Chapter 102. Respiratory Distress and Failure

In healthy individuals, respiratory and circulatory functions are linked to tissue metabolic activity by a responsive regulatory system that translates biochemical and neural signals from the tissues into adjustments in cardiac output, vascular tone, and minute ventilation. The purpose of this system is to assure that all the cells in the organism receive a supply of O2 commensurate with their metabolic needs without accumulating excessive amounts of CO2. The system relies both on local circulatory reflexes, which alter the caliber of the supplying blood vessels in accordance with tissue metabolic activity, and on central circulatory and respiratory reflexes, which adjust the pumping function of the heart and the intensity of the respiratory effort in response to changes in the concentration of the respiratory gases in the blood.1

Chemoreceptor reflexes play a singularly important role in the genesis of the manifestations of respiratory disease (Fig. 102-1). Alterations in the blood’s PO2, PCO2, and pH are sensed by specialized chemo-sensitive cells located in the carotid bodies (peripheral chemoreceptors, PO2) and reticular nuclei of the medulla oblongata (central chemoreceptors, PCO2 and pH). These cells relay the information to a medullary neuronal network of premotoneurons, which also receives inputs from mechanical and chemical sensors distributed throughout the lungs, airways, and chest wall. Chemical and mechanical inputs are integrated to define the amount of ventilation needed to sustain adequate gas exchange and the manner in which the respiratory muscles are applied to achieve this ventilation most efficiently. The physiologic basis of respiration is reviewed in more detail in Chapter 503.

A variety of developmental factors render the infant, and especially the newborn at increased risk for difficulties with respiration. These include immaturity of the neural control of breathing, the small caliber of the airways, the increased flexibility of the rib cage, and a limited respiratory muscle reserve. Furthermore, infants may present with congenital anomalies that impair respiration. Thus, it is not surprising that respiratory dysfunction frequently always occurs when a child becomes critically ill, making it essential that clinicians who care for infants and children recognize the manifestations of respiratory disease. The emergency evaluation of any patient—child or adult—should always begin by assessing the adequacy and characteristics of the patient’s respiratory effort. The phrase respiratory distress denotes an abnormal increase in the effort of the respiratory muscles, typically to overcome an impairment in the mechanical function of the lungs or the chest wall. The increased effort may be noticeable only to the child, who perceives it as shortness of breath (dyspnea), or may also be apparent to an external observer as physical signs. Depending on the severity of the impairment, these usually include a high breathing frequency (tachypnea), nasal flaring (from visible activation of the normally inconspicuous alae nasae muscles), retractions of the intercostal spaces (from the negative swings in pleural pressure generated by the diaphragmatic contractions), and recruitment of accessory muscles (muscles whose primary function is not respiratory but that can help pump air in and out of the lungs). The abdominal muscles (rectus abdominis, oblique muscles, and transversus abdominis) are the most frequently used among the latter. They take advantage of their insertions on the costal cartilages and ribs to stabilize the abdominal wall during inspiration and to accelerate lung emptying during expiration. The scaleni and sternocleidomastoids are also accessory muscles, but, in contrast to the abdominal muscles, they act only to augment rib-cage volume through their distal insertions on the first and second rib and the sternal manubrium, respectively.

Increases in the production of CO2 (eg, when metabolic rate is increased by fever or exercise) or in the concentration of hydrogen ions in the blood (eg, metabolic acidosis) also augment neural output to the respiratory muscles, which increases respiratory effort and may mimic respiratory distress. In patients with neuropathy or myopathy signs of respiratory distress, can sometimes be manifested by dyspnea and activation of the respiratory muscles spared by the disease. In patients with severe forms of neuromuscular dysfunction or following the administration of neuromuscular blockers an increased respiratory effort may not be recognized since the respiratory muscles will fail to respond appropriately to the increased respiratory drive.

On occasion, either the neural mechanisms that regulate ventilation fail or, more frequently, the compensatory effort of the respiratory muscles is not sufficient to restore gas exchange to normal, and arterial hypoxemia (an abnormally low arterial PO2) and hypercapnia (an abnormally high PCO2) ensue. This situation, known as respiratory failure, is discussed later in this chapter. It may occur when the effort expended to sustain adequate ventilation cannot be maintained, or when the neural control of breathing is disabled. Respiratory failure is evident when abnormally low respiratory rate (bradypnea), shallow breathing movements, or no respiratory movements at all (apnea) are observed. The breathing control becomes disabled when the central nervous system, especially the brain stem, has suffered a direct injury (eg, cranial trauma, compression by an expanding cerebral tumor or hemorrhage) or is functionally impaired by hypoxia, hypercapnia, or depressants (eg, opioids or barbiturates).

During normal respiration forces generated during breathing originate from two types of physical phenomena: one relates to lung inflation; the other relates to expiration and gas flow.

During lung inflation volume-dependent forces arise from elements within the lungs, which oppose inflation and create a tendency for the lungs to recoil. The elastic fibers contained in the alveolar interstitium and in the airways and blood vessels of the lung are a good example. When stretched during inspiration, these fibers behave very much like a rubber band, accumulating energy in their molecular structures, which is then released as they recover their shape during expiration. The more the elastic fibers are stretched, the greater is their tendency to return to their original state or the greater is the overall recoil of the lung. The alveolar gas-liquid interface also acts as a recoil element. When a liquid like the water-based solution lining the alveoli contacts air, water molecules within the liquid phase experience a net push to leave the solution. This push translates into a net force, known as surface tension, which acts to reduce the volume of the alveolus. In the healthy lung, the tendency is relieved greatly by the presence of a lipid monolayer (alveolar surfactant), which separates the water and gas phases and makes the recoil generated from surface tension manageable. When disease or immaturity interferes with surfactant function (see Chapter 54), the increase in surface tension increases the lung recoil, often to levels that prevent adequate inflation. Following inflation, the lung recoils but this recoil is balanced by the outward recoil of the chest wall which distends the lung at rest (Fig. 102-2). Because lung inflation requires the active contraction of the inspiratory muscles (expiration is passive and therefore facilitated by recoil), the physical findings of the diseases in which lung or chest-wall recoil is increased are always more prominent during inspiration than during expiration.

Flow-dependent forces are the result of brief molecular interactions between either the gas flowing through the airways and the airway walls or among tissue components as the lungs change volume. The forces generated by these interactions oppose both inflation and deflation. Just like the velocity of a vehicle determines the fuel consumption incurred in overcoming road and air drag, the magnitude of these forces (often grouped as resistive forces) depends primarily on the speed (or flow rate) at which the lungs inflate or deflate.

