∗ We wish to acknowledge gratefully the important contribution of Brian Wood, MD, who was the author of this chapter in the previous editions of this book.To provide individualized care that optimizes pulmonary and neurodevelopmental outcomes, it is essential to have a good working knowledge of the unique physiology and pathophysiology of the newborn respiratory system.
It is the responsibility of those who care for critically ill infants to have a sound understanding of respiratory physiology, especially the functional limitations and the special vulnerabilities of the immature lung. The first tenet of the Hippocratic Oath states, “Primum non nocere” (“First do no harm”). That admonition cannot be followed without adequate knowledge of physiology. In daily practice, we are faced with the difficult task of supporting adequate gas exchange in an immature respiratory system, using powerful tools that by their very nature can inhibit ongoing developmental processes, often resulting in alterations in end-organ form and function.
In our efforts to provide ventilatory support, the infant’s lungs and airways are subjected to forces that may lead to acute and chronic tissue injury. This results in alterations in the way the lungs develop and the way they respond to subsequent noxious stimuli. Alterations in lung development result in alterations in lung function as the infant’s body attempts to heal and continue to develop. Superimposed on this is the fact that the ongoing development of the respiratory system is hampered by the healing process itself.
This complexity makes caring for infants with respiratory failure both interesting and challenging. To effectively provide support for these patients, the clinician must have an understanding not only of respiratory physiology but also of respiratory system development, growth, and healing.
Although the lung has a variety of functions, some of which include the immunologic and endocrine systems, the focus of this chapter is its primary function, that of gas exchange.
Basic Biochemistry of Respiration: Oxygen and Energy
The energy production required for a newborn infant to sustain his or her metabolic functions depends upon the availability of oxygen and its subsequent metabolism. During the breakdown of carbohydrates, oxygen is consumed and carbon dioxide and water are produced. The energy derived from this process is generated as electrons, which are transferred from electron donors to electron acceptors. Oxygen has a high electron affinity and therefore is a good electron acceptor. The energy produced during this process is stored as high-energy phosphate bonds, primarily in the form of adenosine triphosphate (ATP). Enzyme systems within the mitochondria couple the transfer of energy to oxidation in a process known as oxidative phosphorylation .
For oxidative phosphorylation to occur, an adequate amount of oxygen must be available to the mitochondria. The transfer of oxygen from the air outside the infant to the mitochondria, within the infant’s cells, involves a series of steps: (1) convection of fresh air into the lung, (2) diffusion of oxygen into the blood, (3) convective flow of oxygenated blood to the tissues, (4) diffusion of oxygen into the cells, and finally, (5) diffusion into the mitochondria. The driving force for the diffusion processes is an oxygen partial pressure gradient, which, together with the convective processes of ventilation and perfusion, results in a cascade of oxygen tensions from the air outside the body to intracellular mitochondria ( Fig. 2-1 ). The lungs of the newborn infant transfer oxygen to the blood by diffusion, driven by the oxygen partial pressure gradient. For gas exchange to occur efficiently, the infant’s lungs must remain expanded, the lungs must be both ventilated and perfused, and the ambient partial pressure of oxygen in the air must be greater than the partial pressure of the oxygen in the blood. The efficiency of the newborn infant’s respiratory system is determined by both structural and functional constraints; therefore, the clinician must be mindful of both aspects when caring for the infant.
The infant’s cells require energy to function. This energy is obtained from high-energy phosphate bonds (e.g., ATP) formed during oxidative phosphorylation. Only a small amount of ATP is stored within the cells. Muscle cells contain an additional store of ATP, but to meet metabolic needs beyond those that can be provided for by the stored ATP, new ATP must be made by phosphorylation of adenosine diphosphate (ADP). This can be done anaerobically through glycolysis, but this is an inefficient process and leads to the formation of lactic acid. Long-term energy demands must be met aerobically, through ongoing oxidative phosphorylation within the mitochondria, which is a much more efficient process that results in the formation of carbon dioxide and water.
