Conventional Mechanical Ventilation

Abstract

Respiratory failure is one of the most common reasons for admission to the pediatric intensive care unit (PICU), and about half of the patients admitted to PICU need ventilator support. Development of mechanical ventilation has dramatically changed the outcome of pediatric respiratory failure from a commonly fatal condition to a one with very low mortality. Mechanical ventilation has evolved from negative pressure ventilation to modern positive ventilation delivered with multiple options of ventilation modes. This chapter reviews the basic principles of conventional mechanical ventilation and the pathophysiological impact of ventilation to provide readers with essential keys for the management of children with respiratory failure.

FormalPara Learning Objectives

• Describe the differences between negative and positive pressure ventilation.

• Describe the effects of positive pressure ventilation on the respiratory system.

• Describe the effects of positive pressure ventilation on preload and afterload.

• Describe how the main ventilation characteristics impact oxygenation.

• Describe how the main ventilation characteristics impact ventilation and CO₂ clearance.

• Describe the differences between pressure and volume modes of ventilation.

• Detail the mechanisms of ventilator triggering.

• Detail the different methods of ventilator cycling.

• Describe the methods of delivering assisted breaths.

• Understand how to determine the main ventilator settings.

• Describe the benefit and risk of the patient’s spontaneous contribution during mechanical ventilation.

• Describe ventilator strategies to minimize ventilator-induced lung injury and diaphragm dysfunction.

• Detail weaning strategies and extubation readiness assessment.

• Describe the monitoring of mechanical ventilation.

1 Introduction

Respiratory failure is one of the main reasons for admission to a pediatric intensive care unit (PICU). About half of the patients admitted to the pediatric intensive care unit (PICU) need a ventilator support. Before the 1930s, respiratory failure was uniformly fatal due to the lack of equipment and techniques for airway management and ventilator support. The use of mechanical support for respiratory failure began with negative pressure ventilation during the poliomyelitis epidemics. Technological progress over time has allowed the widespread use of positive pressure ventilation. This vital support technology has dramatically changed the outcome of pediatric respiratory failure, which now carries a low mortality rate. The performance of ventilators has improved, allowing a better control of delivered ventilation, increased safety, and the development of many ventilation modes. The complexity of the mechanical ventilation options has increased in parallel with the complexity of the PICU population, and the management of ventilator support can sometimes seem overwhelming. In that context, it is essential for the PICU clinicians to understand the functioning and the impact of conventional mechanical ventilation on the critically ill children.

This chapter reviews the main physiological features of respiratory failure and the basic principles of conventional mechanical ventilation to understand the pathophysiological impact of mechanical ventilation. Based on this background, we will describe a logical approach of the initial setup and adjustment of mechanical ventilation in children. The importance of monitoring for mechanically ventilated children will then be described.

2 Pulmonary Physiology and Conventional Mechanical Ventilation

2.1 Indications for Mechanical Ventilation and Mechanisms of Respiratory Failure

The primary goals of mechanical ventilation are the maintenance of adequate oxygenation and clearance of CO₂ from the body in the amount needed to maintain cellular homeostasis. A diverse group of disease processes, involving the respiratory, cardiovascular, and central nervous systems, may lead to respiratory failure. The primary indications for endotracheal intubation and mechanical ventilation in the PICU are described in ◘ Table 12.1. The objective of the institution of mechanical ventilation varies depending on the disease categories and the respective importance of hypoxemia, hypercarbia, hemodynamic, and neurological status.

Table 12.1 Main indications and objectives for airway control and mechanical ventilation

2.2 Pathophysiology of Hypoxemia and Application to Mechanical Ventilation

When confronted with the hypoxemic patient, understanding the etiology of hypoxemia is crucial to tailor the management of mechanical ventilation. The primary cause of hypoxemia in the PICU patient is ventilation-perfusion inequality related to acute lung injury and pulmonary parenchymal disease. Both ventilation and perfusion are not uniform in the normal lungs due to pleural pressure gradient and gravity. The ventilation-perfusion (V/Q) ratio is also not homogenous, higher in the apical and lower in the basal regions, leading to a normal global V/Q ratio of about 0.8. In acute pulmonary parenchymal disease, regions with low V/Q ratio generate hypoxemia, when the protective mechanism of hypoxic pulmonary vasoconstriction is overwhelmed.

Other mechanisms can be involved less frequently in hypoxemia: right-to-left shunt, impairment of diffusion capacity, and hypoventilation. Limitation of diffusion capacity is not frequent in acute disease but could occur in complex lung disease. Hypercapnia is usually uncommon in that context in that CO₂ diffuses much faster than oxygen and minute ventilation is normally increased by the hypoxemia-mediated ventilation response. Hypoventilation alone does not produce significant hypoxemia in a healthy lung. However, hypoxemia can occur in case of lung disease or in profound hypoventilation. Another issue that must be considered when caring for hypoxemic patients with ventilation-perfusion mismatch or true shunt is the impact of low mixed venous blood saturation. Improvement in mixed venous oxygen saturation may also improve arterial oxygen saturation in that context.

In the setting of hypoxemia, the severity of the lung disease may be estimated by the difference between the PaO₂ and the partial pressure of oxygen in the alveoli, the A-a or alveolar-arterial oxygen gradient. The A-a gradient is normally <10 mmHg, and it reflects the integrity of the alveolocapillary membrane and effectiveness of gas exchange. Widened gradient is observed in case of hypoxemia due to V/Q mismatch, diffusion limitation, and shunt, whereas it is normal in case of hypoventilation. Several indices are also used to diagnose and assess the severity of pediatric acute respiratory distress syndrome (PARDS). The oxygenation index (OI = mean airway pressure × FiO₂/PaO₂ × 100) and the ratio of the PaO₂ to FiO₂ (P:F ratio) are both associated with PARDS severity and mortality, and both require an arterial blood gas. In absence of arterial blood gas, the noninvasive oxygenation saturation index (OSI = mean airway pressure × FiO₂/SpO₂ × 100) and SpO₂ to FiO₂ ratio (S:F ratio) can be used as good alternatives of OI and P:F ratio, respectively, provided a SpO₂ of 97% or lower.

Mean airway pressure is the primary determinant of oxygenation, FiO₂ being the second evident determinant. Mean airway pressure can be calculated as the area under the curve of the pressure-time curve. It is therefore mainly determined by the positive end-expiratory pressure (PEEP), the peak inspiratory pressure (PIP), and the inspiratory time. Increasing the mean airway pressure by manipulation of any of these variables will recruit alveoli, improve ventilation-perfusion matching, and decrease intrapulmonary shunting. In addition, increasing mean airway pressure may also result in a significant improvement in respiratory compliance, thereby allowing lower driving pressure (difference between inspiratory and expiratory pressures) to obtain a similar tidal volume, as illustrated in ◘ Fig. 12.1.

Fig. 12.1

Schematic representation of the respiratory pressure-volume curve. The larger curve represents the entire excursion of the lung volume from a pressure of zero to a maximal pressure. The smaller loops illustrate pressure-volume curves generated with similar and physiologic tidal volumes (ΔV), at different levels of PEEP. The respiratory system compliance (ΔVolume/ΔPressure) can be illustrated as the local slope of the pressure-volume curve. The compliance decreases at each extreme of the ascending limb of the curve, reflecting the overinflation (right extreme, loop 3) and the atelectasis (left extreme, loop 1), as illustrated by the larger ΔPressure observed for similar ΔVolume. The fourth small loop illustrates the interest of ventilating the lung on the deflation limb of the pressure-volume curve, with both a good compliance (relatively small ΔP4) and a large lung aeration, while the level of pressure is smaller than for the loop 1. To go from loop 2 to loop 4 requires a recruitment maneuver

The restoration or maintenance of functional residual capacity (FRC) above the closing capacity (CC, the volume at which small airway closure occurs during expiration) is critical in the maintenance of a normal V/Q ratio. Conditions that decrease FRC below CC (or increase CC above FRC) result in a maldistribution of ventilation-perfusion and adversely affect the mechanics of breathing (► Box 12.1). In conditions associated with a decreased FRC (i.e., pulmonary edema, pneumonitis, PARDS), increasing PEEP is the most logical measure to increase FRC (see below for a full discussion regarding PEEP). In situations associated with increased CC (i.e., bronchiolitis, reactive airway disease), strategies to reduce CC should also be considered (i.e., etiological treatment, bronchodilators, secretions clearance).

Box 12.1 Factors affecting functional residual capacity and closing capacity

Reduced functional residual capacity:

• Acute lung injury/acute respiratory distress syndrome (ARDS)

• Near drowning

• Acute pneumonia

• Pulmonary edema

• Radiation pneumonitis

• Atelectasis

• Supine position

• Abdominal distention

• Obesity

Increased closing capacity:

• Bronchiolitis

• Asthma

• Bronchopulmonary dysplasia

• Smoke inhalation with thermal injury to airway

• Cystic fibrosis

• Tracheomalacia

• Infancy

2.3 Pathophysiology of Ventilation Failure (Hypercarbia) and Application to Mechanical Ventilation

One major aim of mechanical ventilation is to provide minute ventilation (respiratory rate [RR] x the tidal volume [V_T]) that is adequate for CO₂ removal. The PaCO₂ is the result of the equilibrium between CO₂ clearance (by the ventilation) and the body’s production of CO₂ during the energy metabolism. In most clinical circumstances, the control of PaCO₂ will rely on the control of the minute ventilation. However, some control of the body’s endogenous CO₂ production is possible through the increase in the use of fats versus carbohydrates for nutrition, avoidance of overfeeding, or by control of body temperature. Sedation, prevention of hyperthermia, and even induction of mild hypothermia can also facilitate the control of hypercarbia and limit mechanical ventilation requirements.

Although minute ventilation is defined as RR times V_T, not all of the V_T is involved in effective in gas exchange. That part of V_T that does not participate in gas exchange is referred to as physiologic dead space. Total or physiologic dead space is composed of anatomic dead space (that area of the conducting areas or the trachea and bronchi that do not participate in gas exchange) and alveolar dead space (those alveoli which are ventilated but not perfused). In the healthy state, the alveolar dead space is minimal so that anatomic and physiologic dead space are approximately the same. Although anatomic dead space, representing approximately 30% of a normal tidal breath or 150 mL in an average-sized adult, does not generally change regardless of the disease process, alveolar dead space may change significantly in patients with pulmonary parenchymal disease, with pulmonary vascular disease, or with changes in cardiac output resulting in alterations in pulmonary perfusion. The latter principle is clearly demonstrated by the abrupt decline in end-tidal CO₂ (ETCO₂) that occurs with cardiac arrest, a decrease in cardiac output, or pulmonary embolism. Of note, anatomic dead space tends to augment with younger age.

Dead space ventilation (V_D) refers to ventilation that does not participate in gas exchange. Since the anatomic dead space is relatively constant in patients with healthy lungs, increasing the V_T decreases the ratio of V_D to V_T. In effect, the increased V_T increases alveolar ventilation. It is also the case in most patients with mild to moderate lung disease that alveolar dead space is relatively fixed and that changes in V_T primarily affect alveolar ventilation. Thus, in most cases, a 10% increase in V_T will result in a greater than 10% improvement in effective minute ventilation, such that small increases in V_T are more effective at ventilating than the same proportionate increases in rate. However, in some patients with severe lung disease, ventilation of poorly perfused regions of the lungs (alveolar V_D) can be significant. In this setting, increases in V_T may not decrease V_D/V_T since higher alveolar pressures as a result of larger V_T may result in a further decrease in pulmonary perfusion and increase in alveolar V_D. An estimation of the effect of changes in V_T on V_D/V_T in such clinical scenarios can be provided by estimating V_D/V_T using capnography with ETCO₂ measurements. The alveolar dead space fraction (AVDSf) can be estimated using the following equation:

$$ \mathrm{AVDSf}=\left({\mathrm{P}\mathrm{aCO}}_2-{\mathrm{P}}_{\mathrm{ET}}{\mathrm{CO}}_2\right)/{\mathrm{P}\mathrm{aCO}}_{2\kern1em }\ \left({\mathrm{P}}_{\mathrm{ET}}{\mathrm{CO}}_2=\mathrm{end}-\mathrm{tidal}\ {\mathrm{P}\mathrm{CO}}_2\right) $$

The alveolar dead space fraction is an important severity marker and is indepedently associated with mortality in pediatric acute respiratry distress syndrome.