Observation of the clinical manifestations of respiratory distress are often sufficient to establish a distinction between two types of disorders that impair respiration: restrictive and obstructive.

Restrictive Impairments

Restrictive impairments are characterized by an abnormal increase in lung or chest-wall recoil, and therefore they interfere primarily with lung inflation. Examples include pulmonary edema, pneumonitis, interstitial lung disease, and chest-wall deformities that limit lung and chest-wall expansion (see Section 27). In some of these conditions, the presence of fluid (pulmonary edema or pneumonitis), inflammatory cells (pneumonitis), or scar tissue (lung fibrosis) in the interstitial spaces diminishes the ease with which the lung scaffolding can accommodate stretch. In others, the presence of water (advanced pulmonary edema) or exudate (pneumonia) in the alveolar spaces raises surface tension at the gas-liquid interface and reduces the space available for gas in the alveoli. Chest-wall deformities limit lung inflation by making the rib cage more rigid, often asymmetrically. Space-occupying lesions such as pneumothoraces, pleural effusions, and cystic lung anomalies are by definition restrictive, although frequently they distort neighboring airways, thereby creating simultaneous obstructive manifestations.

Restrictive impairments increase the effort required to produce an increase in lung volume. This relationship is best represented as a pressure-volume graph, as seen in Figures 102-2 and 503-1. As shown, there is a curvilinear relationship, such that the upper and lower ends of these curves are relatively flat. Large increases in pressure are required for small increases in volume. This type of curve is used to quantitate the severity of restrictive lung disorders.

Healthy individuals breathe in the range of lung volumes corresponding to the steep portion of the relationship, where small changes in pressure promote relatively larger changes in lung volume. However, disease, or sometimes therapeutic interventions, can force the lungs toward one of the flat portions, considerably increasing the effort the respiratory muscles must make to generate a given tidal volume. Mathematically, the volume-pressure relationship is often described by the compliance of the respiratory system, which is defined as the quotient of lung volume and pressure changes (also the slope of the volume-pressure relationship). A reduction in respiratory system compliance for any given lung volume is the fundamental characteristic of restrictive disease. It is important to realize, that because the slope of the volume-pressure relationship changes with lung volume, compliance can be decreased without underlying abnormality of the lung or chest-wall tissue. A child with status asthmaticus, for example, may not have a primary restrictive disease of the lung but will likely have a decreased lung compliance because of overinflated and collapsed alveoli in the lungs that result from the small airway obstruction.

In newborns and infants, incomplete ossification of the rib cage causes the chest wall to be considerably more compliant than in adults.2 This manifests with less outward recoil force from the chest wall, and a predominance of lung recoil in determining lung volumes. Consequently, when the respiratory muscles are fully relaxed, the lungs of the newborn or small infant tend to adopt a smaller volume relative to the lungs of older children or adults. The newborns of most species, including humans, confront this potential mechanical limitation by adopting several strategies to delay expiration, all in an effort to maintain the functional residual capacity of the lungs (the volume of gas contained in the lungs at the end of a normal expiration) above the relaxation volume of the respiratory system.3 These strategies include partially closing the glottis during expiration, prolonging the activation of the diaphragm during part of the expiratory phase, and initiating the next inspiration before exhalation is complete. Neurological dysfunction, anesthesia, and sedation or neuromuscular blockade render these strategies ineffective, making the newborn vulnerable to alveolar collapse and hypoxemia even in the absence of preexisting lung disease.

Clinical Findings Inrestrictive Disease

The manifestations of restrictive respiratory disease can be attributed to the mechanical impairment itself, the ensuing increase in the respiratory drive, or the alterations of pulmonary gas exchange that develop if the impairment is not compensated.

Restrictive diseases always increase the inward recoil of the thorax as a whole, independent of whether they primarily affect the lung or chest wall. Therefore, both tidal volume and functional residual capacity tend to decrease (Fig. 102-3). Auscultation of the chest wall demonstrates diminished breath sounds, frequently accompanied by inspiratory and expiratory crackles or rales created by reopening of collapsed air spaces as the effort of the inspiratory muscles exceeds the critical opening pressure. To mitigate the decrease in lung volume, newborns and young infants often close their glottis before exhalation is complete. This may result in a characteristic grunting noise, which is a sign of unstable lung volume and a warning that oxygenation is endangered.4 Chest radiographs are characterized by small sized lung fields and a general appearance of reduced aeration of the lungs in restrictive lung disease. Alveolar collapse may occur in areas of the lung fields where an inefficient cough promotes mucus plugging or the effects of gravity neutralize the net forces that leave alveoli open.

An increase in respiratory drive is triggered by alterations in the blood-gas tensions (chemoreceptor reflexes) or by disease-related changes in the tensile forces within the tissues of the lungs and chest wall (mechanoreceptor reflexes). Patients with restrictive disease typically attempt to increase alveolar ventilation by adopting a pattern of rapid and shallow respirations that compensates for the reduced tidal volume of the shallow breaths. This minimizes the average force that the inspiratory muscles must generate over a given period of time (remember that elastic recoil increases proportionally to lung volume) and the energy cost of breathing.5 In patients with respiratory distress the presence of tachypnea as a prominent sign suggests that the disease is primarily restrictive except in patients with severe respiratory dysfunction, for whom tachypnea may simply be a sign of respiratory muscle fatigue, independent of whether the original impairment is restrictive or obstructive.

Other manifestations of increased respiratory drive are less specific. For example, recruitment of accessory inspiratory muscles such as the scalene and sternocleidomastoid muscles can be detected in children with both restrictive disease and upper-airway obstruction, because in both instances the increased effort is mainly inspiratory. Conversely, recruitment of abdominal muscles during expiration is particularly prominent in patients with intrathoracic obstruction but may also be found in patients with restrictive disease or upper-airway obstruction. In these situations, the contraction of the abdominal muscles helps stabilize the lower rib cage and increases respiratory frequency by accelerating expiration.