There is a hierarchy of how energy is used by the infant. During periods of high energy demand, tissues initially draw upon the limited stores of ATP, then use glycolysis to make more ATP from ADP, and then use oxidative phosphorylation to supply the infant’s ongoing energy requirements. Oxidative phosphorylation and oxygen consumption are so closely linked to the newborn infant’s energy requirements that total oxygen consumption is a reasonably good measure of the total energy needs of the infant. When the infant’s metabolic workload is in excess of that which can be sustained by oxidative phosphorylation (aerobic metabolism), the tissues will revert to anaerobic glycolysis to produce ATP. This anaerobic metabolism results in the formation of lactic acid, which accumulates in the blood and causes a decrease in pH (acidosis/acidemia). Lactic acid is therefore an important marker of inadequate tissue oxygen delivery.
Ontogeny Recapitulates Phylogeny: A Brief Overview of Developmental Anatomy
The tracheobronchial airway system begins as a ventral outpouching of the primitive foregut, which leads to the formation of the embryonic lung bud. The lung bud subsequently divides and branches, penetrating the mesenchyma and progressing toward the periphery. Lung development is divided into five sequential phases. The demarcation of these phases is somewhat arbitrary with some overlap between them. A variety of physical, hormonal, and other factors influence the pace of lung development and maturation. Adequate distending pressure of fetal lung fluid and normal fetal breathing movements are some of the more prominent factors known to affect lung growth and development.
Phases of Lung Development
Embryonic phase (weeks 3 to 6)
Pseudoglandular phase (weeks 6 to 16)
Canalicular phase (weeks 16 to 26)
Terminal sac phase (weeks 26 to 36)
Alveolar phase (week 36 to 3 years)
Embryonic Phase (Weeks 3 to 6): Development of Proximal Airways
The lung bud arises from the foregut 21 to 26 days after fertilization.
Aberrant development during the embryonic phase may result in the following:
Pulmonary sequestration (if an accessory lung bud develops during this period)
Pseudoglandular Phase (Weeks 6 to 16): Development of Lower Conducting Airways
During this phase the first 20 generations of conducting airways develop. The first 8 generations (the bronchi) ultimately acquire cartilaginous walls. Generations 9 to 20 comprise the nonrespiratory bronchioles. Lymph vessels and bronchial capillaries accompany the airways as they grow and develop.
Aberrant development during the pseudoglandular phase may result in the following:
Congenital lobar emphysema
Congenital diaphragmatic hernia
Canalicular Phase (Weeks 16 to 26): Formation of Gas-Exchanging Units or Acini
The formation of respiratory bronchioles (generations 21 to 23) occurs during the canalicular phase. The relative proportion of parenchymal connective tissue diminishes. The development of pulmonary capillaries occurs. Gas exchange depends upon the adequacy of acinus–capillary coupling.
Terminal Sac Phase (Weeks 26 to 36): Refinement of Acini
The rudimentary primary saccules subdivide by formation of secondary crests into smaller saccules and alveoli during the terminal sac phase, thus greatly increasing the surface area available for gas exchange. The interstitium continues to thin out, decreasing the distance for diffusion. Capillary invasion leads to an increase in the alveolar–blood barrier surface area. The development and maturation of the surfactant system occurs during this phase.
Birth and initiation of spontaneous or mechanical ventilation during the terminal sac phase may result in the following:
Pulmonary insufficiency of prematurity (due to reduced surface area, increased diffusion distance, and unfavorable lung mechanics)
Respiratory distress syndrome (due to surfactant deficiency and/or inactivation)
Pulmonary interstitial emphysema (due to tissue stretching by uneven aeration, excessive inflating pressure, and increased interstitium that traps air in the perivascular sheath)
Impairment of secondary crest formation and capillary development, leading to alveolar simplification, decreased surface area for gas exchange, and variable increase in interstitial cellularity and/or fibroproliferation (bronchopulmonary dysplasia [BPD]).
Alveolar Phase (Week 36 to 3 Years): Alveolar Proliferation and Development
Saccules become alveoli as a result of thinning of the acinar walls, dissipation of the interstitium, and invagination of the alveoli by pulmonary capillaries with secondary crest formation during the alveolar phase. The alveoli attain a polyhedral shape.
