Pattern of and time course of disease

Table 9.1 summarizes a retrospective analysis of 1000 infants and children (aged older than 28 days and younger than 17 years) who required endotracheal intubation for acute respiratory failure complicating acutely acquired medical, rather than surgical, disease (Tasker 2000). The three major categories relate to the system or problem underlying respiratory failure (respiratory tract disorder, central nervous system disorder or systemic disorder) and the subcategories relate to the clinical diagnostic entities commonly encountered in intensive care. Respiratory tract problems due to infection are, not surprisingly, the most common problems seen. The time course of recovery in survivors is influenced by the site within the airways that infection has reached. This is reflected by an increase in the length of stay in the intensive care unit (ICU) with more distally affected tissues (i.e upper airway compared with lower airway). In relating such information to clinical practice one can use the expected time course to decide on an agenda for treatment or ‘care pathway’. For example, given that the expected time course for intensive care recovery in pneumonia necessitating intubation is around 8 days (interquartile range 5–12 days) one can then predict when certain targets should be met. The same applies to the other 11 distinct diagnosis-related entities. This idea will be revisited later in this section where three clinical examples are discussed.

Table 9.1 Diagnostic distribution of 1000 children requiring endotracheal tube intubation during acute medical illness ordered by number (n), age and length of stay on the intensive care unit in survivors

Acute hypoxaemic respiratory failure

Acute hypoxaemic respiratory failure (AHRF) signifies respiratory failure at the more severe end of the pathophysiological spectrum, irrespective of underlying aetiology. For paediatric practice we identify this state by using diagnostic criteria that have been modified from the American-European Consensus Conference diagnostic criteria for acute respiratory distress syndrome (ARDS) (Bernard et al 1994). These criteria include:

acute onset of respiratory failure over less than 48 hours

evidence of a severe defect in oxygenation (PaO₂/F_iO₂ of less than 26.7 kPa, 200 mmHg) for at least 6 consecutive hours on the day of admission

no evidence of left atrial hypertension

four-quadrant interstitial shadowing on chest radiograph.

Children meeting all the above criteria except the characteristic chest radiographic appearances of ARDS (last criterion) are described as cases of AHRF. The significance of AHRF is that it implies a certain severity of illness and risk of mortality, factors which are important when it comes to deciding which ventilatory strategy should be adopted and the use of adjunctive therapies (see section on ‘ventilation strategies for specific disease’). For example, in a prospective, epidemiological study, Peters and colleagues (1998) found that out of 850 mechanically ventilated infants and children, AHRF occurred in 118 patients (14%, 95% confidence interval (CI) 12–16%). Of these 118 patients, 52 met the criteria for ARDS (44%, 35–53%). In all 850 patients, mortality was four times higher than the mortality seen in those patients without AHRF. In the AHRF patients, mortality was three times higher for those with ARDS (Peters et al 1998). In a study from North America, the PALISI network (Pediatric Acute Lung Injury and Sepsis Investigators) reported a mortality from ARDS of 4.3% (Randolph et al 2003). This coincides with the adoption of protective modes of ventilation for this condition. Therefore, identifying these entities (i.e. AHRF and ARDS) at an early stage is important so as to institute the most appropriate method or mode of ventilation.

Hypoxia

Hypoxia must be treated first; give oxygen. At the same time attempts should be made to correct the underlying problem. Local processes in the lung, such as atelectasis and bronchopneumonia, can result in a portion of the pulmonary blood flow perfusing unventilated alveoli (i.e. intrapulmonary shunt), which in some cases may be effectively treated by pulmonary toilet and postural change. With a large shunt fraction – greater than 25% of pulmonary blood flow – PaO₂ is not significantly improved by solely increasing the FiO₂. In these cases a diffuse pulmonary process is usually present and a form of assisted positive airway pressure is required. Such assistance may also be required for severe impairment of chest wall mechanics (e.g. rib fractures, pain, weakness, etc.) even in the absence of pulmonary parenchymal disease.

