response to minor fluctuations in airway pressure caused by patient movement, airway manipulation, etc., and not from patient inspiratory effort.
A number of ventilators allow the threshold for cycling from inspiration to expiration during pressure support ventilation to be adjusted. The expiratory phase is usually trigged by a fall in inspiratory flow rate to 25% of peak inspiratory flow rate. However, in patients with obstructive airways disease peak flow rates are low and have not fallen to 25% before the patient starts to exhale. As a result, patients may have to exhale actively to initiate cycling to expiration, which may contribute to respiratory distress and failure to wean. This active exhalation is recognized by noting a brief end-inspiratory increase in airway pressure. Increasing the flow rate at which cycling to expiration occurs (e.g. 50%) may improve patient comfort. Patents with restrictive lung disease may be more comfortable with a longer inspiratory phase by reducing the flow rate at which expiratory cycling occurs (e.g. 10%).
The alarm parameters vary according to the individual ventilator. Usually there will be alarms to monitor the following:
• Exhaled tidal volume.
• Minute ventilation (high and low).
• Airway pressure (high and low).
• Respiratory rate (high and low).
An appropriate alarm setting for each parameter is when the measured value deviates by 25% from the desired setting. Patient disconnection will activate a number of alarms, including low pressure, low tidal volume, low minute ventilation, or low rate (or apnoea). The high-pressure alarm limit is usually set to activate at between 30 and 40 cmH2O. If set too close to the peak airway pressure, it will activate frequently, which could result in an inadequate minute ventilation as the ventilator will cycle to expiration as soon as the alarm is triggered. When using dual control modes, the high-pressure alarm limit will set the maximum working pressure. Depending on the ventilator, it will be 5–10 cm lower than the set high-pressure alarm limit. For most patients, the high airway pressure alarm should be set to allow a maximum working pressure of 30 cmH2O.
The ventilator should be checked before connecting to the patient. Many ventilators include an automatic preuse test that occurs whenever first switched on. Typically, this includes a circuit check (leak and compliance) and sensor (pressure, flow, and oxygen) calibration. Once the initial settings have been set, the ventilator can be connected to the patient. Clinical assessment after commencing ventilatory support should include:
• Adequacy and symmetry of chest wall movement.
• Synchronization of the ventilator with the patient’s efforts.
• Vital signs, including heart rate and blood pressure.
• Gas exchange (pulse oximetry, capnography, and arterial blood gases).
The expired tidal volume displayed by the ventilator should be checked and should be like the set tidal volume. Airway pressures should be monitored with the aim of keeping inspiratory plateau pressure as low as possible and certainly <30 cmH2O in the majority of patients.
Only two parameters directly influence arterial oxygen tension: the inspired oxygen concentration and the mean airway pressure.
Mean airway pressure can be adjusted by prolonging the inspiratory time (adding an inspiratory pause) or by applying PEEP. Increasing mean airway pressure may increase haemodynamic instability as it may reduce venous return. Increasing PEEP will increase peak airway pressures and increase the risk of barotrauma. Increasing inspiratory time may result in gas trapping owing to the reduction in expiratory time. Prolonged exposure of normal lungs to high inspired oxygen concentrations is damaging and, although it is unknown whether oxygen has the same effect on abnormal lungs, it would seem prudent to minimize the inspired oxygen concentration. A PaO2 of >8 kPa (SaO2 >92%) is a satisfactory target in the majority of patients and is achieved by a combination of selecting an appropriate FiO2 and level of PEEP.
When assessing a patient with hypoxaemia during mechanical ventilation, reversible causes should always be considered (such as endobronchial intubation, atelectasis secondary to sputum plugs, and pulmonary oedema).
The CO2 tension is influenced by minute ventilation, dead space, and CO2 production. Minute ventilation is adjusted by changing the tidal volume and/or respiratory rate to maintain pH within the normal range (7.35–7.45). If this cannot be achieved without exposing the patient to excessive tidal volumes (>6–10 ml/kg depending on underlying lung diagnosis) or high airway pressures (plateau pressure >30 cmH2O), it is invariably safer to limit tidal volumes and accept the associated respiratory acidosis. This permissive hypercapnia is well tolerated unless the patient has raised intracranial pressure (ICP) (e.g. head injury) and is associated with an improved outcome in acute lung injury and acute severe asthma.
An increase in dead space will raise PaCO2. Reversible causes of increased pulmonary dead space include low cardiac output, hypovolaemia, and high ITPs (secondary to externally applied PEEP or intrinsic PEEP from gas trapping). Equipment dead space may be minimized by removing the catheter mount and using a water bath humidifier rather than an HME. Reducing CO2 production with therapeutic hypothermia combined with deep sedation and muscle relaxation may be of value in managing severe respiratory acidosis in the difficult-to-ventilate patient (e.g. severe asthma).
When the high airway pressure alarm is activated, the ventilator immediately cycles to expiration that will reduce the inspired tidal volume and rapidly results in the patient receiving inadequate ventilation. With the increasing use of pressure and dual modes of ventilation, the high-pressure alarm is rarely activated. However, the ventilator will be unable to deliver the desired tidal volume and will alarm to indicate a low tidal or minute ventilation.
Causes of high airway pressures/reduced tidal volume include:
• Low lung compliance (e.g. ARDS).
• gas trapping
• excessive tidal volumes.
• Low chest wall compliance
• morbid obesity
• intra-abdominal distension
• chest wall rigidity (e.g. secondary to high-dose opiates).
• Increased airway resistance
• airway obstruction (secretions)
• airway occlusion due to compression/kinking.
• Patient ‘fighting the ventilator’
• agitation, coughing, straining.
Reversible causes of high airway pressures should be treated if possible (e.g. removal of secretions, administration of bronchodilator).
Patient ventilator asynchrony
This describes poor synchronization between the patient’s inspiratory efforts and inspiration applied by the ventilator. It is common and has several adverse effects, including increased work of breathing, impaired gas exchange, increased requirement for sedation, and prolonged weaning from mechanical ventilation.
When severe, it presents as the patient ‘fighting the ventilator’ with failure to settle, frequent coughing, straining, and agitation. It is more common when first commencing a patient on ventilatory support and results in poor gas exchange as minute ventilation is not maintained owing to frequent activation of the high-pressure alarm.
A number of factors may contribute to asynchrony:
• Inadequate trigger sensitivity (missed breaths).
• Inspiratory flow waveform (e.g. constant flow).
• Inspiratory flow rate does not match patient effort.
• Rise time setting (time to reach maximum inspiratory pressure).
• Lung pathology (e.g. high resistance and compliance).
• Auto PEEP (impairs ability to trigger).
• High inspiratory effort/drive (e.g. metabolic acidosis).
• Sedation level, inadequate pain control.
Asynchrony may occur at different times in the respiratory cycle:
• Triggering of inspiration.
• During active inspiratory flow.
• Termination of inspiratory flow.
Treatment of patient–ventilator asynchrony requires recognition and correction of the underlying cause(s) with optimal setting of the ventilator. However, increasing the levels of sedation and the administration of a non-depolarizing muscle relaxant may be required, particularly when first commencing respiratory support.
Barotrauma is damage caused to body tissues when exposed to cross-tissue differences in pressure. Whilst such pressure-induced injury can affect many areas of the body, this review will focus on pulmonary barotrauma.
Mechanical ventilation, although life-saving in the setting of critical illness, has been shown also to cause significant lung injury. Normal physiology sees lung tissue exposed to cyclical negative pressures to effect ventilation. The use of intermittent positive airways pressure, particularly to abnormal or diseased tissue, is unphysiological and causes further tissue injury as a result of pressure differences, excessive stretching, or sheer stresses.
The significance of such ventilator-induced lung injury (VILI) has become more widely acknowledged in recent years, with increasing interest in lung protection strategies. Published literature would suggest that the incidence of VILI in ARDS is decreasing; data from Anzueto et al., the ARDS Network, and Brochard et al. suggesting 6.5%, 11%, and 13%, respectively, contrast with data of 10–15 years ago suggesting an incidence of 40 and 60%.
Barotrauma presents both macroscopically and microscopically; manifestations of macroscopic barotrauma include pneumothorax, pneumomediastinum, and pulmonary interstitial emphysema. Microscopic pressure-related lung injury, presenting as clinically similar to a non-specific diffuse acute lung injury, appears related to overpressure, overdistension, or shear stresses, through repetitive mechanical ventilation. This has received increased attention in recent years with the development of more advanced ventilator technology and a better understanding of the importance of lung protection strategies.
New terminologies have been developed to better describe microscopic barotrauma, directed toward the underlying mechanisms. Barotrauma remains the broad umbrella term for pressure-related injury. Volutrauma seeks to better describe tissue injury induced predominantly by excessive or repetitive stretching of tissues, albeit induced by cross-tissue pressure differences. Similarly, atelectrauma is a term used to describe the lung injury seen where there is repetitive opening and collapse of alveoli with associated sheer or tearing stresses.
Pulmonary barotrauma results when gas escapes from the distal fragile lung tissue, the alveolus. Overdistension of a normal alveolus can result in rupture with consequent release of gas. Equally, lesser distension of a pathological alveolus such as in acute or chronic disease states can result in rupture. Hence diseased lungs may be at greater risk of barotrauma than healthy lung, even when ventilated within parameters that may be regarded as normally safe. Alveolar rupture can allow diffusion of gas into the perivascular adventitia, with consequent pulmonary interstitial emphysema. Frank release of gas into adjacent potential spaces will manifest as pneumothorax, pneumomediastinum, etc. The key pathological process is an imbalance of pressure across the distal lung tissue, producing stretch that is sufficient to tear the normal or abnormal alveolus. Factors associated with barotrauma include high plateau pressures, elevated positive end-inspiratory pressures, both external and intrinsic, high end-inspiratory lung volumes, and repetitive opening and collapse of alveoli and terminal airways. High peak airway pressures, if not reflected down to the terminal airways sufficiently to produce the above, such as with high resistance states, are not usually associated with an increased risk of barotrauma.
Patients will usually be those with diseased lungs for whom ventilation has been challenging. Those at highest risk of barotraumas from mechanical ventilation are those with acute lung injury or ARDS.
Coexisting lung pathology such as interstitial lung disease, COPD, Pneumocystis carinii pneumonia, or blunt thoracic injuries increases the risk.
Patterns of presentation range from the asymptomatic to full cardiac arrest from an unrecognized tension pneumothorax. The severity depends on the degree of extra-alveolar air present. Signs of respiratory distress, e.g. ventilator–patient dyssynchrony, use of accessory muscles, etc., may be the earliest clinical manifestation in the patient unable to communicate. The earliest clinical signs of a pneumothorax may be decreased breath sounds and hyperresonance on percussion. These signs are often less apparent in the mechanically ventilated compared with the conscious self-ventilating patient.
A systemic gas embolus is the most dramatic extrathoracic manifestation of barotrauma, presenting as cerebral air emboli, myocardial infarction, or livedo reticularis.
The increased ITPs from mechanical ventilation affect venous drainage from other sites, with reduction in venous return and increased venous pressures from the brain and abdomen.
Biotrauma in the lung increases leukocytes, tumour necrosis factor, interleukin (IL)-6, and IL-8. These are the same cytokines implicated in the systemic inflammatory response syndrome and in sepsis.
A high clinical index of suspicion with radiological confirmation is required for diagnosis in an asymptomatic patient.
Radiological findings in pulmonary interstitial oedema (PIE) include:
• Parenchymal cysts.
• Lucent lines directed towards the hilum.
• Subpleural air cysts.
• Presence of gas around large vessels.
• Pneumomediastinum outlining the great vessels.
• Pneumopericardium outlining the pericardium and contiguous diaphragm.
Pneumothoraces, especially small ones, may be difficult to diagnose on portable chest radiographs, particularly in the supine position common in critically ill ventilated patients.
Computed tomography (CT) scanning is rarely indicated to establish the diagnosis of barotraumas, but it may be helpful in determining the size of a pneumothorax and is often an incidental finding when imaging for other indications. It is also easier to appreciate pneumothoraces that are primarily anterior or basilar than on a two-dimensional chest radiograph. CT scans may be useful in guiding the placement of thoracostomy tube(s) in loculated pneumothoraces or where draining more than one tube is required.
Protective mechanical ventilation
Lung protection strategies are essential to avoid overdistension of alveoli and to minimize barotrauma. Useful parameters include a plateau pressure <30 cmH2O, a tidal volume 6–8 ml/kg (IBW), together with carefully titrated external PEEP and avoidance of intrinsic PEEP or hyperinflation. Modest hypercapnia in the absence of significant acidaemia (pH >7.2–7.25) is regarded as a lesser evil and should be tolerated to keep within such mechanical parameters. The effect of such hypercapnia and modest acidaemia on inflammation is unclear; the evidence available to date is conflicting.
The pressure–volume curve of the lung has both upper and lower inflection points. The upper inflection point is the pressure at which the lung volume ceases to increase sharply with rising airway pressure. The lower inflection point is the pressure at which the lung volume begins to decline sharply with falling airway pressure. Lung injury might be avoided by ventilating the lung at a PEEP above the lower inflection point to prevent atelectrauma while restricting tidal volumes so that the end-inspiratory pressure does not exceed the upper inflection point.
Eisner et al. in the ARDS Network trial showed that, by decreasing tidal volume from 12 to 6 ml/kg, the mortality rate fell from 40% to 30% in ARDS patients. An important component of this trial was the selection of PEEP and FiO2 parameters.
This provides the best approximation of transalveolar pressure. Amato et al. have shown that failure to limit plateau pressures is associated with a high incidence of barotraumas and an increased mortality. It is widely agreed that plateau pressures >30–35 mmH2O lead to increase incidence of barotraumas.
Only rarely is surgical repair of the lung required for the treatment of barotrauma. However, the effective management of pneumothorax requires evacuation of pleural air and placement of a pleural drain to permit the gas to escape. The urgency and type of tube placement depends on the patient’s clinical status. In most instances, leaks associated with ventilator-induced barotrauma are small, and tension pneumothoraces develop slowly.
Many commercial kits are available using a Seldinger technique as an alternative to the traditional intercostal drain insertion using blunt dissection.
Emergency needle thoracostomy
This is indicated for patients with a tension pneumothorax and cardiovascular compromise requiring immediate decompression. In mechanically ventilated patients, the ventilator should be removed and replaced with a bag valve device connected to oxygen. In this way, the clinician can assess the lung compliance and eliminate the deleterious effect of PEEP on the cardiovascular system. Following emergency needle decompression, a thoracostomy tube placement is required.
Weaning is the gradual reduction in ventilator support that leads to re-establishment of spontaneous breathing. Weaning occupies >40% of the ICU stay and is therefore of both clinical and economic importance. There are several steps in the weaning process and patients can be classified into simple, difficult, and prolonged according to ease and time taken to pass through that process. Whilst the majority of patients progress steadily through the weaning process it is important to identify those who are at high risk of delay/failure early so that additional support can be implemented if required.
The weaning process
The progressive reduction in the degree of ventilatory support that leads to the re-establishment of spontaneous breathing (SB) is conventionally termed weaning. Liberation might be a better descriptive as we now better appreciate the potential risks of invasive mechanical ventilation (IPPV). These include VILI and VAP, and these risks of delayed extubation need to be balanced with those attributed to failed extubation, with reintubation being associated with up to 50% mortality. The start of weaning coincides with clinical stability and the beginning of recovery from the critical illness that precipitated ICU admission. When weaning is prolonged, it extends into the period of general rehabilitation, e.g. sitting out of bed or even ambulatory mechanical ventilation. Weaning is unnecessary when mechanical ventilation is employed, for instance, to manage major elective surgery. In such circumstances, controlled ventilation may be employed to allow for the initial recovery from surgery and is followed by cessation of sedation, establishment on low level pressure support, and rapid extubation.
Weaning can be classified according to difficulty and duration. Simple weaning (about 60% of patients) accounts for those who extubate successfully after the first weaning attempt and SBT. Difficult weaning (30–40% of patients) incorporates patients who require up to three SBTs (or up to 7 days) prior to successful extubation. Prolonged weaning (6–15% of patients) represents those exceeding the limits of difficult weaning. Patients still requiring at least 6 h a day of ventilation after day 21 are termed long-term weaners. Weaning delay is associated with a high mortality, with those in the prolonged category having the worst outcomes in terms of length of stay and survival when compared to those in the simple or difficult group. Skeletal myopathy is almost universal in critical illness and involvement of the respiratory muscles is a major factor in weaning delay. The role of disuse atrophy versus diaphragmatic dysfunction is uncertain. The link between paralytic agents, steroids, and critical illness neuropathy/myopathy (CINM) may be more an indicator of severity of illness rather than purely a causative factor. The balance between reduced respiratory pump function and an increased ventilatory load, arising as a result of airflow obstruction and reduced lung, chest, and abdominal wall compliance, will determine whether weaning can successfully proceed or whether improvement is required before progress can be made.
Causes of weaning delay/failure
• Unresolved primary illness.
• Premorbid disease (neuromuscular disease, obstructive sleep apnoea, cardiac failure or severe COPD).
• Weak muscles ±ineffective cough (bulbar disease or depressed respiratory drive, CINM, malnutrition).
• CNS (brain injury, sedation, anxiety depression).
• Physician-related (failure to recognize imbalance between pump and load, ventilator–patient dyssynchrony, or fluid overload).
Two contrasting philosophies lie behind the strategies that can be adopted. When the load/capacity ratio is unfavourable, a slowly progressive reduction in pressure support is appropriate. If disuse atrophy and diaphragm retraining is required, intermittent tracheostomy mask SB or CPAP trials might be a better strategy. Direct measurements of diaphragm strength and ventilatory load are, however, technically demanding and have not yet been shown to influence outcome, so what evidence is there for choosing the right strategy?
Trials of SIMV with pressure support (and a gradual reduction in the number of mandatory breaths and in degree of pressure support) versus either progressive reduction in pressure support or intermittent CPAP/T-tube SBTs have demonstrated that SIMV is a poor weaning strategy. The relative merits between progressive pressure support reduction and intermittent SBTs have not yet been established. However, among long-term weaners duration of weaning was shorter among those weaned using unsupported SB via a tracheostomy (up to 12 h if tolerated) when compared to those weaned with pressure support (gradual reduction attempted three times daily).