We can summarize that a change in the metabolic rate with an alteration in CO₂ production, a change in minute ventilation (RR or V_T), or a change in V_D may affect PaCO₂. Importantly, ventilation to normocarbia is not necessary and may in fact be harmful, especially in patients with severe lung disease. Current practice includes the use of permissive hypercarbia. The Pediatric Acute Lung Injury Consensus Conference (PALICC) recommended maintaining pH 7.15–7.30 within lung protective strategy in moderate to severe PARDS, except in patients with intracranial hypertension, severe pulmonary hypertension, select congenital heart disease lesions, hemodynamic instability, or significant ventricular dysfunction.

2.4 Impact of MV on the Respiratory and the Cardiovascular Systems

Although a lifesaving technology, mechanical ventilation may also have a negative impact on the respiratory system. Ventilation-induced lung injury (VILI) originates from several mechanisms, including volutrauma (the injury caused by alveolar overdistention), atelectrauma (injury due to repeated opening/closure of lung units), barotrauma (consecutive to elevated transpulmonary pressures), and biotrauma (following the release of mediators that can aggravate preexisting lung injury). A review about VILI incidence and mechanisms is beyond the scope of this chapter. However, it is important to acknowledge that although there is clinical evidence that VILI exists in the pediatric population and that protective lung volume ventilation should be recommended, it remains a field which has been much less studied than in adults. In particular, no large randomized controlled trial assessing the impact of low V_T has been conducted in the PICU. Most components of the ventilation strategy in pediatric patients are therefore based on concepts originating from studies conducted in the adult population. This is particularly important since experimental animal data tends to suggest that the pediatric population may be less sensitive to VILI mechanisms.

One goal of mechanical ventilation is to support the respiratory muscles in case of respiration failure, usually associated with a high work of breathing and a risk of diaphragm fatigue. It is, however, clear that complete assistance, with suppression of the spontaneous respiratory activity, is associated with rapid diaphragm dysfunction (ventilation-induced diaphragm dysfunction, VIDD). In adult populations, decreased diaphragm force and atrophy of diaphragm fibers are observed as early as in the first 24 hours and deteriorates further in the absence of spontaneous breathing activity. The dynamic of VIDD in the PICU has been less extensively studied, although diaphragm atrophy has also been reported in infants. The diaphragmatic force is one major factor of extubation success, and maintaining diaphragm function and activity should therefore be an important focus during mechanical ventilation. Importantly, the clinical assessment of the patient breathing activity during mechanical ventilation is not easy. Physiologic studies based on monitoring of esophageal pressure or diaphragm electrical activity have shown that blunted respiratory drive is highly prevalent in the PICU, even at very low levels of support and when the clinical team thought that the patient was actively breathing.

The effects of mechanical ventilation on the cardiovascular system are complex and highly dependent on a patient’s underlying condition. The basic hemodynamic effects of mechanical ventilation are as follows. During positive pressure ventilation, the output of the right ventricle decreases during inspiration while at the same time, the output of the left ventricle increases. The opposite occurs during a spontaneous or negative pressure breath; right ventricular output transiently increases, and left ventricular output decreases. The effects on left ventricular output are felt to be primarily a consequence of changes in afterload. A positive pressure breath decreases left ventricular transmural pressure, while a negative pressure breath does the opposite, increasing LV transmural pressure. Overall, positive pressure ventilation often results in decreased cardiac output due to decreases in systemic venous return and LV preload and an increase in RV afterload. The effects of positive pressure ventilation on preload are most frequently not wanted, and they may be compensated by increasing intravascular volume through fluid loading. Preload reduction may, however, be a positive effect in patients with fluid overload or congestive heart failure. Patients with normal cardiovascular function can tolerate the effects of positive pressure ventilation with little compromise in the absence of dehydration or intravascular volume depletion. However, the initiation of positive pressure ventilation in the face of hypovolemia, including warm septic shock, can have catastrophic hemodynamic consequences, and one should always be ready to volume load such patients at the time of intubation. Patients with cavo-pulmonary anastomosis who are dependent on passive pulmonary blood flow can be highly sensitive to the application of positive pressure ventilation due to decreased systemic venous return and pulmonary blood flow.

The normal pulmonary vascular bed has low pressure and resistance. Critical illness and especially pulmonary disease can increase pulmonary vascular resistance (PVR) through hypoxic vasoconstriction, acidemia, and the release of vasoactive mediators. Variables relating to lung inflation may also affect pulmonary vascular resistance through maintenance of normal lung volumes, ventilation-perfusion matching, and minimizing hypoxemia. Pulmonary vascular resistance is increased when the lung is collapsed and is minimized (optimized) when the lung is inflated to normal functional residual (FRC) capacity. The increased PVR at low lung volumes is a result of the combination of hypoxic pulmonary vasoconstriction and some compression of extra-alveolar blood vessels. With lung inflation, the extra-alveolar vessels are held open by adjacent connective tissue. If the lung is inflated much above normal FRC, pulmonary vascular resistance increases as a result of the compression of the alveolar capillary bed within overdistended alveoli. From a purely mechanical perspective, the initiation of positive pressure ventilation with high levels of PEEP can result in decreased cardiac output by virtue of both mechanisms: decrease in systemic venous return and increased pulmonary vascular resistance. These effects are well tolerated in the presence of normal RV function but may be significant in the patient with RV dysfunction or single ventricle with the absence of a right-sided pumping chamber. To the extent that the institution of positive pressure ventilation with therapeutic levels of PEEP can result in the correction of hypoxemia and respiratory acidosis, these beneficial effects on right ventricular performance can counteract the negative mechanical effects. Thus, even in the patient with severe RV dysfunction or Fontan physiology, improving oxygenation and ventilation with the normalization of functional residual capacity can result in decreased PVR and improved cardiac output.

The influence of positive pressure ventilation on the left ventricular performance is multifactorial. As mentioned above, during positive pressure inspiration, LV afterload is affected by intrathoracic pressure acting on the external wall of the LV. The transmission of positive airway pressure to the mediastinum and the external surface of the left ventricle decreases the transmural pressure of the LV, thus decreasing LV afterload. In addition, during the inspiratory phase of a positive pressure breath, LV preload is increased as the positive pressure applied to the lungs aids in emptying the pulmonary veins into the left atrium. The effect of improved LV systolic function during inspiration is most pronounced in patients with LV dysfunction. These patients also benefit from redistribution of limited cardiac output and oxygen delivery by virtue of decreased work of breathing and thus decreased need to provide oxygen for the respiratory muscle work. The proportion of cardiac output which is normally dedicated to the respiratory muscles is small, close to 5%. In case of respiratory failure (including in patients with cardiogenic pulmonary edema), the respiratory muscles consumption increases, and the proportion of cardiac output used by the respiratory system can exceed 30%. Reducing the work of breathing in these patients is therefore essential.

The complex cardiopulmonary dynamics during positive pressure ventilation can be demonstrated at the bedside through the observation of systolic pressure variation in the arterial line waveform occurring during the respiratory cycle.

3 Basics of the Ventilator Functioning

3.1 Negative Pressure Ventilation

With the poliomyelitis epidemics of the 1930s, negative pressure ventilation was introduced to support patients with neuromuscular weakness leading to acute and chronic respiratory failure. The negative pressure ventilators (“iron lungs”) were large tanks into which the patient’s entire body was placed (◘ Fig. 12.2). The patient’s neck was surrounded by a rubber mat with a small opening in the center through which the patient’s head protruded. The driving force was a piston located at the bottom of the tank. A downward movement of the piston caused an increase in the volume of the container, creating a negative pressure around the patient’s thorax. The negative pressure resulted in the expansion of the chest wall, with air entry through the patient’s airway into the lung. Although somewhat effective in patients with respiratory failure related to muscle weakness, there were significant limitations in the amount of negative pressure that could be generated, and as such, these devices were not effective in patients with significant alterations in respiratory compliance or resistance. Additionally, the devices were bulky, restricted access to the patient, and could not be used in patients with airway disease.

Fig. 12.2

Photograph of a negative pressure ventilator otherwise known as the “iron lung.” These devices were used during the poliomyelitis epidemics of the 1930s and 1940s for the treatment of acute and chronic respiratory failure. The rubber sheet with a small opening in the middle (arrow) was placed over the patient’s head to ensure an airtight seal. The amount of subatmospheric pressure was indicated on the pressure gauge on the top of the tank (circle)

Iron lungs hold a place in the medical history, but negative pressure ventilation is still being used and available in the modern era. Large containers have been replaced by cuirass or vests that fit over the patient’s thorax and are sealed at the waist and neck. The air within the jacket is intermittently evacuated, thereby creating a negative pressure (compared to the atmosphere). The support can be delivered as a constant negative pressure or with intermittent positive pressure cycles to facilitate exhalation. These devices could be used for home care and found their greatest use in patients with chronic respiratory insufficiency due to neuromuscular weakness but are also sometimes used in acute conditions. Negative pressure ventilation can be used alone or in combination with other type of support, like high-flow nasal cannula or noninvasive ventilation.

The advantages of negative pressure ventilation are that it does not require endotracheal intubation, it can be applied intermittently, it can be used at home without the need for tracheostomy, and it avoids the impact of positive pressure ventilation on cardiovascular system (preload reduction and right afterload increase). This may be particularly interesting in children following cavo-pulmonary anastomoses. In the postoperative setting of these surgeries, positive pressure ventilation can significantly decrease the passive pulmonary blood flow. Although the ideal strategy is to attempt early tracheal extubation, negative pressure ventilation could be an interesting alternative when extubation is not possible.

Negative pressure ventilation also has disadvantages. The delivered support is not synchronized with the patient breathing. In case of abnormal airway with inspiratory resistance, the negative pressure ventilation can exacerbate the respiratory distress. Finally, the fact that no positive pressure is applied in the airway can be perceived as better for the impact on the lung and VILI, but this is not absolutely true. Indeed, the transpulmonary pressure (difference between alveolar pressure and pleural pressure) is the variable that best reflects the lung strain and the risk of VILI. Increasing transpulmonary pressure either by a positive pressure in the alveoli (positive pressure ventilation) or by a negative pressure in the pleura (negative pressure ventilation) can both increase the lung strain and expose to the risk of VILI.

3.2 Positive Pressure Ventilation

As opposed to negative pressure ventilation, during positive pressure ventilation, the inspiratory flow is achieved by the increase in the airway pressure, which generates a pressure gradient from airway to the alveoli. This increases the transpulmonary pressure by increasing the alveolar pressure rather than by decreasing the pleural pressure. Positive pressure ventilation therefore requires an airway interface, mostly endotracheal tube or tracheostomy for conventional ventilation. Noninvasive positive pressure ventilation can also be delivered with a nasal, oronasal, or facial mask, nasal cannula or prongs, or a helmet. In this chapter, we focus on conventional invasive positive pressure ventilation.

When the decision has been made to initiate mechanical ventilation, the clinician will be faced with the following decisions: (a) the mode of ventilation (how the ventilator cycles are determined); (b) the controlled variable (pressure or volume) which will control the tidal breath; (c) the specific ventilator settings, including the magnitude of the controlled variable (inspiratory pressure or tidal volume), the PEEP, the inspiratory time and the ventilator rate, the trigger sensitivity, and the FiO₂; and (d) the alarm settings.