In infants and children, there are several methods of administering oxygen (Table 9.2). Young patients do not usually tolerate nasal catheters and cannulae. Facemasks with a reservoir and a non-rebreathing valve can be used to increase the FiO₂. Alternatively, high-flow oxygen via the appropriate Venturi-valve mask can be used. Oxygen delivered via the oxygen inlet of an incubator rarely exceeds an FiO₂ of 0.4. When supplemental oxygen is delivered into an tent the concentration varies depending on any leaks in the system. Regardless of the technique, it is essential that the administered oxygen is warmed and humidified. To avoid damage to the lungs, oxygen administration should be discontinued as soon as possible (as indicated by blood gas measurements). An FiO₂ below 0.6 is preferred, to minimize the risk of oxygen toxicity. Reduction in the FiO₂ should be carried out cautiously in a stepwise manner. To facilitate this process both the concentration and duration of oxygen therapy must be recorded accurately. A well-calibrated oxygen analyser is used to check the inspired concentration every 2 hours when using a tent, head box or when adding oxygen into an incubator. The need for monitoring PaO₂ in preterm newborn infants is related to the potential for pulmonary oxygen toxicity and the danger of retrolental fibroplasia. So, in any patient, oxygen should be administered at the lowest concentration sufficient to maintain the PaO₂ between 6.7 and 13.3 kPa (50–100 mmHg). Continuous measurement or monitoring of transcutaneous oxygen or pulse oximetry arterial oxygen saturation (SpO₂) are essential additions to the direct, and intermittent, measurement of arterial blood gases. In some instances, supplemental oxygen may cause respiratory depression if there has been chronic respiratory failure and the patient has a hypoxic-drive to ventilation (as opposed to the normal situation where the PaCO₂ is the most important factor in the control of ventilation). This phenomenon is generally uncommon in paediatric practice, but has been encountered in children with cystic fibrosis, cerebral palsy and bronchopulmonary dysplasia.

Table 9.2 Methods of oxygen administration

FiO₂, fractional inspired oxygen concentration

In addition to oxygen, some type of positive pressure may be useful in the management of hypoxia. Mask and nasal continuous positive airway pressure (CPAP) or bilevel positive airway pressure (BiPAP) increase lung compliance by recruiting additional areas of the lung for ventilation. Also, lung recruitment improves oxygenation by decreasing intrapulmonary shunt. These modes of non-invasive ventilation (NIV) are being used more frequently as the step before invasive mechanical ventilation. When invasive mechanical ventilation is used, the addition of some positive end-expiratory pressure (PEEP) is a common practice in maintaining adequate functional residual capacity. However, PEEP may adversely affect the patient lung mechanics if hyper-inflation occurs. This problem results in impaired pulmonary perfusion and further accentuates any ventilation–perfusion (/) mismatch. Therefore PEEP above 4 cmH₂O should be used judiciously if there is already regional hyperinflation, such as occurs in bronchopulmonary dysplasia (Box 9.1). In this context a strategy for treating hypoxia is outlined in Box 9.2.

Hypercarbia

When shallow (or ineffectual) breathing is present, the dead space (i.e. ventilated but non-perfused regions) becomes a larger fraction of each breath. This change results in a decrease in alveolar ventilation, even if the lung parenchyma is normal. When hypercarbia has been found and its cause considered, NIV can be tried a step before invasive mechanical ventilation. This may be useful when hypercarbia is secondary to neuromuscular weakness. If the patient needs invasive mechanical ventilation, then increasing tidal volume or respiratory rate can bring about increased alveolar ventilation. However, these changes may increase mean airway pressure that may generate detrimental effects on pulmonary vascular resistance and / matching.

If the patient is treated with full mechanical ventilation, the first step is to make sure that the patient is receiving an appropriate tidal volume and minute ventilation. Ventilatory system leaks and loss of a portion of the tidal volume through compressive loss in the tubing, as well as abnormalities in endotracheal tube function, are common problems that need rectifying. Having excluded mechanical factors, the other causes of hypercarbia may be related to an increase in CO₂ production or an increase in dead-space ventilation. An increase in dead space may be due to excessive PEEP (particularly when there is already hyperinflation or hypovolaemia) and it may be corrected by intravenous volume to increase preload to the heart.