Non-ventilatory aspects of weaning
Optimal fluid balance, treatment of heart failure, ensuring adequate nutrition, and avoiding, or treating, nosocomial infection are all important. Psychological and rehabilitation aspects are equally so. For instance, improving communication, with a speaking valve in the ventilator tubing, managing anxiety and delirium, and re-establishing the sleep/wake cycle are crucial steps in rehabilitation. Feelings of dependency and fear may need to be addressed.
IPPV/assessment of weaning
In 2007 six stages of weaning were proposed Figure 1.11. Non-invasive ventilation has been added. In stage one (treatment of ARF) the emphasis is on treatment of the underlying process that prompted intubation. Whilst weaning is not the main focus it does not preclude titration of ventilation if this is possible. The next stage (clinical suspicion) is when the clinician thinks that liberation from the ventilator is a possibility and coincides with evidence of a response to treatment such as a reduction in inotrope requirement or FiO2. At this point supported SB should be encouraged. VILI remains a consideration. Pressure support should be titrated to limit the tidal volume to between 6 and 8 ml/kg whilst monitoring for signs of fatigue as indicated by a rising respiratory rate, blood pressure, or heart rate, or the development of agitation or a change in conscious level. As the patient improves there is a period during which pressure support and PEEP are reduced. This may be a continual and progressive process or be associated with setbacks when more support will be required, e.g. a septic episode. Initially controlled ventilatory modes may be better at night to ensure that the gains by day are not compromised by inadequate support during sleep.
A daily readiness screen (stage 3) is encouraged to identify both those ready to be liberated from the ventilator (thus avoiding delayed extubation) and those not suitable for extubation, thereby preventing the risks of premature extubation. Ventilator diagnostics/weaning predictors are useful at this stage to help determine readiness but are beyond the scope of this chapter.
Suggested readiness to wean criteria:
• Resolution of the underlying cause for respiratory failure.
• Cooperative patient requiring no sedation (score ≥ – 2), adequate analgesia.
• PaO2/FiO2 ≥26.6 kPa or SpO2≥90% on FiO2 ≤0.4 and PEEP ≤8 cmH2O.
• Haemodynamic stability (no or low-dose inotropes).
• Fluid balance optimized.
• Adequate cough; secretions not excessive.
• Able to initiate an inspiratory effort.
• Haemoglobin ≥70–80 g/l.
• No uncorrected metabolic abnormalities.
• Core temperature between 36°C and 38°C.
In those that fulfil the discussed criteria a trial of SB is undertaken (an SBT Stage 4). This involves assessment for extubation or self-ventilation in the case of a patient with a tracheostomy. Clinical stability over 30 min in the absence of signs of fatigue or distress is required before proceeding to extubation (stage 5). Whilst the SBT is valuable, ~10% of patients will fail postextubation, usually because of an inability to clear secretions, CNS factors, or unexpected upper airways obstruction due to glottic oedema. By contrast, many COPD patients will wean despite a respiratory rate during the SBT of >35 breaths/min.
‘Facilitative’ non-invasive ventilation (NIV) speeds weaning by acting as a bridge to SB in COPD and is probably effective in neuromuscular causes of respiratory failure. The role of NIV to facilitate weaning in other groups of patients who fail an SBT has not been shown to improve time to liberation from all forms of ventilation. NIV to provide postextubation support ‘Prophylactic NIV’, can reduce reintubation rates when used early in patients who are at high risk of requiring reintubation (e.g. previous extubation failure, hypercapnia, chronic heart failure). NIV used in postextubation respiratory failure (‘rescue NIV’) is not recommended in the general ICU population as it may lead to delayed reintubation and increased mortality.
Some patients capable of proceeding quickly are held back inappropriately. Weaning protocols are effective by the earlier detection of such patients. These should at least include a daily sedation hold and assessment of readiness to wean. ‘Smart’ ventilators capable of adjusting the degree of pressure support to the arterial CO2 tension and tidal volume targets may be more effective than simple protocols.
Animal observations and experiments in the early and mid-twentieth century demonstrated that adequate gas exchange could be maintained using modes of ventilation quite outside the physiological range. Ventilatory rates above 2 Hz (120 breaths/min) are generally termed high-frequency ventilation (HFV). Most HFV is delivered at rates greater than 4 Hz. Tidal volumes are generally <1–2 ml/kg (less than physiological dead space). Both positive- and negative-pressure HFV systems have been developed. The Hayek Cuirass has been used to provide HFV by applying negative pressure to the anterior chest and upper abdomen at frequencies of between 0.5 Hz and 3 Hz but is now used at more conventional frequencies except for sputum clearance.
Positive-pressure HFV is generated using pulses of high-velocity gas injected into an endotracheal tube (high-frequency jet ventilation [HFJV]) or by means of the forward and backward movement of a diaphragm or piston within a high bias-flow humidified CPAP circuit (high-frequency oscillatory ventilation [HFOV]). HFOV has an active expiratory phase owing to the backward movement of the diaphragm or piston and is the more commonly used system in critical care.
HFJV is now mostly used to provide ventilation during endobronchial procedures. Gas is entrained by the injected high-velocity jet from a ventilatory circuit (which may be that of a conventional mechanical ventilator, a T-piece, or CPAP system).
HFOV is well established in paediatrics but has never been in widespread use in adult critical care and the indications for its use remain controversial. In adults, positive-pressure HFOV is provided via an endotracheal tube (translaryngeal or through a tracheostomy) using one of two marketed devices: the SensorMedics 3100B or the Novalung R100 (which can provide conventional IPPV as well). Both devices consist of a high bias-flow humidified CPAP circuit, which has an incorporated diaphragm that is oscillated electromagnetically (SensorMedics) or a piston driven pneumatically (Novalung). The pressure in the CPAP circuit is maintained and controlled with the volume of bias flow and leak from the circuit exhaust valve.
A variety of other similar modes of ventilation have appeared over the past 40 years, which have included high-frequency (conventional) positive-pressure ventilation and high-frequency percussive ventilation, and a combination of HFV and conventional IPPV, but these are rarely used.
Rationale for using HFV
HFV, particularly HFOV, has been advocated as a lung-protective form of ventilation because lungs are not exposed to the repetitive stress (or shear forces) of overinflation and atelectasis and re-expansion. Continuing maximum recruitment of lung units (and oxygenation) is maintained by a high value of PEEP and small, frequent tidal volumes (oscillations) provide dead-space replenishment and CO2 removal.
Two recent large, multicentre, randomized trials of HFOV against controls in whom, now standard, lung-protective ventilatory strategies were used for the management of ARDS failed to show any benefit. Moreover, one of the trials was stopped early because of an excess of mortality in the HFOV group. Both studies were published in the same issue of the New England Journal of Medicine and an accompanying editorial recommended, not unreasonably, that clinicians should be cautious about using HFOV routinely in the management of adult patients with ARDS. Its use in paediatric practice has also recently been questioned after a review of just over 900 patients treated with HFOV from a database of more than 9000 patients between the ages of 1 month and 18 years in North America. However, the study has the limitations of being retrospective and excluded patients under 1 month.
Advances in extracorporeal gas exchange interventions, such as venovenous ECMO and other pumped and unpumped forms of extrapulmonary gas exchange, the availability of more conventional ventilatory modes that aim to maintain maximum alveolar recruitment (such as APRV), and the resurgence in the use of prone positioning have all reduced the enthusiasm for HFV.
The question of whether HFOV is beneficial for specific patients with specific pathophysiological conditions remains unanswered by these trials. A recent meta-analysis of trials where HVOV was compared with conventional ventilatory management suggests a small mortality benefit for those with severe ARDS (PaO2/FiO2 <100 mm Hg) but harm in those with mild to moderate disease.
HFOV and HFJV have been successfully used to manage patients with bronchopleural fistula following lung resection. Both modes of ventilation provide a method for maintaining alveolar recruitment and gas exchange in the presence of a large-volume airway leak.
Mechanisms of gas-exchange in HFV
As the bulk movement of gas into and out of the lungs with HFV is generally much less than with conventional ventilation, other mechanisms of gas transport come into play. Seven different mechanisms are described. Essentially, proximal dead space is exchanged with fresh gas by agitation which sets up complex currents of gas flow and molecular diffusion accounts for exchange of gas between the terminal bronchioles and the alveoli. Cardiac movement also causes dead space gas mixing.
Improved oxygenation is a result of a decrease in shunt and is directly related to the proportion of perfused versus non-perfused lung units. Recruitment and prevention of de-recruitment is maintained by a high and almost constant mean airway pressure, which must be adjusted to prevent overdistension. In HFJV, the mean airway pressure is maintained by the jetting gas and entrainment. The pressure is increased by increasing the frequency and increasing the I:E ratio. In HFOV, the mean airway pressure is adjusted relatively independently by changing the bias gas flow or the leak from the circuit.
The elimination of CO2 is less dependent on ventilation–perfusion matching and is (counterintuitively) inversely proportional to the ventilatory frequency. This is because the gas jet (in HFJV) or oscillation generator (in HFOV) creates greater forward and backward flow of gas at lower frequencies. Increasing the power of the jet or oscillations also increases alveolar ventilation. An increased I:E ratio reduces CO2 removal. Occasionally, cuff deflation, which allows greater dead space washout, is useful in resistant hypercapnia.
Practical aspects of HFV
HFOV is normally started in response to a perceived failure in conventional ventilation. Occasionally it may be instigated as the initial mode in patients with respiratory failure following pulmonary resection with or without bronchopleural fistula. Patients on HFV and especially HFOV need to be sedated and usually paralysed initially as respiratory efforts interfere greatly with the efficacy of the ventilation.
If a patient is to be transferred from conventional ventilation to HFV the mean airway pressure is set to approximately 5 cmH2O above the previous pressure or 20 cmH2O. The frequency is set to between 5 and 6 Hz and the power adjusted to achieve visible oscillatory movements at the level of the upper thigh. Progressive recruitment is achieved with gradual increases of the mean airway pressure. Overdistension of the lungs leads to an increase in the proportion of non-perfused lung units (hence, an increase in dead space). High ITPs can drastically reduce venous return and compromise right heart function. This is further compromised by the need for paralysis and the loss of the pump effect of inspiratory effort.
HFV requires either high bias flows of gas in the ventilatory circuit (HFOV) or involves the injection of a high-velocity gas jet. Humidification of airway gases is, therefore, essential and involves heated humidifiers, heating of the circuit tubing, and sometimes (for HFJV) incorporation of a water feed into the injector.
Whilst HFJV offers a very low resistance to breathing if it is used with an appropriate breathing circuit, HFOV does not allow for patients to take significant breaths as it leads to a loss of the recruiting pressure. Thus methods to suppress spontaneous breathing are usually required when HFOV is used, particularly if the patient is relatively hypercapnic.
Complications of HFV
The two 2013 trials of HFOV against conventional ventilation in ARDS patients did not demonstrate an increase in pulmonary complications such as pneumothorax or refractory barotrauma. In the trial that used higher mean airway pressures in the HFOV group there was a greater use of inotropic and pressure supportive drugs and an increased mortality. In both studies, the HFOV groups received more neuromuscular blockade and sedation.
Care and experience are needed to guide recruitment strategies and to prevent overdistension, which may be indicated by decreased CO2 removal. High ITPs may severely compromise venous return. Desiccation of respiratory secretions (and, with HFJV, desiccation of tracheal mucosa and necrosis) can occur. This can lead to occlusion of the endotracheal tube or large airways and may also produce worsening CO2 removal. Because monitoring of intrapulmonary oscillatory volumes is problematic, airway occlusion may occur without warning, although it is usually predicted by a progressive restriction in visible abdominal and upper thigh oscillatory movement.
APRV has been implemented variably by ventilator manufacturers, which has resulted in significant confusion. The precise definition of APRV is a pressure-limited, time-cycled, open mode of ventilatory support conceived of and proven to prevent, arrest the development, and/or reverse early acute lung injury from evolving into ARDS. It achieves this by minimizing ventilator induced lung injury (VILI) (perhaps better called physician-induced lung injury!). It is best considered as a form of high level continuous positive airway pressure (CPAP) augmented by periodic, very short, pressure releases to wash out dead space gas. These releases should not be considered as tidal, mandatory/ventilator breaths nor should this mode be considered extreme inverse ratio ventilation. Conceptually, it is closer to high frequency oscillatory ventilation (HFOV) than conventional ventilation. Gas exchange is achieved by enhancing diffusive gas transport such that less convection (alveolar minute ventilation) is required for equivalent CO2 clearance. It is intended and, indeed, highly beneficial that the patient breathe spontaneously during this open mode of ventilation, with no patient–ventilator synchrony. Even if such efforts generate very small tidal volumes, these efforts create valuable intrapulmonary gas mixing and dependent area recruitment and retention. In addition, spontaneous breathing efforts have a myriad of cardiovascular and respiratory muscle benefits. Although arguably most efficacious in the presence of spontaneous breathing, APRV can be successfully employed in its absence.
APRV evolved from three complementary observations:
1. The expiratory limb of the lung hysteresis curve demonstrates that ‘expiratory’ ventilation (respiratory system recoil) is more efficient (requires less energy) than ‘inspiratory’/tidal ventilation.
2. High levels (sufficient to move functional residual capacity [FRC] toward total lung capacity) of CPAP improve oxygenation by recruitment and retention of functional units, thereby improving ventilation–perfusion matching. In addition, high levels of CPAP decrease both the static and the dynamic heterogeneity of respiratory unit volumes, resulting in less mechanical biotrauma. Furthermore, this reduces pulmonary endothelium to epithelium fluid flux and surfactant deactivation.
3. CPAP applied at two levels (e.g. 5 and 15 cmH2O) in a time-cycled manner independent of a patient’s spontaneous breathing efforts (where the time cycle is long e.g. 5–15 seconds) is a very comfortable and effective mode of ventilatory support that improves both oxygenation and CO2 clearance.
APRV is indicated in patients requiring ventilatory assistance, especially if they are at risk of developing, or have already started to develop, ARDS. APRV has proven efficacy in preventing the development/evolution of ARDS and is most effective when used at the earliest possible time point in the patient’s journey, i.e. as the first mode of ventilatory support to be applied. It can be used as a rescue modality but, like all such interventions, it cannot reverse established pathology; it may, however, minimize further injury. Although surprising to many, it is very well tolerated by the majority of patients and usually requires minimal sedation.
How to initiate APRV
APRV can be delivered either invasively or non-invasively. Having selected an appropriate fraction of inspired oxygen for the clinical circumstances, there are only three variables to set on the ventilator, the distending pressure, Phigh, the number of releases per minute, Thigh, and the duration of each release, Tlow. Plow, the pressure during the release, should always be set as zero. De-recruitment is prevented by the very short duration of Tlow. No tube compensation or spontaneous inspiratory pressure support should be set (if made available by the ventilator’s software) as this is an open mode of ventilatory support and thus the ventilator will provide up to its maximal inspiratory flow to maintain the set Phigh in response to any spontaneous effort.
Typical starting values of Phigh are 20–30 cmH2O. If converting from a pressure control/support mode, set Phigh at the peak airway pressure; if from a volume control mode, set at the plateau pressure. Starting with these pressures may unmask relative hypovolaemia and/or right ventricular failure. Appropriate monitoring/investigations and interventions should be prepared in advance to deal with this contingency.
Typical starting values of Thigh are 4–6 s. The effectiveness of Phigh is often enhanced by longer values of Thigh, such that to gain the same benefit with shorter values may require higher pressures. However, the optimal target is that Phigh should be maintained for 90% of the time, i.e. Thigh/(Thigh + Tlow) × 100 = 90. Hence the same effect can be achieved for a given Phigh and a shorter Thigh by shortening Tlow.
Tlow should be set to 0.2 s and titrated up/down until the termination of the release coincides with 75% of the peak expiratory flow rate (PEFR). To determine this value, record/freeze the ventilator screen after a series of releases and use the cursor function on the expiratory flow-against-time graph to determine the value of the PEFR. Multiply this value by 0.75 and set Tlow such that the ventilator cycles back to Phigh at (or before) the PEFR has decayed to this value.
How to improve oxygenation/optimize the Phigh setting
Start by titrating the FiO2 to achieve a stable peripheral capillary oxygen saturation (SpO2) of 87–90% (i.e. a steeper portion of the oxygen saturation curve). Perform a recruitment manoeuvre by increasing Phigh by 4–10 cmH2O then, if tolerated, Thigh to 15–30 s for 1–5 min. Decrease Thigh back to its previous value, then perform a decremental Phigh trial by reducing the level of Phigh in a stepwise fashion (e.g. 2 cmH2O every 5–10 min). Ensure the Tlow is shortened as the Phigh is reduced such that the transition from release to Phigh never occurs beyond 75% of the PEFR. Establish the lowest Phigh value associated with the highest SpO2. Consider repeating the recruitment manoeuvre, then set the Phigh to this optimal level or 2 cmH2O above. Subject to patient comfort and/or acidosis, increasing the Thigh and/or decreasing the Tlow (to achieve the 90% Phigh time target) may also improve oxygenation.
Consider permissive hypercapnia. If problematic (e.g. patient agitation and/or cardiovascular instability), start by ensuring Tlow is not <75% of PEFR. Then progressively decrease Thigh, i.e. increase the number of releases per minute. If feasible, reduce sedation/analgesia to increase SB efforts. If efforts are present but tidal volumes are very low, consider starting automatic tube compensation and/or minimal inspiratory pressure support (ensure the trigger and inspiratory termination settings are optimized) to increase spontaneous tidal ventilation. If this fails, consider switching to HFOV or a conventional mode of ventilatory support.
How to wean
As oxygenation improves, repeat the Phigh optimization assessment until the level required is ≤10–20 cmH2O. Remember to reassess/reduce Tlow every time you reduce Phigh. As spontaneous tidal volume/PaCO2 improves, gradually increase Thigh to reduce the number of airway releases. Once achieved, perform a trial of conventional CPAP and inspiratory pressure support, progressing to unsupported breathing.
PEEP is the positive pressure applied at the end of expiration during mechanical ventilation. When PEEP is applied to an SB cycle it is called CPAP. Physiological consequences of the application of PEEP depend on its effects on gas exchange, pulmonary compliance, and systemic haemodynamics.
Effects of PEEP on gas exchange
Hypoxaemic ARF is characterized by an acute reduction of lung volume due to pulmonary oedema, atelectasis, and pneumonia; alveolar units tend to collapse at the end of expiration, particularly in gravitationally dependent regions. The use of PEEP improves gas exchange by recruiting functionally closed alveoli, redistributing lung water, and therefore reducing ventilation–perfusion (V/Q) mismatch.
