Acute lung injury has a wide range of causes and a broad spectrum of severity. In its mildest form the only evidence for lung dysfunction may be an increase in the alveolar–arterial oxygen gradient (A-a)Do&sub2;, without clinical signs of lung congestion, radiographic changes or reduction in lung compliance. The more severe forms of acute lung injury are generally referred to as the adult respiratory distress syndrome. The most common conditions associated with the development of this syndrome are listed in Table 1 62.

The severity of the disease process depends upon the nature, severity, and duration of the insult to the lungs. Adult respiratory distress syndrome associated with aspiration, fat embolism, pancreatitis, or sepsis may be more severe than the lung injury associated with a brief episode of hypovolaemic shock or mismatched blood transfusion. In view of the heterogeneity of disease a simple scoring system, such as that shown in Table 2 63, should be used to quantify the severity of acute lung injury.

The incidence of adult respiratory distress syndrome is uncertain, but it may effect 5 to 10 per cent of patients at risk. Clearly, the criteria by which the condition defined will be reflected in the incidence, morbidity, and mortality reported in any particular series. The observation that mortality has changed little over the past two decades may indeed reflect a trend towards a stricter definition. It is evident that patients do not usually die of hypoxaemia but from the complex disturbances that result from multiple organ system failure. The management of patients with adult respiratory distress syndrome is therefore aimed at many facets of organ system support.

Recognition of moderate to severe adult respiratory distress syndrome is not difficult. The patient has acute respiratory distress, requires an increasing Fio&sub2; (greater than 0.5) to maintain a Pao&sub2; of more than 7.0 kPa (50 mmHg) and usually has extensive lung infiltrates on the chest radiograph. With mechanical ventilation, the lungs are stiff and require high inflation pressures. Calculated effective, static, and dynamic compliance (C&subd;&suby;&subn; and C&subs;&subt;&suba;&subt;) are low usually being less than 20 ml/cmH&sub2;O (30 ml/mmHg). Cardiogenic pulmonary oedema should be excluded by measurement of the pulmonary artery wedge pressure (<18 mmHg). The Murray lung injury score will usually be more than 2.5. Calculated venous admixture reveals a shunt fraction between 30 and 50 per cent.

There is no specific therapy at present for acute lung injury, except to suppress or remove the injurious agent, avoid aggravating the condition, and supporting the patient until lung function recovers. Adult respiratory distress syndrome may require prolonged periods of mechanical ventilation and can be associated with a mortality of more than 50 per cent. Above all, the clinician must anticipate its development, with the aid of an awareness of the precipitating causes (Table 1) 62. A proactive approach to the management of these underlying disease processes may help prevent or ameliorate progression to respiratory failure.

The pathogenesis of adult respiratory distress syndrome is complex, being the product of several processes such as complement activation, neutrophil accumulation, platelet aggravation, and mediator release. In patients with trauma and sepsis, endotoxaemia is responsible for initiating complement activation and other components of the immune–inflammatory cascade. Complement activation causes leucocyte sequestration in pulmonary capillaries which, in turn, injures endothelial cells through the release of toxic superoxide radicals, arachidonic acid metabolites, cytokines, and proteases.

Unfortunately, monitoring complement activation appears to have little value in predicting the development or the clinical course of adult respiratory distress syndrome. It is also clear that the condition can still develop in the most severely neutropenic patients. Bronchoalveolar lavage studies have detected the presence of neutrophils, proteolytic enzymes, chemotactic factors, antiproteases, and cytokines in lung washings. Activation of the clotting system may further aggravate the pathophysiological disturbances by causing intravascular coagulation and platelet aggregation within the pulmonary vascular bed. Furthermore, the platelet and fibrin aggregates release vasoactive substances, such as serotonin, histamine, and prostaglandins that, in combination with fibrinogen degradation products, damage the endothelium and pulmonary microvasculature.

Complicating the picture further is the possibility that supportive measures such as oxygen and positive pressure ventilation may themselves promote or prolong lung injury.

Adult respiratory distress syndrome is characterized by diffuse alveolar epithelial and endothelial damage. In the early stages, the alveolar interstitium becomes infiltrated with inflammatory cells and the alveoli spaces filled with proteinaceous and haemorrhagic fluid. Hyaline membrane formation and capillary microembolism are common features. Leakage of plasma through damaged pulmonary endothelium results in interstitial and alveolar oedema and may alter the properties of surfactant. Alveolar type I cells are replaced by cuboidal microvillous Type II cells, disturbing alveolar wall architecture and exacerbating the effects of pulmonary capillary injury. There may be progression to pulmonary fibrosis with obliteration of the pulmonary alveolar and microvasculature architecture. Despite the extensive pathological changes that may be found on biopsy, good physiological recovery is possible.

The pathological changes result in reduced functional residual capacity, increased venous admixture (shunt), reduced lung compliance, and refractory hypoxaemia. The role played by abnormal surfactant in the pathophysiology of adult respiratory distress syndrome is unclear, but its severity seems to be related to the proportion of the abnormal surfactant. Bronchoalveolar lavage fluid obtained from patients with adult respiratory distress syndrome contains aggregated and inactive surfactant. Unlike the experience with respiratory distress in the neonate, surfactant replacement in the adult condition has produced inconsistent results.

The aim of respiratory support is to achieve adequate oxygen on-loading at the lungs without impeding cardiac output. To help maintain oxygen delivery in the haemoglobin concentration should be maintained above 11 g/100 ml of blood. Oxygen, intermittent positive pressure ventilation and positive end-expiratory pressure form the basis of respiratory support in adult respiratory distress syndrome. Although constant positive airway pressure can be delivered by a tight-fitting anaesthetic face mask in the spontaneously breathing patient, endotracheal intubation is generally required. The endotracheal tube should be of the high volume/low pressure cuffed variety and the cuff inflated to a pressure not exceeding 30 cmH&sub2;O.