Endotracheal intubation

There are four absolute indications for controlling the airway by endotracheal intubation:

maintaining the patency of the airway where problems are present or anticipated (e.g. direct airway trauma, oedema or infection)

to protect the airway from aspiration in states of altered consciousness, where airway-protective mechanisms may be lost or impaired

to facilitate pulmonary toilet and avoid airway obstruction when there is marked atelectasis and pulmonary infection – an inadequate cough might necessitate more direct access to the airways for suctioning

when positive pressure breathing is indicated because of inadequate spontaneous ventilation.

In practice, experienced staff should carry out establishing airway and respiratory support for the acutely ill child, because such patients can deteriorate rapidly, particularly at the time of inducing anaesthesia. Following preoxygenation with 100% inspired oxygen, a variety of agents are used to facilitate endotracheal intubation, including intravenous induction with drugs such as fentanyl, midazolam and suxamethonium, or inhalational induction with gases such as halothane or isoflurane. Table 9.3 provides a guide to the appropriate endotracheal tube size, length and suction catheter used in the paediatric age range and Figure 9.1 illustrates a commonly used method of endotracheal tube fixation.

Table 9.3 Endotracheal tube size and suction catheters

Figure 9.1 Endotracheal tube fixation for (A) nasal and (B) oral tubes.

General ventilatory care

A variety of ventilators can be used in paediatric mechanical ventilation (Fig. 9.2). There are specific ventilators designed for neonates and small infants that are used mainly in neonatal units to ventilate premature babies. Fortunately most modern ventilators can be used across the whole age spectrum and they can deliver different modes of pressure control ventilation (PCV) and volume control ventilation (VCV).

Figure 9.2 (A) The SLE 2000 infant ventilator and (B)the Servo 300 ventilator.

One of the main goals during mechanical ventilation is to minimize and optimize patient–ventilator interaction. Adequate patient comfort will potentially decrease the need for pharmacological sedation and may help to minimize the duration of mechanical ventilation. The addition of PEEP at 3–5 cmH₂O above atmospheric pressure is routinely used unless there is a contraindication. This helps to avoid alveolar collapse during mechanical ventilation. The amount of PEEP may need to be increased in conditions associated with low lung volume, such as hyaline membrane disease, atelectasis, severe pneumonia (e.g. viral, Pneumocystis carinii) or ARDS.

Complications of ventilator therapy occur frequently and all intensive care staff should be continually aware of the potential hazards (Box 9.3). Aseptic technique is important for tracheal airway care because nosocomial infection constitutes a large and preventable problem. The application of PEEP, increased tidal volumes and increased airway pressure can also produce complications. Potential disruption of normal / matching seen with spontaneous breathing can occur with lung over-expansion and leads to regional hypoperfusion. A decrease in venous return, an increase in pulmonary vascular resistance and a decrease in left ventricular output can impair cardiac output and oxygen delivery. The more compliant the lung or the less compliant the chest wall, the greater the transmission of positive airway pressure to the mediastinum and the greater the negative effect on cardiac function. Volume loading or inotropic support can overcome the concomitant decrease in cardiac output, in large part.

The significance of ventilator-induced lung injury (VILI) has been appreciated more during the last few years. It now seems clear that the pathogenesis of respiratory failure is greatly influenced by the way the lungs are ventilated. Barotrauma and volutrauma have been used to describe VILI. However, there is some controversy as to which of these is the more damaging. Both can produce overdistension and it is this problem that appears to cause the injury. To complicate matters, another factor appears to be ‘atelectrauma,’ i.e. where the repeated collapse and re-expansion of areas of the lung produce shearing injury that contributes to inflammation and lung damage. The best ventilator strategy should therefore aim to keep tidal volume to a minimum with optimum PEEP and, sometimes, neuromuscular blockade. Theoretically, these manoeuvres should reduce atelectasis and patient–ventilator interactions.

In regard to other problems, pulmonary interstitial emphysema, pneumomediastinum, pneumoperitoneum and subcutaneous emphysema do not require specific treatment unless there is significant haemodynamic impairment. Poor renal function, as exhibited by decreased glomerular filtration rate, urine production and sodium excretion, can be a consequence of hypoxia and hypercarbia. This may be further compounded by the effects of mechanical ventilation with PEEP on producing an antidiuretic hormone-mediated salt and water-retaining effect (probably secondary to decreased cardiac output), an increased renal vein pressure and a neural reflex from the pressure-distorted atrial wall.