Alveolar recruitment, defined as the amount of non-inflated tissue that is re-expanded by a given level of PEEP, results in a more homogeneous distribution of tidal volume and in an increased FRC. Moreover, application of PEEP leads to the redistribution of lung water from the alveolar space to the perivascular interstitial space and improves V/Q mismatch by diverting bloodflow from shunt regions to normal V/Q regions.
All these effects are strongly conditioned by whether or not PEEP recruits previously collapsed alveoli. If PEEP induces alveolar recruitment, a reduction in shunt with improvement in oxygenation and reduction in dead space is expected. If hyperinflation of normal alveoli is the predominant effect, a rise in PaCO2 owing to the increase in dead space is expected, while PaO2 will slightly change according to the variations in cardiac output.
Effects of PEEP on respiratory mechanics
The analysis of volume–pressure curves (V/P) of the respiratory system is a key point to understand the physiological effects of PEEP on respiratory mechanics. In patients with ARF, reduced compliance of the respiratory system corresponds to small variations in volume occurring for each unit of change in pressure.
The inflation limb of the V/P curve shows: 1) a ‘lower inflection point’ (LIP) where compliance suddenly improves; and 2) an ‘upper inflection point’ (UIP) corresponding to the pressure where compliance starts to deteriorate. In normal subjects, the LIP occurs below the volume at FRC and the UIP occurs at a volume close to total lung capacity. In patients with ARDS, both occur within tidal ventilation and indicate the pressure at which recruitment of collapsed alveoli and at which hyperinflation starts to occur, respectively. Ventilation using levels of PEEP lower than the LIP may exacerbate lung injury by shear forces applied during repeated opening and closing of the alveoli; therefore, tidal excursion during mechanical ventilation should occur between LIP and UIP.
Compliance of the respiratory system (i.e. tidal volume divided by plateau pressure minus PEEP) represents a good estimate of the elastic properties of the respiratory system.
Effects of PEEP on compliance
• Increase in compliance suggests recruitment.
• No change in compliance indicates that ventilation is occurring on the linear portion of the V/P curve.
• Decrease in compliance indicates lung overdistension and risk of barotrauma.
The interpretation of the effects of PEEP requires assessment of both lung and chest wall mechanics. In patients with a stiff chest wall because of increased intra-abdominal pressure, underestimation of LIP is expected and, consequently, relatively higher levels of PEEP should be applied. In the presence of a stiff chest wall, the pressure required to expand the lung may exceed the limit of 30 cmH2O without risk of alveolar overdistension.
Effects of PEEP on systemic haemodynamics
Application of PEEP increases ITP, diminishing venous return to the right heart and increasing the right ventricular afterload. These haemodynamic consequences depend on previous ventricular loading conditions, ventricular function, and respiratory system compliance. An adequate circulating blood volume is required before application of PEEP to prevent this depression of right ventricular output.
Application of PEEP decreases left ventricular stroke volume, while heart rate does not change significantly. The reduction in stroke volume is mainly because of a decrease in left ventricular preload. This effect may be explained by different mechanisms: decreased venous return to the right heart; increased right ventricular afterload; decreased ventricular compliance; and decreased ventricular contractility.
In patients with normal cardiac function, the main consequence of increased ITP is the reduction of venous return; in patients with poor left ventricular function or congestive heart failure, the increase in ITP reduces the left ventricular transmural pressure leading to a reduction in left ventricular afterload and improving left ventricular function. In these patients, cardiac output is relatively insensitive to reduction in venous return because diastolic volume is elevated.
PEEP in the clinical setting
The Berlin definition of ARDS, assessing PaO2/FiO2, includes a minimal level of 5 cmH2O of PEEP. The application of PEEP in ARDS is aimed at preventing the end-expiratory collapse of the lung to reverse severe hypoxaemia resulting from pulmonary shunting. Moreover, since mechanical ventilation with a high tidal volume and a low level of PEEP is associated with pulmonary injury (VILI) indistinguishable from ARDS, the use of PEEP may reduce VILI by preventing the cyclical opening and closing of the alveoli. The potential benefit of the ‘open lung approach’ lies in the avoidance of recruitment–de-recruitment of partially consolidated areas, avoiding their exposure to shear stress and miminizing the local inflammatory reaction.
A protective ventilatory strategy, limiting the tidal volume to 6 ml/kg and the plateau pressure to <30 cmH2O, is the gold standard of ventilatory strategy in patients with ARDS. Titration of the optimal level of PEEP must be a compromise between the minimal PEEP providing maximum oxygen delivery at the lowest airway pressure and high PEEP keeping the lung fully recruited at end-expiration.
Patients with moderate or severe ARDS (PaO2/FiO2 ratio <200 mmHg) may benefit from the application of higher levels of PEEP.
In patients with an exacerbation of COPD, the rate of lung emptying is impaired because of increased respiratory resistance and expiratory flow limitation. End-expiratory lung volume is expected to be higher than resting volume, a condition called dynamic hyperinflation. As a consequence, a positive pressure called intrinsic PEEP (PEEPi) is present at the end of expiration. PEEPi represents an inspiratory threshold load to be counterbalanced by the patient’s inspiratory muscles to initiate inspiration or to trigger the ventilator. In the presence of PEEPi owing to expiratory flow limitation, externally applied PEEP below measured PEEPi (85% of total PEEP) does not cause hyperinflation or an increase in ITP and does not affect respiratory mechanics and haemodynamics. During the assisted mode of ventilation, application of PEEP reduces inspiratory muscle effort and improves patient–ventilator interaction.
Acute severe asthma
The main cause of dynamic hyperinflation and PEEPi during acute exacerbation of asthma is the increased expiratory resistance in the absence of flow limitation. Consequently, flow continues to the very end of expiration, driven by the difference in pressure between the alveoli and the airway opening. In this situation, use of external PEEP provides a back pressure to respiratory airflow, causing parallel increases in lung volume and airway, alveolar and thoracic pressure. In these patients undergoing mechanical ventilation, the use of external PEEP is often detrimental and should only be used if expiratory flow limitation is proved.
Adverse effects of PEEP
Hepatic and renal perfusion
PEEP decreases splanchnic bloodflow and may compromise hepatic perfusion and portal venous drainage. This effect may be explained by a reduction in cardiac output, coupled with elevation of central venous pressure and an increased outflow resistance because of a direct compressive effect of the diaphragm on the liver. The resulting passive hepatic congestion can cause moderate elevation of bilirubin and hepatic enzymes. PEEP may also interfere with renal function, even when cardiac output is preserved, by reducing renal bloodflow. The impact of PEEP on renal bloodflow is dependent on the volume status and applied pressure.
The application of PEEP may affect the cerebral circulation by haemodynamic and CO2-mediated mechanisms. The haemodynamic mechanism may alter cerebral circulation both on the arterial side, reducing arterial pressure, and on the venous side, reducing cerebral venous drainage. According to the concept of the vasodilatory cascade, the decrease in arterial pressure caused by PEEP may diminish cerebral bloodflow in patients whose cerebral autoregulation is impaired. It may cause a compensatory vasodilation if autoregulation is preserved. In the latter case, vasodilation will lead to an increase in cerebral blood volume and ICP, given reduced intracranial compliance. However, application of PEEP does not induce a significant reduction in arterial and cerebral perfusion pressure if euvolaemia is maintained.
CPAP is the application of positive pressure throughout the respiratory cycle. It can be delivered non-invasively or invasively in spontaneously breathing patients. It is functionally similar to PEEP, but the ventilator does not cycle and generate added pressure above CPAP, thus the patient must be able to generate an adequate tidal volume independently. CPAP is applied by generating a constant flow of an oxygen/air mixture, with a resistance applied in the circuit to create PEEP. It is primarily used to improve oxygenation in patients with distal airways collapse for a variety of reasons, thereby improving oxygen exchange in selected patients whose hypoxaemia does not respond to merely increasing FiO2. Through recruitment of atelectatic/collapsed lung, CPAP increases FRC and improves lung compliance, resulting in reduced work of breathing, an increase in minute ventilation through lung expansion, and improved oxygenation.
Alternatively, CPAP is also effective for maintaining patency of the collapsible upper airway in obstructive sleep apnoea hypopnoea syndrome (OSAHS) and related sleep-disordered breathing (SDB).
CPAP increases intra thoracic pressure (ITP), which may reduce venous return and preload to the heart. However, this can be beneficial in offloading the left atrium/ventricle in cardiogenic pulmonary oedema associated with acute left ventricular failure. Furthermore, CPAP reduces ventricular transmural pressure and reduces left ventricular afterload. Conversely, in patients with relative intravascular depletion, the application of CPAP may cause a further reduction in ventricular preload, stroke volume, cardiac output, and blood pressure.
1. ARF (usually normocapnic but even in acute hypercapnic ARF if because of reversible basal atelectasis/collapse).
a. Postoperative—for assisted ventilation as an adjunct to lung expansion manoeuvres and analgesia (i.e. physiotherapy, incentive spirometry, intermittent positive pressure breathing).
b. Acute respiratory insufficiency (e.g. pneumonia, inflammatory lung diseases, acute lobar collapse)—when simply increasing FiO2 is insufficient to maintain adequate oxygenation.
c. Inadequate oxygenation but adequate ventilatory drive.
d. Chest wall trauma with persistent hypoxaemia, despite high FiO2 and analgesia (care with potential pneumothorax).
2. Cardiogenic pulmonary oedema—unresponsive to initial medical management (i.e. increased FiO2, nitrate-containing venodilators, diuretics).
3. OSAHS and related conditions of SDB.
4. Weaning from mechanical ventilation—prior to extubation/tracheostomy decannulation or prior to step down to oxygenation only.
1. Coma/impaired consciousness.
2. Inability to maintain/protect own airway.
3. Fixed upper airway obstruction.
4. Haemodynamic instability/severe comorbidity (relative).
5. Head/facial trauma.
6. High risk of aspiration—including patients with gastrointestinal bleeding. A nasogastric tube reduces the risk.
7. Mechanical bowel obstruction.
8. Recent upper gastrointestinal surgery.
9. Copious secretions—risk of inability to clear with tight-fitting mask.
The equipment required is a CPAP machine, circuit, PEEP valves, HME filter, heated humidifier, full facial mask, straps, oxygen analyser and alarm, power supply, and oxygen source (Figure 1.12).
Venturi devices (i.e. WhisperFlow 2™ or Vital Signs™ CPAP systems; Figures 1.13 and 1.14) use an oxygen supply in conjunction with entrained air to generate an output. It can generate flows >140 l/min with a minimal FiO2 of 28%. These systems have internal pressure relief safety valves for pressures >28 cmH2O. The use of pressure and oxygen monitors ensures measured pressures approximate to the specified PEEP and FiO2.
Portable CPAP flow generators used in the treatment of OSAHS produce lower flow rates (i.e. 30 l/min).
Masks and headgear
In the acute hospital/ICU setting, full face masks are preferred to maintain a closed system. In contrast, the outpatient population with OSAHS generally prefer nasal interfaces with mouth closure. Careful daily inspection is required to prevent skin sores. Use of a sponge or Granuflex® is advised. Dentures are maintained in situ if possible. Special attention should be paid to leaks into the eyes and around the mouth. Pressure monitoring is advisable to check the delivered PEEP.
High-flow nasal oxygen
High-flow nasal oxygen therapy involves the delivery of heated and humidified oxygen through nasal cannulae at high flow rates of up to 60 l/min. By changing the fraction of oxygen in the delivery gas an accurate FiO2 can be set. The high flow rates generate low levels of PEEP, which may generate alveolar recruitment. Other mechanisms through which oxygenation is improved include reduction of oxygen dilution, dead space washout, and reduction in inspiratory nasopharyngeal resistance. High-flow nasal oxygen is better tolerated than oxygen via face mask due to comfort and the ability to communicate and cough more easily. In observational studies, when compared to standard oxygen therapy through a face mask, high-flow nasal oxygen was associated with decreased work of breathing, improved oxygenation, and decreased respiratory rate in adults with hypoxaemic respiratory failure owing to a variety of causes.
Checklist prior to commencement
• Refractory hypoxaemia? No hypercapnia?
• Optimization of standard management.
• Reversible component?/Need for further step-up in care.
• Correct setting for CPAP delivery, i.e. high dependency unit (HDU)/ICU.
• Check and set up equipment.
• Initial settings, e.g. PEEP 5 cmH2O, FiO2 maximum.
• Nasogastric tube necessary?
• Explanation to patient.
• Monitoring clinical/physiological parameters.
Application of CPAP in clinical settings
Acute respiratory insufficiency as a bridge to recovery
Early clinicophysiological improvement, based upon subjective parameters, respiratory rate, and oxygenation improve, as compared with face mask oxygen alone. However, CPAP may not reduce intubation rates in acute lung injury. There remains some debate as to the minimum effective PEEP level to produce a CPAP effect on oxygenation that is not merely a high-flow phenomenon. This is particularly so with the availability of high–low humidified oxygen systems through nasal cannulae. Indeed, a study in adults with acute hypoxaemic respiratory failure without hypercapnia found those treated with high-flow nasal oxygen had an improved survival rate and required fewer days of mechanical ventilation when compared to those treated with standard oxygen therapy or non-invasive ventilation. However, these results should be interpreted with caution as there was no difference in the primary outcome intubation rate.
Weaning from mechanical ventilation
The time spent on weaning as a percentage of total ventilator time remains about 40–50% and can vary from one to several days. CPAP has been used as part of SBTs prior to extubation. Flow-based CPAP is deemed superior to ventilator-delivered CPAP, although both provide a degree of inspiratory support to compensate for tube-related airway resistance during SBT (however, it may be argued that this reduces the sensitivity of the SBT). Apart from a small trial in morbidly obese patients post gastric bypass, there is no published robust evidence demonstrating an outcome benefit of CPAP used pre-emptively in postoperative patients. Nevertheless, intuitively and anecdotally, successes and reassurance are apparent, so the adage ‘absence of evidence does not mean evidence of absence’ is applicable in this setting.
CPAP as an adjunct to physiotherapy
A number of techniques are used to support the active cycle of breathing by physiotherapists and to promote lung expansion manoeuvres. These include IPPV, CPAP, positive expiratory pressure (PEP) devices (e.g. flutter valve), and cough in-exsufflators. The literature would suggest there is little to choose between them and personal preference/patient comfort are their determinants.
CPAP in OSAHS
OSAHS is the periodic reduction (hypopnoea) or cessation (apnoea) of airflow through repetitive closure of the upper airway during sleep. CPAP, delivered by a portable flow generator and mask, overcomes this periodic obstruction, thereby reducing episodic hypoxia and fragmented sleep. Since its first use in this setting in 1981, it has become the standard of care for domiciliary management of laboratory-diagnosed OSAHS. There is compelling evidence emerging of its short- and mid-term benefits in those with moderate to severe OSAHS in improving daily function and cardiovascular/cerebrovascular/metabolic risk profiles. The evidence in mild OSAHS is less apparent, although a trial of treatment is suggested if clinical symptoms persist in spite of a low apnoea/hypopnoea index. Patients’ tolerability of mask interfaces is sometimes problematic. This has driven the technical development of newer machines and masks for ease of increased nocturnal use.
The mechanism underlying hypoxaemia in acute lung injury is intrapulmonary shunt due to non-ventilated but perfused alveoli. Any manoeuvre that recruits lung volume and increases the amount of aerated lung can be expected to improve gas exchange. However, the challenge is to recruit atelectatic alveoli without overdistending normal lung units. Although PEEP is the most commonly used method to achieve or maintain recruitment, other techniques have been evaluated, including volume recruitment manoeuvres, prone position, and HFOV. Volume recruitment manoeuvres and periodic inflations to airway pressures of 35–45 cmH2O sustained for 5–30 s are used to reverse alveolar de-recruitment and to improve gas exchange while allowing acceptable ventilator pressure. In ARDS, inadequate levels of PEEP may cause tidal recruitment/de-recruitment of part of the consolidated lung and may expose these regions to shear stress. The combined use of low tidal volume and PEEP may minimize VILI. Driving pressure (ΔP = VT/CRS, where VT is tidal volume and CRS is respiratory system compliance) is more strongly associated with survival than VT or PEEP in patients who are not actively breathing.
Reductions in VT or increases in PEEP are beneficial only if associated with decreases in ΔP. Recruitment manoeuvres are procedures specifically designed to reach an opening pressure sufficient to open the collapsed lung regions that, in association with an increased level of PEEP postrecruitment manoeuvre, may improve gas exchange, reducing intrapulmonary shunt. However, the effectiveness of recruiting manoeuvres to improve oxygenation is also influenced by elastic properties of the chest wall since part of the pressure applied to the respiratory system during the recruiting manoeuvres can be dissipated against a stiff chest wall. Moreover, the response to recruitment manoeuvres is a function of the potential for recruitment (unstable units), the phase and extent of lung injury, the level of PEEP, and the characteristics of the recruiting technique.
In clinical practice, recruitment manouvre (RM) may play a role in the lung-protective ventilatory strategies based on the so-called open lung approach; however, their utility has not been confirmed by large clinical trials. Lung recruitment manoeuvres have also been successfully used in patients during general anaesthesia to restore the decreased CRS and oxygenation.
Several approaches have been used to perform the recruitment manoeuvres:
• Sustained inflation by application of CPAP at 30–40 cmH2O for 40 s.
• Incremental PEEP on maximum pressure: PEEP is increased in increments of 5 cmH2O from a baseline PEEP to 35 cmH2O, reducing VT to limit peak inspiratory pressure to 35 cmH2O. CPAP of 35 cmH2O is maintained for 30 s.
• Intermittent higher tidal volume in pressure-controlled ventilation applied with escalating PEEP and constant ΔP: peak pressure of 45 cmH2O, I:E ratio of 1:2, and PEEP level of 16 cmH2O for 2 min.
If gas exchange and lung mechanics improve substantially with recruitment manoeuvres, patients should be considered to have ‘high potential for recruitment’. However, in gauging response to PEEP, it is important to consider the oxygenation response as well as CO2 exchange: when PEEP is applied, PaO2 tends to increase and CO2 exchange may improve, reflecting increased alveolar ventilation.