The respiratory system is composed of millions of air sacs that are connected to the outside air via airways. The lung behaves like a balloon that is held in an expanded state by the intact thorax and will deflate if the integrity of the system becomes compromised. The interior of the lung is partitioned so as to provide a large surface area to facilitate efficient gas diffusion. The lung is expanded by forces generated by the diaphragm and the intercostal muscles. It recoils secondary to elastic and surface tension forces. This facilitates the inflow and outflow of respiratory gases required to allow the air volume contained within the lung to be ventilated. During inspiration the diaphragm contracts. The diaphragm is a dome-shaped muscle at rest. As it contracts, the diaphragm flattens, and the volume of the chest cavity is enlarged. This causes the intrapleural pressure to decrease and results in gas flow into the lung. During unlabored breathing, the intercostal and accessory muscles serve primarily to stabilize the rib cage as the diaphragm contracts, countering the forces resulting from the decrease in intrapleural pressure during inspiration. This limits the extent to which the infant’s chest wall is deformed inward during inspiration.
Although the premature infant’s chest is very compliant, the rib cage offers some structural support, serves as an attachment point for the respiratory muscles, and limits lung deflation at end expiration. The elastic elements of the respiratory system—the connective tissue—are stretched during inspiration and recoil during expiration. The air–liquid interface in the terminal air spaces and respiratory bronchioles generates surface tension that opposes lung expansion and promotes lung deflation. The conducting airways, which connect the gas exchange units to the outside air, provide greater resistance during exhalation than during inspiration, because during inspiration, the tethering elements of the surrounding lung tissue increase the airway diameter, relative to expiration. The respiratory system is designed to be adaptable to a wide range of workloads; however, in the newborn infant, several structural and functional limitations make the newborn susceptible to respiratory failure.
Differences between the shape of a newborn infant’s chest and that of an adult put the infant at a mechanical disadvantage. Unlike the adult’s thorax, which is ellipsoid in shape, the infant’s thorax is more cylindrical and the ribs are more horizontal, rather than oblique. Because of these anatomic differences, the intercostal muscles in infants have a shorter course and provide less mechanical advantage for elevating the ribs and increasing intrathoracic volume during inspiration than do those of adults. Also, because the insertion of the infant’s diaphragm is more horizontal than in the adult, the lower ribs tend to move inward rather than upward during inspiration. The compliant chest wall of the infant exacerbates this inward deflection with inspiration. This is particularly evident during rapid eye movement (REM) sleep, when phasic changes in intercostal muscle tone are inhibited. Therefore, instead of stabilizing the rib cage during inspiration, the intercostal muscles are relaxed. This results in inefficient respiratory effort, which may be manifested clinically by intercostal and substernal retractions associated with abdominal breathing, especially when lung compliance is decreased. The endurance capacity of the diaphragm is determined primarily by muscle mass and the oxidative capacity of muscle fibers. Infants have low muscle mass and a low percentage of type 1 (slow twitch) muscle fibers compared to those of adults. To sustain the work of breathing, the diaphragm must be provided with a continuous supply of oxygen. The infant with respiratory distress is thus prone to respiratory muscle fatigue leading to respiratory failure.
During expiration the main driving force is elastic recoil, which depends on the surface tension produced by the air–liquid interface, the elastic elements of the lung tissue, and the bony development of the rib cage. Expiration is largely passive. The abdominal muscles can aid in exhalation by active contraction if required, but they make little contribution during unlabored breathing. Because the chest wall of premature infants is compliant, it offers little resistance against expansion upon inspiration and little opposition against collapse upon expiration.
This collapse at end expiration can lead to atelectasis. In premature infants the largest contributor to elastic recoil is surface tension. Pulmonary surfactant serves to reduce surface tension and stabilize the terminal airways. In circumstances in which surfactant is deficient, the terminal air spaces have a tendency to collapse, leading to diffuse atelectasis. Distending airway pressure in the form of positive end-expiratory pressure (PEEP) or continuous positive airway pressure (CPAP) may be applied to the infant’s airway to counter the tendency toward collapse and the development of atelectasis. The application of airway-distending pressure also serves to stabilize the chest wall.