Chest-wall retractions are a very common manifestation of respiratory distress in children, but their presence has little value in distinguishing the mechanism responsible for the respiratory distress. The retractions occur during inspiration, when the increased effort of the diaphragm and other inspiratory muscles to overcome lung recoil creates subatmospheric pressures in the pleural space (eFig. 102.1). These pressures cause inward deformation of the cartilages and soft tissues of the chest wall, especially at the intercostal, subcostal, and suprasternal areas. The compliant chest wall of the newborn and infant undergoes considerably more deformation for the same degree of lung mechanical impairment than that of older children and adults. In premature and term newborn infants, entire sections of the cartilaginous rib cage can cave during inspiration, adding considerably to the work the diaphragm must perform to ventilate the lungs.6

The abnormalities of gas exchange caused by restrictive disease depend on the ability of the respiratory system to maintain alveolar ventilation and on the effectiveness of the local vascular and airway reflex mechanisms that preserve regional ventilation-perfusion ratios within the lungs. These are discussed in more detail later in this chapter.

Obstructive Impairments

Obstructive impairments are characterized by an increase in the flow-dependent forces generated through interactions among moving gas and tissue molecules during breathing. The largest portion of these forces originates from friction between air and the relatively narrow airway passages of the infant or child. A smaller proportion originates from molecular interactions within the tissues of the lung and chest wall or in the gas-liquid interface. (These forces may increase in some forms of pulmonary disease, but their contribution to the overall mechanical dysfunction of the respiratory system is not well characterized.)

From a mechanical perspective, obstructive impairments have two distinguishing characteristics. First, any work that the respiratory muscles perform to overcome the obstruction is ultimately dissipated as heat (unlike restrictive disease, where the work done by elastic forces is stored during inspiration and used to facilitate expiration). Because energy that is transformed into heat leaves the system immediately, the relationship obtained by plotting the pressure generated by the respiratory muscles against the volume change of the lungs follows a different trajectory during inspiration and expiration (indicating that pressure diminishes even if volume is the same), a phenomenon known as hysteresis. Second, the magnitude of the energy losses incurred at the obstruction (and thus the pressure necessary to compensate these losses) depends on the gas’s velocity. Gas velocity, in turn, is determined by the gas-flow rate and thus depends on the speed with which the lungs inflate and deflate.

The pressure that the respiratory muscles generate to overcome obstruction increases as respiratory rate rises. However, the exact terms of the relationship between pressure and gas flow are defined by the manner in which the gas molecules travel in the flow stream. During normal breathing, flow in most airways adopts a layered or laminar pattern, whereby the molecules in the central layers travel faster than those in the outer layers, which are slowed down by the drag of the airway wall. Under such conditions, frictional pressure losses are proportional to the length of the airways and the viscosity of the gas and are inversely related to the fourth power of the airway’s radius (Poiseuille’s law). The laminar flow pattern is disrupted when flow velocity increases, or when irregularities develop in the airway wall. Gas molecules start to move randomly, in a pattern known as turbulent flow, which dissipates more energy than laminar flow and therefore requires increased respiratory muscle effort to move less air. Furthermore, because pressure losses during turbulence are caused by molecular collisions, the magnitude of these losses depends on gas density rather than on gas viscosity. This is one of the reasons why patients with obstructive airway disease may improve when breathing high concentrations of helium, which has a lower density than air or O2 (even though its viscosity is slightly greater).

In normal airways, viscous friction predominates. The loss of energy dissipated as heat causes the pressure inside the airways to decline gradually in the direction of flow. For this reason, alveolar pressure is always lower (or more negative) than the pressure measured elsewhere in the airway tree during inspiration and is higher (or less negative) during expiration. When there is a discrete obstruction, the gradual decline in pressure turns into a sudden step. Gradual or sudden, the changes in the pressure inside the airways have important effects on the size of the airway lumen. Indeed, unlike rigid pipes, airways vary their caliber depending on the net balance of pressures acting on their inside and outside wall surfaces (known as airway transmural pressure). A positive transmural pressure (inside pressure is greater than outside) increases airway caliber; a negative transmural pressure (outside pressure greater than inside pressure) decreases airway caliber and, depending on the rigidity of the airway wall, may cause the airway lumen to collapse altogether.

Airway obstruction can reduce the pressure distending the inside of the wall by two different mechanisms. First, as discussed earlier, increased friction and, when present, turbulence cause a step reduction in pressure downstream from the obstruction. Second, to accommodate flow as the airway narrows, the gas molecules must accelerate and gain kinetic energy. This kinetic energy gain can only occur with a reduction of the potential energy stored from the force distending the airway wall. As the velocity increases, the pressure on the wall decreases and the airway caliber decreases.

These effects of the obstruction on airway caliber further diminish the amount of flow the obstructed airway can accommodate. To compensate, the respiratory muscles need to contract during expiration to force increased flow through the obstruction. However, there is a limit at which the increase in driving pressure is offset by a further reduction in airway caliber caused by the increase in pressure applied on the airway’s outside surface (viscous flow limitation) or by the decrease in the inside pressure as flow needs to be accelerated even more (wave-speed flow limitation). The maximum flow achievable under those circumstances depends on the location and severity of the obstruction and on the rigidity of the airway wall. Flow limitation is common in diseased airways and contributes substantially to the manifestations of airway obstruction.7

Clinical Findings in Obstructive Disease

The manifestations of obstructive respiratory disease depend on the hierarchy of the obstructed airways and the severity of the obstruction. Obstructions of the larynx and trachea affect gas flow to both lungs and therefore represent a much greater threat than localized bronchial obstructions.

Because gas flow defines the rate at which the lungs change volume, some delay in inspiration, expiration, or both is always detectable. The delay lengthens the total duration of each breath. Thus, tachypnea is a less prominent sign of respiratory distress (when not absent altogether) in patients with airway obstruction than in patients with restrictive disease.

Determining whether the impediment to gas flow is predominantly inspiratory or expiratory helps considerably in the differential diagnosis of airway obstruction. As a rule, obstructions of the extrathoracic airway (nose, pharynx, larynx, and cervical segment of the trachea) are exacerbated during inspiration; obstructions of the intrathoracic airway (thoracic segment of the trachea, bronchi, and bronchioles) are exacerbated during expiration. These are an exaggeration of the normal fluctuations of the airway caliber during unobstructed breathing. They result from differences in how breathing affects the transmural pressure of the extra- and intrathoracic airways (eFig. 102.1).