If gas exchange and lung mechanics do not improve substantially with recruitment manoeuvres, patients are considered to have a ‘low potential for recruitment’, and the use of higher level PEEP may provide little benefit and actually be harmful since application of PEEP may cause overdistension.
Data suggest that patients with ‘high potential for recruitment’ are those ventilated for not more than 48–72 h, while patients with ‘low potential for recruitment’ are those ventilated for >3 days; the underlying disease responsible for ARDS (primary versus secondary ARDS) seems not to have a role in identifying responders or non-responders to recruitment manoeuvres.
Clinical evaluation of the response to recruitment manoeuvres and of their potential use can be evaluated at the bedside. More accurate evaluation of the percentage of potentially recruitable lung may be obtained by analysis of CT images or of respiratory mechanics.
After performing a recruitment manoeuvre, its efficacy can be evaluated at the bedside by looking at the values of plateau pressure (or VT if the patient is ventilated in a pressure-controlled mode), PaO2, and PaCO2. If the recruitment manoeuvre: 1) decreases plateau pressure (or increases VT if the patient is ventilated in a pressure-controlled mode); 2) increases PaO2; and 3) decreases PaCO2 (even if only by a few mmHg), patients may be considered responders.
Analysis of CT findings can identify the distribution of normally aerated lung regions in non-dependent regions and poorly aerated lung regions distributed in the dependent lung region. The CT analysis can assess the effects of the recruitment manoeuvres as well as their safety, and, most importantly, the adequate PEEP levels to keep the lungs opened after the recruitment manoeuvres. With an inspiratory and expiratory pause image acquisition of the lung, the regional VT distribution can be assessed.
Pressure–volume (P/V) curve
Analysis of P/V curves can confirm that the LIP and UIP correspond to CT scan evidence of atelectasis and overdistension. Alveolar recruitment is confirmed to occur continuously and along the inspiratory limb of the P/V curve, while the critical point for lung de-recruitment is identified below the point of maximum curvature of the deflation limb.
The analysis of the dynamic pressure/time (P/T) curve during constant flow ventilation (stress index) is a new parameter to identify the best compromise between alveolar recruitment and overdistension. The stress index is the exponent of the equation correlating the airway pressure profile and time during each VT. A stress index <1 is associated with recruitment, assessed by CT scan, whereas a stress index >1 is associated with hyperinflation. Modern ventilators are able to deliver square-wave inspiratory flow profiles and are equipped with monitoring that provides online dynamic P/T curves.
The potential risks of recruitment manoeuvres are barotrauma (high transpulmonary pressure) and haemodynamic derangement (high pleural pressure). In patients with low potential for recruitment (non-responders), application of recruiting manoeuvres may result in a substantial haemodynamic impairment. These effects are because of a reduced preload secondary to transmission of pleural pressure to intrathoracic structures and an increased afterload owing to increased lung volume. The degree of pleural pressure transmitted is higher in patients with a stiff chest wall than in patients with a normal chest wall. Therefore, the use of recruitment manoeuvres may be indicated only in patients with high potential for recruitment, keeping in mind that, after the manoeuvre, the level of PEEP has to be increased, otherwise its beneficial effects on gas exchange and respiratory mechanics will be lost in a few minutes.
ARDS is a syndrome characterized by impaired oxygenation, non-cardiogenic pulmonary oedema, decreased lung compliance, and reduced CO2 elimination that follows a direct pulmonary or systemic insult.
ARDS occurs in 10.4% of all critically ill patients and in 23% of patients who require mechanical ventilation. Seventy per cent of patients have moderate (46.6%) or severe (23.4%) ARDS and are at higher risk of death even when lung protective ventilation is applied with limitation of VT to 6 ml/kg of IBW and airway plateau pressure below 30 cmH2O.
The lung involvement in ARDS is heterogeneous and affects the dependent lung regions predominantly. This is because of three main factors: 1) gravitational forces; 2) the superimposed lung hydrostatic pressure; 3) the mismatch between the shape of the lungs and of the thoracic cavity that must contain them. This shape mismatch causes the ventral lung regions (with less alveolar mass; i.e. smaller) to overexpand to occupy the entire space offered by the chest wall, and the dorsal lung regions (with greater lung mass i.e. larger) to collapse to fit into the thoracic cavity.
Ventilation in prone position refers to the provision of mechanical ventilation with a patient lying in the bed on the sternum and abdomen facing downwards. Prone position is one of several adjunctive strategies utilized to improve gas exchange in refractory hypoxaemia or as a lung-protective strategy when elevated airway pressures are required to maintain adequate gas exchange.
Prone position improves oxygenation in patients with moderate or severe ARDS (PaO2/FiO2 <20 kPa or <150 mmHg) through changes in the distribution of ventilation and pulmonary blood flow, and a reduction in the harm caused by mechanical ventilation. The latter mechanism is consequent on a more even distribution of the alveolar sizes and a decreased stress–strain production during inspiration.
Physiological effects of prone position on gas exchange and lung mechanics
The size of the alveoli during lung inflation—once the resistive and frictional forces are dissipated—is determined by the difference between the airway pressure (which causes inflation) and pleural pressure (which opposes lung inflation). This difference is called transpulmonary pressure. Gravity is the major determinant of pleural pressure and the greater the gravitational force and the hydrostatic pressure of the weight of the lung above a given alveolar region, the lower the probability that those alveoli will be open and ventilated. In other words, because of gravity the size of the ventilated alveolar units decreases progressively going from the non-dependent regions to the dependent regions.
In supine position the non-dependent regions (ventral) are ventilated (often overinflated) and the dependent regions (dorsal)—subjected to compressive forces—are collapsed but perfused (areas of shunt). Prone positioning reverses the ventrodorsal gravitational vector and the effects of gravity on pleural and hydrostatic pressure.
The improvement in oxygenation and respiratory mechanics observed during prone positioning result from a synergistic interaction of the effects of the prone position on the lung parenchyma, chest and abdominal wall, the pulmonary circulation, and the postural drainage of secretions.
The gravitational forces exerted on the lung in prone position cause a redistribution of the regional alveolar ventilation, which favours a dorsal alveolar recruitment and limits overdistension of the ventral region. These effects synergistically increase the overall lung homogeneity, regional end-expiratory lung volume, and improve gas exchange. The increase in lung homogeneity is a consequence of five factors:
1. Reduction in the pleural pressure of the dorsal regions (i.e. increase in transpulmonary pressure for the same airway pressure).
2. Resolution of compression atelectasis in the dorsal regions caused by the superimposed hydrostatic pressure of heavy oedematous lungs and by the weight of the heart. In the prone position, the heart is dependent and the diaphragm is caudally displaced, reducing posterior compression of the lung parenchyma and improving regional ventilation in these areas.
3. Progressive and sustained lung recruitment. The more uniform distribution of alveolar pressure also prevents collapse of vulnerable lung units on expiration, maintaining alveolar recruitment.
4. Better shape matching between the lung and the thoracic cavity.
5. Increased postural drainage of secretions.
In patients with ARDS, the distribution of regional pulmonary blood flow is influenced by four main factors:
1. The extrinsic compression of the vessels by the pressure exerted by the weight of the lung and the pressure delivered by mechanical ventilation.
2. The geometry of the vascular and airway tree (fractal model).
3. The hypoxic vasoconstriction.
4. The hydrostatic pressure gradient due to gravitational forces.
Although gravity does not appear to be the dominant factor in determining the distribution of blood flow in humans (only 1–25% of pulmonary perfusion is mediated by gravitational forces), gravity causes compression of the dependent lung regions under the weight of the lung above, which decreases the size of the alveoli and compacts the alveoli and vessels together, increasing their density. Moreover, the dorsal regions have greater vascular capacitance and conductance to receive a greater proportion of the blood flow. In prone position the recruitment of the dorsal alveoli coupled with a relatively preserved dorsal perfusion leads to a reduction in the shunt fraction (i.e. proportion of pulmonary blood flow perfusing non-ventilated alveoli) and a more homogeneous distribution of perfusion. In addition, prone position seems to decrease pulmonary resistances and, in patients with preload reserve, increases cardiac output. The overall effect is a reduction in the intrapulmonary shunt and a reduction of the regional ventilation/perfusion heterogeneity.
1. Patients with moderately severe or severe ARDS with PaO2/FiO2 <20 kPa (150 mmHg), with FiO2 ≥0.6, PEEP ≥5 cmH2O, and a VT of 6 ml/kg of predicted body weight despite attempted optimization.
2. Requirement of higher inspiratory pressures (>30 cmH2O) to maintain acceptable gas exchange.
3. CT chest appearance (i.e. dependent consolidation) may encourage the use of prone position, but CT scan appearance alone does not reliably predict response to prone position and therefore should not in itself preclude prone position.
4. Right heart failure, which contraindicates recruitment manoeuvres.
Contraindications to prone positioning detailed in previous studies include few absolute contraindications: head injury—with documented or suspected raised ICP; severe facial trauma or facial surgery in the previous 15 days; unstable spinal, pelvic, or femoral fractures. The majority are, however, relative contraindications, which require additional precautions and monitoring. For example, tracheal surgery or sternotomy in the previous 15 days; cardiac pacemaker within the previous 2 days; severe haemodynamic instability (owing to the inherent difficulties in immediate access for cardiopulmonary resuscitation in the prone position); pregnancy (second/third trimester); recent abdominal surgery, recent stoma formation, open abdomen; recent cardiothoracic surgery/unstable mediastinum; new tracheostomy (<24 h).
The codelivery of lung-protective ventilation, low ΔP, and use of prone position has been shown to improve oxygenation (increased PaO2/FiO2 of 25–36%) in 60–70% of patients with ARDS.
The effects on oxygenation and mortality are significant in:
• Moderate and severe ARDS.
• When prone position is maintained for >16 h/day.
• Earlier in the course of ARDS (<3–7 days).
• Particularly if PEEP is >10 cmH2O.
The timing of oxygenation response is variable. Most patients respond within 1 h of prone positioning, while others may require up to 6 h.
The effects of prone positioning on CO2 depend on the combined changes of chest wall compliance, alveolar recruitment, and improvement in pulmonary vascular resistance. Responders to prone positioning tend to decrease PaCO2.
Three oxygenation response patterns have traditionally been described:
Non-responders (20%)—no improvement in oxygenation.
• Persistent: improvement in oxygenation is maintained on turning supine.
• Non-persistent: improvement in oxygenation is not sustained on returning to supine position.
A more recent randomized controlled trial (PROSEVA) has demonstrated clear mortality benefits at 28 days (16% prone positioning versus 33%) and at 90 days (21% prone positioning versus 41%). Patient-level meta-analyses have shown consistent reduction in mortality for the prone position group.
Technique of prone position
Prone positioning requires a concerted effort from all members of the multidisciplinary team. One airway consultant and six members of the medical staff are required for a safe procedure, and an ECMO nurse or perfusionist may be necessary if the patient is on ECMO. Specialist proning beds exist, but are uncommonly used and generally unnecessary.
Prone positioning requires no additional monitoring, although electrocardiograph (ECG) leads must be placed on the back. The need for endotracheal suctioning may increase in frequency and there needs to be strict attention to potential pressure areas, including regular repositioning of the patient, padding to vulnerable areas, and specific attention to ETT and catheter entry sites. As with the standard supine position, elevation of the head of the bed (reverse Trendelenburg) may reduce the risk of pharyngeal aspiration as well as help minimize ocular and facial oedema whilst prone. Despite this, however, enteral feeding may be more problematic in the prone position. In the presence of adequate levels of sedation, paralysis is not usually required and may have detrimental effects on diaphragm-related benefits. Having a prepared checklist may be useful to standardize the procedure and increase safety (Table 1.5).
Table 1.5 Pre-proning checklist
Pre-proning check list
Pressure area assessment
Eye and mouth care
Insert/perform closed suctioning
Check/change intravenous lines
Check/change wound dressings
Secure endotracheal tube/tracheostomy—check and document length at the teeth
Ensure chest X-ray is taken
Document grade of intubation
Stop nasogastric feed and aspirate nasogastric tube
Ensure safe movement of: chest drains, vascular catheters, essential intravenous lines
Disconnect non-essential intravenous lines
Disconnect electrocardiograph electrodes
Inform patient/next of kin
Careful attention must be kept on patient position during and after proning to minimize the risk of damage on joints and tendons and to prevent pressure-correlated lesions.
Swimmer’s position (swapping arm positions and turning head away from lifted arm) should be alternated 4-hourly to avoid pressure sores to the ears, cheeks, and neck.
After prone positioning, arterial blood gases should be taken for an initial assessment of effectiveness of treatment (rise in PaO2 and reduction in PaCO2). SpO2 should be monitored.
The optimal duration of prone positioning is still debated. Recent studies have preferred more prolonged periods (up to 20 hours per day), with a minimum of 16 consecutive hours a day until prone positioning stopping criteria. Although onerous, with appropriate equipment, guidelines, and training in place, prone positioning is a relatively simple and safe procedure with a low economic burden.
Complications observed with prone positioning
• Loss of airway and vascular access.
• Injury to cervical spine, shoulders.
• Increased sedation requirements.
• Transient hypoxaemia.
• Cardiac arrhythmia.
• Haemodynamic instability.
During prone position ventilation
Discontinuation of prone positioning
The prone position is terminated in any of the following conditions:
a. Termination owing to patient improvement: the daily realization of prone position sessions is interrupted if PaO2/FiO2 is ≥20 kPa with PEEP ≤10 cmH2O and FiO2 ≤60% in supine position (SP) at least 4 h after the end of a prone position session. If oxygenation is worsening in SP up to the point at which the original inclusion criteria are again present, the patient is turned to prone position. This strategy is followed every day until clinician decision to stop prone position. A minimum duration of 2–3 days proning is recommended—even if gases have improved—to allow lung homogenization and stabilization and therefore achieve greater lung protection.
b. Emergency termination of a session: any ongoing prone position session can be stopped, upon the clinician decision, because of a complication (unplanned extubation, tube displacement, ETT obstruction, SpO2 <85% for more than 5 min under FiO2 100%, cardiac arrest, bradycardia (heart rate <30/min for more than 1 min), or severe hypotension (systolic blood pressure <60 mmHg for more than 5 min) or for any other consultant decision.
Non-invasive positive pressure ventilation (NIPPV) is the delivery of mechanically assisted or generated breaths through a facial interface without the placement of an artificial airway such as an endotracheal tube (ETT) or a tracheostomy. It is an established alternative to invasive mechanical ventilation, for selected patients with acute respiratory failure (ARF), and as a weaning mode following extubation or tracheostomy decannulation. However, it does not replace invasive mechanical ventilation in patients requiring emergent endotracheal intubation.
NIPPV is a safe, effective technique that can avoid the side-effects associated with endotracheal intubation. It is primarily used to avert invasive mechanical ventilation in patients with early ARF and to prevent reintubation in patients with recurrent weaning failure. NIPPV preserves upper airway defence mechanisms, speech, and swallowing.
NIPPV improves alveolar ventilation, decreases work of breathing, and reduces intubation rates, length of hospital stay, morbidity (i.e. pneumonia), and mortality. These are greatest for patients with acute exacerbations of COPD associated with hypercapnic respiratory failure. There is also benefit in cardiogenic pulmonary oedema. Conversely, failure of NIPPV to prevent intubation has been associated with a higher mortality in patients with respiratory failure owing to other causes, emphasizing the need for careful patient selection and monitoring for markers of early treatment failure.
Other than the need for emergent intubation, contraindications for NIPPV include cardiorespiratory arrest, non-respiratory system organ failure (e.g. severe encephalopathy, haemodynamic instability, severe gastrointestinal bleeding), facial or upper airway trauma, postneurosurgery to the head, loss of airway patency, excessive airway secretions, or lack of cooperation with a high aspiration risk.
Severe respiratory acidosis is not a contraindication to NIPPV, so long as intubation is readily available in an ICU setting if the NIPPV trial fails. In a case–control study of 64 patients with COPD and severe hypercapnic respiratory failure (mean pH 7.18) who received NIPPV, 38% never required an ETT. Those who failed NIPPV and required intubation were not harmed by the delayed intubation and prolonged acidaemia.
Pressure-cycled modes (pressure support or bilevel positive airway pressure [BiPAP]) are preferred for patient comfort, although volume-cycled modes may further reduce the work of breathing. Delivery through standard ICU ventilators can offer time-cycled options to improve synchrony, precise O2 concentrations, and enhanced CO2 clearance. Full face masks, nasal masks, or alternatives such as helmets or nasal pillows/prongs are available. Initiation requires dedicated staff, awareness of mask-related complications, and troubleshooting skills. Early clinicophysiological assessment of success/failure by blood gases and respiratory rate at 1 hour is vital.
Complications associated with invasive ventilation (e.g. nosocomial pneumonia, barotrauma, haemodynamic instability) are less common in NIPPV. Local skin damage is related to pressure effects of the mask and straps. Cushioning the forehead and the bridge of the nose helps. Mask leaks are common and do not preclude NIPPV. Consider using different masks or ventilator settings. Eye irritation and sinus congestion may occur and may necessitate lower inspiratory pressures or the use of a facial mask rather than a nasal mask. Gastric distension occurs with some frequency but is rarely significant. Routine use of a nasogastric tube is not warranted. Barotrauma is uncommon in NIPPV when administered in the pressure support or bilevel modes.
The success of NIPPV depends on several factors, such as type of ARF, the underlying disease, location of treatment, and experience of the team. The timing of initiation and duration of use are also important for outcomes. If a trial of NIPPV is commenced there should be a clear plan of action should the patient not respond within a predetermined time, usually 1 hour. Guidelines for the use of NIPPV improve the utilization and process of care without changing clinical outcomes.
Acute hypercapnic respiratory failure
Patients with acute hypercapnic respiratory acidosis secondary to an exacerbation of COPD benefit the most from NIPPV, and are the best studied group in the context of randomized controlled trials and systematic reviews of NIPPV for ARF. Medical treatment failure rates vary between 27% and 74% from studies. The successful use of NIPPV can reduce mortality (relative risk [RR] 0.52), intubation rates (RR 0.41), and treatment failure rates (RR 0.48). It can also reduce length of hospital stay and may be cost-effective through reduced ICU admissions/tracheostomy rates. These benefits are for patients with mild to moderate ARF (i.e. pH <7.3 and >7.25) who can be managed on a dedicated general ward/intermediate care setting with trained staff. The benefits are less and potentially harmful (i.e. through delayed intubation) in more severely ill patients, who should be managed in a higher dependency setting with greater staffing and monitoring, thus allowing intubation without delay if deemed necessary. There is no evidence to suggest that NIPPV can prevent acute respiratory distress/failure in mild exacerbations (i.e. pH >7.35, respiratory rate RR <20).