3.3 The Different Ventilation Modes

The era of positive pressure ventilation began with controlled mandatory ventilation (CMV) which provided intermittent positive pressure breaths to the patient without the ability to sense the patient’s own respiratory efforts and no gas flow in between the ventilator breaths. This provided no means to allow the patient to breath spontaneously, resulting in significant patient-ventilator asynchrony unless deep levels of sedation or neuromuscular blockade were used. CMV was followed by intermittent mandatory ventilation (IMV) which provided a set number of breaths/min provided at a specific interval but also allowed for spontaneous ventilation through the use of a continuous gas flow or a demand valve. Despite the ability to allow spontaneous ventilation, the IMV mode did not synchronize the ventilator breath with the patient’s effort, and there was no assistance during the spontaneous breaths. With the development of technology for sensing the patient’s respiratory efforts, strict IMV has been replaced by modes such as assist control (AC) and synchronized intermittent mandatory ventilation (SIMV). Both AC and SIMV modes deliver a set number of mandatory ventilator “full” breaths, determined by the set respiratory frequency and the set volume or pressure, synchronized with the patient’s effort. If the patient breathes above the set minimal frequency, there will be additional minute ventilation from this spontaneous ventilation.

In SIMV, the additional breaths can be allowed with no added support (SIMV alone) or can be assisted according to pressure support (SIMV-PS) or volume support (SIMV-VS) principle (see below). In assist control, every additional detected breath will lead to a full ventilator breath. The theoretical advantage of AC ventilation is that the patient receives a similar support for each breath. However, if the spontaneous respiratory frequency is excessively high or in case of auto-triggering, the risk of overassistance and of expiratory flow limitation with overdistension may be higher. Of note, those risks also exist in SIMV, and in both modes, it is essential to avoid auto-triggering and limit unnecessary tachypnea. There is also a belief that with SIMV, weaning may be facilitated by the possibility of decreasing either the frequency or the level of support. However, studies in adults rather suggest that this complexity prolongs the weaning duration.

In its earliest forms, triggering was accomplished by detecting a pressure change in the ventilator circuit (usually −1 to −3 cm H₂O). Further refinement of patient effort sensing relies on detection of flow differences between the inspiratory and expiratory limbs of the ventilator circuit (flow triggering). Available now in all new critical care ventilators, flow triggering requires less patient work and is more comfortable than pressure triggering. The trigger setting can be difficult. Setting the threshold too high may lead to failure to sense the patient’s spontaneous efforts (“wasted efforts’”). Setting the trigger threshold too low (too sensitive) can lead to auto-cycling of the ventilator due to flow changes caused by cardiac oscillations, turbulence from condensation in the circuit, or a leak around the endotracheal tube. Even with modern ventilators, patient-ventilator synchronization remains a challenge, and mechanically ventilated children spend on average 27% of the time in conflict with their ventilator. An alternative mode of synchronization has been developed with neurally adjusted ventilatory assist (NAVA), in which the ventilator is triggered by the diaphragm electrical activity, a rapid signal of patient’s breathing, leading to markedly improved patient-ventilator interactions. NAVA is not considered a conventional ventilation mode and will not be discussed further in this chapter.

3.4 The Control Variable: Volume-Controlled Ventilation, Pressure-Controlled Ventilation, and Pressure-Regulated Volume Control

The control variable (pressure or volume) is that parameter which is set to determine the magnitude of the tidal breath.

With volume-controlled ventilation , a specific V_T is set by the clinician and an inspiratory time is chosen. The flow provided is then integrated based on the tidal volume and inspiratory time. For example, if a V_T of 500 mL with an inspiratory time of 1 s is chosen, 500 mL will be delivered over 1 s using a gas flow of 30 L/min if flow is constant during the entire inspiratory phase (500 mL/1 s = 30 L/min). Commonly, a decelerating flow pattern is used such that the flow is higher in early inspiration and lower toward the end of inspiration. Volume-controlled ventilation may be best used in patients with relatively normal resistance and compliance of the respiratory system. The advantage of volume-controlled ventilation is that a constant V_T is delivered even with changing resistance and compliance. When volume-controlled ventilation is used, the peak inspiratory pressure (PIP) and plateau pressure (P_Plat) should be monitored as changes in PIP or P_Plat reflect changes in resistance and compliance of the respiratory system. The peak inspiratory pressure (PIP) is strongly influenced by the resistance of the tubing and of the airways. The plateau pressure better reflects the impact of ventilation on the poorly compliant lung. To measure the plateau pressure, a no-flow end-inspiratory pause (0.2–0.5 s) should be applied to interrupt the flow and allow pressure equilibration in the absence of patient effort. Use of the plateau pressure eliminates the resistance imposed by the ETT and airways and thereby approximates the pressures that occur within the alveoli. Rising pressures require an investigation which should start at the ventilator and work toward the patient including a check for kinking of the circuit or endotracheal tube and obstruction to the endotracheal tube or major airways by mucus, auscultation to rule out mainstem intubation or bronchospasm, and a radiograph to evaluate for deteriorating alveolar disease (i.e., pneumonia, ARDS) or external factors impeding respiratory excursion (i.e., pneumothorax, restrictive diseases of the thorax, abdominal distention) or agitation. Increased resistance of the respiratory system can be suspected when observing a significant difference between peak and plateau pressures. In such cases, increasing inspiratory time decreases the inspiratory flow rate and can be used to decrease the PIP. However, the I:E ratio should be considered when increasing inspiratory times at high ventilator rates or in diseases with high expiratory resistance (asthma) because of the risk of insufficient expiratory time. In setting the ventilator rate, tidal volume, and inspiratory time (or peak flow), one must be aware of the dynamic interplay of these variables. If the peak airway pressure is unacceptably high, the pressure-controlled mode may be chosen, although a decrease in V_T may be a relatively equivalent solution.

With pressure-controlled ventilation , a preset pressure above PEEP is delivered over a selected inspiratory time. The inspiratory flow rate will be somewhat dependent upon the airway resistance and respiratory system compliance, achieving high levels initially and decelerating toward zero near the end of inspiration. Most ventilators allow manipulation of inspiratory gas flow rate through selection of the percentage of the inspiratory phase devoted to developing peak airway pressure (how quickly the plateau is achieved). Because inspiratory pressure is the controlled variable, changes in respiratory system mechanics (i.e., compliance and/or resistance) will result in changes in the delivered V_T and minute ventilation. During pressure-controlled ventilation, the delivered V_T is determined by the pressure level above PEEP (sometimes referred to as the delta pressure, ΔP, or the driving pressure), the inspiratory time, loss of V_T from a leak around an uncuffed ETT, and the patient’s resistance and compliance. Because there is often flow cessation with pressure equilibration at the end of inspiration in pressure-controlled ventilation, the PIP is usually close to the plateau pressure. But significant differences can occur, especially when the inspiratory time is short or when the resistance of the respiratory system is high. It is usually considered that for a given minute ventilation, the mean airway pressure is higher and the peak pressure is lower in pressure-controlled than in volume-controlled mode, although this remains debated. An additional potential advantage is the decelerating flow pattern, which may help in the recruitment of alveoli with long time constants (pendelluft effect). This advantage must be balanced against the increased shearing forces of the high early inspiratory flow rates. Also, decelerating flow is now also used in most volume-controlled modes.

Given that the PIP is controlled, the risk of barotraumas is assumed to be less than with volume-controlled ventilation. Another potential advantage is that in case of deteriorating compliance, the V_T will automatically decrease for a given delta pressure, applying the principle of a more protective ventilation in the sickest patients. Pressure-controlled ventilation may be particularly beneficial in patients with decreased compliance or alveolar space disease such as pneumonia or ARDS since the higher mean airway pressure may improve oxygenation. On the other hand, the risk of large V_T is higher in pressure-controlled ventilation, as compared to a volume-controlled ventilation set with a low V_T.

Pressure-controlled ventilation has also historically been preferentially used in neonates and small infant because of the lack of precision of V_T delivery by older ventilators. A discrepancy of a few mL is a significant issue when the set V_T is 20 mL. Now, critical care ventilators are relatively accurate to very small tidal volumes in neonatal modes.

As with volume-controlled ventilation, inspiratory time is set with pressure-controlled ventilation. Since most pressure modes are time cycled (end inspiration based on the inspiratory time), increasing the inspiratory time will increase the mean airway pressure. Increasing inspiratory time will increase the delivered V_T if inspiratory time is shorter than that required for all lung units to fill and come to pressure equilibration. This situation is different from volume-controlled ventilation where lengthening the inspiratory time may decrease the PIP but does not affect V_T. With pressure-controlled ventilation, the exhaled V_T should be monitored to assess ongoing changes in the compliance of the respiratory system. A decrease in the exhaled V_T should prompt a thorough investigation into its cause, which includes the same steps as outlined above for investigating an increase in PIP during volume-controlled ventilation.

Pressure-regulated volume control (PRVC) is an option which combines features of both volume- and pressure-controlled ventilation. The tidal volume is set by the clinician, as with volume-controlled ventilation. However, the pressurization pattern follows the principle of pressure-controlled ventilation, with high early inspiratory flow rates, decelerating flow, and inspiratory pressure relatively constant during a given inspiration (square pressure-time curve). The key difference with pressure-controlled ventilation is that the level of inspiratory pressure is not set, but adjusted breath by breath by the ventilator. The ventilator continuously monitors the delivered inspired V_T and adjusts the next inspiratory pressure accordingly in order to deliver the prescribed title volume. As a safety feature, an internal software algorithm restricts the magnitude of pressure changes so that the patient cannot be markedly overinflated in response to rapid changes in compliance (e.g., tracheal tube kinking). The putative advantage of PRVC is that it combines the advantages of both variable controls: it delivers a relatively fixed V_T (as with volume-controlled) but with the advantageous flow pattern of pressure-controlled ventilation. However, the benefit of this mode has not been demonstrated in the PICU.

From the discussion thus far, it should be apparent that six basic types of ventilation can be considered: AC-pressure-controlled, AC-volume-controlled, SIMV-pressure-controlled, SIMV-volume-controlled, AC-PRVC, and SIMV-PRVC. These are the basic modes of mechanical ventilation used in the pediatric ICU today, and most modern ICU ventilators can provide all of these modes and options. Importantly, the advantages of the different options have not yet been evaluated nor compared in pediatric studies. Clinicians must decide based on the patient condition and the knowledge of the theoretical principle of these different options.

3.5 Supported Ventilation

Supported ventilation is defined as a breath that is triggered by the patient, assisted by the ventilator (volume or pressure), and cycled by the patient (the patient determines the inspiratory time). It is used with SIMV ventilation to support spontaneous breaths that occur in between ventilator breaths, or it can be used alone with no set minimal frequency usually as a means of weaning. In that situation, the patient determines the ventilator pattern (i.e., frequency, inspiratory, and expiratory times) by initiating and terminating each breath. Therefore, supported ventilation is used only in patients with an intact ventilatory drive. The patient provides the work to trigger the breath, and then the remaining work for the breath is shared by the ventilator action and the patient’s contribution.

In pressure support ventilation (PSV), once triggered by the patient’s effort, the ventilator delivers a rapid and decelerating flow of gas to provide a preset pressure support level. The breath is terminated when inspiratory flow decreases to a percentage (generally 25%) of its peak value. At that point, the exhalation valve opens, and the circuit pressure returns to the expiratory pressure (PEEP). Therefore, the patient retains control of the cycle length and flow characteristics. The V_T is determined by the patient’s inspiratory effort, the preset pressure support level, and the respiratory system impedance (resistance and compliance). Traditionally, PSV has been used to compensate for the supposed inspiratory work imposed by the ETT, but this belief has been challenged. Several studies have shown that even minimal levels of pressure support decrease excessively the work of breathing. Even though the internal diameter of pediatric ETTs is small, the flow generated through these tubes are too slow to induce important resistance. The resistance imposed by an ETT is in fact lower than the physiological resistance of the normal turbulent flow through non-intubated pediatric airways. However, PSV may help to compensate for an excessive work of breathing in a patient not yet ready for extubation.

Multiple methods of weaning ventilation with PSV have been used. One approach involves setting the pressure support level high enough to achieve delivery of the typical mechanical tidal breaths (8–10 mL/kg) with no backup SIMV rate and then gradually decrease the pressure support down till extubation criteria are met. Another method involves the combined use of SIMV and PSV in which the pressure support during assisted breaths is set to a minimal level, and the SIMV rate is gradually decreased.