Acute deterioration and ‘troubleshooting’

In mechanically ventilated patients, the adequacy of gas exchange and ventilation should be assessed frequently. Changes in therapy should then be titrated against expected parameters or targets (Table 9.5). When there is an acute deterioration during mechanical ventilation, problem solving or ‘troubleshooting’ should begin with making the patient safe. In the first instance this means disconnecting the patient from the ventilator and support breathing with bag-ventilation using FiO₂ of 1.0. Easy ventilation and patient stabilization with the bag suggests a ventilator problem that should be systematically addressed (e.g. check the circuit for leaks, check ventilator function, check gas flow). However, it should be remembered that ‘hand-bagging’ might result in increased tidal volume, which can also be responsible for the patient’s improvement. Patients with stiff lungs are frequently dyspnoeic, despite adequate gas exchange. Increasing the tidal volume will correct this subjective feeling and may also account for patient improvement. Difficult bagging at the time of disconnecting the ventilator strongly suggests a problem with the endotracheal tube or the lung to chest-wall complex. A suction catheter (Table 9.3) should be passed down the endotracheal tube to check for narrowing or blockage. Chest examination, blood gases and chest radiography should be ordered. A blocked endotracheal tube should be replaced. A pneumothorax requires chest tube placement. If neither of these is the cause for deterioration, then the possibilities may include new problems such as an increased oxygen demand due to sepsis, impaired oxygen delivery due to heart failure or acute pulmonary injury due to gastric aspiration. These and other causes need to be sought and treated appropriately.

Table 9.5 Normal values

Commonly used modes of mechanical ventilation in children

High-frequency oscillatory ventilation

High-frequency ventilation techniques, including high-frequency positive pressure ventilation, high-frequency jet ventilation and high-frequency oscillatory ventilation (HFOV), achieve adequate ventilation by employing tidal volumes that are often less than actual dead space and respiratory rates of 60–3000 cycles/minute (Fig. 9.3). The most widely used of these modes in children is the HFOV. The theoretical advantage of HFOV is that it keeps the lungs open by using a relatively high mean airway pressure and it delivers very low tidal volumes, thereby minimizing lung damage caused by high volumes and high inspiratory pressures. The high mean airway pressure allows lung-volume recruitment and improves oxygenation.

Figure 9.3 Endotracheal tube fixation for (A) nasal and (B) oral tubes.

The most common indication for this form of ventilation is refractory hypoxaemia. However, HFOV has been used in bronchopleural fistula and some types of obstructive disease and bronchiolitis (Kneyber et al 2005, Slee-Wijffels et al 2005). Unfortunately there is a paucity of published clinical trials on the use of HFOV in the paediatric population and its benefits have not been clearly established as compared with conventional mechanical ventilation. Despite this lack of evidence, HFOV is frequently used in the treatment of hypoxemic respiratory disease. When used, early institution of this therapy seems to be more beneficial in this group of patients (Arnold et al 1994). The experience reported by Watkins and colleagues (2000) in 100 courses of such ventilation would suggest that, in the presence of AHRF or ARDS, a threshold mean airway pressure of 16 cmH₂O is an appropriate indication.

Extracorporeal membrane oxygenation

Extracorporeal membrane oxygenation (ECMO) is designed to provide a variable degree of cardiopulmonary support for a predetermined period of time over which the underlying pulmonary disorder is expected to recover. Potentially, ECMO allows recovery without subjecting the lungs to the risks of VILI or oxygen toxicity. Venoarterial systems may be used to completely take over the child’s own heart and lung function (Fig. 9.4), although in practice extracorporeal flows may be limited by venous drainage (usually from the right internal jugular vein). Venovenous systems have been used for CO₂ removal, although a retrospective study concluded that venovenous ECMO could effectively provide adequate oxygenation for children with severe acute respiratory failure (Pettignano et al 2003). In the complete absence of pulmonary function, venovenous ECMO will provide SpO₂ of 80%. Using extracorporeal support, success has been achieved in some children with acute lung injury (Pearson et al 1993). However, appropriate patient selection is a critical and contentious issue. One method, proposed by Bartlett (1982, 1990), is to identify neonates at high risk for failing to respond to conventional therapy by applying an index of oxygenation, which is related to the mean airway pressure and the FiO₂ during mechanical ventilation, and the achieved PaO₂:

Figure 9.4 Extracorporeal membrane oxygenation.