The ventilator selected to support the patient should have the facility to deliver the basic modes of ventilation, control mechanical ventilation, assist control, synchronized intermittent mandatory ventilation, and pressure support. The best mode of ventilation is not known but selection should be based upon whichever achieves ventilation goals for the lowest peak and mean airway pressure.

Most patients will require a Fio&sub2; of more than 0.6 to achieve an acceptable Pao&sub2; (one that satisfies the oxygen delivery requirements of the patient). For example a patient on 100 per cent oxygen and 20 cmH&sub2;O positive end-expiratory pressure, with a haemoglobin concentration of 12 g/100 ml and a Pao&sub2; of 7.0 kPa (53 mmHg) might appear dangerously hypoxaemic. Yet this low Pao&sub2; may satisfy oxygen requirements provided the cardiac output is adequate. Oxygen extraction by the tissues drives the mixed venous oxygen content to lower than normal values. The clinician must therefore be alert to evidence of tissue hypoxaemia, such as metabolic acidosis, mental obtundation, or cardiac arrhythmias. If such evidence is not present it may be reasonable to assume that oxygen delivery is adequate. If doubt exists about the adequacy of oxygen delivery, a flow directed pulmonary artery catheter can be inserted to measure cardiac output and mixed venous saturation and to calculate oxygen delivery.

This recruits lung units partly to restore the reduced functional residual capacity to normal. Positive end-expiratory pressure of 5 to 20 cmH&sub2;O (0.5–2.0 kPa) is added in increments of 5 cmH&sub2;O until an acceptable Pao&sub2; is reached. Current practice encourages the use of the lowest pressure that will meet oxygenation goals. ‘Best’ or ‘optimal’ positive end-expiratory pressure is the pressure that maximizes oxygen delivery, with the lowest Fio&sub2;, lowest venous admixture, and highest effective compliance. Concern over barotrauma has led to a more conservative approach to the use of positive end-expiratory pressure in recent years.

Ventilation settings of the mechanical ventilator may be as crucial as the level of positive end-expiratory pressure with regard to causing barotrauma. Conventional high tidal volume settings of 10 to 12 ml/kg may be inappropriate in the patient with adult respiratory distress syndrome and non-compliant lungs. Increasing the ventilation rate and reducing the tidal volume will reduce peak airway pressures at the cost of higher VD/VT and higher mean airway pressure. A deliberate policy of ‘permissive hypercapnia’, allowing the Pco&sub2; to rise to between 7 and 9 kPa, can be achieved by reducing both the tidal volumes and ventilator rate. This results in lower mean and peak airway pressures. The attendant respiratory acidosis is usually well tolerated and ultimately compensated for.

A feature of the patient with adult respiratory distress syndrome is the extensive gravity dependent lung collapse that is best visualized by CT scan. To a large degree this is a consequence of nursing the patient for prolonged periods in the supine position. Sitting the patient as upright as possible minimizes the volume of dependent lung. In the persistently hypoxic patient, significant improvement in oxygenation may be obtained by rolling the patient into a prone position every 6 to 8 h.

Non-conventional forms of ventilation such as high frequency ventilation have potential advantages in supporting patients with adult respiratory distress syndrome. In randomized studies peak airway pressures are lower but mean airway pressure and mortality are unchanged. Inverted I:E ratio ventilation (inspiration to expiration ratio) has been claimed to reduce the positive end-expiratory pressure level for a given Pao&sub2;. However, the mean airway pressures are inevitably raised by such techniques.

Extracorporeal membrane oxygenation for respiratory failure was abandoned over a decade ago following controlled, randomized investigation. With the availability of improved oxygenators there has been renewed interest in combining extracorporeal partial CO&sub2; removal and low frequency conventional ventilation. Whether this approach offers significant advantages over permissive hypercapnia techniques remains to be seen. Both techniques are aimed at reducing the deleterious effect of positive pressure upon the alveolar epithelium and may therefore hasten recovery from adult respiratory distress syndrome.

Intravenous gas exchange devices (IVOX® ) have been developed that may offer the safest alternative to conventional ventilation and gas exchange techniques. These devices will require extensive evaluation before they replace or supplement conventional ventilation techniques.

Fluid administration should be regulated to ensure adequate cardiac output without aggravating the pulmonary oedema. Measurement of pulmonary capillary wedge pressure using a flow directed pulmonary artery catheter may provide a more reliable assessment of the filling status of the patient than measurement of central venous pressure. The information provided by measures of cardiac output and derived indices of vascular resistance rationalizes the use of inotropes, pressors, and afterload reduction. Pulmonary hypertension is not an uncommon feature of adult respiratory distress syndrome, and although attempts to reduce the pulmonary arterial pressures have not reduced mortality, significant increases in oxygen delivery can be obtained with agents such as prostacyclin. Despite an increase in venous admixture with prostacyclin, the increases in cardiac output and mixed venous saturation produce enhanced oxygen delivery and improved Pao&sub2;.

There is no strong evidence to favour colloidal over crystalloid replacement fluid. The volume of fluid is more critical: the type of fluid should be selected to ensure electrolyte equilibrium and defend plasma colloid osmotic pressure. Diuretics should be used to reduce intravascular volume as needed. If the patient becomes oliguric, early haemofiltration will be needed to maintain fluid balance. Nutritional support in the form of enteral feeding or total parenteral nutrition can be started early in the course of adult respiratory distress syndrome.

Specific treatments to inhibit mediator cascades are largely experimental. Multicentre randomized studies have failed to show significant benefit from corticosteroids, non-steroidal anti-inflammatory agents, or vasodilator prostaglandins. Anecdotal reports of high dose steroids (40–60 mg prednisone/day) in ‘chronic’ respiratory distress suggest some benefit in terms of oxygenation and lung compliance. However, the use of steroids in the acute phase afforded no benefit in multicentre, controlled studies.