There are no robust, easily reproducible clinical predictors of success/failure in acute exacerbations of COPD. Poor signs include an increase in respiratory rate and/or worse pH at 1–2 hours post initiation. A Glasgow Coma Score below 11 and/or an APACHE II index of >29 can predict increasing need for intubation. Persistent deterioration in physiological parameters despite optimal settings of NIPPV should prompt intubation if deemed appropriate.
Hypoxaemic respiratory failure
Although there is conflicting evidence, the efficacy of NIPPV in patients with hypoxaemic respiratory failure has been demonstrated in pneumonia (provided that patients can manage their secretions), immunosuppression, and following single lung resection. However, the studies have generally enrolled patients with moderate ARF, in whom emergent endotracheal intubation was not necessary. In a heterogeneous population of 105 ICU patients, bilevel NIPPV decreased the need for intubation (25% versus 52%) and the incidence of septic shock (12% versus 31%), improved ICU mortality (18% versus 39%), and increased 90-day survival versus high-concentration oxygen. The benefits of NIPPV in ARF (PaO2/FiO2 <33 kPa) due to pneumonia may be limited to those with underlying COPD although a trial in all-comers is justified.
Alternative to invasive ventilation
Weaning off invasive ventilation
Patients with COPD receiving invasive mechanical ventilation for >48 h and failing SBTs can be safely extubated to NIPPV, shortening ICU stay and reducing mortality. The recently published Breathe trial found that whilst patients who failed an SBT can safely be extubated onto protocolised weaning via non invasive ventilation, there was no difference in time from liberation from all forms of ventilation when compared to protocolised weaning via invasive ventilation.
NIPPV reduces reintubation rates when used early in patients at risk of postextubation failure (i.e. previous extubation failure, high APACHE score >12, hypercapnia, chronic heart failure, poor cough, stridor, comorbidities). However, it is ineffective and potentially harmful in postextubation respiratory failure owing to delayed reintubation. Thus, the early use of NIPPV is once again emphasized for best efficacy. It has been used successfully to preoxygenate patients prior to endotracheal intubation, compared with bag–valve–mask ventilation.
Other specific indications
Cardiogenic pulmonary oedema
NIPPV reduces the incidence of intubation (RR 0.52) and mortality (RR 0.66) compared with standard medical therapy and oxygen alone. It is equivalent to CPAP in terms of mortality and frequency of intubation, but may cause respiratory distress to resolve quicker. There appears to be no additional risk of acute myocardial infarction associated with the use of NIPPV in this setting.
Patients with ‘do not intubate’ orders
In patients unsuitable for, or declining intubation, for ARF, NIPPV can act as a bridge to recovery in >40%. It may also be used as palliation for breathlessness if tolerated.
Only one prospective trial has adequately assessed the value of NIPPV in acute asthma (not requiring emergent intubation). Using sham (subtherapeutic BiPAP) as the control arm, NIPPV improved lung function and reduced admission rates from the emergency room. A more recent meta-analysis highlights the paucity of data in this area.
In one small (n = 40) randomized controlled trial, NIPPV, when compared to high-concentration oxygen therapy, improved oxygenation and reduced the need for intubation, but there were many limitations. Therefore, there is insufficient information to recommend the use of NIPPV in this setting. If used, then an ICU setting with early recognition of failure and intubation is advisable.
ECMO is a technique that allows deoxygenated venous blood to be drained from a central vein and then reinfused—fully oxygenated and decarboxylated—back into either a central vein (venovenous [VV-ECMO]) or a central artery (venoarterial [VA-ECMO]).
Depending on its configuration (VV or VA) ECMO is used to support respiratory function, cardiac function, or both. When peripheral cannulae are used, the access cannula is placed percutaneously in a larger vein, either the femoral or jugular vein, and advanced into the superior or inferior vena cava to allow blood to be withdrawn from the vein into the extracorporeal circuit by a centrifugal pump and actively pumped through a membrane oxygenator. Within the membrane oxygenator, there is a capillary network with a blood path, a gas path, and a water path (to facilitate temperature control). Blood travels through the capillaries separated from the gas path via a semi-permeable membrane. The gas and the blood flow in different directions to create an effective countercurrent flow which enhances overall gas exchange. The oxygenated blood is then returned to the systemic circulation through a second, or return, cannula. The cannulae can be inserted percutaneously using a Seldinger technique (peripheral ECMO) or surgically implanted through sternotomy (central ECMO).
In VV-ECMO, the return cannula can be placed either in the femoral or jugular vein. Therefore, there are three possible configurations using two single-lumen cannulae: femorofemoral, femorojugular, or jugulofemoral; there is also a double-lumen cannula which is placed via the right internal jugular vein, passes transatrially, and the tip is in the inferior vena cava. During VV-ECMO, the extracorporeal pump works functionally in series with the patient’s heart and, as the volume of blood drained from and reinfused into the venous system are equal, there is no change in central venous pressure or ventricular filling. Although VV-ECMO does not directly support cardiac function, the pulmonary artery pressure can decrease owing to the increased venous oxygen tension and reduced venous CO2. This may in turn improve right heart function and hence overall cardiac function.
In VA-ECMO, the return cannula is positioned in the femoral (artery common iliac) in adults or in the carotid artery in children. VA-ECMO provides additional cardiac output, thereby preserving organ function, although at the cost of increasing afterload on the failing heart.
More complex configurations for peripheral ECMO include two cannulae providing drainage of the venous system and return into either a central vein or central artery, which is used when there is inadequate access, and a single venous access with a return divided into both venous and arterial cannulae, which is used to support the partially recovered heart in the absence of pulmonary recovery, thereby maintaining oxygen delivery to the brain.
ECMO does not provide any direct benefit to the heart or lungs; rather it provides support to other organs by adequate oxygen delivery and CO2 clearance. ECMO may provide an indirect benefit to heart or lung recovery by reducing the need for intrinsically harmful therapies (mechanical ventilation or inotropic therapy), which are traditionally used to ensure adequate organ perfusion in the presence of the failing heart or lungs. Hence ECMO is essentially a bridge, either to healing of natural hearts and/or lungs (bridge-to-recovery), or to long-term devices (bridge-to-destination), or to organ replacement (bridge-to-transplantation). ECMO may also be used for stabilization of the patient when the aetiology is unclear and a delayed evaluation is necessary (bridge-to-decision). For some patients, ECMO is a bridge to palliation when it becomes apparent that no therapeutic options are available.
Extracorporeal CO2 removal (ECCO2R) is a form of extracorporeal support that utilizes the fact that CO2 is more soluble in blood and therefore clinically significant amounts can be removed at lower blood flows (0.4–1 l/min). Because oxygen is predominantly attached to haemoglobin with very little being dissolved in blood, to provide significant oxygenation high blood flows are required. Hence ECCO2R has limited impact on oxygen delivery. Though CO2 removal is relatively independent of blood flow rate, it is dependent on the surface area of the gas-exchanging membrane. When the effective gas exchange area decreases, CO2 clearance is affected before oxygenation. The ECCO2R configuration includes an AV cannulation or a double-lumen VV cannula. In AV-ECCO2R flow is maintained by arterial blood pressure (pumpless system), whilst in the VV configuration there is a mechanical pump. By removing a proportion of total CO2 production, ECCO2R may allow lower tidal volumes and more protective mechanical ventilation (‘lung rest’) or avoid intubation in patients with severe exacerbations of COPD.
The ECMO circuit
ECMO has four main components:
1. Mechanical blood pump. ECMO pumps were initially base on roller pumps which utilized the physical compression of blood to allow it to move through a circuit. Modern ECMO pumps are centrifugal pumps, driven by electromagnetic induction motors to generate blood flow. Flow is entirely dependent upon unrestricted access to a blood reservoir (the venous system); should there be inadequate venous filling, blood flow will cease. The blood flow generated depends on the speed of rotation per minute (RPM) of the pump, the haemodynamic conditions (preload and afterload), and resistance through the cannulae and membrane oxygenator resistance.
2. Modern extracorporeal membrane oxygenators are made of capillaries formed from polymethylpentene with three paths—blood, gas, and water. Gas and blood are separated by a semi-permeable membrane that allows passage of O2/CO2 but not blood, thereby acting in a similar manner to the native lung, where air is separated from blood by the alveolar epithelium, interstitium, and capillary endothelium. The blood and gas move in opposite directions through the membrane (countercurrent), which enhances efficiency.
3. Heat exchanger, which uses a flow of temperature-controlled water to modify the temperature of blood through the membrane oxygenator.
4. Vascular cannulae and a system of tubing connecting the cannula drawing venous blood (inflow, access, or drainage cannula) and the one reinfusing oxygenated and decarboxylated blood into the systemic circulation (return cannula). Modern circuits have heparin bonding and a proprietary lining to reduce the likelihood of thrombus formation and activation of the immune system.
5. Circuit monitoring is usually pressure-based—measurement of access pressures, which are always negative, measurement of pressure between the pump and the oxygenator, and measurement after the oxygenator. The absolute pressure is dependent upon the RPM and blood flow through the circuit. Different information can be obtained from each of the pressures and the relative relationship between pressures. One of the key relative pressures is the transmembrane pressure, the difference between the preoxygenator and postoxygenator pressures. Sustained increases in transmembrane pressure at a constant RPM or flow indicate obstruction of the membrane oxygenator with thrombus.
Systemic oxygenation is complex. Bulk blood flow is a primary determinant of oxygen delivery per minute. Hence, as access to venous blood is one of the key determinants of flow through the circuit, it is also a key determinant of systemic oxygenation. However, with VV-ECMO, both deoxygenated (true venous) and oxygenated blood (from the circuit) will be drawn through the circuit. The fraction of oxygenated blood passing through is termed the recirculation fraction and results in inefficiency in the system. The key determinant of systemic oxygenation is the net blood flow (total—recirculation) from the ECMO circuit as a proportion of cardiac output. To achieve systemic saturations of 90% in the absence of any native lung function, net blood flow needs to be at least two-thirds of cardiac output. Clearly, if there is native lung function, then this will also contribute to overall systemic oxygenation. This fact can be used to assess native lung recovery by measuring the changes in systemic oxygenation with alterations in ventilator FiO2.
ECMO is indicated as a support for patients with severe but potentially reversible respiratory or cardiac failure. There is no widely accepted international indication for the use of ECMO for respiratory failure. In the UK, the national indication is a Lung Injury Score of 3 or more with potentially reversible respiratory pathology not improving with other means of respiratory support (for example, prone position and lung-protective ventilation) or a pH <7.20 due to respiratory acidosis (for example, severe asthma). Other potential benefits of ECMO include a bridge-to-heart or -lung transplantation, peritransplantation cardiorespiratory failure or severe bronchopleural fistulae.
The reversibility of respiratory failure is difficult to determine in adults and patients placed on ECMO, or where possible before ECMO, should have a diagnosis secured using imaging, microbiological, and serological investigations. Most prognostic indicators have come from studies of outcomes of patients on ECMO rather than prospectively for patients prior to consideration of ECMO. Although the generalizability of these studies is unclear, there is a consistent theme that longer duration of less protective ventilation (peak pressures >30 cmH2O, high FiO2 (>0.8–0.9), low PEEP, and not using prone position) is detrimental to patient outcomes.
ECMO patient management
Patients who receive ECMO are at risk of both thrombosis and bleeding. Once a patient is considered a candidate for ECMO, the risk of bleeding and thrombosis should be assessed based on clinical criteria (e.g. presence of active bleeding, particularly CNS) and baseline laboratory test (e.g. clotting, platelets, fibrinogen, activated clotting time, antithrombin activity). The most common anticoagulant is unfractionated heparin.
At the time of cannulation an initial heparin bolus of 50 units/kg body weight is administered, then heparin is continued as an infusion to achieve a target of:
• Activated partial prothrombin time (aPPT) 50–70 s.
• Activated partial thromboplastin time (aPPTr) 1.5–2.0.
• Activated clotting time 180–220 s.
• Or anti-Xa 0.3–0.7 IU/ml.
Platelets should be maintained at a value of >45 × 109 with low risk of bleeding or >80 × 109 with patients at high risk of bleeding, whilst fibrinogen target should be >200 mg/dL; both levels should be monitored as they can fall during ECMO. In patients with heparin-induced thrombocytopaenia, other agents (e.g. argatroban or bivalirudin) can be used in place of heparin, and anticoagulation level can be monitored with anti-Xa or APPTr. Cryoprecipitate can be given if the fibrinogen level is <100–150 mg/dl. In the presence of bleeding, ECMO can be run without additional anticoagulation. Careful monitoring of the oxygenator function is essential to anticipate circuit clotting and failure.
Mechanical ventilation during ECMO
ECMO allows a decrease of mechanical ventilation to ultraprotective settings. The complete removal of CO2 allows the potential to decrease the harm of mechanical ventilation significantly.
• FiO2 is reduced to 0.3 (or the lowest possible).
• Tidal volume is decreased to 2–4 ml/kg of predicted body weight; many patients, however, have tidal volumes of <1 ml/kg predicted body weight (generally because extremely high elastance limits the tidal volumes that can be achieved).
• Respiratory rate 6–10/min.
• ΔP is reduced to <10 cmH2O.
• PEEP can be gradually reduced to 10–15 cmH2O.
It needs to be understood that an ultraprotective strategy of ventilation can lead to reabsorption atelectasis and significant reduction in end-expiratory lung volume and lung collapse due to compression atelectasis.
Protective and ultraprotective ventilation strategies used in the management of patients with ARDS lead to hypercapnia with its consequent side-effects. ECCO2R can be used to remove CO2. ECCO2R can remove enough CO2 to allow a 50% reduction in minute alveolar ventilation, with a significant reduction of PaCO2, and the possibility of using an ultraprotective ventilation without hypercapnia. Although there is good evidence supporting the impact of ECCO2R on blood gases, there is no currently available evidence supporting an improvement in mortality using ECCO2R.
As the underlying condition improves, the ECMO flow rate is reduced to a minimum of 2.5–3 l/min (to minimize risk of thrombus formation with lower flows). There is no benefit in reducing the circuit FiO2 from 1.0. To reduce ventilator support, the sweep gas flow rate can be gradually reduced until it is turned off. In this phase, the patient is essentially ‘off ECMO’ and can be decannulated if gas exchange is stable and ventilation is lung-protective after 12–24 h.
Advances in component technology have significantly reduced the rates of adverse events associated with ECMO. Common patient complications include bleeding, hospital-acquired infection, and thrombosis. Bleeding is the most common adverse event (e.g. intracranial, gastrointestinal, haemothorax, etc.). The incidence of deep vein thrombosis is estimated at about 20% following ECMO. Other complications include those related to the circuit, including circuit thrombosis, air embolism, power failure, haemolysis, and disseminated intravascular coagulopathy.
Tracheostomy is a common procedure in intensive care, with an estimated 15 000 insertion procedures per year in the UK. The most common problems, in both general wards and critical care, are related to obstruction or displacement. The indications for temporary tracheostomy in intensive care include treatment for upper airway obstruction, the avoidance of the laryngeal complications of prolonged endotracheal intubation, and the continued need to protect and maintain the airway in patients with severe neurological injury. Insertion may be elective or in an emergency, as an open surgical procedure or using a percutaneous dilation technique.
Indications for a tracheostomy
• Aid to weaning from invasive ventilation.
• Tracheal access to suction thick pulmonary secretions (easier than via an endotracheal tube).
• Long-term airway management.
• Bypass of upper airway obstruction (e.g. patients with trauma, infection, malignancy, laryngeal or subglottic stenosis, bilateral recurrent laryngeal nerve palsy, severe sleep apnoea).
• Prevention of pulmonary aspiration (e.g. patients with laryngeal incompetence, bulbar dysfunction [e.g. cerebrovascular accidents, Parkinson disease]).
• Neuromuscular disorders (e.g. Guillain–Barré syndrome, critical illness neuromyopathy).
• Severe brain injury, reversible or irreversible.
• Trauma or surgery in the face/neck region.
Relative contraindications for percutaneous dilatational tracheostomy:
• Children under 12 years of age.
• Significant coagulopathy.
• Active infection over the anterior neck.
• Local tracheal pathology.
• Unstable cervical spine fracture.
• Morbid obesity (BMI >35).
• Gross anatomical distortion of the neck or aberrant vessels.
• Recent neck surgery, radiotherapy, or trauma.
• Requirement for high PEEP >10 cmH2O or FiO2 >0.6.
• Potential instability (e.g. patients with raised ICP).
• Patient unlikely to survive >48 h.
Provision of information and consent/assent
Few patients within intensive care have the capacity to give informed consent, but attempts should be made to seek their understanding and approval where this is possible. It is the responsibility of medical staff to act in the best interests of patients lacking capacity, rather than pass the responsibility for consent to the next of kin. The role of the next of kin in healthcare decision-making is increasingly formalized under the new Mental Capacity Act (England and Wales) and the Adults with Incapacity Act (Scotland). Current directives from the General Medical Council and Department of Health specify their involvement using Consent Form 4—‘Form for Adults who are Unable to Consent to Investigation or Treatment’. This process requires provision of information on the nature of the procedure, proposed benefits, potential hazards, and alternatives.
A number of commercial percutaneous dilatational tracheostomy kits and tubes are available. Patients with significant obesity or distorted anatomy are likely to require a longer adjustable flange tube, as highlighted by the NAP4 (4th National Audit Project of the Royal College of Anaesthetists) and NCEPOD (National Confidential Enquiry into Patient Outcome and Death) reports.
A dedicated ‘difficult airway’ trolley must be immediately available in case of loss of the airway during tracheostomy insertion.
Routine monitoring of ECG and SpO2 is assumed for all patients. Invasive blood pressure monitoring should be in place given the potential for abrupt changes in blood pressure with either administration of anaesthetic agents or the stimulation of the procedure. Capnography is mandatory during tracheostomy insertion. An arterial line is indicated for rapid serial arterial blood gas analysis, since capnography may be unreliable in the presence of inadequate ventilation due to obstruction of the ETT by the bronchoscope, or loss of tidal volume by the inevitable leak as the stoma is created.