Oxygenation index (OI) = (mean airway pressure × FiO₂ × 100)/PaO₂ (mmHg)

In Bartlett’s proposal (1990), OI greater than 25 predicted 50% mortality and OI greater than 40 predicted 80% mortality. Anecdotally, in children, Goldman and colleagues (1997) have found in meningococcal sepsis that ARDS that fails to respond to HFOV may be reasonably treated with ECMO.

Acute bronchiolitis

The typical features of bronchiolitis are:

acute generalized peripheral airway obstruction (‘air trapping’) with tachypnoea, decreased breath sounds and low hemidiaphragms on chest radiography

infant less than 2 years of age

little or no evidence of past similar episodes.

Respiratory syncytial virus (RSV) is the most frequent cause of bronchiolitis. Other viral causes include adenovirus, influenza and parainfluenza viruses and rhinovirus. Cytomegalovirus can produce a bronchio-litis or pneumonitis-like illness in immunocompromised children. Rare non-viral causes of the bronchiolitis syndrome include Mycoplasma pneumoniae and Bordetella pertussis infection.

Sixteen per cent of infants hospitalized for RSV have apnoea and its course is usually short-lived. Clinically these episodes are diaphragmatic or non-obstructive with complete absence of respiratory effort. In these cases, endotracheal intubation with minimal support is required until the problem resolves. In patients with worsening respiratory distress due to pulmonary parenchymal changes, mechanical support does not necessarily require endotracheal intubation. In some instances, nasopharyngeal prong CPAP, which maintains positive transpulmonary pressure during spontaneous breathing, can be used to avoid mechanical ventilation. When infants with RSV infection require mechanical ventilation, there are many similarities with mechanical ventilation of adults with status asthmaticus (Box 9.4). Clinical observation of inspiratory and expiratory chest excursion as well as regular auscultation is important if overventilation, with associated hyperinflation and barotrauma, is to be avoided. The aim is to maintain or achieve adequate arterial oxygenation and control of respiratory acidosis. This may even necessitate ventilating at slow rates with prolonged expiratory times to permit adequate CO₂ clearance. Low levels of PEEP are used to decrease airway resistance and improve gas exchange, although in studies of lung mechanics this has not been verified. The presence of inadvertent or auto-PEEP requires that extrinsic PEEP be applied to the same level in order to maintain expiratory flow.

In patients with hyperinflation the ventilatory strategy should aim to limit ventilator-associated dynamic hyperinflation and impaired minute ventilation. The ventilator should be set at a slow rate (10–15 breaths/minute) and with a prolonged expiratory time. Time-cycled, PCV is used in this instance while aiming for an arterial pH >7.25 and a SpO₂ 88–92%. When indicated, neuromuscular blockade and antibiotics will need to be prescribed. All patients should receive adequate analgesia and sedation during mechanical ventilation. Bronchodilators can be administered if patients demonstrate a therapeutic response to an initial trial dose. In the acute phase of illness, fluids, electrolytes and hydration must be closely monitored while generally restricting fluid to 67–75% of maintenance requirements. In the weaning phase, patients can be removed from ventilatory support when safe to do so. In regard to blood gas parameters this means, in general, adequate oxygenation in an FiO₂ <0.4 and normal pH, with good respiratory drive, in the absence of hypercarbia. Discharge from the intensive care unit can then be considered once the patient has managed at least 12–24 hours without any respiratory assistance. Overall these patients will, on average, spend about 7 days (interquartile range 4–8 days) on the intensive care unit (Tasker et al 2000).