In volume support ventilation (VSV), the supported breaths are triggered and cycled off in a similar way as during PSV. However, the level of pressure assist is not set but regulated by the ventilator to deliver the preset volume, provided that a maximum pressure limit is not exceeded. This mode therefore combines the theoretical benefits of PSV with the capability of providing a guaranteed minimum volume.

3.6 How to Set the Control Variable (Tidal Volume or Delta Pressure)

It has become increasingly clear over the last 20–30 years of mechanical ventilation that setting the mechanical ventilator to achieve “normal” values for pH and PaCO₂ can cause harm. While this is of utmost importance for nonhomogeneous injured lungs (e.g., ARDS), the principles extend even to those with seemingly healthy lungs. Given the discussion above on dead space ventilation, it may seem more efficient to prioritize increasing V_T to normalize pH, particularly in circumstances where alveolar dead space is increased. However, the use of supraphysiologic V_T (or pressure) has been clearly implicated in the pathogenesis of ventilator-induced lung injury, either further harming an already damaged lung or inducing injury in previously healthy lungs. While there is evidence that pediatric patients (and animals) are less sensitive to high-tidal volume ventilation than adult patients (and animals), we believe that it is still important to stay in or below the resting physiologic range of tidal volume (generally 5–8 ml/kg based on predicted body weight) and consider using lower than this (3–6 ml/kg) for children with more severe impairments in lung compliance.

Of course, the tidal volume cannot be decoupled from the ventilator pressures, particularly the delta pressure (peak pressure, PEEP) which is either set (in a pressure control mode) or achieved (in volume set modes). This is because the underlying disease state of the patient alters respiratory system compliance and airway resistance, which in turn affects the pressure-volume relationship. In truth, the mechanisms of injury to the lung are largely dependent on regional transpulmonary pressure gradients and regional tidal volume. The transpulmonary pressure reflects the alveolar pressure minus the pleural pressure (which may be increased if the chest wall compliance is poor), either at end exhalation or end inspiration. We can estimate the alveolar pressure using airway pressure under static conditions (no flow), both at end inspiration (plateau) and end expiration (PEEP). The pleural pressure is more difficult to measure in routine practice but can be estimated with esophageal pressure using a variety of assumptions. However, airway pressure, esophageal pressure, and V_T are estimating what is happening in the respiratory system in general. There may be marked variation in regional pressures and volumes, which ultimately may lead to lung injury.

As a general principle, peak inspiratory pressure is increased under circumstances of impaired lung compliance, impaired chest wall compliance (or increased chest wall elastance), and increased airway resistance. The risk of barotrauma and VILI is most closely tied to the transpulmonary pressure at end inspiration under static conditions, estimated by the plateau pressure minus the pleural (or esophageal) pressure and the transpulmonary driving pressure [(Pplat-P_Esplat) − (PEEP-P_EsPEEP)]. If the peak pressure is high from increases in airway resistance (such as asthma), then there will be a larger gradient between the peak and plateau pressure as the peak pressure applied during airflow is not all being transmitted to the alveoli. In contrast, if the peak pressure is high from decreased chest wall compliance (such as chest wall edema from fluid resuscitation), the plateau pressure will also be high, but some of that plateau pressure is being dissipated across the chest wall to move it out of the way. As a result, the transpulmonary driving pressure will be lower than the set ventilator driving pressure. These principles are crucial to remember when selecting pressure and volume targets for individual patients as they must be linked to the underlying respiratory system physiology. While pediatric specific evidence is weak, the general consensus is that plateau pressure should be limited to 28 cm H₂O, although there is not clear evidence for this particular target. This can be higher (up to 32 cm H₂O) if there is thought to be significant impairment in chest wall compliance (if a direct estimate of pleural pressure like esophageal manometry is not available). However, the risk of VILI is probably more closely tied to the driving pressure (plateau-PEEP), although there are no clear pediatric specific recommendations regarding optimal limits of this pressure. Adult data has advocated driving pressure targets less than 15 for clinical practice.

Setting the optimal targets for V_T, inspiratory pressures, and ventilator rate must be individualized for patients but embrace the general principles of limiting these pressures and volumes as much as possible. This implies, particularly for patients with more severe lung disease, that normocarbia is not expected and that pH will be lower than 7.4 during the acute disease. There is no clear standard for the lower limit of “acceptable” pH, and the decisions regarding these limits need to weigh the risks of lung injury against other potential risks of low pH such as hemodynamic status and cerebral and pulmonary blood flow. This will likely mandate heavy sedation and/or muscle relaxation to suppress the patient’s respiratory drive. In fact, nearly all of these recommendations regarding limiting driving, plateau pressure, and V_T are meant for the fully controlled patient. This is because when the patient is breathing spontaneously, the airway pressures underestimate the total pressure generated (sum of patient and ventilator contributions) (see section on spontaneous versus controlled ventilation).

3.7 How to Set the Positive End-Expiratory Pressure (PEEP)?

PEEP refers to the positive pressure applied during expiration phase until the initiation of the next breath. PEEP maintains the patency of injured lung units which may collapse during exhalation. Although physiologically accomplishing the same thing, it should be differentiated from CPAP or continuous positive airway pressure, which is the term used during spontaneous ventilation. In normal adults, the functional residual capacity (FRC, the volume at which lung recoil inward is balanced by chest wall recoil outward, obtained when no respiratory effort is ongoing and the glottis is open) and the end expiratory lung volume (EELV, the volume at which the next inspiration begins) are equal and exceed the closing capacity (CC, the lung volume at which airway closure begins to occur). Thus, spontaneously breathing healthy adolescents and adults require little PEEP to prevent atelectasis or de-recruitment. In contrast, newborns and infants younger than 1 year with their highly compliant chest wall will have an FRC that approaches and, in some cases, may be less than CC under passive (i.e., sedated or paralyzed) conditions. Infants, therefore, actively increase and maintain their EELV during spontaneous ventilation using several mechanisms: a rapid respiratory rate with short expiratory times (i.e., there is insufficient time for expiratory flows to reach zero; intrinsic or auto PEEP is present), a braking of the expiratory airflow by the laryngeal muscle contraction during exhalation (which does not occur with a tracheal tube in place) and by the persistence of diaphragm activity during expiration (tonic activity), and increased intercostal muscle tone that stabilizes the chest wall, thereby increasing elastic recoil. This active control leads to the concept of “physiologic PEEP” (typically 3–5 cm H₂O) to avoid airway closure and lung volume de-recruitment with ventilation-perfusion inequalities. Thus, sedated or intubated infants generally require the use of PEEP to overcome the loss of these dynamic compensatory mechanisms. Studies on teens and young adults demonstrated that even patients with apparently normal lungs benefit from the addition of PEEP during prolonged mechanical ventilation.

Higher levels of PEEP may be required in patients with alveolar space disease or other pathologic conditions that increase CC and/or decrease FRC, such as conditions associated with poor chest wall compliance (i.e., obesity, fluid overload, restrictive syndrome) or abdominal distention. Change in PEEP is frequently the first method used for regulating mean airway pressure. PEEP increases lung volume at expiration, restoring FRC and preventing the injury associated with the repetitive circle of collapse-opening-recollapse (atelectrauma). The application of PEEP also prevents airway closure during expiration, redistributes pulmonary edema fluid from alveoli to the interstitium, maintains alveolar surfactant activity, and improves ventilation to low V/Q lung units. When applied in the proper amount, PEEP should improve lung compliance so that a given change in pressure results in a greater V_T (◘ Fig. 12.1). This positive impact of PEEP supposes an improvement in lung aeration (prevention of lung atelectasis, reopening of non-ventilated area). However, in some patients with very severe parenchymal disease or with heterogeneous disease (complete atelectasis), the PEEP may not be sufficient to reopen the non-aerated area while being sufficient to over-distend the aerated lung segments. In these conditions, PEEP can be counterproductive and be associated with a worsening of lung compliance and of dead space ventilation. Other side effects of excessive PEEP include the depression of cardiovascular function and increase in pulmonary vascular resistance.

Several different methods of determining the optimal PEEP for patients with parenchymal lung disease have been suggested. Lung imaging (i.e., chest X-ray, tomodensitometry, lung ultrasound) can help to evaluate the expansion and aeration of the lung fields. Identification of lower and upper inflection points on the sigmoidal pressure-volume curves has also been used to determine the pressure window between overdistension and atelectasis, but it has been progressively abandoned because of its technical challenge and poor reproducibility. Titration maneuver of PEEP has been described, aiming to determine the opening pressure and the closing pressure, based on the optimal lung compliance. The monitoring of esophageal pressure may also help to determine the PEEP needed to overcome the pleural pressure during expiration, preventing the occurrence of expiratory negative transpulmonary pressure and its risk of lung segment collapse. Finally, FiO₂/PEEP grids are simple tools to guide the PEEP management based solely on the oxygenation requirements. All these methods have been studied primarily in adults, and none of them has been proven superior to simple grids. Evidence in pediatric patients is even more limited. The Pediatric Acute Lung Injury Consensus Conference (PALICC) recommended moderately elevated levels of PEEP (10–15 cm H₂O) in patients with severe PARDS, titrated to the observed oxygenation, markers of oxygen delivery, respiratory system compliance, and hemodynamic response. PEEP levels greater than 15 cm H₂O may be needed for severe PARDS although attention should be paid to limiting the plateau pressure.

The use of PEEP is also discussed in patients with serious airflow obstruction and resulting high intrinsic PEEP. In this condition, the patient’s respiratory effort needs to overcome the intrinsic PEEP before inducing a pressure or flow change in the endotracheal tube, leading to marked patient-ventilator asynchrony by triggering delay or wasted efforts. The addition of PEEP in that context can improve ventilator triggering, although application of excessive levels of PEEP may dangerously increase FRC. A logical approach is to use levels of PEEP somewhat below the level of measured auto-PEEP (e.g., 80% of auto-PEEP) to facilitate triggering and to avoid increasing the already markedly elevated FRC.

3.8 How to Set the Respiratory Rate and Inspiratory and Expiratory Times?

The rate is commonly set primarily based on the patient’s age, the desired PaCO₂ level, and the V_T that is delivered. In patients with severe lung injury, higher rates are frequently needed to compensate for lower V_T. In patients with less severe lung injury, higher rates and lower tidal volumes may not always be the most appropriate approach since dead space is relatively constant, causing the ratio of dead space to tidal volume (V_D/V_T) to increase as V_T is decreased. Guidelines for starting ranges of respiratory rates include 12–15 breaths/min for an adolescent, 15–20 breaths/min for an older child (6–10 years of age), 20–30 breaths/min for a toddler, and 30–40 breaths/min for a neonate. Higher rates may be needed in patients with more severe degrees of lung injury, when hyperventilation is used to treat increased intracranial pressure or pulmonary hypertension or if endogenous CO₂ production is elevated. As previously mentioned, one may be able to limit rate since ventilation to normocapnia is frequently not required in most conditions with the permissive hypercapnia concept.

The setting of inspiratory time and inspiratory/expiratory ratio (I:E ratio) is frequently overlooked. Depending on the type of ventilation (i.e., pressure-controlled or volume-controlled), the effect of changing the inspiratory time has different effects. With pressure-controlled ventilation, lengthening out the inspiration time will increase the V_T in cases where the inspiratory time is shorter than that required to achieve filling of lung units with the highest time constants. More importantly, the inspiratory time along with PEEP and PIP determines the mean airway pressure. Lengthening the inspiratory time increases the mean airway pressure and will commonly increase oxygenation. Lengthening the inspiratory time can also be used as a therapeutic maneuver to help recruit alveoli with long time constants and help the resolution of atelectasis. With volume-controlled ventilation, change in inspiratory time has little impact on the V_T. Increasing inspiratory time helps to decrease the inspiratory flow rate and thereby reduce the PIP, especially in cases where airway resistance is significant. Depending on the ventilator, the inspiratory time may be set as a fixed time, by adjusting the inspiratory flow rate, as an I:E ratio, or as a percentage of the respiratory cycle. If the I:E ratio or the inspiratory percentage of the respiratory cycle is set, change in respiratory rate affects the actual inspiratory time. Since such changes could result in changes in the peak airway pressure during volume-controlled ventilation or V_T during pressure-controlled ventilation, most ventilator manufacturers have abandoned this practice. If the inspiratory time is fixed, increases in respiratory rate can result in an increase of the I:E ratio and sometimes an inverse of the I:E ratio, especially in cases of auto-triggering or when the patient has a rapid spontaneous ventilation.