Antiendotoxin monoclonal antibodies significantly reduce mortality in septic shock and may therefore prevent or moderate adult respiratory distress syndrome in this group of patients. As yet, this hypothesis and others aimed at alternative components of the immunoinflammatory cascade have not been formally tested.

Provided that multiple organ system support can be maintained, a positive attitude towards final recovery is justified. Even after periods of mechanical ventilation, a high Fio&sub2; and positive end-expiratory pressure for up to 3 to 4 months a good functional outcome is possible. Biopsy proven pulmonary fibrosis does not inevitably mean that there is fixed, irreversible pathology. A mortality rate of over 50 to 60 per cent is generally quoted for the last decade, although most recent reports have shown a fall in mortality to about 20 per cent. Pulmonary function testing 1 year following recovery may show a reduction in vital capacity with a mild obstructive defect. In many patients the only abnormality may be a reduction in carbon monoxide transfer.

The provision of efficient and safe mechanical ventilation is a skill that must be mastered by all physicians practising critical care. The basic principles still pertain despite the introduction of complex and sophisticated mechanical ventilators and the overabundance of studies claiming superiority of certain techniques over others. The application of common sense and sound physiological doctrine will serve better than devotion to an attractive technical innovation.

Although mechanical ventilation is not to be undertaken lightly since it is associated with much morbidity and some mortality, failure to intervene promptly can have catastrophic consequences for the patient. The indications for mechanical ventilation fall into two broad categories: inadequate alveolar ventilation with increasing Pco&sub2;, and inadequate gas exchange with increasing alveolar–arterial oxygen gradient and arterial hypoxaemia. Guidelines for mechanical ventilation in acute respiratory failure are shown in Table 3 64. The physician must always exercise clinical judgement in the interpretation of these guidelines and anticipate problems before they arise. For example, one of the simplest criteria for mechanical ventilation is a respiratory rate of 35 breaths per minute or more. If a patient with a respiratory rate of 30 breaths per minute is clearly becoming fatigued an early elective intubation is preferred to an emergency procedure an hour or so later. Similarly, a progressive fall in vital capacity in a patient with myasthenia gravis receiving full medication may indicate a need for ventilatory support, although the critical value of less than 15 ml/kg is not broached.

Endotracheal intubation will be the preferred technique in most cases. Orotracheal intubation is particularly suited to emergency intubation while nasotracheal intubation requires a little extra time. A coagulation defect or thrombocytopenia makes nasotracheal intubation inadvisable due to risk of serious haemorrhage. Whatever technique is selected, intubation should be performed in a safe and expeditious manner by the most experienced clinician available: this will usually be an anaesthetist or trained specialist in intensive care. Neuromuscular relaxant drugs should only be used to facilitate intubation by experienced personnel. Complications of endotracheal intubation are due to occlusion or displacement of the tube, and airway trauma. The appropriate endotracheal tube size for most adult males is 8 to 9 mm internal diameter; and for women, 7 to 8 mm. For children, a rough calculation using the child's age in years divided by 4, plus 4.0 will provide the tube internal diameter in mm. These smaller tubes are generally uncuffed.

It is essential that the endotracheal tube be securely anchored and the cuff inflation pressure restricted to less than 30 cmH&sub2;O. The latter can be achieved by always using high volume, low pressure cuffed tubes and allowing a small cuff leak during each ventilator cycle (minimal leak technique). Alternatively, cuff inflation pressures can be measured periodically using an anaeroid manometer and adjusted accordingly. Contrary to popular belief, higher cuff pressures do not improve airway protection against aspiration but only serve to damage the tracheal mucosa and risk later stenosis.

Tracheostomy should replace endotracheal intubation for specific indications and not merely after the elapse of a predefined time interval. Using modern endotracheal tubes and techniques, endotracheal intubation can be tolerated without permanent harm to the airway for months if necessary. It has been shown that most mucosal damage is caused in the first week of intubation with little additional change thereafter.

However, much can be gained by the judicious selection of patients for tracheostomy either as the preferred primary route for airway access or as a replacement for endotracheal intubation. The common indications for replacement include the need for chronic or permanent ventilation, to help weaning after previously failed attempts at extubation, to facilitate oral nutrition, or the presence of upper airway complications of endotracheal intubation. Indications for primary tracheostomy are considered elsewhere.

The same principles of cuff pressure management apply to tracheostomy tubes as to endotracheal tubes. Tracheostomy is associated with fewer but more serious complications than endotracheal intubation. These include tube displacement, pneumothorax, severe haemorrhage, and wound infection.

Minitracheostomy tubes are 3.5- to 4.0-mm diameter, cuffless tubes inserted percutaneously through the cricothyroid membrane, usually under local anaesthesia. A Seldinger technique for introduction of the minitracheostomy tube offers an alternative to the direct trochar method. Minitracheostomy allows suctioning lung secretions without the need for formal endotracheal intubation or tracheostomy.

A cricothyroidotomy may be preferred in life-threatening, upper airway obstructions where endotracheal intubation is not feasible and there is insufficient time to perform tracheostomy. Performed under local anaesthesia, a full sized tracheostomy tube can be inserted (6–8 mm internal diameter) to facilitate mechanical ventilation. In emergency conditions, temporary oxygenation (30 min) can be aided by the intermittent insufflation of 100 per cent oxygen at 50 PSI (15 l/min) via a 14-gauge cannula inserted through the cricothyroid membrane. Interruption of oxygen flow is regulated by a simple Y piece adaptor.