Ultrasound scanning of the neck prior to percutaneous tracheostomy allows visualization of anterior neck structures, particularly the assessment of blood vessels, depth and level of tracheal rings, and angulation of the trachea. Useful information about adjacent structures helps with the risk-benefit analysis of an open versus percutaneous tracheostomy.
A flexible fibreoptic scope passed through the endotracheal tube may be used to guide correct placement of the introducer needle, guidewire, and tracheostomy tube in the anterior midline. Direct visualization should reduce tracheal wall damage and tube misplacement. Tracheostomy tube position within the trachea and distance from the carina should be confirmed by passing the bronchoscope through the tracheostomy at the end of the procedure. The presence of a scope may hinder ventilation, increasing the risk of hypoxia and hypercarbia with associated increase in ICP in susceptible patients. It should be appreciated that bleeding, distortion of structures, and obstruction of the visual field with larger dilators may prevent endoscopic visualization of damage until after it has occurred. Also, the section of trachea adjacent to the tracheostomy tube cannot be easily visualized after tube insertion.
Adequate anaesthesia is required for a tracheostomy and, although they can be performed under local anaesthesia, better operating conditions are achieved under general anaesthesia, with concurrent use of neuromuscular blocking agents and assisted ventilation. Care should be taken to prevent awareness and local anaesthesia should be used at the site of incision. An assessment of the ease of tracheal reintubation should be performed and plans formulated in case of loss of the airway during the procedure.
Environment and staffing
The environment must have adequate space and lighting, facilitate an aseptic approach, and contain drugs and equipment required for resuscitation.
Two trained operators are required, one to administer anaesthesia and related airway care and a second to perform the procedure. Trainees should be under the direct supervision of a competent doctor as part of an established training programme. A third member of staff (e.g. a nurse familiar with the procedure, equipment, and environment) must also be present to support clinicians in the management of any complications that may develop.
Consider using a World Health Organization surgical safety checklist, including details of consent, coagulation status, and discontinuation of nasogastric feed.
• Preoxygenate with 100% oxygen, maintain anaesthesia, paralyse, and ventilate.
• Optimize patient position by extending the head using a pillow or ‘sandbag’ under the shoulders.
• Check airway using direct laryngoscopy, aspirate secretions in the pharynx and via the ETT, and assess difficulty of reintubation.
• Pull back the ETT under direct vision and resecure when the cuff lies within the larynx and tip at the level of the cricoid cartilage.
• Assess tube position and tracheal anatomy using a fibreoptic bronchoscope.
• The operator should wear a surgical mask, sterile gown, and gloves. Eye protection is recommended for both anaesthetist and operator.
• Disinfect skin with 2% chlorhexidine or iodine and apply surgical drapes (ensuring good visualization and upper airway access for the anaesthetist).
• Ensure all essential equipment is available, functional, and ready for use.
• Choose an incision site according to the shape of the patient’s neck rather than rigidly adhering to the quoted optimal level of T2–3 (avoiding the cricoid and first rings due to the risk of stenosis or low stomas, which can be problematic).
• Inject local anaesthetic with adrenaline 1 in 200 000 into the pretracheal tissues, avoiding distortion of the anatomy.
• The choice of skin incision, cannulation of the trachea, and blunt dissection are a matter of operator preference. The objective is to achieve midline cannulation between rather than through the tracheal rings, avoiding excessive force and trauma to the posterior wall.
• Needle entry into the trachea can be confirmed by aspiration of air or pulmonary secretions. If bronchoscopy is not used then capnography should be used for confirmation.
• Subsequent dilatation should apply progressive, controlled pressure between the tracheal rings to avoid stenosis.
• Correct size of tracheostomy tube is determined by clinical examination prior to the procedure, but occasionally the trachea will be deeper than expected and the tube length will need to be revised.
• Correct placement should be confirmed by bronchoscopy or capnography to avoid surgical emphysema or a pneumothorax. An assessment of tube length can also be made at this stage.
• Secure in position, preferably with a commercial tracheostomy holder.
• Document the procedure, including personnel, technique, size and type of tube, and any difficulties or immediate complications, including bleeding.
Postprocedure chest X-ray?
The usefulness of a chest X-ray (CXR) is debatable. It should be appreciated that tracheal placement cannot be inferred from a plain X-ray. A tube that is partially kinked or too short may be identified from plain film.
• Bleeding (may lead to total airway obstruction)
• Tube misplacement or dislodgement
• Surgical emphysema
• Mucus plugging/obstruction
• Stomal infection.
• Tracheal stenosis
• Tracheo-oesophageal fistula
• Skin tethering/scarring
• Haemorrhage from innominate vessels.
Cricothyroidotomy is a life-saving procedure used to provide emergency access to the airway (e.g. following obstruction of the upper airway) in a ‘can’t intubate, can’t ventilate’ scenario. It involves the insertion of a small tube through the cricothyroid membrane, through which oxygen/ventilation can be provided until a definitive airway is obtained. A cuffed device is desirable, and a 6-mm rather than 4-mm internal diameter greatly improves suctioning and ventilation capacity. Cricothyroidotomy kits are commercially available.
Meticulous skin care at the stoma site has been suggested to decrease bacterial contamination and the inflammatory response leading to granulation tissue. Adequate humidification, tracheal suctioning, and physiotherapy are essential to avoid obstruction of tracheostomy tubes. Obstruction can be static, owing to thick tenacious secretions, or dynamic, owing to partial obstruction by the membranous posterior wall encroaching on the tracheostomy tube lumen. The degree of dynamic obstruction appears to increase when the ITP increases. Dynamic obstruction can be prevented by using a properly designed tracheostomy tube with optimum length and angle to ensure correct tube positioning within the trachea.
Tracheostomy tubes with an inner cannula need to be regularly removed for cleaning to maintain tube patency. Tubes without inner cannulae should be exchanged every 7–14 days, or more frequently if secretions build up. A tracheostomy tube blocked with tenacious secretions renders the patient at risk of progressive hypoxia and possibly cardiorespiratory arrest. Resuscitation attempts will be unsuccessful unless the airway obstruction is recognized and treated promptly. Removal of the tracheostomy tube may be required if suctioning fails to clear the obstruction. In the short term, spontaneously breathing patients will usually manage to breathe through their own upper airway or the stoma. If the tracheostomy is more than 7 days old, the stoma is generally well established to allow early tube replacement if required.
For patients dependent on assisted ventilation, reintubation by the oral route may be needed in the interim if difficulties occur in replacing the tracheostomy tube.
Choice of tracheostomy tube
Soft and flexible tubes provide maximum patient comfort, minimizing any trauma to the trachea and associated structures. Rigid tubes are used more commonly in the longer term as they are thought to keep the stoma open and are easier to change.
Cuffed tubes provide airway protection and facilitate IPPV. Disadvantages are risk of excessive cuff pressure and difficulty in swallowing and communication. High-volume low-pressure cuffs reduce the incidence of cuff-related mucosal damage by providing a wider surface area of the trachea for the pressure to be dissipated. The cuff pressure should not exceed 25 cmH2O to reduce the risk of impaired mucosal perfusion, tissue necrosis, and tracheal stenosis.
Adjustable flange longer length tubes are designed for patients whose trachea is deeper than usual below the skin and soft tissues in the neck, e.g. obese patients. The depth of the trachea should be considered at the outset, during the insertion procedure, and by visualizing the tube position at endoscopy via the glottis and through the tube.
Tracheostomy tubes with an inner tube may remain in place up to 30 days or more as the inner cannula can be cleaned and changed regularly.
A fenestrated tube allows air flow through the vocal cords when the tube is occluded or a speaking valve is attached. A disadvantage is that the diameter of the inner lumen will be reduced by 1–2 mm, increasing work of breathing and potential for aspiration of gastric contents. It is unsuitable for patients dependent on positive pressure ventilation unless a non-fenestrated inner cannula is used, and is usually reserved for those who are imminently weaning from ventilation.
Tracheostomy tubes with a subglottic suction channel can be used to reduce the risk of aspiration from collected subglottic secretions.
Essential bedside equipment for tracheostomy patients:
• Suction unit, Yankauer suckers, and appropriately sized suction catheters.
• Gloves, aprons, and eye protection.
• Tracheostomy tubes (same size as in situ and one size smaller) and spare inner tubes.
• Tracheal dilators.
• Self-inflating bag-valve mask device and tubing.
• Catheter mount or connection.
• A 10-ml syringe for cuff inflation and deflation.
• Resuscitation and translaryngeal intubation equipment.
• Portable oxygen.
• Consider a ‘tracheostomy box’ plus emergency algorithms.
Humidification of respiratory gases is essential to prevent blockage of the tracheostomy tube by tenacious sputum, keratinization, and ulceration of the tracheal mucosa, atelectasis, impaired gas exchange, and secondary infection. It should be complemented by physiotherapy, patient mobilization, and appropriate suctioning.
An inflated cuff can compress the oesophagus and make swallowing difficult, increasing the risk of aspiration. Oral intake may be permitted for psychological wellbeing and to help establish enteral feeding early. The decision to allow feeding with an inflated cuff should be made on an individual patient basis following a swallowing assessment or after referral to a speech and language professional.
Initiation of oral intake
• Confirm that the patient can tolerate cuff deflation.
• Sit the patient up with head slightly flexed, place a suction catheter just to the end of the tracheostomy tube, and deflate cuff while suctioning.
• Start with sips of water, fluids, and then soft diet, providing that the patient shows no signs of respiratory distress (coughing, desaturation, increased tracheal secretions, increased respiratory rate, etc.).
In problematic cases, consider referral to speech and language therapy.
Risk factors for swallowing problems in patients with a tracheostomy:
• Neurological injury, e.g. bulbar palsy.
• Head and neck surgery.
• Evidence of aspiration of enteral feed or oral secretions on tracheal suctioning.
• Increased secretion load, or persistent wet/weak voice, when cuff is deflated.
• Coughing and/or desaturation following oral intake.
• Patient anxiety or distress during oral intake.
The inability to speak with a cuffed tracheostomy can cause great distress to patients. Communication aids should be available along with a nurse call bell.
Speaking valves are one-way valves that fit over the end of a tracheostomy tube, allowing patients to breathe in through the tracheostomy but not out. Exhaled airflow must go up through the larynx and out of the mouth, allowing a patient to talk. They should only be used with a deflated cuff and can be tiring because of the increased resistance to air flow. They may have a useful role in ‘training’ the larynx after prolonged disuse and improve swallowing and secretion management.
Changing tracheostomy tubes
The basic principles for changing a tracheostomy tube are:
• Tracheostomy tubes without an inner cannula should be changed every 7–14 days, the frequency then decreasing once the patient is free of pulmonary secretions and has a well-formed clean stoma.
• A European Economic Community Directive (1993) states that tracheostomy tubes with an inner cannula can remain in place for a maximum of 30 days.
• The first routine tracheostomy tube change:
• Should not be performed within 4 days following a surgical tracheostomy and 7–10 days after a percutaneous tracheostomy to allow the stoma to become established (preferably over a gum elastic bougie or airway exchange catheter).
• The decision to change the tube is usually a multidisciplinary one, considering weaning, swallowing, ventilation, speaking, and the ongoing need for a cuff.
• Must be carried out by a person competent to do so with advanced airway skills and appropriate equipment immediately available. Technique used and ease should be recorded.
• Subsequent changes can be made by experienced personnel trained in tracheostomy tube changes (e.g. specialist tracheostomy nurse).
In practice, the frequency with which the tube needs to be changed will be affected by the individual patient’s condition and the type of tube used. Elective changes are inherently safer than those done in an emergency. An emergency algorithm should be immediately available at the bedside of all tracheostomy patients.
Decannulation should be considered when patients demonstrate a satisfactory respiratory drive, a good cough, and the ability to protect their own airway. Patients who show no signs of tiring on CPAP or a T-piece with low-flow oxygen therapy are potential candidates for decannulation. Coughing secretions up into the tracheostomy tube is a good sign, whereas generalized weakness and inability to hold the head up are negative predictors of successful decannulation. An impaired conscious level also reduces the chance of success. There is a common tendency to leave tubes in too long while clinicians await the perfect time to decannulate, and time-consuming referrals to speech or physiotherapy are made. If left too long, it can stimulate mucus production and affects the mucociliary system. It should be appreciated that an effective cough relies on build-up of positive pressure within the trachea against a closed glottis and then sudden release to generate a cough. This cannot be achieved with a large-bore cannula open within the trachea, and a tracheostomy should be removed as soon as it is no longer required.
Following decannulation patients should be closely monitored for 24 h. Most tracheostomy stomas can granulate without suturing. They achieve a functional seal within 2–3 days. These partially healed wounds can be quickly reopened with artery forceps in the first few weeks after closure if necessary. Occasional patients will require ear, nose, and throat (ENT) referral for tethered scars or a sinus. At longer-term follow-up, clinicians should be aware of the rare significant tracheal or laryngeal stenosis giving rise to respiratory symptoms of stridor, persistent cough, and voice changes. Such cases require specialist ENT or thoracic referral.
• Treatment of a pneumothorax in a patient requiring positive pressure ventilation.
• Treatment of a large pneumothorax/hydrothorax/haemothorax.
• Following needle decompression of a tension pneumothorax.
• Management of bronchopleural fistula.
• Management of empyema.
• Management of localized pneumothorax causing ventilatory compromise (this usually requires CT guidance).
Depending upon the size, site, and anticipated nature of the pleural collection, select an appropriate drain. Smaller drains are preferable for most indications. The exceptions requiring larger drains with multiple (more than three) holes are:
• Bronchopleural fistulas with a large gas leak.
• Haemothoraces with ongoing bleeding.
• Viscous/highly purulent empyemas.
This procedure should always be performed with strict adherence to aseptic precautions.
• 2% chlorhexidine-based skin cleaning fluid and reservoir.
• Sterile gauze.
• Local anaesthetic, syringe, and needle. Consider adjunctive systemic analgesia and sedation.
• For Seldinger technique:
• Bedside ultrasound.
• Needle and syringe.
• Drain with stiffener.
• Three-way tap.
• For blunt dissection/thoracostomy technique:
• Blunt dissection forceps (e.g. curved Robert’s).
• Collection bag/underwater seal drainage bottle/Heimlich valve.
• Appropriate dressing.
Whenever practical, inform the patient about the proposed procedure and gain consent.
Position patients such that they are comfortable and the area in which the procedure is to be performed is easily accessible and, for fluid collections, is gravitationally dependent. This may be difficult in sedated and intubated patients on positive pressure ventilation. Sitting patients in as near an upright posture as possible with unhindered access to the posterior and/or lateral chest walls is ideal. Ensure that whatever the position, the operator is in an ergonomic position with access to the equipment.
Although not obligatory, it is undoubtedly best practice to perform a thoracic ultrasound immediately prior to aspiration to define the anatomy and avoid visceral injury. When aspirating small or complex collections, continuous ultrasound guidance is essential.
Clean the area and apply sterile towels.
Infiltrate a small volume of local anaesthetic into the subcutaneous and intradermal spaces, avoiding the neurovascular bundle, which runs along the inferior border of the rib.
Insert the needle, whilst aspirating using a syringe, until air/fluid is freely withdrawn. Having entered the pleural space, disconnect the syringe and gently insert the J-wire through the needle. Be aware that the J-wire, despite its soft tip, can puncture and damage visceral organs, in particular consolidated lung. Withdraw the needle, leaving the wire in situ. Make a small stabbing incision through the skin at the exit site of the wire. This is most easily performed by placing the flat surface of a no. 11 blade on the wire and sliding it into the skin. Insert the dilator into the pleural cavity over the wire (be cautious not to insert this too far) and angle the tract formed towards the desired location (apical for simple pneumothorax, posterior-basal for fluid). Leave the dilator in place for a few moments, then remove, again leaving the wire in situ. Depending upon the drain design (straight, curved, or pigtail), ensure it is mounted on its stiffener (if required) and gently insert, over the wire, directing the tip as required. Withdraw the wire and stiffener and then connect to a closed three-way tap. Aspirate via the tap to ensure adequate placement and anchor with a holding suture. Connect to the appropriate drainage device, having first obtained any desired specimens. Pad and stick to the skin using a small dressing, ensuring that the drain will not kink and that the three-way tap is accessible and will not cause a pressure injury. One possible technique is to stick the drain in line with the ribs directed anteriorly.
Blunt dissection/thoracostomy technique
Clean the area and apply sterile towels.
The site of drain insertion should usually be in the so-called safe triangle. This is made up of the anterior border of latissimus dorsi, the lateral border of pectoralis major, a line superior to the horizontal level of the nipple, and an apex below the axilla.
Infiltrate a small volume of local anaesthetic into the subcutaneous and intradermal spaces, avoiding the neurovascular bundle, which runs along the inferior border of the rib. As an alternative, consider performing an intercostal nerve block in the relevant space and the spaces above and below. Perform a diagnostic aspiration to ensure air/fluid can be drained from the intended insertion site.
Make a 2–3-cm incision along the centre of the rib below the desired intercostal space. Using forceps, bluntly dissect over this rib into the pleural cavity. Insert a finger into the pleural cavity and perform a sweep. This acts as a diagnostic examination, enhances the blunt dissection, and can potentially break down loculations, if present. Take hold of the drain tip with the forceps by placing them in through the distal side hole and out through the end hole. Prior to drain insertion consider disconnecting the patient from any positive pressure ventilation to reduce the chance of intrapulmonary lung placement. Gently insert the drain with the forceps and release. Try to position the drain apically for a pneumothorax and posterior-basally for fluid. Be cautious of intrapulmonary and mediastinal drain placement. To ensure adequate position and to obtain any desired specimens, aspirate using a bladder-tipped syringe and/or connect to an underwater sealed drainage bottle. Suture one end of the incision and place an anchoring suture around the drain. Avoid purse-string sutures and use monofilament suture material. Pad and stick to the skin using a small dressing, ensuring that the drain will not kink. One possible technique is to stick the drain in line with the ribs directed anteriorly.
Following insertion, obtain a CXR to assess the position. If functionally inadequate, regardless of radiological position, manipulate the drain accordingly or remove and, if necessary, reinsert. In difficult circumstances, seek radiological advice and consider CT guidance or thoracic surgical assistance.