Lung compliance and airway resistance are related to lung size. The smaller the lung, the lower the compliance and the greater the resistance. If, however, lung compliance is corrected to lung volume (specific compliance), the values are nearly identical for term infants and adults. In term infants, immediately after delivery, specific compliance is low but normalizes as fetal lung fluid is absorbed and a normal functional residual capacity (FRC) is established. In premature infants, specific compliance remains low, due in part to diffuse microatelectasis and failure to achieve a normal FRC, because the lung recoil forces are incompletely opposed by the excessively compliant chest wall.
The resistance within lung tissue during inflation and deflation is called viscous resistance . Viscous resistance is elevated in the newborn. In immature small lungs, there are relatively fewer terminal air spaces and relatively more stroma (cells and interstitial fluid). This is manifested by a low ratio of lung volume to lung weight. Although in absolute terms airway resistance is high in the newborn infant, when corrected to lung volume (specific conductance, which is the reciprocal of resistance per unit lung volume), the relative resistance is lower than in adults. It is important to remember that because of the small diameter of the airways in the lungs of the newborn infant, even a modest further narrowing, will result in a marked increase in resistance. That the newborn’s bronchial tree is short and the inspiratory flow velocities are low are teleologic advantages for the newborn because both of these factors decrease airway resistance.
Overcoming the elastic and resistive forces during ventilation requires energy expenditure and accounts for the work of breathing. The normal work of breathing is essentially the same for newborns and adults when corrected for metabolic rates. When the work of breathing increases in response to various disease states, the newborn is at a decided disadvantage. The newborn infant lacks the strength and endurance to cope with a significant increase in ventilatory workload. A large increase in ventilatory workload can lead to respiratory failure.
Elastic and resistive forces of the chest, lungs, abdomen, airways, and ventilator circuit oppose the forces exerted by the respiratory muscles and/or ventilator. The terms elastic recoil , flow resistance, viscous resistance, and work of breathing are used to describe these forces. Such forces may also be described as dissipative and nondissipative forces. The latter refers to the fact that the work needed to overcome elastic recoil is stored like the energy in a coiled spring and will be returned to the system upon exhalation. Resistive and frictional forces, on the other hand, are lost and converted to heat (dissipated). The terms elasticity, compliance, and conductance characterize the properties of the thorax, lungs, and airways. The static pressure–volume curve illustrates the relationships between these forces at various levels of lung expansion. Dynamic pressure–volume loops illustrate the pressure–volume relationship during inspiration and expiration ( Figs. 2-2 to 2-4 ).
Elastic recoil refers to the tendency of stretched objects to return to their original shape. When the inspiratory muscles relax during exhalation, the elastic elements of the chest wall, diaphragm, and lungs, which were stretched during inspiration, recoil to their original shapes. These elastic elements behave like springs ( Fig. 2-5 ). The surface tension forces at the air–liquid interfaces in the distal bronchioles and terminal airways decrease the surface area of the air–liquid interfaces ( Fig. 2-6 ).
At some point, the forces that tend to collapse are counterbalanced by those that resist further collapse. The point at which these opposing forces balance is called the resting state of the respiratory system and corresponds to FRC ( Fig. 2-7 ; see also Fig. 2-2 ). Because the chest wall of the newborn infant is compliant, it offers little opposition to collapse at end expiration. Thus the newborn, especially the premature newborn, has a relatively low FRC and thoracic gas volume, even when the newborn does not suffer from primary surfactant deficiency. Clinically, this manifests as a mild degree of diffuse microatelectasis and is referred to as pulmonary insufficiency of prematurity . This low FRC and the relative underdevelopment of the conducting airway’s structural support explain the tendency for early airway closure and collapse, with resultant gas trapping in premature infants.