Most adult patients are supported on volume/time cycled, pressure limited ventilators (volume ventilator or flow generator). The volume/time cycled ventilators deliver preset tidal volumes regardless of changes in lung compliance or impedance. The price paid for this desirable characteristic is that the inflation pressures must rise to overcome the mechanical load. To protect the patient against inadvertently high pressures, a pressure limit must be set. When this limit is reached the ventilator terminates inspiration regardless of the volume delivered and triggers an alarm.

Neonates and infants may be satisfactorily ventilated using time cycled, pressure limited devices (pressure ventilator or pressure generator). The pressure limited paediatric ventilator offers simplicity and reliable ventilation although the delivered tidal volume is difficult to measure. In the premature neonate these are not serious limitations and pressure limited ventilation is the preferred technique.

Specifically designed, compact, lightweight ventilators, driven by cylinder oxygen and using fluid logic circuits are available for transporting ventilator dependent patients. They are pressure generators and can be used for both adults and children. By entraining air, a choice of either 60 per cent or 100 per cent oxygen is available.

All ventilators possess a drive mechanism that propels the air/oxygen gas mixture into the patient. Some deliver the gas to the patient directly (single circuit) or indirectly through a dual circuit. There are several types of drive mechanisms that determine the flexibility, the available ventilation modes, and the ability to deliver preprescribed tidal volume in the face of abnormal lung mechanics.

Schematics of a simple positive pressure ventilator and ventilator circuit during inspiratory and expiratory phases are shown in Figs. 1 and 2 59,60.

The control mechanism that cycles the ventilator from inspiration to expiration may be electromechanical, or electronic (using microprocessor technology). Modern mechanical ventilators tend to fall into the latter category and offer a degree of sophistication that has greatly improved the safety and efficiency of mechanical ventilation.

Depending upon the indications for ventilation, the clinician must select the mode of ventilation, choose the ventilation parameters, and adjust the ventilator alarms. The most commonly available ventilator modes include control mechanical ventilation, assist control (triggered ventilation), intermittent mandatory ventilation (IMV or SIMV), and pressure support, although some others are used.

This provides time and volume cycled, pressure limited breaths at preset rates, but does not allow the patient to breathe spontaneously. This mode is suitable for the paralysed or heavily sedated patient.

This synchronizes the ventilator to the patient's own respiratory rhythm, delivering a volume preset, pressure limited tidal volume. A trigger sensitivity must be selected (usually −0.5 to −2.0 cmH&sub2;O) by which the patient can initiate volume preset breaths above the set rates. Patients have a tendency to hyperventilate on assist control. As a safety requirement, a high respiratory rate alarm is needed and a ‘back up’ ventilation rate must be set in the event of apnoea. Assist control is better tolerated than control mechanical ventilation and the patient requires less sedation.

This was originally devised for weaning but is now widely adopted as a maintenance mode (see Fig. 3 61). It provides the opportunity for the patient to breathe spontaneously and supplement the positive pressure minute ventilation. In the standard intermittent mandatory ventilation mode there is a theoretical risk of stacking a ventilator breath on top of a spontaneous breath. However, this does not appear to be a significant problem and an appropriately set pressure limit should prevent this causing inadvertent overinflation of the lungs. More modern ventilators use the triggering or assist facility to synchronize the machine breaths with the patient's own spontaneous breathing pattern (synchronized intermittent mandatory ventilation). This technique is intended as partial ventilation support. With the patient taking spontaneous breaths, it is better tolerated than control mechanical ventilation, results in lower mean airway pressures, has less effect on the cardiovascular system, and allows the patient to regulate their own Pco&sub2; to at least some degree.

This uses a triggering facility to deliver, not a volume preset breath as in assist control, but a pressure limited breath (as with paediatric pressure ventilation). The inspiratory flow rate is usually high to minimize phase lag and the work of breathing. Pressure support may be used alone or in conjunction with synchronized intermittent mandatory ventilation when it assists spontaneous breaths. Pressure support provides an efficient maintenance and weaning mode that is well tolerated by the patient.

Pressure release, high frequency and inverse I:E ratio ventilation have their proponents. Evidence to indicate significant superiority of these modes over conventional methods of ventilation is not convincing. High frequency ventilation in its several forms has been recommended for use following reconstructive laryngeal, tracheal, or bronchial surgery or for patients with bronchopleural or bronchocutaneous fistulae. Even these applications may not offer much advantage, if any, over conventional modes. Mandatory minute ventilation is an innovative mode whereby the combined spontaneous and mechanical ventilation must reach a minimum preset level. As the patient's spontaneous ventilation increases the mechanically assisted breaths become fewer. Individual ventilators vary in their ability to achieve successful mandatory minute ventilation.

Restriction of the expiratory gas flow through a flow-dependent resistance prolongs expiration and delays the airway pressure drop to atmospheric levels. Expiratory retard was claimed to prevent premature airway closure in patients with chronic obstructive pulmonary disease and improve lung emptying. It is doubtful whether expiratory retard contributes any additional benefit beyond that provided by positive end-expiratory pressure or continuous positive airways pressure and may increase the work of breathing.

Inspiratory pause or hold prolongs the inspiratory phase and delays the onset of expiration. As such it is thought to increase oxygenation by improved ventilation distribution and reduced V/ Q mismatch. Like positive end-expiratory pressure and continuous positive airways pressure, it increases mean airway pressure, and may have cardiovascular effects. It is doubtful whether inspiratory pause contributes significantly to patient management.

Before the advent of high tidal volume ventilation and positive end-expiratory pressures and continuous positive airways pressure, sighs were added to the ventilation protocol to prevent progressive atelectasis. Each sigh was delivered 2 to 6 times per hour and was equivalent to about twice the conventional tidal volume. The risks of barotrauma probably outweigh the theoretical benefits.