Draining fluid off at too high a rate can result in re-expansion pulmonary oedema. This is rare, especially in patients receiving positive pressure ventilation. The risk can be minimized by limiting drainage to a maximum of 1500 ml/h by clamping the drain. In all other instances, clamping of the drain, except transiently, should be avoided. Clamping has no place in the management of pneumothoraces.
For small drains, consider flushing 6–12-hourly with 5–10 ml of 0.9% saline to assess and maintain patency.
If connected to an underwater seal drainage bottle, the meniscus in the tube should transduce intrapleural pressure and swing with respiratory phase. If it does not, the drain is blocked, kinked, or has become mislocated. Examine, flush, and reimage as necessary.
Remove the drain as soon as it is no longer required or has failed. Close the drain site with a suture if required. If there continues to be a leak through the drain site, place a stoma bag over it. If this fails to contain the situation, then either insert a new drain or seek surgical advice.
• Bleeding from an intercostal vessel or a damaged viscus/organ. This is rare, but can be fatal.
• Trauma to lung (including formation of bronchopleural fistula), heart, liver, spleen, or kidney. Again, serious trauma is rare, but can be fatal.
• Infection: from superficial drain site infection to empyema and lung abscess.
• Diagnosis of the nature of a pleural effusion.
• Treatment of a simple pneumothorax.
• Treatment of pleural effusion of sufficient size to impair or compromise respiratory mechanics.
• Treatment of an empyema.
Depending upon the size, site, and anticipated nature of the pleural collection, select an appropriately sized needle or cannula.
This procedure should always be performed with strict adherence to aseptic precautions.
• Bedside ultrasound.
• 2% chlorhexidine-based skin cleaning fluid and reservoir.
• Sterile gauze.
• Local anaesthetic, syringe, and needle.
• Chosen needle/cannula, three-way tap on a short extension, and an appropriately sized syringe for sampling/aspiration (20–60 ml).
• Sterile sample pots, blood gas syringe, glucose testing strip or blood glucose (fluoride) sample bottle, blood culture bottles (aerobic and anaerobic), blood chemistry/enzyme assay sample bottle.
• Small dressing.
Whenever practical, inform the patient regarding the proposed procedure and gain consent.
Position patients such that they are comfortable and the area in which the procedure is to be performed is easily accessible and, for fluid collections, is gravitationally dependent. This may be difficult in sedated and intubated patients on positive pressure ventilation. Sitting the patient in as near an upright posture as possible with unhindered access to the posterior and/or lateral chest walls is ideal. Ensure that, whatever the position, the operator is in an ergonomic position with access to the equipment.
Although not obligatory, it is undoubtedly best practice to perform a thoracic ultrasound immediately prior to aspiration to define the anatomy and avoid visceral injury. When aspirating small or complex collections, continuous ultrasound guidance is essential. A classic ultrasound image of a pleural effusion is shown in Figure 1.15.
Clean the area and apply sterile towels.
Infiltrate a small volume of local anaesthetic into the subcutaneous and intradermal spaces, avoiding the neurovascular bundle, which runs along the inferior border of the rib.
Insert the needle/cannula, whilst aspirating using a syringe, until air/fluid is freely withdrawn. Having entered the pleural space, if there is a significant volume of air/fluid to be aspirated, attach a three-way tap on a short extension to facilitate syringe change over.
If no fluid/air is aspirated, remove the needle/cannula and reimage with ultrasound. If necessary, reposition the patient.
For a simple pneumothorax, aspirate as much air as possible, making note of the volume. If, having drained 1000 ml, air is still freely aspiratable, consider inserting a pleural drain. If in doubt, reimage with X-ray or ultrasound.
For a pleural effusion, aspirate sufficient fluid for all diagnostic tests. Continue to aspirate further fluid if there is a significant residual volume. Measure the total volume removed and be conscious of the possibility of re-expansion pulmonary oedema.
Once completed, remove the needle/cannula and cover the puncture site with a simple dressing if required. If fluid starts to leak through the puncture site, cover with a small stoma bag or consider inserting a drain.
For microbiology, send raw fluid for microscopy, culture, and sensitivities. If suspicious, also request staining and culture for mycobacteria. To increase the sensitivity of bacterial culture, inoculate a set of blood culture bottles with 10 ml of fluid per bottle.
Unless the fluid is frankly purulent, take a specimen in a blood gas syringe and put it through a blood gas analyser to measure the pH. A pH <7.20 is consistent with an empyema and is probably the most sensitive test. The two other assays consistent with this diagnosis are a fluid glucose of <3.35 mmol/l and/or a lactate dehydrogenase level more than three times the upper limit of normal for serum.
Determination of the fluid total protein to establish whether the effusion is a transudate or exudate is not of any great value and is unreliable in diagnosing empyema.
Complications and their management
The most common complication of pleural aspiration is a small pneumothorax. This requires no action other than vigilance for increasing size, which is rare unless the lung has been punctured. If this has occurred, it is usually obvious, as air is unexpectedly aspirated during the procedure. The risk of an enlarging/significant pneumothorax is increased if the patient is on positive pressure ventilation. Under these circumstances, watch for an evolving tension pneumothorax and be prepared for immediate decompression and chest drain insertion.
The second most common complication is damage to the intercostal neurovascular bundle. To avoid this, use the superior border of the rib as the landmark for the puncture site and maintain an insertion angle that minimizes the risk of damage to the bundle associated with the rib above. Be aware that damage to the intercostal artery can result in a significant haemothorax and require surgical ligation.
Puncture of and/or damage to adjacent structures is easily avoided by ultrasound guidance. There are case reports of serious complications from pleural aspiration procedures in which the liver, spleen, kidneys, diaphragm, and myocardium have been injured. Such iatrogenic injuries are rare with this procedure but have a significantly higher incidence associated with pleural drain insertion.
Flexible bronchoscopy is an essential diagnostic and therapeutic procedure in the ICU. The majority of procedures are performed in intubated patients receiving mechanical ventilation.
• Direct visualization of the upper airway in difficult endotracheal intubation.
• Direct endotracheal visualization and guidance during percutaneous tracheostomy.
• Inspection of the distal portion of an ETT or tracheostomy tube to assess patency and position.
• Inspection of the distal trachea and proximal bronchial tree for mucosal pathology/extrinsic compression.
• Removal of material obstructing one or more major bronchi.
• Performing sampling of distal airways for microbiological and/or cytological specimens.
• Guiding the placement of an endobronchial blocking/isolation catheter.
Equipment preparation and aftercare
Prior to use, the bronchoscope should be thoroughly cleaned and disinfected. Cross-infection between patients is a serious hazard. If stored, this should be in a dedicated clean environment. The scope should be handled aseptically and transported in an appropriate enclosed container. All the remaining equipment should be clean or sterile single use. Essential equipment:
• Light source.
• Camera/videoscope stack (optional).
• Cleaning brush and/or solution for working channel.
• A packet of sterile gauze.
• A sterile 1-litre jug.
• A 500-ml bag of 0.9% sodium chloride.
• A 20-ml syringe.
• Sputum traps.
• Dedicated suction.
• A bronchoscopy catheter mount.
The scope should be checked for clear vision and a functioning suction system immediately prior to use. The outside of the scope should be wiped with saline-soaked gauze. Avoid water-based gel lubricants as these can dry out and become sticky rather than lubricate. In situations where there is problematic sticking of the scope to the inside of an ETT or tracheostomy tube, use sterile liquid paraffin.
The bronchoscopist and any assistants should wear full protective clothing, including gowns, gloves, and face/eye cover. Appropriate and sensible precautions should be made regarding the potential aerosolization of infected material.
At the end of the procedure, the suction channel of the scope should be immediately brushed and/or rinsed. The outside of the scope should be cleaned. The scope should then be sent for full decontamination. Any other reusable equipment should be cleaned and stored appropriately.
Whenever practical, inform the patient regarding the proposed procedure and gain consent. Depending upon the indication, likely duration, and clinical condition of the patient, an appropriate plan regarding topical anaesthesia, sedation, and neuromuscular blockade should be made. At a minimum, continuous ECG and SpO2 monitoring together with intermittent, automated non-invasive blood pressure should be used. The FiO2 should be increased to ≥70% (or as high as possible). If ventilated, a pressure control mode is preferable, although a strictly pressure-limited volume control mode can be used. Hypoventilation is an inevitable occurrence during the procedure. In patients in whom even transient hypercapnia needs to be avoided, a series of timed, short bronchoscopies can usually be safely performed. In such circumstances, continuous end-tidal capnography is essential. Alternatively, consider using high frequency oscillatory ventilation (HFOV).
Before starting, ensure that all necessary equipment is available and working, the patient is comfortable, and the bronchoscopist is ergonomically positioned. Consider slowly injecting 3–5 ml of sterile saline into the ETT/tracheostomy tube to lubricate the passage of the scope. At least one assistant must be present to watch the patient, the ventilator, and the monitoring, in addition to being able to assist the bronchoscopist. When performing a bronchoscopy via the mouth/through an oral ETT, it is recommended that a bite guard is inserted to protect the scope.
First navigate the ETT/tracheostomy tube and ascertain whether the tube is encrusted with secretions. If so, this may both inhibit the procedure and present the patient with unwanted additional resistance. It may be possible to clean the tube effectively using the scope; however, electively changing the tube is sometimes required.
Next, consider whether the tube opens centrally within the trachea and is a sufficient distance from the main carina. If this is not so, reposition the tube under bronchoscopic guidance and make note of the optimal position using the visible reference markers on the tube. Always consider whether moving the patient will adversely affect the position of the distal end of the tube, and leave a detailed description in the patient’s notes.
Then go on to examine the distal trachea, in particular looking for mucosal trauma caused by the distal tip of the tube and blind suction catheter insertion. Continue by inspecting the remainder of the accessible bronchial tree in a logical order. It should be possible to visualize the first two to five divisions of each lobar bronchus.
For visible secretions, try to sample/remove without causing trauma to the mucosa and without using saline lavage. If required, collect a specimen for microbiological examination. Difficult-to-clear secretions can often be removed piecemeal. Large pieces can often be held against the tip of the scope by application of continuous suction. The scope, together with the offending mass, can then be removed en masse by slowly withdrawing the scope whilst maintaining suction. Obstructing aggregations of inspissated secretions, blood clots, or mucosal sloughing may require prolonged or multiple procedures. The use of biopsy forceps and cytology brushes may be useful but requires skill and patience. Should flexible bronchoscopy fail, consider using a rigid scope. Nebulized hypertonic (3–7%) saline, unfractionated heparin (10 000–25 000 IU 4–12-hourly), 4.2% sodium bicarbonate, and dornase alfa have all been described as useful adjuncts, but none has been proven to have superior efficacy over the others in ventilated patients. There is also no evidence to suggest a superior efficacy for direct instillation versus nebulization.
For visible mucosal bleeding, haemostasis will usually occur spontaneously. Haemostasis can be augmented by topical vasoconstriction using adrenaline (epinephrine). Gently instil/irrigate with a 1 in 10 000 (0.1 mg/ml) solution. Topical antifibrinolytics, such as neat tranexamic acid (100 mg/ml) can also be useful.
If the bleeding is distal to the main carina and unstoppable and/or distal to the limit of visualization, temporary isolation and tamponade can be achieved by wedging the tip of the bronchoscope into the origin of the identified bronchus. The efficacy of instilling vasoconstrictors and/or antifibrinolytics is uncertain. The value of prolonged continuous suction is also debatable. If this fails to achieve haemostasis, then a balloon-tipped bronchial isolation catheter can be inserted parallel to the bronchoscope, which can then be used to guide catheter placement. This can be a difficult procedure owing to the aerosolization of blood within the airway masking any vision. Be careful not to dislodge the blocking catheter when withdrawing the scope. Consider paralysing the patient and applying a high level of PEEP. If practical, turn the patient bleeding side down. As a further adjunct, connect and instil oxygen through the bronchoscope suction channel at a high flow rate. Definitive treatment for persistent haemorrhage is either selective bronchial angiography and embolization or surgery.
Whenever possible, send undiluted secretions for microbiological investigation. To obtain a specimen from a region of interest beyond visualization, first locate the nearest lobar, segmental, or subsegmental division, then gently wedge the tip of the bronchoscope into it. Before any sampling, ensure that the suction channel is clear of any proximal secretions, which might contaminate the specimen. This may require the scope to be fully withdrawn, the channel cleaned with a brush and rinsed with clean saline, and the scope reinserted and positioned. Next, apply the specimen trap as close to the bronchoscope as possible. Slowly instil 20–60 ml of 0.9% sterile saline, wait for a few seconds, then apply continuous suction. If the airway completely collapses, ask an assistant to turn the strength of the suction down gently until some fluid is drawn into the specimen trap. When no further fluid flows, slowly withdraw the scope whilst maintaining continuous suction. Make a note of the volume instilled and the volume of the specimen. Examine the specimen for adequacy, looking for the presence of mucoid or infected airway secretions. It is worth discussing the optimal handling of specimens with the labs receiving them, in particular if qualitative or semi-quantitative microscopy is required or specific pathogens are suspected. If a good-quality, large-volume specimen is obtained, this can often be divided in the laboratory for both microbiological and cytological examination if required. This prevents repeated saline lavage and scope trauma, both of which are injurious.
Blind brush specimens may be useful in both cytological and microbiological testing. Brush and biopsy specimens of visible lesions can also be taken. In patients receiving positive pressure ventilation, transbronchial biopsy and transbronchial needle aspiration specimens carry a significant risk of pneumothorax and pneumomediastinum, and are best avoided.
Protected lavage and brush catheters are available, but are of questionable value. There is conflicting evidence regarding the value of bronchoscopic specimens over and above those obtained by blind endotracheal suctioning, most especially in the diagnosis of ventilator-associated pneumonia.
The following complications can occur and should be prepared for:
• Displacement of the ETT/tracheostomy tube out of the trachea.
• Obstruction of the ETT/tracheostomy tube.
• Obstruction of the trachea or major bronchus.
• Hypoxia/de-recruitment/increasing ventilatory requirements postprocedure.
• Sepsis secondary to translocation (bacteraemia).
• Hypotension or hypertension/cardiac arrhythmias.
Critical illness is associated with high morbidity and mortality rates, and the associated care is a major determinant of healthcare costs. Critically ill patients cared for on the ICU can have prolonged periods of immobility, with critical illness lasting from hours to months, depending on the underlying pathophysiology and the patient’s response to treatment.
Respiratory dysfunction is one of the most common causes of critical illness necessitating ICU admission. The aims of chest physiotherapy (CPT) are to clear secretions, to prevent pulmonary complications, to improve ventilation and/or regional ventilation and lung compliance, and to reduce airway resistance and the work of breathing.
CPT in mechanically ventilated patients
Such patients are especially at risk of complications. They are generally sedated, with an artificial airway and sometimes inadequate humidification. These three components alter mucociliary clearance.
Increase of expiratory flow
CPT manoeuvres involve inspiratory and expiratory techniques, with and without the aid of positive pressure devices. The physiotherapist applies external forces during expiration, inducing an increase in expiratory flow and therefore an increase in mucus transport. These techniques of increased expiratory flow can be used passively in ventilated, sedated patients or actively in pressure support mode in more alert, cooperative patients. Recently, a study in animals has specified that hard manual rib cage compression improved mucus clearance while soft manual rib cage compression was not effective and was potentially unsafe. These findings corroborate the predominant role of peak expiratory flow on mucus clearance.
Patients are placed in semi-recumbent or in lateral decubitus position for these manoeuvres.
To evacuate the sputum, physiotherapists perform tracheal suctioning. Some may also add manual hyperinflations, manual vibrations, etc.
Manual hyperinflation or ventilator hyperinflation
The aims of hyperinflation are to prevent pulmonary atelectasis, re-expand collapsed alveoli, improve oxygenation, improve lung compliance, and facilitate movement of pulmonary secretions towards the central airways. Manual hyperinflation is performed by delivering a large tidal volume combined with an expiratory plateau and a fast release of the resuscitator bag. The quick release of the bag enhances expiratory flow and mimics a forced expiration. Ventilator hyperinflation is delivered by the ventilator with a greater control of the pressure. Hyperinflation can result in marked haemodynamic changes associated with a decreased cardiac output, which result from large fluctuations in ITP. A pressure of 40 cmH2O has been recommended as an upper limit. Hyperinflation can also increase ICP and mean arterial pressure, which has implications for patients with brain injury. These increases are usually limited, however, so that the cerebral perfusion pressure commonly remains stable.
In a recent systematic review, authors reported that all four included studies concluded no significant differences in sputum wet weight, dynamic and static pulmonary compliance, oxygenation, and cardiovascular stability between ventilator hyperinflation and manual hyperinflation.
Prevention of pulmonary complications in mechanically ventilated patients
A decrease in VAP in mechanically ventilated patients treated with physiotherapy and standard nursing care in comparison to a group managed with standard nursing care alone has been demonstrated. After trauma brain injury, CPT does not prevent, or hasten recovery from, VAP.
CPT in patients without endotracheal intubation
Inspiratory breathing exercises coupled with mobilization and body positioning are used to increase lung volumes and improve ventilation for patients with reduced inspiratory volumes, for example, following surgery. Expiratory breathing techniques (forced expirations, huffing, and coughing) are used to increase expiratory flow rates and thereby enhance airway clearance and mobilize secretions from the peripheral to upper airways. Thus, expiratory breathing exercises may need to be accompanied by interventions that increase inspiratory volume if reduced inspiratory volumes are contributing to an ineffective cough. To evacuate the sputum, physiotherapists help the patient with cough techniques or perform tracheal suctioning. Forced expiratory manoeuvres should be used with caution in patients with bronchospasm to avoid exacerbation of spasm or cardiac dysfunction.
Manually assisted cough using thoracic or abdominal compression may be indicated for patients with expiratory muscle weakness or fatigue (e.g. neuromuscular conditions).
The inspiratory/expiratory insufflator is a device that delivers inspiratory pressure followed by a high negative expiratory force, via a mouthpiece or facemask. It is indicated when a patient is unable to clear secretions using other interventions. It has been applied with success in the management of non-intubated patients with retained secretions secondary to respiratory muscle weakness (e.g. muscular dystrophy). Cough assist should be initiated when peak cough flow is <270 l/min.
Incentive spirometry (IS) can be used to encourage improved lung volumes and flow rates. IS is believed to improve the distribution of ventilation by increasing lung volumes and encouraging prolonged slow inspiration. IS may be prescribed for patients after major surgery either to prevent or to treat postoperative pulmonary complications; however, it has not been shown to be of added benefit (beyond body position and early mobilization) in the management of routine postoperative patients. Because the effect of IS depends on patient cooperation, critically ill patients may be unable to perform IS effectively.