During normal spontaneous ventilation, I:E ratio is 1:3 or 1:4. During conventional ventilation with AC, SIMV, or PRVC ventilation, the inspiratory time is preset during the ventilator-programmed breaths as opposed to supported breaths where the patient sets the inspiratory time. In clinical practice, the use of rate and inspiratory time is very variable. Some centers choose to use inspiratory times as low as 0.3–0.5 s for infants and up to 0.7–1 s in adolescents. Other centers commonly used lower rates and longer inspiratory times for all age groups, but it should be kept in mind that non-physiologic long inspiratory times may induce patient-ventilator asynchrony (cycling off delay), which can be uncomfortable.

The underlying disease process should also be considered. Relatively longer inspiratory times can be considered in patients with alveolar space disease and poor compliance to increase mean airway pressure and improve oxygenation. In usual practice, most clinicians restrict the inspiratory time to limit the I:E ratio at 1:1. Reversal of the I:E ratio has been used in the management of patients with severe ARDS in attempts to augment oxygenation and allow weaning of the FiO₂; however, this practice is less frequently used since the PEEP can achieve a similar results on mean airway pressure with more physiological timing. In heterogeneous lungs, prolongation of inspiration facilitates the recruitment of alveoli with long time constants (high resistance and low compliance), encourages collateral ventilation via pores of Kohn and canals of Lambert, reverses atelectasis, and improves matching of ventilation and perfusion. Importantly, the reversal of the I:E ratio may result in inadequate exhalation times, a risk that is exacerbated in case of significant spontaneous ventilation. This may result in air trapping, the stacking of one breath on another (inspiration for the next breath starts before exhalation is completed), thereby resulting in auto-PEEP. An evaluation for the presence of auto-PEEP can be performed by holding the ventilator breath at the end of exhalation. When performing this maneuver, one will observe that the airway pressure will initially be equivalent to the set level of PEEP. However, when the ventilator expiratory valve closes at the end of the expiratory time, the pressure measured in the circuit will then rise above the PEEP (= auto-PEEP) because of the exhalation of the air which was “trapped” within the lung due to the insufficient length of the expiratory phase. Auto-PEEP can also be suspected when looking at the expiratory limb of the flow-time curve on the ventilator. At the end of exhalation prior to the next breath, end expiratory flow that has returned to zero generally indicates a relatively complete exhalation (i.e., absence of significant auto-PEEP).

In patients with bronchospasm and air trapping, a focus on expiratory time is important, favoring prolonged expiration to prevent air trapping. However, a common mistake is to overlook the inspiratory time in these patients with very compliant lungs but very high airflow resistance. Increased airway resistance is an inspiratory as well as an expiratory problem. Shortening inspiratory time may lead to poor distribution of inspiratory volume with failure to involve large areas of lung in gas exchange.

Most ventilators allow the addition of an end-inspiratory pause, which can be set as an absolute time or as a percentage of inspiratory time or cycle time. It can also simply be the result of the flow rate settings. The inspiratory pause holds the inspiratory volume at the end of inspiration without ongoing gas flow. This maneuver also aims to promote the recruitment of alveoli with long time constants which are secondarily filled by the air that had previously reached the alveoli with short time constants (pendelluft effect). During volume-controlled ventilation, the inspiratory pause is associated with a typical decrease on the pressure-time curve, from the peak pressure to a plateau pressure, immediately before the expiratory drop to the PEEP. During pressure-controlled ventilation, this pattern is not observed (squares pressure-time pattern), although this does not imply that the inspiratory pressure is equivalent to a plateau pressure (the flow rate is frequently not zero). To accurately measure the plateau pressure as a reflection of the lung compliance, removing the resistance component with a zero-flow condition is essential. An inspiratory hold should be manually applied (0.5 second) while the patient has no activity (i.e., deep sedation or paralysis), and the resulting pressure at the end of the pause can be recorded as the plateau pressure.

3.9 Balancing the Contribution of the Patient to the Ventilation: Benefit/Risk of Spontaneous Breathing

For the patient with respiratory failure, there is an important balance to attempt to maintain between controlled and spontaneous ventilation. With improvements in noninvasive ventilation as well as modern modes of ventilation to synchronize with the ventilator, increasingly, patients are being maintained with spontaneous ventilation. Maintaining spontaneous ventilation with diaphragm contraction in a physiologic range during the course of mechanical ventilation may prevent ventilation-induced diaphragm dysfunction (VIDD). This is crucial because diaphragm dysfunction is thought to occur in up to 40% of mechanically ventilated patients, results in more difficulty weaning from mechanical ventilation and higher rates of extubation failure, and is associated with higher post-ICU mortality. While there are several risk factors for VIDD which are related to the underlying disease state of the patient and medication choices, higher degrees of controlled ventilation, higher ventilator driving pressures, and low diaphragm contractile activity are highly implicated in VIDD. Beyond its impact on the respiratory muscles, the patient’s contribution to the ventilation also has theoretical advantages for lung recruitment, for the ventilation of lung-dependent zones, for improving the ventilation-perfusion ratio, and for the sedation requirements.

On the other hand, there is increasing evidence to suggest that patients with severe lung injury can exacerbate the lung injury with spontaneous breathing, even when they are not being ventilated (patient self-induced lung injury or SILI). Hence, it may be prudent to interrupt this SILI cycle with fully controlled ventilation with lung protective strategies and sedation and neuromuscular blockade, especially in the sickest patients. Neuromuscular blockade, in particular with cisatracurium, has been shown to decrease mortality in adults with moderate to severe ARDS, with a variety of proposed mechanisms of action including prevention of SILI, relieving patient-ventilator dyssynchrony as a source of VILI, as well as modulating inflammation. These findings have yet to be reproduced in another cohort of adults with ARDS or in children. Thus, the findings should be interpreted with caution.

Understanding this balance between controlled and spontaneous ventilation seems particularly important although a matter of debate. There are no data which reliably demonstrate if or when patients cross from “safe” to “unsafe” levels of spontaneous ventilation. It is highly possible that this balance strongly depends on the patient’s illness severity. It seems logical to hypothesize that in the sickest patients, the risk of patient self-inflicted lung injury is particularly high, and its prevention should be prioritized, while in patients with less severe pulmonary disease, the benefits of maintaining the spontaneous ventilation may outweigh the risks associated with deep sedation and full ventilation control.

It is difficult to estimate the patient’s contribution to mechanical ventilation when they are spontaneously breathing on a mechanical ventilator using only measures of airway pressure and flow. This is because the patient augments the change in airway pressure with a reduction in pleural pressure, resulting in a net larger transpulmonary pressure gradient to achieve the tidal volume. The delivered pressure indicated on the ventilator screen informs on the ventilator action but does not account for the patient contribution, leading to a risk of underestimation of the actual transpulmonary pressure. The use of continuous measures of esophageal pressure, as a reflection of the pleural pressure, can probably help to estimate the net total transpulmonary pressure shared between the patient and the ventilator. The utility of this monitoring is currently under evaluation. More details on esophageal pressure monitoring are provided in the monitoring section below. The patient ventilatory drive can also be estimated using the continuous recording of diaphragm electrical activity (Edi). Using a specific nasogastric tube equipped with an array of microelectrodes, some ventilators can continuously provide the level of Edi. The Edi reflects the patient ventilatory drive (not the mechanical action of the diaphragm) and is closely related to the output of the brain stem respiratory center. Edi monitoring has been primarily developed to control the ventilator, with the mode NAVA (neurally adjusted ventilatory assist). During NAVA, the ventilator support is triggered and proportional to the Edi signal. Beyond NAVA ventilation, the Edi monitoring can also be used to assess the patient ventilatory drive. Edi monitoring has demonstrated that children spend a large proportion of time with no spontaneous activity during conventional ventilation, even when they are thought to be spontaneously breathing by the clinical team.

3.10 Weaning the Mechanical Ventilation and Extubation Readiness Test

The course of MV begins with intubation and the provision of ventilator support. As the acute phase of the disease subsides, weaning begins which may take up 50% or more of the total time on assisted ventilation. The end of weaning is defined as the time at which the patient’s spontaneous breathing alone can provide effective gas exchange. How this point can best be determined is unclear. At the end of weaning is extubation or the act of removal of the endotracheal tube (ETT). An extensive review on weaning and extubation readiness in pediatric patients was published in 2009. It described that there were many myths, unique practices, little consensus, and less objectivity surrounding these important pediatric critical care activities.

The length of weaning depends on a number of factors, among them fluid status. When total body water increases, lung compliance decreases due to increased lung water, chest wall, and diaphragm edema. In adults with acute respiratory distress syndrome (ARDS), patients managed with a conservative fluid regime had fewer MV days and a quicker return of normal lung function than those receiving a more liberal regime. The importance of fluid balance in children is not as clear although observational data in pediatric acute respiratory distress syndrome (PARDS) is also supportive of a conservative fluid approach. Positive end-expiratory pressure (PEEP) management is another factor that may affect the length of weaning. While institution and escalation of PEEP generally improves oxygenation in PARDS, pediatric practitioners are generally conservative in its application and infrequently change PEEP levels even after oxygenation improves. This may result in failure to recognize that the patient is actually ready for extubation. It is generally recommended that PEEP levels should be physiologic at the time of extubation (≤ 5 cm H20) although extubation from higher PEEP levels may be important to maintain lung recruitment for certain types of patients (e.g., obesity). Sedation further complicates weaning and extubation. Oversedation may depress central respiratory drive, whereas undersedation can leave a child restless which can result in airway trauma from the ETT. Sedation assessment tools have been developed for this purpose which may be helpful to target a particular level of sedation, although implementation of a nurse-driven protocol to achieve these targets does not appear to have a significant impact on weaning times. Differences in diaphragmatic function may relate to longer weaning times in infants and young children. Accessory respiratory muscles are not as developed as in older children. As diaphragmatic dysfunction develops with prolonged MV, the duration of weaning can increase. Steroids may play a role in preventing extubation failure and reintubation by reducing tracheal inflammation associated with tracheal injuries from the ETT, as they do in another cause of subglottic edema in children, croup. Successful randomized controlled trials in both adults and children have started steroids 6–24 hours before extubation, whereas the unsuccessful ones have started the drug under 6 hours before extubation. A Cochrane Review on the role of steroids concludes, “Using corticosteroids to prevent (or treat) stridor after extubation has not proven effective for neonates, children or adults. However, given the consistent trend toward benefit, this intervention does merit further study.” Finally, other factors are likely important to the weaning process, but there is a dearth of research in these areas, and they are not further discussed. These include cardiac function, postoperative, neurologic, and nutritional status.

Predictive Indices for Weaning . Several indices have been developed to predict success in weaning and extubation. Although these indices have been variably used in research, they have not found common use in clinical care, some because of their complexity and others due to lack of proven benefit over clinical judgment in pediatric practice.

Rapid Shallow Breathing Index (RSBI = f / V_T). The RSBI was devised by Yang and Tobin and found to be a good discriminator of weaning success and failure in adults. This test has become more widely used in adult practice and research with varying success. Some investigators have demonstrated that RSBI normalized to actual body weight (f/V_T (ml/kg)) has some predictive ability in pediatrics, although these values are often very low by the time extubation readiness testing occurs, reinforcing the concept that these should be used early in the course of weaning. Both an RSBI <8 or <11 breaths/min/mL/kg body weight have been promoted as good predictors of successful extubation.

Compliance, Respiratory Rate, Oxygenation, Pressure Index (CROP Index) [Dynamic Compliance × (PaO₂/PAO₂) × Maximal Negative Inspiratory Pressure) / Respiratory Rate]. In pediatrics, a CROP Index of >0.15 or >0.1 mL/kg body weight/breaths/min have been recommended as good predictors of successful extubation.