Once a ventilation mode has been selected (at least temporarily), ventilatory parameters must be set before attaching the patient to the ventilator. The ventilator parameters include:

1.Tidal volume;

2.Ventilation rate;

3.Inspiratory/expiratory (I:E) ratio;

4.Flow waveform;

5.Fio&sub2; (0.21 to 1.0);

6.Pressure limit;

7.Positive end-expiratory pressure/continuous positive airway pressure (0 to 20 cmH&sub2;O).

The delivered, inspiratory tidal volume may be set at 10 to 12 ml/kg body weight. This should be reduced if the patient has restrictive lung disease or has undergone lobectomy or pneumonectomy. Using respiratory rates of more than 10 breaths per minute with such tidal volumes will provide full ventilatory support. If the patient is breathing spontaneously, an intermittent mandatory ventilation mode will be preferred at rates between 4 to 8 breaths per minute. If assist control or pressure support is chosen the respiratory rate will be the patient's spontaneous rate.

The rate of inspiratory to expiratory time (I:E ratio) will generally range from 1:2 to 1:4. This provides sufficient time for full passive exhalation. In patients with obstructive lung disease failing to allow adequate time for exhalation results in hyperinflation (auto- or intrinsic positive end-expiration pressure). The higher the set respiratory rate the shorter expiration becomes and the I:E ratio falls. This may lead to the paradoxical situation in the patient with chronic obstructive pulmonary disease where the Pco&sub2; rises as the ventilator rate is increased.

The I:E ratio can be adjusted in several ways depending upon the make of ventilator. In some, a ratio can be selected directly, while in others, the inspiratory flow rate determines the duration of inspiration. An acceptable range for inspiratory flow rates is between 30 and 60 l/min (0.5–1.0 l/s).

Many volume and time cycled (flow generator) ventilators allow the choice of several waveforms. Although there is little evidence to favour one over the other, a square waveform delivers the tidal volume in the least time and with higher peak pressures. A decelerating flow pattern results in lower peak pressures, longer inspiratory intervals, and lower I:E ratios.

This should be constantly adjusted to provide adequate arterial oxygenation without hyperoxia. Too high a Fio&sub2; is frequently the cause of failure to wean patients with chronic obstructive pulmonary disease from a mechanical ventilator.

Setting a pressure limit about 10 cmH&sub2;O above the peak pressure reached during each ventilator cycle protects the patients against inadvertently high pressures experienced during coughing or straining. Hitting the pressure limit terminates inspiration and sounds an alarm.

Maintaining airway pressure above barometric pressure in a spontaneously breathing patient is called constant positive airway pressure. The same pressure applied to a patient on intermittent positive pressure ventilation is called positive end-expiratory pressure. This technique is used to correct lung volume (FRC) in conditions characterized by reduced lung volume such as adult respiratory distress syndrome or cardiogenic pulmonary oedema. It may also be of benefit in patients with flail chest segments since it acts to splint the chest wall.

Positive end-expiratory pressure/continuous positive airways pressure is achieved by the inclusion of a resistance at the expiratory end of the breathing circuit. Ideally this resistance should be as close to a threshold resistor as possible, such as an underwater column. In practice most of the valves produce some flow-dependent retardation of expiration that increases the work of breathing in a spontaneously breathing patient.

Bypassing the oropharynx by an endotracheal tube is known to cause a fall in FRC in both adults and children. The application of low levels of continuous positive airways pressure (3–5 cmH&sub2;O) reverses this effect and has therefore been suggested as part of routine management of the intubated patient (‘physiological continuous positive airways pressure’). The usual indication for positive end-expiratory pressure is the presence of refractory hypoxaemia due to acute lung injury. Starting at 5 cmH&sub2;O the pressures are increased progressively until satisfactory oxygenation is achieved for a Fio&sub2; ideally less than 0.6. It is rarely necessary to exceed levels of 20 cmH&sub2;O. Assessment of the effects of such treatment can be made in several ways by calculating the venous admixture or shunt fraction, oxygen delivery or effect static compliance of the lungs. Clearly, following the Pao&sub2; alone takes no account of the effects of positive pressure upon cardiac output. Continuous measurement of mixed venous Sao&sub2; with a suitable flow directed pulmonary artery catheter is a particularly good method of evaluating the response to a change in airway pressure.

1.Exhaled tidal volume, exhaled minute ventilation;

2.Airway pressure/disconnect;

3.Peak and mean airway pressures;

4.Fio&sub2;;

5.Inhaled gas temperature.

The ability to measure exhaled tidal volume and minute ventilation is a great asset and aid in evaluating the efficiency of ventilation. By setting low limit alarms the safety of mechanical ventilation is greatly enhanced and supplements the airway pressure/disconnect alarm (Fig. 1(d)) 59.

Careful monitoring of peak and mean airway pressures is essential in the patient supported with a flow generating (volume/time cycled) ventilator. A step increase in peak pressure may indicate a change in lung compliance, airway resistance, displacement of the endotracheal tube into a main bronchus, or even a pneumothorax. The pressure alarm should therefore be set 5 to 10 cmH&sub2;O above the normal peak pressure (Fig. 4) 62.

An oxygen sensor just proximal to the endotracheal tube ensures that any sudden fall in Fio&sub2; is detected immediately and before the patient can desaturate.

Strict control of the temperature of the inhaled gases is essential to avoid thermal injury and to ensure effective humidification of inspired gases. Modern humidifiers are usually of the ‘wick’ variety that efficiently humidify high gas flows. Heat/moisture exchangers placed between the ventilator circuit Y-piece and the endotracheal tube provide an alternative method of humidification particularly for short-term ventilation. Inspired gas should leave the humidifier at about 40°C and reach the endotracheal tube at close to 37°C (Fig. 4) 62. A low voltage heating element can be incorporated into the inspiratory limb of the circuit to reduce gas cooling and water condensation.