The pressure acting on the inside surface of the extrathoracic airways is transmitted from the alveoli through the gas column. It is therefore slightly negative (subatmospheric) during inspiration and slightly positive (above atmospheric) during expiration. However, the pressure acting on the outside surface of these airways is similar to atmospheric pressure throughout the respiratory cycle. Consequently, extrathoracic airways are exposed to a negative (collapsing) transmural pressure during inspiration and to a positive (dilating) transmural pressure during expiration. Although small, the negative swings of the transmural pressure during inspiration are sufficient to collapse the lumen of the pharynx. The well-timed inspiratory contraction of the pharyngeal and laryngeal dilator muscles prevents this collapse, providing a protective mechanism that is absent in patients with brain-stem dysfunction, who sometimes require tracheal intubation or tracheostomy to bypass the upper segment of the airway. When the extrathoracic airway is obstructed, increased friction and turbulence at the level of the obstruction force the respiratory muscles to augment their contraction force. Pleural and alveolar pressures become more negative during inspiration, as does the pressure inside the airways downstream from the obstruction point. If the effort is sufficient, flow is restored to the levels needed to maintain alveolar ventilation. As a consequence, however, transmural pressure becomes substantially more negative during inspiration in the airway segment comprised between the obstruction point and the thoracic inlet. This, coupled with the Venturi effect created by the acceleration of the gas in the narrow segment, causes a reduction of the airway lumen, which further aggravates the obstruction and may create the conditions for flow limitation.

In patients with airway obstruction, inward deformation of the chest wall is present and frequently severe during inspiration as contraction of the diaphragm and recruitment of accessory inspiratory muscles decrease pleural pressure well below atmospheric levels. Turbulence in the gas column and vibrations induced in the airway wall by the rushing gas create a high-pitched inspiratory noise known as stridor, which often can be heard without a stethoscope. However, the most prominent clinical feature of extrathoracic airway obstruction is the prolongation of the inspiratory time. Expiration proceeds normally or may also be lengthened if the obstruction is severe enough to impede air exit despite the relative dilation of the airway lumen during expiration. Any circumstance that shortens the duration of inspiration compounds the negative effects of the obstruction, because a shorter inspiration can be accomplished only by increasing gas velocity and by making the pressure inside the airways even more negative. Moreover, if viscous or wave-speed flow limitation is present, gas flow cannot be increased, and the patient’s extra effort is wasted.

The inside surface of the intrathoracic airways is also exposed to negative pressure during inspiration and to positive pressure during expiration. The pressure acting on the outside surface of these airways, however, is similar to pleural pressure and therefore varies substantially during the breathing cycle. To understand how this variation affects airway transmural pressure, it is useful to keep in mind the two premises that define the relationship between the pressures in the pleural space and inside the airways. First, regardless of effort and breathing phase, pleural pressure must remain lower than alveolar pressure (the difference being the recoil pressure of the lungs). Second, the pressure inside the airways must be higher than alveolar pressure during inspiration and lower during expiration; otherwise, flow cannot proceed. Based on these premises, the transmural pressure of the intrathoracic airways can be positive only during inspiration. During expiration, however, the pressure inside these airways decreases rapidly as friction steals energy from the flow stream and may, beyond a certain point, become lower than pleural pressure. Consequently, the transmural pressure of the intrathoracic airways can be positive or negative—depending on the location of an airway obstruction and the magnitude of the frictional losses—during expiration. When there is an obstruction in the intrathoracic airways, the decrease in the pressure applied to the inside surface is more than offset by the decrease in pleural pressure during inspiration, and the obstruction is partially relieved. However, during expiration, the steep decrease in the pressure inside the airways causes the transmural pressure to become negative and the caliber of the airways smaller somewhere between the obstruction point and the thoracic inlet. The decrease in caliber depends on the severity of the obstruction, the velocity of the flow (which draws the airway wall farther into the flow stream), and the stiffness of the airway wall. Airways with abnormally soft cartilage or poor muscle tone are particularly vulnerable and may become obstructed during normal breathing (tracheobronchomalacia) or crying. Activation of the expiratory (abdominal) muscles may accelerate flow, but it also increases pleural pressure. Flow limitation (viscous or wave speed) develops eventually, and the muscle effort becomes wasted or counterproductive.

The clinical signs of extrathoracic airway obstruction are more prominent during inspiration, which is prolonged and obviously more strenuous than expiration. A high-pitched noise, or stridor, produced by the vibration of the airway mucosa at the level of the obstruction can usually be heard without the help of a stethoscope. When the obstruction is severe, a child may sit in a forward-leaning position, resting on both arms to facilitate the use of the neck accessory muscles during inspiration.

The physical findings of intrathoracic obstruction are more prominent during expiration. The expiratory phase is prolonged, and wheezing (a whistling noise produced by high-frequency vibrations of the airway wall and gas column) can be auscultated over the obstructed area or, if the obstruction is diffuse (eg, asthma), over the entire lung fields. Inspiration is less affected than expiration, because the obstruction is partially relieved by the decrease in pleural pressure during this phase of the breathing cycle. The preferential impairment of expiratory gas flow may not be compensated entirely by the prolonged expiratory phase. If so, the alveolar spaces subtended by the obstructed airways do not empty entirely before the next inspiration starts, and the volume of the affected alveoli at end-expiration increases (a condition easily identified on chest radiographs that is frequently referred to as gas trapping). However, the alveolar volume increase is limited by the effects of distention on lung recoil, which dictate an equilibrium whereby the increased recoil limits tidal volume and accelerates exhalation enough to compensate for the low expiratory flow. The development of lung distention in patients with intrathoracic airway obstruction adds a restrictive component to the manifestations of their lung disease. This may be one of the reasons why children with asthma or other forms of bronchial obstruction have tachypnea as a prominent sign. The increase in lung volume is evident on radiographs as decreased density denoting overinflation in the affected areas. When the disease is diffuse, the diaphragmatic curvature is reduced. Shortening of the phrenic fibers further diminishes the patient’s ability to sustain ventilation. Complete bronchial obstruction also causes alveolar collapse due to reabsorption of the alveolar gas.

The main functions of the respiratory system are to replenish the venous blood’s content of O2 while removing its excess CO2. Respiratory failure describes the inability to carry out this function commensurate with the needs of the organism. This definition may be misleading. For example, to state that respiratory failure exists when the partial tensions of O2 (PO2) and CO2 (PCO2) in the arterial blood remain persistently outside the range found in normal humans (values of PO2 < 50 mm Hg or PCO2 > 45 mm Hg are often cited). However, this definition ignores the fact that breathing O2-enriched gas can by itself restore the arterial PO2 to the normal range, even though the gas-exchanging mechanism is faulty and may not sustain the needs of the individual while breathing air. Conversely, the same definition would categorize children with congenital cyanotic heart disease (who are hypoxemic) or with diuretic-induced metabolic alkalosis (who, as a compensatory mechanism, are hypercapnic) as suffering from respiratory failure, even though their respiratory system may be perfectly functional.