Intrapulmonary percussive ventilator (IPV) is a mode of ventilation that can be used in the management of intubated or non-intubated patients using a mouthpiece. IPV is believed to increase mucociliary clearance but two reviews have shown that this technique has no effect on sputum weight, mucociliary transport, or pulmonary function. Adverse effects include bronchospasm, haemoptysis, hypoxaemia, and bradycardia, and thus appropriate monitoring is indicated.
Positive expiratory pressure
PEP refers to a device to augment oxygenation and airway clearance in non-intubated patients and can be applied with a facemask or mouthpiece. PEP is used mostly for medical patients with excessive airway secretions (e.g. cystic fibrosis). PEP in critical care could assist in secretion removal in COPD patients requiring NIPPV.
Flutter® valve device
The Flutter® valve is contained within a pipe-shaped device with an opening at the mouthpiece and small outlets at the top of the bowl. As the patient exhales, a steel ball is displaced, generating PEP and oscillating waves of pressure within the airways (high-frequency oscillations). The Flutter® device is indicated in spontaneously breathing patients who have excessive airway secretions (e.g. cystic fibrosis). The role of the Flutter® device in critically ill patients has not been studied but may be limited by the patient’s capacity to cooperate.
Inspiratory muscle training
Comparable to other skeletal muscles, respiratory muscles can be trained to enhance contraction performance (force, endurance, velocity of shortening, and metabolic efficiency). Because ventilatory failure can be related to respiratory muscle dysfunction, improving the function of the respiratory muscles is a rational treatment goal. Uncontrolled trials of inspiratory muscle training in critically ill patients suggest that inspiratory muscle training improves inspiratory muscle function and, hence, may contribute to successful weaning. A recent review found evidence that preoperative inspiratory muscle training was associated with a reduction of postoperative atelectasis, pneumonia, and duration of hospital stay in adults undergoing cardiac and major abdominal surgery. However, the potential for overestimation of treatment effect needs to be considered when interpreting the present findings.
Patient positioning can be used to increase gravitational stress and associated fluid shifts, through head tilt and other positions that approximate the upright position. The upright position increases lung volumes and gas exchange, stimulates autonomic activity, and can reduce cardiac stress from compression.
The semi-recumbent, prone, and lateral positions are used, respectively, to decrease the risk of nosocomial pneumonia and to improve gas exchange in ARDS and in unilateral lung disease.
Stiller underlined the beneficial effect of good positioning combined with CPT in the treatment of acute lobar atelectasis, indicating that this treatment is as effective as fibreoptic bronchoscopy.
Although part of physiotherapy, mobilization cannot be included as part of CPT but it can influence ventilation and can play a role in pulmonary complications. Mobilization has been part of the physiotherapy management of acutely ill patients for several decades and refers to physical activity sufficient to elicit acute physiological effects that enhance ventilation, central and peripheral perfusion, circulation, muscle metabolism, and alertness.
CPT clears secretions, prevents pulmonary complications, and improves ventilation and lung compliance.
Physiotherapists must choose their therapeutic interventions as a function of a preliminary clinical examination to be sure that their treatment will be therapeutic and safe. Unstable patients should be monitored continuously during physiotherapy. Finally, physiotherapists can contribute to the patient’s overall wellbeing by providing emotional support and enhancing communication.
Water vapour is contained in the air and its amount is influenced by atmospheric conditions.
While the normal temperature of the atmosphere never reaches the boiling point of water (100°C), variation in air temperature is certainly the most important parameter influencing water vapour content. Water vapour capacity is the maximum amount of water that a gas can hold at a particular temperature. Two other concepts are important concerning water vapour content: the first is the absolute humidity, defined as the weight of water vapor contained in a given volume of gas; the second is the relative humidity (RH), defined as the relationship between the content of water in air at a specific temperature and the capacity of water that air can hold at the same temperature:
Humidification during spontaneous breathing and mechanical ventilation
With natural airways
Heat and moisture exchange (HME) is one of the most important functions of the respiratory system.
The upper airways ensure that inspired gases are heated or cooled to body temperature (37°C) and humidified to a relative humidity of approximately 100% at body temperature.
More specifically, when inspired gases arrive at the conchae, turbulent flow is created. The architecture of the conchae consists of three turbinates covered by a folded mucous membrane, representing a volume of only 20 ml but a surface area of 160 cm2. Therefore, each gas molecule is likely to come into contact with the surface area of the vascular nasal mucous membrane. The moist mucous membrane heats inspired gases to body temperature and, producing up to 650–1000 ml of H2O per day, brings inspired gases to a relative humidity of 80% on leaving the nose to enter the nasopharynx.
Mucus produced in the nasal cavity is also responsible for the humidification of inspired gases.
With artificial airways
The gas arriving from the ventilator is cold (15°C) and dry (2% RH), and the ETT bypasses the physiological humidification and heating of inspired gases. Absence of humidification induces alterations in mucus transport, decrease of mucus clearance, lesions of the epithelium and mucosa, and changes in ventilatory parameters (decreased compliance and FRC, increased resistance). Therefore, artificial devices must be installed during mechanical ventilation to replace the humidifying function of the upper airways by increasing the water vapour content of dry gas. In the absence of these, Branson et al. have shown alterations in tracheobronchial structure and function resulting from exposure to dry gases and amplifying by duration of exposure. These damages to the respiratory epithelium are manifested by increased work of breathing, atelectasis, thick and dehydrated secretions, and cough and/or bronchospasm.
There are two types of humidifier: active and passive.
Active humidifiers add water vapour to inspired gases, whereas passive humidifiers exchange heat and moisture by preserving the moisture and heat in the gas exhaled by the patient.
Active humidifiers consist of a humidity generator (or water reservoir) and humidity delivery system (or breathing circuit). An ideal system generates the required amount of humidity, in the form of water vapour, at the correct temperature, and transports it to the patient without the loss of either heat or moisture.
The most effective way to achieve this is to use a large heated water surface for the generator and heating elements within the delivery system to prevent condensation.
There are two types of active humidifiers:
Bubble: gas is forced down a tube into the bottom of a water container. The gas escapes from the distal end of the tube under the water surface, forming bubbles, which gain humidity as they rise to the water surface. Some of these humidifiers have a diffuser at the distal end of the tube that breaks gas into smaller bubbles. Bubble humidifiers may be unheated or heated. Typically, unheated bubble humidifiers are used with low-flow oral–nasal oxygen delivery systems. At body temperature, the relative humidity of the inspired gas never exceeds 40%. Humidification with this device is a function of the temperature in the room, the volume of water in the humidifier, and the gas flow (most efficient with flows less than 5 l/min). Heated bubble humidifiers provide higher absolute humidity. They are designed to work with flow rates as high as 100 l/min.
Passover: this humidifier is used for patients being treated by mechanical ventilation. It is designed to bring the gas in the airways to a temperature of between 34°C and 37°C with a relative humidity of 100%. Nevertheless, the performance of humidifiers may be affected by room temperature, as well as patient minute ventilation.
Condensation in ventilator circuits is limited by using heated circuits.
Passive humidifiers, which are placed close to the patient at the end of the dead space, preserve the water vapour contained in the gas exhaled by the patient. During the subsequent inspiration, the water maintained in the HME is used to humidify the inspired gas.
HMEs can be hydrophobic, combined hydrophobic hygroscopic, and pure hygroscopic. In hydrophobic HMEs, the condenser is made of a water-repelling element with low thermal conductivity. In combined hydrophobic hygroscopic HMEs, a hygroscopic salt (calcium or lithium chloride) is added inside the hydrophobic HME. These salts have a chemical affinity to attract water particles and thus increase the humidification capacity of the HME. Pure hygroscopic HMEs have only the hygroscopic compartment. During exhalation, vapour condenses in the element as well as in the hygroscopic salts. During inspiration, water vapour is obtained from the salts, obtaining an absolute humidity ranging between 22 and 34 mg H2O/l.
The appropriate HME should have at least 70% efficiency, providing at least 30 mg/l of water vapour.
In a recent study, Lellouche and colleagues independently assessed the humidification capacity of 32 HMEs. Strikingly, 36% of tested HMEs had an absolute humidity of 4 mg H2O/l lower than that listed by the manufacturer. In fact, in some of them the difference was higher than 8 mg H2O/l.
The contraindications of HMEs are presented in Box 1.1.
Comparison between active and passive humidifiers
Some authors have shown an increase in ETT occlusion by using HMEs in comparison to HH.
Based on this information, combined hydrophobic hygroscopic HMEs have been compared to hydrophobic HMEs and HH. After 72 h, the mean diameter of the ETT has decreased by 6.5 mm with hydrophobic HME, 2.5 mm with hygroscopic hydrophobic HME, and 1.5 mm with HH.
HMEs also have unfavourable effects on ventilation parameters. They increase the dead space, which in turn decreases alveolar ventilation and leads to increase in arterial CO2 tension. Hence, to keep the same level of alveolar ventilation, tidal volume has to be increased, exposing patients to VILI. In spontaneously breathing patients, the addition of dead space associated with HMEs may increase work of breathing, precluding liberation from mechanical ventilation.
Le Bourdellès et al. observed a significant decrease in alveolar ventilation with HMEs in comparison to HH during weaning from mechanical ventilation in a pressure support mode. These observations should be taken into account during difficult weaning.
HMEs increase dynamic hyperinflation and inspiratory and expiratory resistance, which contributed to the development of intrinsic PEEP in comparison to HH.
On the basis of these observations, one may recommend increasing the level of pressure support (by 5–10 cmH20) to maintain constant work of breathing. For all these reasons, HME is not recommended for patients with COPD.
In non-COPD patients receiving prolonged mechanical ventilation, Ricard et al. compared a group of patients where the HME was changed every 48 h to a group with a change after 7 days. There was no difference in resistance in the ETT or in infection, indicating that the HME can be changed only once a week, which represents a substantial cost saving.
The effect of these systems on the incidence of VAP is controversial. Three reviews or meta-analyses reached different conclusions: Ricard et al. and Niël-Weise et al. concluded that the type of humidification device does not influence the incidence of VAP. Yet Niël-Weise et al. recommended a heated wire circuit in active humidification because less condensate reduces colonization. Kola et al. reported a significant reduction in the occurrence of VAP with HMEs, particularly for patients ventilated for more than 7 days, but they excluded patients at high risk of airway occlusion (COPD).
In 2009, the European Respiratory Society, the European Society of Clinical Microbiology and Infectious Diseases, and the European Society of Intensive Care Medicine issued a joint statement preferring HMEs over HHs for the prevention of VAP. In the same year, the VAP Guidelines Committee and the Canadian Critical Care Trials Group stated that there was no difference in the incidence of VAP between HMEs and HHs. The inclination of the European guidelines toward HMEs coincides with the trend in clinical practice. A cross-sectional survey denoted that HMEs were more commonly used in France than in Canada.
Regardless of their type, humidifiers are essential for mechanically ventilated patients. For at-risk patients (e.g. COPD, copious secretion), one may prefer active humidifiers with heated wire circuits.
During weaning from mechanical ventilation in pressure support mode and in non-invasive ventilation, HMEs are not recommended because of increases in dead space, PaCO2, and work of breathing, unless one significantly increases the pressure support level. Passive and active devices seem to be equivalent in terms of efficacy and prevention of infection.
The respiratory system and the cardiovascular systems are not separate, but tightly integrated. ARF can directly alter cardiovascular function and vice versa. Many of these effects are predictable from knowledge of cardiovascular function. Both lung underinflation and hyperinflation increase pulmonary vascular resistance, heart–lung interactions, and the work of breathing. Both spontaneous inspiratory efforts during acute bronchospasm and acute lung injury induce markedly negative swings in ITP. Artificial ventilatory support increases ITP during inspiration, in contradistinction to spontaneous ventilation, which will decrease ITP for the same tidal breath. Heart–lung interactions involve four basic concepts: inspiration increases lung volume, spontaneous inspiration decreases ITP, positive pressure ventilation increases ITP, and ventilation is exercise, i.e. it consumes O2 and produces CO2.
Haemodynamic effects of changes in lung volume
Lung inflation alters autonomic tone and pulmonary vascular resistance, and, at high lung volumes, compresses the heart, limiting filling, similar to cardiac tamponade. The associated diaphragmatic dissent increases abdominal pressure, compressing the liver and altering venous return. Each of these processes may predominate in determining the final cardiovascular state. Small tidal volumes (<10 ml/kg) inspiration increase heart rate by vagal withdrawal, called respiratory sinus arrhythmia. Larger tidal volumes (>15 ml/kg) decrease heart rate, arterial tone, and cardiac contractility by sympathetic withdrawal.
The major determinants of the haemodynamic response to increases in lung volume are mechanical in nature. Lung inflation, independently of changes in ITP, primarily alters right ventricular preload and afterload and left ventricular preload. First, inspiration induces diaphragmatic descent, which increases abdominal pressure. Venous return is a function of the ratio of the pressure difference between the right atrium and the systemic venous reservoirs and the resistance to venous return. Since a large proportion of the venous blood volume is in the abdomen, abdominal pressure increases should augment venous blood flow. Diaphragmatic descent also compresses the liver, increasing hepatic outflow resistance, thus decreasing venous blood flow. Thus, inspiration shifts venous flow from high-resistance splanchnic circuits draining through the liver to low-resistance systemic venous circuits, making venous return greater for the same blood volume. Thus, increasing lung volume may increase or decrease venous return depending on which factors predominate. Usually, inspiration increases venous return in volume-overloaded states and decreases venous return in hypovolaemic and hepatic cirrhotic states.
End-expiratory lung volume also determines alveolar stability. Alveolar collapse increases pulmonary vasomotor tone by hypoxic pulmonary vasoconstriction. Alveolar recruitment reverses this process, although, during the recruitment manoeuvres, the associated hyperinflation may impair right ventricular function. Increasing lung volume above the FRC also increases right ventricular outflow resistance by increasing transpulmonary pressure more than pulmonary artery pressure. Reversing hyperinflation by any means decreases pulmonary arterial pressure, improving right ventricular ejection. The use of smaller tidal volumes and less PEEP has reduced the incidence of acute cor pulmonale seen in critically ill patients. Left ventricular end-diastolic volume (preload) can be altered by ventilation by decreasing venous return, right ventricular dilation-induced decreased left ventricular diastolic compliance (ventricular interdependence), and by cardiac compression by the expanding lungs.
Haemodynamic effects of changes in ITP
The heart within the chest is a pressure chamber within a pressure chamber. Changes in ITP affect the pressure gradients for systemic venous return to the right ventricle and systemic outflow from the left ventricle independent of the heart itself. Increases in ITP reduce these pressure gradients, decreasing intrathoracic blood volume. Decreases in ITP augment venous return and impede left ventricular ejection, increasing intrathoracic blood volume. Variations in right atrial pressure represent the major factor determining the fluctuation in pressure gradient for systemic venous return during ventilation. Increases in ITP with positive pressure ventilation or hyperinflation during spontaneous ventilation decrease venous return, whereas decreases in ITP during spontaneous inspiration increase venous return.
Left ventricular afterload or systolic wall tension is proportional to the product of transmural systolic left ventricular pressure and left ventricular volume. Thus, increasing ITP will decrease transmural left ventricular pressure if arterial pressure is constant; increases in ITP unload the left ventricle, whereas decreases in ITP have the opposite effect. Increases in ITP actually may increase cardiac output in congestive heart failure states. Spontaneous ventilatory efforts against a resistive (bronchospasm) or elastic (acute lung injury) load decrease left ventricular stroke volume manifest as pulsus paradoxus by ventricular interdependence and increased left ventricular afterload. Increases in ITP have the opposite effect, decreasing left ventricular afterload. Although increases in ITP should augment left ventricular ejection by decreasing left ventricular afterload, this effect is limited because of the obligatory decrease in venous return.
There is no difference from a mechanical perspective between increasing ITP from a basal end-expiratory level and eliminating negative end-inspiratory ITP swings seen in spontaneous ventilation. Removing negative swings in ITP may be more clinically relevant than increasing ITP for many reasons. First, many pulmonary diseases are associated with exaggerated decreases in ITP during inspiration. In restrictive lung disease states, such as interstitial fibrosis or acute hypoxaemic respiratory failure, ITP must decrease greatly to generate a large enough transpulmonary pressure to ventilate the alveoli. Similarly, in obstructive diseases, such as upper airway obstruction or asthma, large decreases in ITP occur owing to increased resistance to inspiratory airflow. Secondly, exaggerated decreases in ITP require increased respiratory efforts that increase the work of breathing, taxing a potentially stressed circulation. Finally, the exaggerated decreases in ITP can only increase venous blood flow so much before venous collapse limits blood flow. The level to which ITP must decrease to induce venous flow limitation is different in different circulatory conditions but occurs in most patients below an ITP of –10 cmH2O. Thus, further decreases in ITP will further increase only left ventricular afterload without increasing venous return. Abolishing these markedly negative swings in ITP reduces left ventricular afterload more than venous return (left ventricular preload). These concepts of a differential effect of increasing and decreasing ITP on cardiac function are illustrated for both normal and failing hearts in Figures 1.16 and 1.17 using the left ventricular pressure–volume relationship during one cardiac cycle to interpose venous return (end-diastolic volume) and afterload (end-systolic volume). Thus, performing an endotracheal intubation in patients with obstructive breathing abolishes the markedly negative swings in ITP without reducing venous return.
Ventilation as exercise
Spontaneous ventilatory efforts are exercise, requiring increased blood flow and O2, and produce CO2. In lung disease states where the work of breathing is increased, the work cost of breathing may increase to 25% or more of total O2 delivery, limiting exercise capacity, inducing coronary ischaemia, and leading to weaning failure. Starting artificial ventilation will decrease O2 extraction, increasing venous oxygen saturation (SvO2) for a constant cardiac output and CaO2. Under conditions in which fixed right-to-left shunts exist, the obligatory increase in SvO2 will result in an increase in the PaO2, despite no change in the ratio of shunt blood flow to cardiac output.
Multiple choice questions and further reading
Interactive multiple choice questions to test your knowledge on this chapter and additional further reading can be found in Appendix Chapter 1 Multiple choice questions and further reading