An essential aspect of monitoring is regular clinical examination of the patient, and inspection of the ventilator and ventilator circuit. Expansion of the chest should be symmetrical with each ventilator cycled breath (control mechanical ventilation, intermittent mandatory ventilation), assisted breath (assist control or pressure support), or unassisted spontaneous breath (intermittent mandatory ventilation). Auscultation should confirm air entry and detect any added sounds. The patient should be sat up or rolled side to side to allow inspection of the whole of the chest. The endotracheal tube should be secure and as comfortable as possible for the patient. The endotracheal cuff pressure should be checked (<30 mmHg) or a small leak should be audible with a stethoscope on the side of the neck. The ventilator circuit should feel warm but be free of significant amounts of condensed water. The humidifier temperature and water level should be checked.

The pulse oximeter has contributed significantly to the monitoring and safety of patients on mechanical ventilation. Not only does it provide a continuous measurement of oxygenation but also reduces the need for arterial blood gas sampling.

Much can be appreciated from watching the ventilator pressure gauge with each cycle. In addition to evaluating peak inspiratory pressure the clinician will be able to judge whether the patient is ‘fighting’ the ventilator. Comparing inspiratory and expiratory tidal volumes may indicate a leak in the circuit, either at circuit connections or at the endotracheal tube cuff. When peak pressures are high, the internal compliance of the ventilator and circuit (about 2 to 2.5 ml/cmH&sub2;O) may account for much of the volume loss. A rough assessment of the compliance of the lung can be made by following the peak inflation pressures. However, it is preferable to calculate effective static and dynamic compliance in the following way.

The effective dynamic compliance is always less than the effective static compliance by about 10 ml/cmH&sub2;O. Greater differences indicate airways obstruction is contributing to the peak airway pressure. Normal values for effective dynamic compliance range between 50 and 80 ml/cmH&sub2;O depending upon the size of the patient. Dynamic compliance will fall with both bronchoconstriction or lung restriction as in adult respiratory distress syndrome. Static compliance is relatively unaffected by bronchoconstriction. In severe adult respiratory distress syndrome, static compliance may fall to as low as 10 to 20 ml/cm H&sub2;O from the normal range (60 to 90 ml/cmH&sub2;O).

Patients with restrictive lung diseases such as sarcoidosis or fibrosing alveolitis should be ventilated with small tidal volumes of between 5 and 8 ml/kg at rates of 15 to 20 breaths per minute. Oxygen need not be restricted in the manner recommended for patients with chronic obstructive pulmonary disease.

Although most patients with acute on chronic respiratory failure can be managed successfully without mechanical ventilation, a small proportion fail to respond to conservative measures and require ventilatory assistance. In many cases, the need for mechanical ventilation is the direct result of injudicious oxygen therapy. Low rate synchronized intermittent mandatory ventilation (6–8 breaths per minute) or low pressure levels of pressure support are ideal for such patients with acute on chronic respiratory failure. The Paco&sub2; should be reduced very slowly towards but not to normal levels. The Fio&sub2; rarely needs to be higher than 0.35. High ventilator rates (>14 per minute) are associated with high values of V&subD;/ V&subT; (>0.5). Paradoxically, as the ventilator rates are increased in an attempt to increase minute ventilation, the Paco&sub2; may rise. To avoid intrinsic or autopositive end-expiratory pressure the I:E ratio should be maintained at 1:2 or higher. Weaning can begin as soon as the precipitating cause of respiratory failure has been corrected. Weaning will be unsuccessful if there is any underlying metabolic alkalosis or if the patient receives sedative or analgesic agents. The Paco&sub2; can be allowed to rise slowly to above normal levels provided sufficient time is given for the blood pH to correct and the Fio&sub2; kept below 0.35. Carbon dioxide production can be minimized by providing balanced nutrition with calories being provided by both lipid and carbohydrate.

Probably less than 1 per cent of acute severe asthma attacks require mechanical ventilation. However, it is apparent that some patients suffer cardiac arrest and die each year because intubation and mechanical ventilation was not performed in time. Hypercarbia alone is generally insufficient indication for ventilation but a combination of a rising Paco&sub2;, fatigue, failure of conservative measures, or arrhythmias does call for elective intubation and mechanical ventilation.

Since asthma has a prevalence of between 4 and 8 per cent in the United Kingdom and the United States, it is not uncommon for a postoperative patient on a mechanical ventilator to develop bronchospasm. Asthmatic patients are difficult to ventilate and usually require high inflation pressures. Hypoxaemia may persist despite the addition of high concentrations of oxygen and is the result of mucus plugging of the airways. A philosophy of ‘permissive hypercapnia’ or ‘controlled hypoventilation’ should be adopted with the Paco&sub2; remaining at elevated levels (7–8 kPa, 50–70 mmHg). Lower tidal volumes and respiratory rates are therefore possible. Lower inspiratory flow rates result in lower peak pressures and reduced risk of barotrauma. Deaths in ventilated asthmatic patients are usually the result of barotrauma, hypotension in volume depleted patients, arrhythmias, or lung infection.

Maximal bronchodilator therapy, including corticosteroids, is continued throughout the period of mechanical ventilation supplemented if necessary with inhalational anaesthetics such as isoflurane or the intravenous anaesthetic ketamine. Both of these agents are potent bronchodilators. Rehydration and adequate humidification of inspired gases will usually mobilize secretions and mucous plugs; if not, bronchoalveolar lavage may be indicated.

The use of extracorporeal membrane oxygenation and CO&sub2; removal has been reported in acute asthma. These must be considered exceptional cases and such techniques cannot be generally recommended.