For these reasons, rather than detecting specific aberrations in blood-gas content, the clinician’s primary concern should be whether the respiratory system can support metabolic demands under all the circumstances the patient is likely to encounter. Fever or exercise often unveil otherwise compensated anomalies of gas exchange by adding to the ventilatory load of the respiratory muscles or by increasing demand for pulmonary blood flow. Hypoxemia and hypercapnia interfere most noticeably with the function of the central nervous and cardiovascular systems. Agitation, somnolence, apathy, combativeness, or even stupor in an infant or child with respiratory distress or decreased respiratory effort should at least prompt the administration of O2, even if blood-gas analysis is unavailable. Increasing tachycardia, arterial hypertension, or, by way of progression, decreased perfusion, arterial hypotension, and bradycardia are worrisome signs in a patient with a respiratory derangement and, under most circumstances, constitute an immediate indication to institute artificial ventilatory support.

Disruption of Gas Exchange

Understanding the mechanisms that lead to blood-gas aberrations in infants and children with respiratory failure is essential to interpreting clinical information and to planning treatment.8 Basic to such understanding is the notion that the exchange of gases between the inspired gas and the blood in the lungs produces two products: the expired gas and the arterial blood (Fig. 102-4). Each of these products is a composite of the contributions of millions of alveolar-capillary units, weighted by the amount of oxygen or carbon dioxide in blood or gas from each unit. In every one of these units, and in the lung as a whole, the gas contents of the alveolar gas and the capillary blood are linked reciprocally by relatively simple laws that establish a framework for understanding the gas-exchanging process and its abnormalities. Moreover, in alveolar-capillary units where there is an exchange of gas, it is generally assumed that the pressure of O2 and CO2 develop equilibrium between the blood and gas phase.

Oxygen and CO2 are highly diffusible gases. Diffusion impairments across the alveolar-capillary membrane play little role in the genesis of hypoxemia or hypercapnia in children with respiratory disease. Accordingly, it is safe to assume that the PO2 and PCO2 of the blood exiting a given pulmonary capillary reflects the PO2 and PCO2 of the gas contained in the corresponding alveolus. The composition of the alveolar gas is determined by the rate at which gases are exchanged across the alveolar-capillary membrane (diffusion and perfusion), and the turnover rate of fresh air entering the alveoli (ventilation). Mathematical modeling of the relationship between alveolar PO2, ventilation, and O2 uptake is relatively complicated,9 because O2 is present in both the inspired and the expired gas. In contrast, modeling of the relationship between PCO2, ventilation, and CO2 elimination is relatively simple, because the inspired gas does not contain CO2. For the whole lung, the average alveolar PCO2 is directly proportional to the amount of CO2 produced by the body and is inversely proportional to the volume of gas that participates in alveolar gas exchange per unit of time. The latter is known as alveolar ventilation (as opposed to minute ventilation, which is the total amount of gas that exits the lungs per unit of time or the product of tidal volume by breathing frequency).

The dead space, or wasted ventilation (Fig. 102-4) is the portion of the total minute ventilation that does not participate in CO2 exchange. This generally represents the gas that is expired from the lungs that remains in the large airways during expiration, until re-inspired into the alveoli during the next inspiration. Since this gas still contains CO2, it does not contribute to elimination of CO2 from the lungs and therefore is shown in Figure 102-4 as not contributing to CO2 gas exchange. Disorders that increase the dead space usually do not cause hypercapnia because the individual is able to increase minute ventilation sufficiently to restore alveolar ventilation to the needed level to maintain CO2 within a normal range (as occurs when breathing through a short piece of tubing). Alveolar CO2 levels determine the pulmonary arterial blood CO2 concentrations.

Arterial blood gas concentrations can be conceptualized as the product of two idealized components. One consists of systemic venous blood that bypasses (or is shunted away from) the alveoli and hence does not participate in gas exchange. The other arises from blood that undergoes perfect exchange with alveolar gas. The mixture of these two components determines the actual arterial blood gas concentrations.

For simplicity, the source of the blood that bypasses alveoli (or venous admixture in the physiological parlance) has conventionally been viewed as arising from several discrete pathways. First, true anatomic shunt follows anatomic communications between the venous and arterial side of the circulation. Some of these communications are found in normal individuals (eg, thebesian veins, which connect the coronary circulation to the left ventricle, or the bronchial vessels, which direct venous blood into the pulmonary veins); others result from cardiac malformations (see Chapter 484) or lung disease resulting in alveolar collapse or consolidation (ventilation-perfusion ratio of 0). Second, diffusion defects, whether caused by true alterations of the alveolar capillary membrane or, more frequently, by blood being forced through the capillaries at a rate that does not permit equilibration with the alveolar gas, reduce the end-capillary PO2 and have an effect similar to mixing venous blood with fully oxygenated blood. Finally, ventilation-perfusion inequality results primarily from incomplete oxygenation of blood circulating through lung units with a low ventilation-perfusion ratio. Diffusion abnormalities and ventilation-perfusion inequality create only a virtual shunt, which decreases if the inspired oxygen concentration is increased. In contrast, when a true anatomic shunt exists, O2 administration can only increase the oxygen content of the arterial blood by raising the volume of O2 dissolved in the pulmonary capillary blood; thus, the calculated size of the shunt is affected minimally by O2 administration.

Ventilation-perfusion inequality is by far the most common mechanism of hypoxemia and hypercapnia,10 both in children and adults with respiratory disease (Fig. 102-5). Ventilation-perfusion differences are a natural consequence of the parallel organization of the bronchial and arterial networks of the lungs, which permits an infinite combination of ventilation-perfusion ratios to coexist in the same lung. Gravity causes a certain degree of ventilation-perfusion inequality in normal lungs by directing a larger share of blood flow to dependent areas. Bronchial obstruction, consolidation or collapse of alveolar spaces, and abnormalities in pulmonary vascular function greatly exaggerate this inequality. The cause of hypoxemia in ventilation-perfusion inequality lies primarily with the alveolar-capillary units that have a low ventilation-perfusion ratio. Because renewal of the alveolar gas through ventilation cannot keep up with O2 uptake by the blood, these units have a low alveolar PO2, usually in a range where the O2-hemoglobin dissociation curve is steep. As a result, the end-capillary blood is not fully loaded with O2, and when mixed with blood from other units, it creates a substantial venous admixture. Units with high ventilation-perfusion ratios have a high alveolar PO2. However, these units cannot compensate for the venous admixture caused by units with low ventilation-perfusion ratios, because at high PO2 levels, the O2-hemoglobin dissociation curve is flat, and the blood cannot increase its O2 content substantially.