In about one-fifth of mechanically ventilated patients with RSV more severe disease is seen (Tasker et al 2000). In this instance, more extensive pulmonary pathology results in a picture of pneumonitis with diffuse alveolar consolidation rather than bronchiolitis with lung hyperinflation. These infants have the clinical features of ARDS as they exhibit four-quadrant consolidation on chest radiograph. The aim of mechanical support should be to recruit lung volume with the addition of PEEP. Sometimes HFOV or ECMO is required if lung injury becomes more extensive with likely development of interstitial emphysema and pneumothorax. The time course of this problem is very different to the usual course of bronchiolitis and, on average, patients spend at least 2 weeks on the intensive care unit (Tasker et al 2000).

Acute hypoxaemic respiratory failure

In patients with AHRF or ARDS the low tidal volume strategy in addition to PEEP helps to minimize VILI. The level of PEEP should be adjusted so as ensure maximum lung volume recruitment. Adequate recruitment improves oxygenation and avoids repeated alveolar collapse and reopening that lead to shear injuries. If oxygenation continues being inadequate with a mean airway pressure of 16 cmH₂O or greater, then HFOV should be considered, particularly if the problem is one of diffuse parenchymal changes or consolidation. A series of adjunctive therapies have been proposed for this condition.

Recruitment manoeuvres

Recruitment manoeuvres are aimed at recovering collapsed areas of the lung during ALI, especially after specific manipulations such a suctioning and disconnection of the ventilator. A variety of manoeuvres have been described including increases in the level of PEEP while maintaining constant inspiratory pressures, prolonged increases in pressure (i.e. 40 mmHg for 40 seconds) as a single manoeuvre or using stepwise increases, and the use of high tidal volumes. There are no controlled studies and there is no consensus regarding these manoeuvres and their safety in children.

Other considerations during mechanical ventilation

Monitoring during mechanical ventilation

Adequate monitoring of the parameters affected by mechanical ventilation is essential. The information helps to characterize the pathophysiology of the underlying condition, to improve ventilation, to enhance patient comfort, and to guide weaning. Blood gas measurement is the gold standard for assessing pulmonary gas exchange, but this technique is limited by the fact that measurements are not continuous, because they require invasive techniques. The use of pulse oximeters and end-tidal carbon dioxide concentration helps to minimize the number of blood gas samples. Transcutaneous carbon dioxide measurement is another way of continuously monitoring the CO₂ status.

Monitoring of the respiratory mechanics and analysis of the patient–ventilator interaction should be part of the routine care during mechanical ventilation. Most modern mechanical ventilators provide a graphical display of gas flow, airway pressures and tidal and minute volumes. In small infants and neonates, the information obtained with a pneumotachometer positioned at the end of the endotracheal tube instead of inside the mechanical ventilator can be more reliable. These data, together with measurements of pulmonary gas exchange, help to assess the adequacy of ventilation.

A new technique, electroimpedance tomography, analyses the distribution of ventilation and can be used at the bedside. This technique used 16 electrodes placed around the chest. Injection of an alternating electrical current between sequential pairs of adjacent electrodes and the repeated measurement of the differences in voltage through the array of electrodes permits the detection of changes in impedance. Since air is a poor conductor of electricity compared with other tissues, computer-assisted use of a mathematical algorithm then allows the reconstruction of a spatial image of a section of the chest. This method has the advantage of being non-invasive, radiation free, portable and dynamic. Different studies have shown its usefulness to assess regional lung ventilation (Victorino et al 2004).

Weaning from mechanical ventilation

Stopping mechanical ventilation as soon as the child is ready to be weaned from support is essential to avoid complications. Prolonged mechanical ventilation has been associated with higher risk of nosocomial pneumonia, progressive ventilator-associated lung injury, airway injury, physiological dependence on sedative and narcotic drugs, and even higher mortality. In children who need mechanical ventilation for respiratory failure, the usual approach is to wean the ventilator settings gradually, until the patient is considered to be ready for extubation. However, this gradual weaning process is now being questioned.