The respiratory system’s resting volume is very close to the closing volume of the lung (the volume at which dependent lung regions cease to ventilate because the airways leading to them have collapsed). In newborns, closing volume may occur even above FRC (see Fig. 2-2 ). Gas trapping related to airway closure has been demonstrated experimentally by showing situations in which the thoracic gas volume is greater than the FRC. For this to occur, the total gas volume measured in the chest at end expiration is greater than the amount of gas that is in communication with the upper airway (FRC).
The main contributor to lung elastic recoil in the newborn is surface tension. The pressure required to counteract the tendency of the bronchioles and terminal air spaces to collapse is described by the Laplace relationship:
P = 2 S T r
Simply stated, this relationship illustrates that the pressure ( P ) needed to stabilize the system is directly proportional to twice the surface tension (2 ST ) and inversely proportional to the radius of curvature ( r ). In infants, the relationship should be modified, because, unlike in a soap bubble, there is an air–liquid interface on only one side of the terminal lung unit, so P = ST / r probably describes the situation more accurately in the lung.
In reality, alveoli are not spherical but polyhedral and share their walls with adjacent alveolar structures, making strict application of Laplace’s law suspect. Nonetheless, the basic concept of the law does apply to both terminal air sacs and small airways, and it provides a crucial framework for the understanding of respiratory physiology. The surface tension in the lung is primarily governed by the presence or absence of surfactant. Surfactant is a surface-active material released by type II pneumocytes. It is composed mainly of dipalmitoyl phosphatidylcholine but contains other essential components, such as surfactant-associated proteins A, B, C, and D, as well.
Surfactant has a variety of unique properties that enable it to decrease surface tension at end expiration and thereby prevent further lung deflation below resting volume and allow an increase in surface tension upon lung expansion that facilitates elastic recoil at end inspiration. In addition, surfactant reduces surface tension when lung volume is decreased. A reduction in the quantity of surfactant results in an increase in surface tension and necessitates the application of more distending pressure to counter the tendency of the bronchioles and terminal air spaces to collapse (see Fig. 2-6, A ).
As can be seen from the Laplace relationship, the larger the radius of curvature of the terminal bronchioles or air spaces, the less pressure is needed to hold them open or to expand them further (see Fig. 2-6, B ). The smaller the radius of curvature (e.g., in premature infants), the more pressure is required to hold the airways open. Surfactant helps this situation throughout the respiratory cycle. As the radii of the air–liquid interfaces become smaller during exhalation, the effectiveness of surfactant in reducing surface tension increases; as the radii become larger, its effectiveness decreases.
Respiratory distress syndrome (RDS) imposes a significant amount of energy expenditure on the newborn infant, who must generate high negative intrapleural pressures to expand and stabilize his or her distal airways and alveoli (see Fig. 2-4 ). In untreated RDS, each breath requires significant energy expenditure because lung volumes achieved with the high opening pressures during inspiration are rapidly lost as the surfactant-deficient lung collapses to its original resting volume during expiration. The burden imposed by this large work of breathing may quickly outstrip the infant’s ability to maintain this level of output and lead to respiratory failure.
The infant with RDS may need relatively high inflation pressure to open atelectatic alveoli, and provision of adequate end-expiratory pressure will help keep the lung open. However, once the lung is expanded, the radii of the bronchioles and terminal air spaces are larger, and, therefore, less pressure is required to hold them open or to expand them further. Attention should be paid to tidal volume and overall lung volume after initial alveolar recruitment to avoid overdistention and volutrauma, which are major factors in the development of BPD. Failure to reduce inspiratory and distending pressures appropriately and thus avoid lung overdistention once normal lung volume has been achieved may lead to air-leak complications such as pulmonary interstitial emphysema (PIE) and pneumothorax (see Chapter 20 ).