Alveolar-capillary units with low ventilation-perfusion ratios cannot decrease their alveolar PCO2 much below the mixed-venous level; thus, their ability to remove CO2 from the blood is impaired. However, units with a high ventilation-perfusion ratio may lower their alveolar PCO2 considerably. This establishes an efficient mechanism of compensation, which makes hypercapnia less prominent than hypoxemia, provided that the infant or child has sufficient respiratory muscle reserve to support the necessary increase in ventilation.

By envisioning that each breath is a mixture of gas from dead space (PCO2 = 0) and the alveolar space (PCO2 = PACO2), the ratio of the dead space ventilation to the tidal volume (VT) can be calculated as

VD/VT = (PACO2 – PECO2)/PACO2

where PECO2 is the PCO2 of the mixed expired gas. Alveolar PCO2 cannot be measured directly but can be estimated using widely available monitoring tools. The end-tidal PCO2, for instance, measures the PCO2 of the gas exhaled at end-expiration on the assumption that it contains only alveolar gas. The arterial PCO2 is similar to the end-capillary PCO2, and barring any shunting of venous blood into the arterial circulation, it should be similar to the alveolar PCO2 (in clinical practice, the end-tidal PCO2 is monitored as a surrogate of the arterial PCO2; see Chapter 106). However, the end-tidal PCO2 and the arterial PCO2 may be quite different, especially in the presence of lung disease. The reason is that alveoli, which are relatively underperfused, have very low alveolar PCO2 values; thus, the gas that exits these alveoli lowers the end-tidal PCO2 considerably. In contrast, alveoli that are relatively underventilated cannot have PCO2 values greater than the mixed-venous blood (which is only a few mm Hg greater than the arterial PCO2); thus, they are underrepresented in the end-tidal PCO2. It follows that substituting the end-tidal or the arterial PCO2 for the alveolar PCO2 in the equation above yields two different values of the dead space. When the end-tidal PCO2 is used, the result estimates the volume of gas contained in the conducting airways (fresh gas during inspiration and a mixture of fresh gas and alveolar gas during expiration), away from the gas-exchanging regions of the lung (anatomic or series dead space).

On the other hand, when the arterial PCO2 is used, the resulting value (physiological dead space) is greater, because it includes the anatomic dead space and all the gas that enters the alveoli without undergoing CO2 exchange (alveolar or parallel dead space). Dead-space calculations are impractical in clinical practice, because they require cumbersome collection of the expiratory gas. However, the difference between arterial and end-tidal PCO2 can be used as a simple bedside tool to assess disease or treatment-related changes in the distribution of perfusion in the lungs. Pulmonary emboli, for example, can be detected as a conspicuous decrease in the end-tidal PCO2 relative to the arterial PCO2 when the expired PCO2 is monitored continuously (see Chapter 106). Similarly, redistribution of pulmonary blood away from overinflated areas of the lung during the application of positive end-expiratory pressure (PEEP) widens the difference between arterial and end-tidal PCO2 in a fashion that can be used to adjust the therapy.

Another factor that contributes to the complex relationship between PO2, ventilation, and oxygen uptake is that O2 and CO2 are often exchanged at an uneven rate; this creates a volume deficit that must be filled with O2-containing fresh gas (typically O2 consumption exceeds CO2 production by 20%). Physiologists have circumvented these difficulties by recognizing that the reduction in PO2 as gas equilibrates with blood in the alveolus is directly related to the increase in CO2. The alveolar PO2 can then be calculated by determining the PO2 of the inspired gas and by subtracting the PO2 decline produced by the CO2-O2 exchange between alveolus and pulmonary capillary. The resultant expression is the alveolar gas equation, in which the inequality of the CO2-O2 exchange is represented by the respiratory exchange ratio (R), or the result of dividing the CO2 production by the O2 consumption:

PAO2 = FIO2 (PB – PH2O) – PACO2 [FIO2 + (1 – FIO2)/R]

where FIO2 is the fractional concentration of O2 in the inspired gas, PB is the atmospheric pressure, and PH2O is the partial pressure of water vapor at body temperature. This formulation illustrates well the point that alveolar PO2 and PCO2 are linked to each other in such a way that, when one of them is specified, only one value can exist for the other in any particular alveolus. The alveolar gas equation can also be used to take a global view of the lung as if gas exchange were homogeneous. Under such ideal circumstances, the alveolar PO2 represents the highest PO2 that could be expected in arterial blood. Thus, any difference between the calculated alveolar PO2 and the actual arterial PO2 (the alveolar-arterial PO2 difference) is an important means for detecting and understanding aberrations in gas exchange.

When alveolar ventilation decreases globally (hypoventilation), alveolar PCO2 and, by extent, arterial PCO2 increase with the reduction in alveolar ventilation. As shown by the alveolar gas equation, alveolar and arterial PO2 decline, and the decrease in arterial PO2 is proportional to the increase in alveolar or arterial PCO2 (the proportionality constant being R). While the alveolar-arterial PO2 difference is useful for detecting an abnormality in gas exchange, its magnitude changes considerably with the inspired O2 concentration, which cannot always be controlled or known with precision. Thus, physiologists have developed an alternative strategy to quantify derangements in gas exchange. Just as dead space is an index of the efficiency (or rather the inefficiency) of CO2 exchange in the lungs, shunt or venous admixture provides a similar quantitative index for the efficiency of O2 exchange by dividing the pulmonary blood flow into two compartments. Shunt is composed of pulmonary capillary blood that undergoes ideal O2 exchange with the alveolar gas, and the venous admixture is composed of venous blood that travels unaltered to the arterial side of the circulation. Venous admixture (Q̇s) is usually expressed as a fraction of the systemic blood flow (Q̇t), calculated by performing an O2 mass balance across the pulmonary circulation:

Q̇s/Q̇•t = (Cc'O2 – CaO2)/(Cc'O2 – C⊢O2)

where Cc'O2, CaO2, and C⊢O2 are the O2 contents (total volume of O2 per 100 mL of blood) of the pulmonary capillary (estimated by the alveolar gas equation), the arterial blood (measured directly in a systemic artery), and the mixed venous blood (measured directly in the pulmonary artery).