A randomized controlled trial comparing weaning protocols versus standard physician-guided weaning in children did not show any impact on the duration of mechanical ventilation. This study also showed that gradual weaning may not be indicated for the majority of children receiving mechanical ventilation due to respiratory failure, and that a large group of children were ready to be extubated when the physicians determined that they were ready to begin a weaning process (Randolph et al 2002). Readiness for extubation can be evaluated using a daily test as soon as the patient meets the criteria for testing (Box 9.5). Different tests for extubation readiness have been proposed. In one paediatric study, patients were extubated if they passed a trial of spontaneous breathing with either pressure support of 10 cmH₂O, or T-piece for up to 2 hours. Mechanical ventilation was reinstituted if any of the following signs appeared:

Use of sedation has an impact on the duration of mechanical ventilation. Improved management of sedatives, including the use of objective sedation scores, should be part of the standard care for children receiving mechanical ventilation.

Non-invasive support

Non-invasive ventilation is the delivery of supportive ventilation without the use of an endotracheal airway. NIV is usually delivered as positive pressure ventilation using CPAP or BiPAP; however, negative pressure ventilation by mean of a cuirass is still useful for some conditions. Initially the use of NIV was mostly used in children with neuromuscular disease. Now, it is used in most types of respiratory failure.

In neonates and small infants the most frequent way of delivering non-invasive positive pressure ventilation (NPPV) is by means of nasal CPAP, although modern technology also incorporates the possibility of delivering BiPAP. The most frequent indication is bronchiolitis and other infections with apnoea. In bigger children NPPV is frequently delivered as BiPAP (Fig 9.5), although CPAP can also be used. In this age it is used for different forms of acute hypoxaemic or hypercarbic respiratory failure including asthma, neuromuscular disease, or after extubation in children that have received long periods of mechanical ventilation. The usual way of applying this form of support is via a nasal mask or a facemask. Nasal masks are better tolerated, but many patients with respiratory distress are mouth breathers and therefore a facemask might be necessary (Fig. 9.6). In order to tolerate NPPV some sedation is needed in the young infant. In older children, explanation, reassurance and familiarity with the therapy should suffice. To increase the tolerance, NPPV can be started at low-pressure settings that are then increased progressively until the target pressures are obtained. Nasogastric tubes can be used when the accumulation of air in the stomach is a concern, especially in children with facemasks. However, many clinicians prefer not to use nasogastric tubes as their placement with tapes can reduce the tightness of the seal between the mask and the face.

Figure 9.5 Respironics BiPAP® Synchrony®, portable.

Figure 9.6 Facemask for non-invasive ventilation.

A review of clinical studies using NPPV has shown that improvement usually occurs after the first 3 hours of therapy (Akingbola & Hopkins 2001). NPPV is not recommended in ARDS (Essouri et al 2006). The contraindications for NPPV include respiratory arrest, inability to use the mask because of trauma or surgery, difficult secretion management, haemodynamic instability, altered mental state, risk of aspiration, intolerance to the therapy and life-threatening refractory hypoxaemia or hypercarbia.

Finally, one technique that has historically been used effectively in children with neuromuscular disease since the poliomyelitis epidemics in the 1950s is extrathoracic negative pressure ventilation with a cuirass (Lassen 1953). In the past decade there has been renewed interest in this form of ventilation (Meessen et al 1994). There are many physiological reasons why negative pressure support should be beneficial, such as its ability to increase tonic activity in the diaphragm and intercostal muscles. In children with neuromuscular disease, who are on positive pressure mechanical support, it has been found that, when it comes to weaning, extubation and the introduction of negative pressure support means that analgesia and sedation can be discontinued quite quickly (Chisakuta & Tasker 1998). This approach should limit the unavoidable iatrogenic worsening of respiratory drive that results from the co-administration of analgesia and sedation (which is invariably necessary for children in order that they may tolerate the endotracheal tube). In myasthenic patients the time course of mechanical ventilatory support can be more than halved using this technique. Regarding bronchiolitis-related apnoeas, a retrospective analysis comparing two centres, one using negative pressure ventilation and the other using conventional therapies, concluded that the use of NPV was associated with a reduced rate of endotracheal intubation, and shorter paediatric intensive care unit (PICU) stay (Al-balkhi et al 2005). As more portable and easier-to-use ventilators become available, this form of respiratory support may be used more frequently.