For both RSBI and CROP, as with adults, conflicting studies found that those indices did not reliably predict extubation outcome in children. The RSBI has become moderately popular in adult ICUs. However, since there is a wide range of age groups with different respiratory rates, it may not be a good predictor of extubation success or failure in the pediatric population. Whether age-specific f/V_T ratios would perform better is currently unknown.

Volumetric Capnography. The slope of an expired, single-breath CO2 waveform can be used to calculate the physiologic dead space (V_D/V_T) and a value <0.50 reliably predicted extubation success, whereas a V_D/V_T > 0.65 identified patients at risk for failure. Not all investigators have been able to reproduce this. Volumetric capnography requires an arterial or a capillary blood gas, and the predictive ability may depend on the types of patients studied, degree of parenchymal lung disease, and tidal volume generated.

3.10.1 Techniques of Weaning

Not all patients require gradual weaning. Both adult and pediatric studies have shown that when patients pass a spontaneous breathing test (SBT) and are subjected to an extubation readiness test (ERT), 50–75% of the patients are deemed ready to extubate and will do so successfully. Standardized weaning protocols have been promoted to minimize the time on a ventilator and provide uniform decisions about weaning. The concept of ventilator-free days as an end point is implicitly based on having a low failed extubation rate for any reason other than the original cause of respiratory failure. This standard may be inappropriate in pediatric trials since there is not only a higher rate of failed extubation in this group, but up to 40% of failures may involve upper airway obstruction (UAO). For research purposes, it may be important to define the end of successful weaning in a manner short of extubation. Overall, it is likely that a consistent approach to ventilator weaning will shorten ventilator time and result in better outcomes.

The most common approach to weaning infants and children is gradual reduction of ventilator support. Weaning with synchronized IMV (SIMV) occurs by reducing the ventilator rate. With pressure support ventilation (PSV alone or in combination with SIMV), the inspiratory pressure is initially set to provide the required support and then reduced gradually. As mentioned above, volume support ventilation (VSV) is a special form of PSV that targets a specified minimal tidal volume. Weaning with VSV is somewhat automatic in that the support level required to maintain the set tidal volume is reduced automatically as respiratory mechanics improve. Its disadvantage is it may allow the patient to work harder than desired. Extubation occurs from a low level of ventilator support or after an extubation readiness test (ERT). It appears that it is common practice to extubate infants and children from a low level of ventilator support though there is no rationale for this.

A second school of thought recommends to continuously provide the ventilator support to rest the patient’s respiratory muscles and to perform a daily extubation readiness test. MV is discontinued if the test is passed. This approach has been more commonly used to wean adult patients than children.

In a small number of patients, weaning is attempted with alternating periods of complete ventilator support and graded spontaneous breathing with assistance. This “sprinting” is performed on the theory that the respiratory muscles can be slowly trained to sustain complete spontaneous breathing. There is currently little evidence that such an approach is an effective way of training muscles. There are also no data comparing such an approach with more traditional approaches of weaning. A multicenter randomized controlled trial comparing three modes of weaning found that there were no significant differences between having no protocol and weaning by PSV or VSV.

3.10.2 Criteria for Readiness for Extubation

Readiness for extubation implies that weaning is completed and the patient is sufficiently awake with intact airway reflexes, is hemodynamically stable, and has manageable secretions. Extubation failure has been variably defined as reintubation within 24–72 hours. Tests commonly used to assess extubation readiness include testing for a leak around the ETT (“leak test”) and assessing respiratory muscle strength by measuring negative inspiratory force (NIF).

Cuffed ETTs have been cited as a cause of increased UAO on extubation, but several studies have found no difference in the incidence of failed extubation over all age groups between those intubated with appropriately sized cuffed or uncuffed tubes. The leak test whereby air is heard (without using a stethoscope) to leak around the ETT at low pressure, usually <20–25 cm H₂O, is commonly used to predict UAO after extubation. However, it is neither a sensitive nor specific test. For cuffed ETTs, cuff leak fraction is calculated as (expiratory tidal volume with cuff inflated minus expiratory tidal volume with cuff deflated)/(expiratory tidal volume with cuff inflated). Evaluating uncuffed ETTs, leak percentage is calculated as the fraction (inspiratory tidal volume minus expiratory tidal volume)/(inspiratory tidal volume). In a study of 409 infants and children immediately prior to their extubation, it was found that a cuff leak fraction less than 10% or a leak pressure (with the cuff deflated) greater than 25 cm H₂O was highly associated with risk of UAO following extubation. The presence or absence of a leak was not associated with UAO for uncuffed ETT.

Negative Inspiratory Force (NIF). In the PICU, the test is usually performed quickly at the bedside with an uncalibrated manometer and with both inspiration and exhalation obstructed. The test has not been hitherto standardized nor validated in children. Nonetheless, it is reassuring if a spontaneously breathing patient has a routinely obtained NIF of at least 30 cm H₂O. Similarly, consistently low values (i.e., below 15 cm H₂O), irrespective of technique, are unlikely to be associated with successful weaning to extubation.

3.10.3 Impact of Endotracheal Tubes on Weaning and Spontaneous Breathing Trials

Many clinicians believe that, for an infant or young child, breathing through a small ETT is akin to breathing through a straw, thereby imposing an unacceptable work of breathing. It is therefore a frequent practice to extubate children from levels of 5–10 cm H₂O PSV above PEEP in order to overcome the presumed increased effort of breathing. This notion is contrary to both clinical observation and physiology. In a large prospective trial, the pressure rate product (PRP) was measured in infants and children during preparation for extubation. PRP has been validated as a surrogate for work of breathing and is defined as respiratory rate times peak-to-trough esophageal pressure. The PRP was recorded on PSV and CPAP prior to and after extubation. No matter how small the ETT (down to 3.0 mm ID), effort of breathing on CPAP of 5 cm H₂O best estimated post-extubation work. Furthermore, the PRP on PSV was almost half the level after extubation, significantly underestimating the post-extubation effort of breathing. The reasonable conclusion from this study is that, all things being equal, if a patient cannot breathe comfortably on CPAP alone, there is little chance he or she will do so when extubated.

A further benefit of PRP is that when evaluated along with the maximum NIF, the combination of the former being >500 and the latter ≤30 cm H₂O predicts high reintubation rates of >20%. These values reflect diminished respiratory muscle strength (aPImax) in the face of high effort of breathing (PRP).

3.10.4 Assessment of Post-extubation UAO

Upper airway obstruction (UAO) is frequent after endotracheal extubation. Definitive data on risk factors and prevention of pediatric post-extubation UAO have been lacking. More objective measures of post-extubation UAO severity in infants and children may help identify risk factors and elucidate optimal treatment or prevention strategies. Inspiratory flow limitation is relatively specific to extrathoracic UAO, characterized by disproportionately large inspiratory effort relative to flow. The most widely accepted method to measure flow is spirometry, which, for noncooperative spontaneously breathing children, requires a tight-fitting mask over the nose and mouth. This may require sedation and results in changed flow dynamics. Respiratory inductance plethysmography (RIP) is a less invasive alternative to spirometry. With RIP, variations in the self-inductance of a coil (wires around the rib cage (RC) and abdomen (ABD)) are measured as a result of changes in the cross-sectional area of the RC and ABD. The combination of calibrated RIP and esophageal manometry has shown promise for providing both an objective measure of the severity of post-extubation UAO and also some insights into risk factors for UAO. With the addition of a pneumotachygraph on the ETT prior to extubation, RIP flow can be calibrated during 3–5 breaths of airway occlusion during an NIF procedure. This allows the construction of noninvasive flow-pressure loops from the RIP and esophageal pressure measurements (◘ Fig. 12.3).

Fig. 12.3

Flow-pressure loop displaying the evolution of flow, obtained from calibrated RIP belts around the thorax and abdomen, and esophageal pressure, obtained using an esophageal balloon catheter

These loops can be inspected for inspiratory flow limitation after extubation to characterize post-extubation UAO. Patients can be observed as having UAO when inspiratory flow limitation is newly observed after extubation with an increase in PRP (◘ Fig. 12.4, panel A). The reversal of inspiratory flow limitation can also be seen after the child receives a UAO-specific intervention such as racemic epinephrine, heliox, or corticosteroids. UAO can be further classified as subglottic if a jaw thrust maneuver does not reduce PRP by at least 50%.

Fig. 12.4

Panel A illustrates the evolution of the flow-pressure loop as recorded in ◘ Fig. 12.3 in a 6-month-old infant before (left) and after extubation (right). The right panel shows inspiratory flow limitation after extubation due to upper airway obstruction, as denoted by the flattened flow-pressure limb above the X-axis and large negative change in esophageal pressure. The pressure rate product also increased tenfold. Panel B illustrates the evolution of inspiratory flow limitation in an infant with UAO before (left) and after racemic epinephrine (right). The narrowing of the loops illustrates the improvement in UAO

3.11 Role of Automation and Clinical Decision Support System

While the basic principles of lung protective ventilation have been embraced by pediatric intensive care physicians, there is still great variability in ventilator management. Most critical care practitioners believe they are being lung protective, but it is likely that consistent, replicable decisions are not made to minimize ventilator support across the duration of mechanical ventilation for patients with lung injury. Since respiratory support is required in the majority of children in the PICU and because complications may occur with its use, it is essential to develop strategies to improve patient outcome and reduce medical errors related to mechanical ventilation. The observed variation in clinical practice is likely due in part to clinicians’ low adherence to guidelines, and this is compounded because many facets of caring for a mechanically ventilated patient in the PICU lack high levels of evidence or evidence is conflicting. There is good evidence, however, that clinical decision-making with a protocol decreases practice variation between clinicians, standardizes patient care, and improves research and patient outcomes. Replicable ventilator management decisions should help decrease practice variability and directly shorten the length of mechanical ventilation for children in pediatric ICUs. A Cochrane systematic review assessing the impact of automated versus nonautomated weaning for critically ill adults and children identified 21 trials, and the meta-analysis suggested a reduction in ventilation duration and ICU length of stay, although a great heterogeneity was observed, and only two trials were conducted in pediatric ICU.

Decision support tools, both paper and electronic, have been demonstrated to improve medical care, reduce errors, and improve patient outcomes. However, paper-based protocols are dependent on caregiver availability and are often written in broad terms so they remain dependent on clinician judgment and local context for interpretation and therefore are difficult to transfer from one PICU or NICU to another intensive care unit. Paper-based tools can be difficult to follow accurately, leading to low adherence rates. Computer decision support (CDS) tools like computer-based protocols are an automated method to reduce medical errors. Such tools aim to ensure replicable, evidence-based clinician decisions for equivalent patient states and to improve protocol compliance. The tools can generate explicit recommendations that can be carried out with little inter-clinician variability while still remaining responsive to the patient’s unique situation. The tools can assist clinicians by standardizing descriptors and procedures, by consistently performing calculations, by incorporating complex rules with patient data, and by capturing data relevant to decision-making. Computer-based protocols can contain more extensive detail than textual guidelines or paper-based flow diagrams while at the same time protecting the user from complexity and information overload.

Decision support tools vary in terms of how dynamic they are, the degree of specificity of their recommendations, and the level of integration into workflow. One end of the spectrum includes general guidelines that consist of a set of broad, static recommendations. At the other end of the spectrum are computerized protocols, which function as a set of standardized orders, with detailed explicit instructions based on dynamic patient-specific parameters, available at the point of care. The latter type of protocol has been called an “explicit computerized protocol” (ECP). A few ventilators have incorporated closed-loop algorithms to automatically adjust the ventilator settings depending on the patient condition and trajectory. These specific modes have been mostly developed for the weaning phase of the ventilation (e.g., SmartCare system), and more recent tools aim to manage the complete ventilation process from the acute to the weaning phase (e.g., Intellivent system).