Clinical Evaluation of Respiratory Failure

The majority of infants and children who develop respiratory failure suffer from some form of mechanical dysfunction of the lungs or chest wall. Respiratory distress is then the leading manifestation of the failure and the usual reason why medical attention is sought. Careful assessment of the incumbent physical signs usually provides helpful insight into the restrictive or obstructive character of the dysfunction and into the exact nature of the disease causing it (see above). Occasionally, however, an infant or child presents with respiratory failure caused by a primary or secondary dysfunction of the neural respiratory control. This circumstance may be difficult to detect, unless the clinician suspects that the respiratory effort is inadequate to support the patient’s ventilatory requirements and analyzes the patient’s gas exchange function.

Increased Respiratory Effort and Mechanical Dysfunction

Severe respiratory disease inevitably interferes with both the mechanical and gas-exchanging functions of the respiratory system. Increased demands on the respiratory muscles combined with arterial blood-gas abnormalities trigger a successful compensatory response if two conditions are met: (1) the respiratory muscles must be able to perform and sustain the necessary work, and (2) the gas-exchange abnormalities must be correctable by an increase in ventilation. The situation in which the respiratory muscles are no longer capable of performing work commensurate with the organism’s ventilatory needs is described as respiratory muscle fatigue, drawing an obvious analogy with the behavior of skeletal muscle when subjected to excessive work. Recognizing its imminence is fundamental to preventing the development of life-threatening hypoxemia and hypercapnia, usually via mechanical ventilatory support (see Chapter 109).

The ability of the respiratory muscles to assume an increased mechanical load depends on the balance between the amount of energy that the muscles can transform into physical work and the magnitude of the work demands imposed by breathing.11 Increased work requires that increased oxygen and energy substrate is supplied to the respiratory muscles. Although these muscles, especially the diaphragm, can increase their blood supply several fold,12 disorders that decrease cardiac output may limit the capacity of the muscle to respond.

The efficiency of the respiratory system13 describes the proportion of energy consumed by the respiratory muscle for pressure-volume work. In adults this is estimated not to exceed 15%, whereas in infants values of 5% have been reported.14 Factors that may affect the ability of the respiratory muscles to perform the work of breathing include the respiratory pattern, the identity and state of conditioning of the respiratory muscles, and the configuration of the chest wall.

Breathing pattern is influenced by the nature and extent of the mechanical derangements produced by the disease. Although the neural mechanism has not been fully identified, the practical reality is that infants and children modify their tidal volume and breathing frequency in a manner that minimizes energy expenditure. Thus, any conditions—external or internal—that interfere with a patient’s ability to adopt an optimal pattern diminish the respiratory system’s efficiency. Rapid breathing caused by agitation, for example, may precipitate respiratory muscle fatigue and failure in a child with croup or epiglottitis (see Chapter 510). Similarly, bradypnea caused by central nervous system depression is a very disadvantageous breathing pattern in a child with pulmonary edema or other form of restrictive lung disease.

Respiratory muscle fatigue appears to dictate respiratory pattern requirements of its own. Patients who are experiencing a decrease in the contraction force of their diaphragm often breathe rapidly and shallowly, regardless of whether their disease is predominantly restrictive or obstructive. It is also common for such patients to alternate the respiratory load between several muscle groups in a fashion that suggests that intermittent resting may make the effort more sustainable.

Which specific muscles or muscle groups are activated in an effort to overcome a mechanical impairment has substantial bearing on breathing efficiency. The diaphragm can increase its work with only limited increases of its O2 consumption and blood-flow requirements. Other accessory inspiratory and expiratory muscles are much less economical relative to the increase in ventilatory output that they generate15; consequently, their energy demands may quickly overburden the patient. Poor nutrition, atrophy from lack of use (eg, in patients whose ventilation is fully supported for long periods of time), and myopathy all decrease the efficiency of the respiratory muscles by raising their energetic demands out of proportion with the work that they perform.

Limited ossification of the rib cage and the short axial dimension of the thorax make the newborn and small infant’s chest wall particularly prone to distortion. This may affect respiratory muscle efficiency by several mechanisms. First, all muscles achieve their maximal ratio of work-to-energy consumption at an optimal length. Overinflation of the lungs and abdominal distension flatten the diaphragmatic dome and reduce the effective length of the phrenic muscle fibers, degrading the maximal force that these fibers are able to develop and the muscle’s overall energetic efficiency.

Second, developmental and disease-induced changes in the chest’s geometry influence the area of contact between the lateral surface of the diaphragm and the internal surface of the rib cage. This area, known as the area of apposition of diaphragm and the rib cage,16 facilitates lung inflation by translating the increase in intra-abdominal pressure produced by the diaphragmatic descent into an outward-directed force acting on the ribs. Because infants have a relatively wide lower chest, the costal insertions of the diaphragm are spread out, making their area of apposition small. Overinflation of the lungs and abdominal distention can further limit the lateral contact between diaphragm and rib cage, thereby wasting the work that the diaphragm does on the abdominal organs.

Finally, inward distortion of the rib cage increases the shortening the phrenic fibers must undergo in order to generate a certain volume change in the lungs. In the absence of distortion, the volume displaced by the diaphragmatic contraction is approximately the same as the volume increase of the lungs. When inward distortion occurs, however, the volume displaced by the diaphragm is divided between the volume increase of the lungs and the volume created by the inward movement of the rib cage (Fig. 102-6). Although the diaphragm performs real work to distort the rib cage, the energy used in the process is wasted in terms of ventilation. In this waste lies one of the major disadvantages that the immature child faces when developing lung disease: Chest-wall distortion can multiply the work performed by the diaphragmatic muscle and may lead rapidly to respiratory muscle fatigue, even in the absence of a serious mechanical derangement of the lungs.7

Decreased Ventilatory Effort and Hypoventilation

Abnormal decreases in the respiratory effort lead to hypoventilation, a condition in which the renewal of the alveolar gas is insufficient to maintain normal CO2 and arterial O2 tensions. The mechanisms underlying the ensuing blood-gas abnormalities are (1) the dependence of alveolar PCO2 on alveolar ventilation and (2) the combined effects of the poor renewal of the alveolar gas and the continued uptake of O2 by the pulmonary capillaries on the alveolar PO2. Hypoventilation may be difficult to detect unless it occurs in a patient whose gas exchange is already being monitored or unless it is accompanied by other clinical findings such as upper-airway obstruction. Hypoxemia is not always evident as cyanosis, especially if the patient has anemia or when the light conditions are unfavorable. Hypercapnia produces nonspecific clinical manifestations, somnolence usually being the most prominent.