Although bronchopleural and bronchocutaneous fistulae can occur after trauma or lung infection, many arise during the postoperative period following lobectomy or pneumonectomy. It is generally appreciated that early weaning and extubation is preferred in these patients. Occasionally, postoperative complications necessitate a longer period of ventilation when there is significant risk of dehiscence of the bronchial stump. To reduce the risk of this, low tidal volume, high respiratory rate ventilation should be adopted to minimize inflation pressures. High frequency ventilation would appear to be ideally suited to the prevention of bronchopleural fistula, although evidence to prove superiority over conventional ventilation is lacking.

The development of bronchopleural fistula is heralded by clinical deterioration, reduced chest wall movement on the affected side, tracheal deviation, and a sudden increase in inflation pressures. Emergency tube thoracostomy must be performed converting the bronchopleural fistula into a bronchocutaneous fistula. Compensation for the loss of tidal volume through the fistula is easily made by adjusting the ventilator but if the leak is large, endobronchial intubation may be necessary. Bronchopleural and bronchocutaneous fistulae are unlikely to close until the patient is weaned from the ventilator.

More than 80 per cent of patients who are ventilated postoperatively can be weaned simply by clinically evaluating their spontaneous ventilation on a ‘T-piece’ or similar circuit. The remainder require a progressive reduction in ventilatory support until measurement of ventilation parameters can be made. These parameters include the negative inspiratory force and vital capacity. A negative inspiratory force greater than −25 cmH&sub2;O or a vital capacity greater than 10 ml/kg usually indicates sufficient ventilatory reserve for spontaneous ventilation. These parameters cannot be applied to patients with severe chronic obstructive pulmonary disease, in whom blood gases have to be followed with each reduction in ventilation support. Failure to wean a patient successfully should prompt the questions addressed in Table 4 65.

Intermittent mandating ventilation or pressure support modes are very suitable for a gradual weaning process. Assist-control can also be used to wean patients by progressively reducing the tidal volume.

Several complications of mechanical ventilation can be attributed to the local effects of the endotracheal tube upon the airway. These include airway obstruction due to endotracheal tube displacement and pressure necrosis leading to vocal cord injury and subglottic stenosis. The risk of nosocomial pneumonia is increased in the intubated patient.

Other complications are the direct consequence of positive pressure ventilation. Haemodynamic effects such as reduced cardiac output, reduced renal perfusion, and salt and water retention are primarily the result of mechanical, neuroreflex, and humoral factors.

Interstitial lung damage may occur due to positive pressure ventilation and this has prompted renewed interest in extracorporeal systems for the management of patients with acute respiratory distress syndrome. However, the greatest concern relates to the risk of pneumothorax, pneumomediastinum, pneumopericarium, or subcutaneous emphysema. Pneumothorax is the most feared complication because it is associated with rapid deterioration unless dealt with quickly. Tube thoracostomy is mandatory since progression to a tension pneumothorax is very likely. Prophylactic thoracostomy tubes are not recommended, even in the presence of pneumomediastinum. Sudden clinical deterioration associated with a rise in inflation pressures and absence of breath sounds should raise the question of pneumothorax. Emergency decompression with a 14-gauge cannula may produce temporary relief and may have a diagnostic role, but tube thoracostomy should be performed without delay and without radiographic confirmation if necessary. Blunt dissection through the parietal pleura with forceps and digital exploration of the pleural space prior to insertion of the thoracostomy tube is essential if lung damage is to be avoided. Thoracostomy tubes with rigid metal stylets must not be used under any circumstances.

Ashbaugh DG, Bigelow DB, Petty TL, Levine BE. Acute respiratory distress syndrome in adults. Lancet, 1967; ii: 319–23.

Barnes PK. Principles of lung ventilators and humidification. In: Scurr, C. Feldman S, eds. Scientific Foundations of Anaesthesia. London: William Heinemann, 1982:533–43.

Bernard GR, et al. High dose corticosteroids in patients with the adult respiratory distress syndrome. N Engl J Med, 1987; 317: 1545–70.

Cameron PD, Oh TE. Newer modes of mechanical ventilatory support. Anaesthesia Intensive Care, 1986; 14: 258–66.

Danek SJ, Lynch JP, Weg JG, Dantzker DR. The dependence of oxygen uptake on oxygen delivery in the adult respiratory distress syndrome. Am Rev Resp Dis, 1980; 122: 387–95.

Downs JB, Block AJ, Vennum KB. Intermittent mandatory ventilation in the treatment of patients with chronic obstructive pulmonary disease. Anesth Analg, 1974; 53: 437–43.

Downs JB, Kelin EF, Desautels D, Modell JH, Kirby RR. IMV: A new approach to weaning patients from mechanical ventilators. Chest, 1973; 64: 331–5.

Downs JB, Olsen GN. Pulmonary function following adult respiratory distress syndrome. Chest, 1974; 65: 92–3.

Duchateau J, et al. Complement activation in patients risk of developing the adult respiratory distress syndrome. Am Rev Resp Dis, 1984; 130: 1058–64.

Elliott CG, Morris AH, Cengiz M. Pulmonary function and exercise gas exchange in survivors of adult respiratory distress syndrome. Am Rev Resp Dis, 1981; 123: 455–92.

Fairley HB. Critique of intermittent mandatory ventilation. In: Kirby R, Graybar GB, eds, Intermittent mandatory ventilation. Boston: Little, Brown and Co., 1980: 79–90.

Fowler, AA, et al. Adult respiratory distress syndrome: risk with common predispositions. Ann Intern Med, 1983; 98: 593–7.

Gattinoni L, et al. Treatment of acute respiratory failure with low frequency positive-pressure ventilation and extracorporeal removal of CO&sub2;. Lancet 1980; ii: 292–4.

Gotloib L, Barzilay E. The impact of using the artificial kidney as an artificial endocrine lung upon severe septic ARDS. Intensive Crit Care Digest, 1986; 5: 3–5.