The systems need to be fed with clinical information and ventilator data. The first parameter to be defined is a surrogate of the child’s lung volume. Actual body weight is not accurate enough as children may be malnourished or obese. The PALICC guidelines recommend using predicted body weight. Height is a better alternative in most cases except when neuromuscular weakness or spinal deformity is present. In such circumstances, ulna length is an excellent surrogate. Ventilator input data may include set ventilator parameters and measured ventilator parameters (i.e., spontaneous respiratory rate, expiratory tidal volume and air leak around the endotracheal tube, dynamic compliance, resistance). CO₂ removal is assessed intermittently by arterial or capillary PCO₂ on blood gas, continuously with end-tidal CO₂ (PETCO₂) and transcutaneous PCO₂ and also indirectly by spontaneous respiratory rate and tidal volume in some assist modes when respiratory control is functioning properly. Oxygenation can be assessed intermittently by PaO₂ on blood gases and continuously by SpO₂. For real-time adjustment of ventilation, ECPs are currently using SpO₂ and end-tidal CO₂ in addition to respiratory rate and tidal volume or predicting PaO₂/FiO₂ ratio from SpO₂/FiO₂ ratio and pH from a previous arterial to end-tidal CO₂ difference utilizing the Henderson-Hasselbalch equation.

Barriers to protocol use are considerable including lack of awareness, lack of familiarity with the protocol, lack of agreement, lack of efficacy, lack of known improved outcome, and lack of ability to overcome the inertia of previous practice. There are also external barriers: protocol-related barriers (i.e., not easy to use, not convenient, cumbersome, confusing) and environment-related barriers (e.g., new resources or facilities not accessible). To ensure acceptance, users must feel that they can count on the system to be available whenever they need it. The amount of downtime needed for data backup, troubleshooting, and upgrading should be minimal. The response time must be fast, data integrity must be maintained, and data redundancy must be minimized. It is also important to assess the amount of training necessary for users to feel comfortable with the system. The ventilator market itself is also a barrier to the implementation of ECPs. There are numerous companies, and the competition between them results in difficulties implementing the same ECP in different ventilators.

3.12 Monitoring of the Mechanical Ventilation

3.12.1 Blood Gas

While arterial blood gas (ABG) and capillary blood gas (CBG) measurements are the standards for supportive management in pediatric intensive care units, frequent use of SpO₂ (and end-tidal CO₂) requires incorporating noninvasive technologies into decision-making and scoring systems. Given the challenges to placing arterial catheters in small children in daily care, practitioners frequently make decisions for children based on oxygen saturations obtained from pulse oximetry. Pulse oximeters are developed to perform optimally in a range of oxyhemoglobin saturation from 70% to 100%. It is not always clear how well pulse oximeters perform when the majority of observations are in infants and children in a hypoxemic range. Nonetheless, in practice, pulse oximetry in the low range is now routinely used for clinical decision-making for children with cyanotic congenital heart disease.

ABG-based measures of hypoxemia such as PaO₂/FiO₂ (PF) ratio have been used in ICU severity of illness scores as diagnostic criteria for acute respiratory distress syndrome (ARDS) and for scoring of lung injury. It is now well demonstrated that noninvasive SpO₂-based oxygenation saturation index OSI (mean airway pressure × FiO₂/SpO₂ × 100) is an adequate surrogate marker for the ABG-based oxygenation index OI (mean airway pressure × FiO₂/PaO₂ × 100), as long as SpO₂ is between 80% and 97%. Similarly, the S/F ratios of 264 and 221 correspond well with the P/F ratio cutoff values of 300 and 200, respectively (◘ Fig. 12.5). While it is clear that there is considerable scatter of SpO₂ values corresponding to a PaO₂ value, the accuracy of these measurements can be improved if targets of SpO₂ are consistently between 88% and 95%.

Fig. 12.5

Association between the PaO₂/FiO₂ (PF) ratio and the SpO₂/FiO₂ (SF) ratio in a cohort of children with hypoxemic respiratory failure and a SpO₂ between 80% and 97%

The most widely used noninvasive sensor to estimate adequacy of ventilation is end-tidal carbon dioxide pressure (PETCO₂). However, the relationship between PETCO₂ and PaCO₂ changes as a function of alveolar dead space. Additionally, estimating pH from PETCO₂ is confounded by changing metabolic acidosis with illness frequently compounded by the use of diuretics. Nonetheless, incorporating noninvasive measures of intrapulmonary shunt, noninvasive continuously available values from the ventilator and previously known values for pH and PCO₂ from an ABG or CBG, allows accurate pH prediction for up to 12 hours in a relatively stable patient, as illustrated in ◘ Fig. 12.6. The advantage of continuous pH projection is that it permits more frequent (lung protective) ventilator changes than from intermittent blood gases. Similarly, a model has been developed to continuously estimate the PaCO₂ based on the PETCO₂, mean airway pressure, FiO2, and the capnographic index (KPIv) derived from volumetric capnography, using the following equation:

Fig. 12.6

pH prediction utilizing the Henderson-Hasselbalch equation, based on the previous pH and the PCO₂-PETCO₂ difference

$$ {\displaystyle \begin{array}{l}{\mathrm{Predicted}\ \mathrm{PaCO}}_2\ \left(\mathrm{mmHg}\right)=0.859+0.827\times {\mathrm{PETCO}}_2\ \left(\mathrm{mmHg}\right)+0.310\\ {}\times \mathrm{MAP}\ \left(\mathrm{cm}\ {\mathrm{H}}_2\mathrm{O}\right)+0.081\times {\mathrm{FiO}}_2\ \left(\%\right)+0.529\times \mathrm{KPIv}\end{array}} $$

This model is shown to achieve a predicted PaCO₂ correct at ±5 mm Hg in 95% of measurements.

3.12.2 Capnography

Capnography describes the continuous display of carbon dioxide partial pressure in the expired breath (PETCO₂). Capnography is noninvasive, accurate, easy to do, and relatively inexpensive and has been studied extensively. Capnography can be measured with the mainstream technique (i.e., sensor placed in line between the proximal end of ETT and ventilator circuit Y-piece) or with the sidestream technique (exhaled gases are aspirated from near the proximal end of the endotracheal tube and analyzed at a sensor nearby). There are two distinct types of capnography. Conventional, time-based capnography allows only qualitative and semiquantitative measurements. Volumetric (volume-based) capnography has emerged as the preferred method to assess the quality and quantity of ventilation. Volumetric capnography combines the measurement of exhaled CO₂ with continuous flow (and hence volume) measurement to provide a continuous measure and visual display of the quantity of CO₂ eliminated and how it relates to expired breath volume. These measurements can be used to estimate anatomic and physiologic dead spaces and ventilation effectiveness, while with both methods, pattern recognition of displayed waveforms can be used to identify pathophysiological problems, as detailed in ◘ Fig. 12.7. In the figure, the area X represents the actual volume of CO₂ exhaled at one breath. Adding all single breaths at 1 minute gives the total elimination of CO₂ per minute (V′CO₂). Area Y represents the amount of CO₂ that is not eliminated due to alveolar dead space. The latter is increased in cases of lung overdistention, pulmonary embolism, pulmonary hypertension, and decreased cardiac output. It decreases when these conditions are ameliorated. An elevation of the baseline during Phase I indicates rebreathing of CO₂, which may be due to mechanical problems with the ventilator, the need for recalibration of the CO₂ sensor, or incomplete reversal of neuromuscular blockade while breathing spontaneously.

Fig. 12.7

Volumetric capnography typical tracing and identification of the different slopes and corresponding physiological variables, where (1) signifies the slope of Phase III, (2) signifies slope of Phase II, (3) the intersection of lines 1 and 2 define the limit between Phases II and III, and (4) a perpendicular line is projected onto the X-axis, and its position is adjusted until the areas p and q on both sides become equal

In pediatric ARDS , the ventilation-perfusion ratio is disturbed, and both types of capnography have been associated with mortality prediction, particularly when added to oxygenation metrics. Using time-based capnography, the alveolar dead space will be increased, but the slope of Phase III is usually almost flat. More distinct changes are shown by the volumetric capnography curve. Here, Phase I may be larger due to increased anatomic dead space caused by PEEP. The slope of Phase II is decreased due to lung perfusion abnormalities, and the slope of Phase III is increased due to lung heterogeneity (◘ Fig. 12.8).

Fig. 12.8

Schematic representation of time-based capnography (left panel) and volumetric capnography (right panel) in patients with pediatric acute respiratory distress syndrome (PARDS)

3.13 Pleural Pressure Monitoring

As mentioned above (sections on setting inspiratory pressure and PEEP), an estimate of pleural pressure has many theoretical implications during mechanical ventilation. Unfortunately, pleural pressure cannot be measured in routine clinical practice, so the most common surrogate relies upon an esophageal pressure. In general, the esophageal pressure is measured in the lower 2/3 of the esophagus and presents a global estimate of the pleural pressure in the thorax. In truth, regional pressure may be considerably different since during normal physiologic conditions, the pleural pressure differs from apex to base of the lung or from ventral to dorsal regions when the patient is lying supine. The pleural pressure estimate may be useful for setting the PEEP, setting the delta pressure, and estimating patient effort and degree of spontaneous ventilation. It may also be useful to quantify patient-ventilator asynchrony and measure intrinsic PEEP and a variety of other applications.

During fully controlled ventilation (i.e., no spontaneous effort), the pleural pressure is used to estimate transpulmonary pressure at end expiration (to set PEEP) and at end inspiration (to set tidal volume or delta pressure). At end expiration, when there is no airflow (expiratory hold), the pressure in the pleural space gives an estimate of the pressure being exerted by the chest wall. Under normal conditions (not mechanical ventilation), the pleural pressure is slightly more negative than the alveolar pressure (i.e., −3 to −5 with alveolar pressure = 0). This results in a slightly positive transpulmonary pressure (Palv-Ppl of +3 to +5) at end expiration, which promotes the alveoli staying open, above the critical closing capacity. If the pleural pressure is increased (i.e., obesity, chest wall edema), then the transpulmonary pressure at end exhalation will become negative, without applying PEEP. This may promote alveolar collapse and can exacerbate atelectrauma as a mechanism of lung injury, particularly for patients with ARDS. Hence, an area of active investigation is to determine whether using esophageal manometry (generally with esophageal balloon catheters) as an estimate of pleural pressure can be used to select optimal PEEP levels in patients with ARDS. A Phase II study has shown some promise in this approach for adults with ARDS compared to the ARDSnet low PEEP/FiO₂ grid. A larger Phase III study recently studied the titration of PEEP with esophageal pressure compared to the ARDSnet high PEEP/FiO₂ grid, in adults with severe ARDS. In this study, there was no difference. While this approach likely has applicability in pediatrics, there are additional limitations with the accuracy of esophageal balloon catheters in small children, which would be important to overcome before widespread use.

The esophageal pressure can be used at end inspiration (during an inspiratory hold) to estimate the transpulmonary driving pressure and transpulmonary pressure at plateau. These variables may be crucially important when considering VILI from barotrauma. This is because under conditions of significant chest wall restriction, much of the inspiratory pressure is dissipated across the chest wall. Having a more precise estimate of the transpulmonary pressure may allow for slightly higher plateau and driving pressures to be applied, in circumstances where the chest wall compliance is poor. This remains an active area of investigation, with limited current evidence to suggest a particular target for this transpulmonary pressure. Similar issues with the accuracy of balloon catheters in pediatrics need to be overcome before this approach can be considered for widespread use in children.