The finding of a reduced respiratory effort, particularly when the patient is hypercapnic and hypoxemic, should raise immediate suspicion about the integrity of the central nervous system’s function. Direct injuries to the brain, such as those caused by ischemia, an expanding intracranial mass, or infection, can lessen the brain stem’s response to chemoreceptor stimulation. Metabolic toxins or exogenous pharmacological agents may have similar consequences. Opioids, in particular, are effective inhibitors of the respiratory drive, a property that is often exploited to reduce spontaneous breathing and facilitate mechanical ventilation. However, their advantages under such circumstances often turn to disadvantage when it comes time to discontinue ventilatory support. Overdosing with opioids and benzodiazepines is a frequent cause of persistent respiratory failure in mechanically ventilated infants and children. Although it is not always easy to distinguish medication-induced hypoventilation from mechanical dysfunction and especially from muscle weakness, the clinician should be alerted to this possibility whenever a patient has a decreased arterial pH (acidemia) without signs of respiratory distress.

The respiratory control appears to be under more dominant inhibitory influences from supramedullary centers in newborn infants, especially if they are born prematurely, than in older children and adults.17 This developmental singularity explains why newborn infants may breathe shallowly or even become apneic in response to alveolar hypoxemia. It may also explain why these infants become apneic when their pulmonary stretch receptors are activated by excessive lung inflation (the basis of the Hering-Breuer reflex) or by stimuli that arise from the lung interstitium and airway walls in the presence of lung disease. An exaggerated inhibitory response to a combination of alveolar hypoxemia and mechanoreceptor stimulation is likely to be responsible for the frequency with which small infants have apnea as the first manifestation of lung diseases such as viral pneumonitis.

In rare circumstances, the anomaly of the respiratory control is isolated to the respiratory premotor network. The only manifestation of the disease is hypoventilation, usually during sleep, when supramedullary excitatory influences on the respiratory control are at a minimum (Ondine’s curse). More frequently, the decrease in respiratory drive is part of a more extensive dysfunction of the central nervous system, involving supratentorial areas of the brain or other centers in the brain stem. Because the medullary neuronal networks that control the inspiratory muscles (eg, diaphragm) and the muscles that dilate the upper airway (eg, genioglossus or cricoarytenoid) are integrated functionally, hypoventilation is usually associated with upper-airway obstruction caused by decreased pharyngeal tone and glottic obstruction. This is manifested as snoring (stertor), stridor, and reduced air entry into the lungs during inspiration. The neuronal reflexes responsible for airway-protective mechanisms such as coughing and gagging are part of these networks, and they are also frequently impaired. In such case, accumulation of mucus and saliva in the upper airway and bronchi compounds the airway obstruction. Because the amplitude of the respiratory excursions and absolute lung volume are decreased, alveolar collapse becomes inevitable. Thus, it is not unusual to find alveolar densities in chest radiographs of patients in whom the primary alteration is a reduction of the respiratory drive, a feature that may create confusion by leading the clinician toward a diagnosis of primary lung disease.

Management of Respiratory Failure

Mechanical ventilation, whether provided through an endotracheal tube or a mask, is often the only viable alternative to restore gas exchange and to unload the respiratory muscles when respiratory failure is imminent or already present. The physiological bases and practical applications of the various techniques of ventilatory support are discussed in Chapter 109.

Patients with respiratory dysfunction are best served when even the initial treatment addresses the cause, or at least the mechanism, of the dysfunction. For example, pulmonary edema in a child with left-ventricular failure is best treated with diuretics and, if appropriate, with inotropic medications. Bacterial pneumonia demands the use of antibiotics selected for the causal organism. Life-threatening upper-airway obstruction should be relieved by bypassing the obstructed airway segment with an endotracheal tube or with another type of artificial airway. However, etiologic or mechanistic approaches require time and may not be possible if the cause of the disease is not apparent. Under such circumstances, the goal of therapy is to guarantee the adequacy of gas exchange with minimal discomfort, pain, and complications for the patient. On occasion, this goal can be achieved by simple measures that increase the efficiency of the respiratory system, while avoiding interventions that may render it inefficient. For example, lifting the head of the bed in a patient with severe orthopnea may reduce upper-airway resistance and increase the initial length of the diaphragmatic fibers, improving the diaphragm’s ability to handle its load enough so as to turn an unstable situation into a more stable one. Removing ascites that are impinging upon diaphragmatic movement may also improve the efficiency of the diaphragmatic contraction and increase lung volume at end-expiration. Avoiding actions that may frighten or upset a severely distressed child with croup or epiglottitis, on the other hand, allows the child to continue using an advantageous breathing pattern while physicians prepare for safe intubation of the trachea.

Of all the alterations found in respiratory failure, hypoxemia is by far the most life-threatening. Every clinician must remember that administering O2 to a child with respiratory distress is inherently safe and should be done immediately until the hypoxemia is corroborated by blood-gas analysis. The only notable exceptions are in patients for whom hyperoxia-induced pulmonary vasodilation may divert systemic blood flow to the pulmonary circulation through a large left-to-right shunt (eg, large ventricular septal defects, hypoplastic left-heart syndrome) or in newborns with ductal-dependent lesions whose ductus arteriosus may constrict in response to an increasing PO2. Oxygen can be administered with a variety of devices. Nasal prongs are widely used at all ages, because they are comfortable and usually well tolerated by infants and toddlers. Unfortunately, they provide only limited O2-enrichment and humidification of the inspired gas and are not useful in patients who breathe through their mouths. Hoods can raise the concentrations of inspired O2 to close to 100%, but they are cumbersome and threatening to small children. Masks and face tents are best tolerated by older patients. When equipped with a bag reservoir and a one-way exhalation valve (non-rebreathing masks), these two types of devices approach the O2 delivery efficiency of a hood.