Hammerschmidt DE, Weaver LJ, Hudson LD, Craddock PR, Jacob HS. Association of complement activation and elevant plasma-C5a with adult respiratory distress syndrome: pathophysiological relevance and possible prognostic value. Lancet, 1980 i: 947–9.

Haynes JB, Hyers TM, Giclas PC, Franks JJ, Petty TL. Elevated fibrinogen degradation products in adult respiratory distress syndrome. Am Rev Resp Dis, 1980; 122: 841–7.

Heenan TJ, Downs JB, Douglas ME, Ruiz BC, Jumper L. Intermittent mandatory ventilation. Is synchronization important? Chest, 1980; 77: 598–602.

Hickling KG. Extracorporeal CO&sub2; removal in severe adult respiratory distress syndrome. Anaesthesia Intensive Care, 1986; 14: 45–53.

Hickling KG, Henderson, SJ, Jackson R. Low mortality associated with low volume pressure limited ventilation with permissive hypercapnia in severe adult respiratory distress syndrome. Intensive Care Med, 1990; 16: 372–7.

Hurst JM, Branson RD, Davis KJ, Barrette RR, Adams, KS. Comparison of conventional mechanical ventilation and high-frequency ventilation. A prospective, randomized trial in patients with respiratory failure. Ann Surg, 1990; 211: 486–91.

Kariman K, Burns SR. Regulation of tissue oxygen extraction is disturbed in adult respiratory distress syndrome. Am Rev Resp Dis, 1985; 132: 109–114.

Kirby RR, et al. High level PEEP in acute respiratory insufficiency. Chest, 1975; 67: 156–63.

Kumar A, et al. Continuous positive-pressure ventilation in acute respiratory failure. N Engl J Med, 1970; 283: 1430–6.

Mohsenifar Z, Tashkin DP, Goldbach P, Campisi DJ. Relationship between O&sub2; delivery and O&sub2; consumption in the adult respiratory distress syndrome. Chest, 1983; 84: 267–71.

Montgomery AB, Stager MA, Carrico J, Hudson LD. Causes of mortality in patients with the adult respiratory distress syndrome. Am Rev Resp Dis, 1985; 132: 485–9.

Murray JF, Matthay MA, Luce JM, Flick MR. An expanded definition of the adult respiratory distress syndrome. Am Rev Resp Dis, 1988; 138: 720–3.

National Heart Lung and Blood Institute, Division of Lung Diseases. Extracorporeal support for respiratory insufficiency. Bethesda, MD:National Institutes for Health, 1979; 243–5.

Ognibene FP, et al. Adult respiratory distress syndrome in patients with severe neutropenia. N Engl J Med 1986; 315: 547–51.

Pepe PE, Hudson LD, Carrico CJ. Early application of positive end expiratory pressure in patients at risk for the adult respiratory distress syndrome. N Engl J Med, 1984; 311: 281–6.

Pepe PE, et al. Clinical predictors of adult respiratory distress syndrome. Am J Surg, 1982: 144: 124–30.

Petty TL, Ashbaugh DG. The adult respiratory distress syndrome: clinical features, factors influencing prognosis and principles of management. Chest, 1971; 130: 66–71.

Petty TL, Silvers GW, Paul GW, Stanford RE. Abnormalities in lung elastic properties and surfactant function in adult respiratory distress syndrome. Chest 1979; 75: 571–4.

Pontoppidan H, Geffin B, Lowenstein E. Acute respiratory failure in the adult. N Engl J Med, 1972; 287: 690–9.

Qvist J, Pontoppidan H, Wilson RS, Lowenstein E, Laver MB. Haemodynamic responses to mechanical ventilation with PEEP: the effect of hypervolaemia. Anesthesiology, 1975; 42: 45–55.

Rinaldo JE. Mediation of ARDS by leukocytes. Clinical evidence and implications for therapy. Chest, 1986; 89: 590–3.

Rinaldo JE, Petty TL. Indicators of risk, course, and prognosis in adult respiratory distress syndrome (ARDS). Am Rev Resp Dis, 1986; 133: 343–4.

Rinaldo JE, Rogers RM. Adult respiratory distress syndrome. N Engl J Med, 1986: 315: 578–80.

Rocker GM, Wiseman MS, Pearson D, Shale DJ. Diagnostic criteria for adult respiratory distress syndrome: time for reappraisal. Lancet 1989; i: 120–3.

Rotman HH, Lavelle TF, Dimcheff DG. Vendenbelt RJ, Weg JG. Long term physiological consequences of the adult respiratory distress syndrome. Chest, 1977; 72: 190–2.

Saldeen T. The microembolism syndrome. Microvasc Res, 1976; 11: 227–59.

Simpson DL, Goodman M, Spector SL, Petty TL. Long term follow-up and bronchial reactivity resting in survivors of the adult respiratory distress syndrome. Am Rev Resp Dis, 1978; 117: 449–54.

Smith RA, Desautels DA, Kirby RR. Mechanical ventilators. In: Kirby RR, Smith RA, Desautels DA, eds. Mechanical Ventilation. New York: Churchill Livingstone, 1985; 327–474.

Tate RM, Repine JE. Neutrophils and the adult respiratory distress syndrome. Am Rev Resp Dis, 1983; 128: 522–9.

Weigel JA, Norcross JF, Borman KR, Sayder WH. Early steroid therapy for respiratory failure. Arch Surg, 1985; 120: 536–40.

Weiland JE, et al. Lung neutrophils in the adult respiratory distress syndrome. Clinical and pathophysiological significance. Am Rev Resp Dis, 1986; 133: 218–25.

Weinberg P, et al. Biologically active products of complement and acute lung injury in patients with the sepsis syndrome. Am Rev Resp Dis 1984; 130: 791–6.

Wood LH, Prewitt RM. Cardiovascular management in acute hypoxemic respiratory failure. Am J Cardiol 1981; 47: 963–72.