Applications related to selection of PEEP and driving pressure using esophageal pressure are highly sensitive to the accuracy of the balloon catheter because an accurate assessment of the actual pleural pressure is needed to set the airway pressure. In contrast, applications related to effort or work of breathing are more robust in pediatrics using esophageal manometry because they simply use the change in esophageal pressure rather than the absolute value for applications. Classically, esophageal manometry has been combined with measures of tidal volume, airway pressure, and flow to measure work of breathing. This work of breathing is characterized by the force needed (i.e., change in pressure, both airway and esophageal) over a distance (i.e., the tidal volume). This work is shared between the patient and the ventilator, with an algorithm used to divide the contribution coming from the patient with what is coming from the ventilator, based on the Campbell diagram. This has advantages of understanding the overall stress on the respiratory system (in this case, shared between patient and ventilator). However, oftentimes, the patient’s effort while on mechanical ventilation is the more important clinical variable to consider when adjusting the ventilator settings. This is frequently estimated with the pressure time product (PTP), given by the area under the esophageal pressure tracing during inspiration times the respiratory rate, where inspiration is typically defined using spirometry to characterize when airflow crosses zero. Therefore, this requires the use of an additional sensor (other than esophageal manometry) to estimate patient effort of breathing. To overcome this limitation, some investigators use the pressure rate product (PRP), which is defined as the peak-to-trough change in esophageal pressure during a breath multiplied by the respiratory rate. This is simpler to calculate and does not require an additional sensor. While these metrics generally provide similar results under conditions of increased respiratory load, the circumstances of when to use one versus the other remains an area of investigation.

3.14 Chest Radiography

A chest radiography is generally obtained after tracheal intubation and, in most situations of clinical deterioration, to detect complications or the displacement of tubes and catheters. It is also required to establish the diagnosis of PARDS, even if its interpretation is difficult with large interobserver variability. The optimal frequency of radiograph control in mechanically ventilated children is not established. Some studies conducted in adult ICUs suggest that an on-demand strategy can safely replace the routine daily radiograph strategy. It is not certain if this approach may be completely transposable to children with acute respiratory failure, in whom half of routine chest X-rays are followed by an intervention. However, routine radiography could probably be avoided when the risk of intervention is low, in particular when the number of devices is low, and in stable and older children.

3.15 Diaphragm Ultrasound

In recent years, diaphragm ultrasound has become an increasingly utilized modality in research and clinical practice to quantify patient effort of breathing as well as identify architectural changes in the diaphragm which may signify ventilator-induced diaphragm dysfunction. There are a variety of approaches which advocate slightly different methods for the orientation of the ultrasound probe and methods for measurement. The general concept is to measure the diaphragm in the zone of apposition, where it is running near vertically along the chest wall. Serial measurement of the diaphragm thickness during exhalation can be used to identify atrophy of the muscle throughout the course of ventilation and is thought to be an indirect surrogate for VIDD. Further investigation is needed to see if these architectural changes truly correspond to direct measures of respiratory muscle strength (such as airway or esophageal pressures during airway occlusion (i.e., aPiMax or ePiMax). In addition, similar measurement of the diaphragm thickness and how this varies during inspiration compared to exhalation can be used to estimate a thickening fraction (or diaphragm contractile activity). Very low thickening fractions have been described as common during controlled ventilation and may be implicated in thinning of the diaphragm during MV. In contrast, very high thickening fractions may be present when work of breathing is high and has been associated with hypertrophy of the diaphragm muscle, which also may have implications on overall diaphragm function. As of now, these tools fit largely in the research domain in pediatric mechanical ventilation.

3.16 Diaphragm Electrical Activity Monitoring

The monitoring of electrical activity of the diaphragm (Edi) is a new, minimally invasive, bedside technology that was primarily developed for ventilation mode NAVA. In addition to its role in NAVA ventilation, this technology provides the clinician with previously unavailable and essential information on diaphragm activity. The Edi signal is acquired using a specific nasogastric catheter equipped with multiple microelectrodes, connected to a specific ventilator, in which a signal processing algorithm allows the detection of the diaphragm position and the construction of an optimal Edi signal not affected by the diaphragm movement. The Edi is a rapid electrical signal, which closely reflects the output of the phrenic nerves. The quantification of the Edi allows easy detection of overassistance or oversedation with absent ventilator drive. This situation occurs frequently during conventional ventilation. Approximately one-third of children had no or very low diaphragm activity in a recent study. Increased inspiratory Edi levels can also suggest insufficient support, while a strong tonic activity may reflect the patient’s efforts to increase lung volume. Edi monitoring also allows the detection and quantification of patient-ventilator asynchrony, which occurs during more than a quarter of the time during pediatric conventional ventilation.

Importantly, Edi reflects the ventilatory drive (i.e., patient “demand”). When correlated to the resultant mechanical action (e.g., the inspiratory force generated by the diaphragm contraction, with no ventilator support), one can obtain an index of the neuromechanical efficiency (NME) of the diaphragm.

$$ \mathrm{NME}=\mathrm{negative}\ \mathrm{inspiratory}\ \mathrm{pressure}/\mathrm{inspiratory}\ \mathrm{Edi} $$

This index has been shown as an interesting predictor of extubation failure in adults, but pediatric studies are lacking. ◘ Figure 12.9 provides an illustrative example of the evolution of the neuromechanical efficiency in an infant with a neuromuscular disorder. Further studies are warranted to demonstrate the impact of this monitoring on important clinical outcomes.

Fig. 12.9

Illustration of the neuromechanical efficiency assessed using diaphragm electrical activity (Edi) and esophageal pressure (Peso) monitoring in an infant with congenital myotonic dystrophy. Panel A illustrates this infant at 3 weeks post-term, with low esophageal pressure swings and relatively high Edi, reflecting a poor neuromechanical efficiency. Panel B shows the same infant 3 weeks later, after a respiratory weaning in NAVA ventilation. Peso swings were larger for relatively similar Edi, suggesting an improved efficiency of the diaphragm

4 Summary

The ability to implement and provide mechanical support to children with respiratory failure may be the single most important maneuver ever added to the ICU armamentarium. Although this technology started with the introduction of endotracheal intubation in the operating room, the more recent advances have focused on the design and function of modern mechanical ventilators which can deliver complex ventilator modes and allow precise adjustments. However, a large part of the management of pediatric mechanical ventilation is not based on strong evidence, more of which is needed. In that context, it is essential for the PICU care physician to understand the functioning and the impact of mechanical ventilation, as well as the main pathophysiological concepts leading to ventilator-induced lung injury and diaphragm dysfunction. Future studies are necessary to better understand how to best titrate the ventilation to each patient-specific condition and needs and to guide the development of future ventilators which will probably include automatic assessment and adjustment of the ventilation provided.

Review Questions

• 1. A 14-year-old, 50 kg male develops post-obstructive pulmonary edema following a near-hanging. He develops profound hypoxia and requires endotracheal intubation and mechanical ventilation at the referring institution. He arrives ventilated in a volume control mode. His settings are FiO₂ 80%, rate 16, tidal volume 400 mL, inspiratory time 1 s, and PEEP 10 cm H₂0. The PIP is measured at 34 cm H₂0. He is lightly sedated but appears agitated. An arterial blood gas reveals pH 7.58, PaCO₂ 21 mmHg, and PaO₂ 98 mmHg. The following is the most likely cause of the hypocarbia:

  • A. High spontaneous rate causing excessive minute ventilation

  • B. High ventilator rate causing excessive minute ventilation

  • C. Large tidal volume causing excessive minute ventilation

  • D. Large tidal volume causing excessive PIP

  • E. Short inspiratory time allowing prolonged expiration

• 2. Which of the following is most correct regarding the inspiratory time in mechanical ventilation?

  • A. Altering the inspiratory time in pressure-controlled ventilation can affect minute ventilation but does not affect mean airway pressure.

  • B. An inspiratory pause added at end expiration does not add to the total inspiratory time.

  • C. Lengthening the inspiratory time in volume-controlled ventilation can decrease the inspiratory flow and thus PIP.

  • D. Longer inspiratory times help recruit alveoli with short time constants.

  • E. Pressure support ventilation requires the inspiratory time be preset depending on the degree of airflow resistance or noncompliance.

• 3. A 9-year-old, 30 kg girl with severe influenza infection develops ARDS and progressive hypoxia. Current ventilator settings are SIMV-volume, FiO₂ 100%, rate 16, tidal volume 200 mL, inspiratory time 1.4 s, and PEEP 10 cm H₂O. The PIP is measured at 40 cm H₂O, and the plateau pressure is 34 cm H₂O. Arterial blood gas reveals pH 7.28, PaCO₂ 51 mmHg, and PaO₂ 48 mmHg. Appropriate application of PEEP is best demonstrated by:

  • A. Improved oxygenation with decreased PIP

  • B. Improved oxygenation with increased PIP

  • C. Improved oxygenation with decreased plateau pressure

  • D. Pressure volume curve demonstrating a decrease in delta volume with no change in delta pressure

  • E. Pressure volume curve demonstrating an increase in delta volume with increased delta pressure

• 4. Pressure support ventilation is a mode that is:

  • A. Patient triggered (flow or pressure), flow limited, and pressure cycled

  • B. Patient triggered (flow or pressure), pressure limited, and flow cycled

  • C. Patient triggered (flow or pressure), time limited, and time cycled

  • D. Patient triggered (flow or pressure), volume limited, and flow cycled

  • E. Time triggered, pressure limited, and flow cycled

• 5. A 16-year-old with chronic renal disease and long-standing hypertension presents with respiratory distress and hypertensive crisis. Admission vital signs are temperature 38.7, heart rate 108, respiratory rate 44, blood pressure 189/102, and oxygen saturation of 88% while breathing 100% FiO₂. He appears mottled with cool extremities and poor pulses. His chest radiograph shows bilateral patchy infiltrates and cardiomegaly. He rapidly progresses to respiratory failure and requires endotracheal intubation and positive pressure ventilation. Thirty minutes after intubation, his vitals are heart rate 90, ventilator rate 12, spontaneous rate 20, blood pressure 159/92, and oxygen saturation of 98% while on 60% FiO₂. He has improved color and has easily palpable pulses. He is comfortable with intermittent benzodiazepine sedation. Arterial lactate has declined to 3.5 mMol/L from 5.7 at admission. The most likely explanation for his hemodynamic improvement is:

  • A. Improved oxygenation due to the application of positive pressure

  • B. Positive pressure reducing afterload to the right heart

  • C. Positive pressure reducing afterload to the left heart

  • D. Positive pressure reducing preload to the left heart

  • E. Positive pressure reducing preload to the volume overloaded right heart

• 6. A 2-year-old, 12 kg female develops respiratory failure due to pneumococcal pneumonia and right empyema. She undergoes video-assisted thoracoscopic surgery and has a patent draining right chest tube. She is ventilated with SIMV-volume, FiO₂ 40%, rate 16, tidal volume 90 mL, inspiratory time 1.2 s, and PEEP 6 cm H₂O. The PIP is measured at 24 cm H₂O, and the plateau pressure is 20 cm H₂O. Her oxygen saturation is 98%, and she appears comfortable on a low-dose midazolam infusion and intermittent morphine. She develops acute hypoxia with oxygen saturation falling to 81%. Her lung exam is significant for equal but diminished breath sounds and no wheezing. The PIP is measured at 34 cm H₂O, and the plateau pressure is 23 cm H₂O. Her perfusion is adequate, and she is progressively agitated. The most appropriate initial intervention would be:

  • A. Increase FiO₂ to 100% and administer a bronchodilator

  • B. Increase FiO₂ to 100% and administer a neuromuscular blocker

  • C. Increase FiO₂ to 100% and increase PEEP to 10 cm H₂0

  • D. Increase FiO₂ to 100% and reduce the tidal volume to 6 ml/kg

  • E. Increase FiO₂ to 100% and suction the endotracheal tube

• 7. The optimal ventilator circuit has:

  • A. Low resistance, high compliance, and nebulizer system for humidification

  • B. Low resistance, high compliance, no servo-controlled heated wire system, but a rainout trap

  • C. Low resistance, low compliance, and extra spacers between the Y-piece and the patient for easy patient positioning

  • D. Low resistance, low compliance, and nebulizer system for humidification

  • E. Low resistance, low compliance, and servo-controlled heated wire humidification system

• 8. Which of the following maneuvers will not increase the mean airway pressure?

  • A. Adding an inspiratory pause

  • B. Increasing the inspiratory time

  • C. Increasing the PIP

  • D. Increasing the PEEP

  • E. Increasing the minute ventilation by increasing the respiratory rate at fixed I:E ratio

Answers

• 1. A

• 2. C

• 3. C

• 4. B

• 5. C

• 6. E

• 7. E

• 8. E