Compliance is a measure of the change in volume resulting from a given change in pressure:
C L = Δ V / Δ P
When measured under static conditions, compliance reflects only the elastic properties of the lung. Static compliance is the reciprocal of elastance, the tendency to recoil toward its original dimensions upon removal of the distending pressure required to stretch the system. Static compliance is measured by determining the transpulmonary pressure change after inflating the lungs with a known volume of gas. Transpulmonary pressure is the pressure difference between alveolar pressure and pleural pressure. It is approximated by measuring pressure at the airway opening and in the esophagus. To generate a pressure–volume curve, pressure measurements are made during static conditions after each incremental volume of gas is introduced into the lungs (see lung curve in Fig. 2-2 ). If one measures the difference between pleural pressures (esophageal) and atmospheric pressures (transthoracic) at different levels of lung expansion, the plotted curve will be a chest wall compliance curve (see chest wall curve in Fig. 2-2 ). This kind of plot shows the elastic properties of the chest wall. In the newborn, the chest wall is very compliant; thus large volume changes are achieved with small pressure changes. Taking the lung and chest wall compliance curves together gives the total respiratory system compliance (see the total curve in Fig. 2-2 ).
If one measures compliance during continuous breathing, the result is called dynamic compliance . Dynamic compliance reflects not only the elastic properties of the lungs but, to some extent, also the resistive component. It measures the change in pressure from the end of exhalation to the end of inspiration for a given volume and is based on the assumption that at zero flow the pressure difference reflects compliance. The steeper the slope of the curve connecting the points of zero flow, the greater the compliance. Dynamic compliance is the compliance that is generally measured in the clinical setting, but its interpretation can be problematic.
At the fairly rapid respiratory rates common in infants, the instant of zero flow may not coincide with the point of lowest pressure. This is because dynamic compliance is rate dependent. For this reason, dynamic compliance may underestimate static compliance, especially in infants who are breathing rapidly and those with obstructive airway disease. Two additional factors further complicate the interpretation of compliance measurements. In premature infants, REM sleep is associated with paradoxical chest wall motion, so pressure changes recorded from the esophagus may correlate poorly with intrathoracic or pleural pressure changes. Chest wall distortion generally results in underestimation of esophageal pressure changes. Also, because lung compliance is related to lung volume, measured compliance is greatly affected by the initial lung volume above which the compliance measurement is made. Ideally, comparisons should be normalized to the degree of lung expansion, for example, to FRC. Lung compliance divided by FRC is called specific lung compliance.
Dynamic pressure–volume relationships can be examined by simultaneous recording of pressure and volume changes. The pressure–volume loop allows one to quantify the work done to overcome airway resistance and to determine lung compliance (see Figs. 2-4 and 2-7 ). Figure 2-3 shows a static lung compliance curve upon which three pressure–volume loops are superimposed. Each of the loops shows a complete respiratory cycle, but each is taken at a different lung volume. The overall compliance curve is sigmoidal. At the lower end of the curve (at low lung volume), the compliance is low, that is, there is a small change in volume for a large change in pressure (see Fig. 2-3, A ). This correlates with underinflation. Pressure is required to open up terminal airways and atelectatic terminal air spaces before gas can move into the lung. The lung volume is starting below critical opening pressure. At the center of the curve, the compliance is high; there is a large change in volume for a small change in pressure. This is where normal tidal breathing should occur (see Fig. 2-3, B ). This is the position of maximum efficiency in a mechanical sense, the best ventilation/perfusion matching and lowest pulmonary vascular resistance. At the upper end of the curve (at high lung volume), the compliance is low; again, there is a small change in volume for a large change in pressure (see Fig. 2-3, C ). This correlates with a lung that already is overinflated. Applying additional pressure yields little in terms of additional volume but may contribute significantly to airway injury and compromises venous return because of increased transmission of pressure to the pleural space. This is the result of the chest wall compliance rapidly falling with excessive lung inflation. Thus it is important to understand that compliance is reduced at both high and low lung volumes. Low lung volumes are seen in surfactant deficiency states (e.g., RDS), whereas high lung volumes are seen in obstructive lung diseases, such as BPD. Reductions in both specific compliance and thoracic gas volume have been measured in infants with RDS.
The rapid respiratory rates of premature infants with surfactant deficiency can compensate for chest wall instability to a certain extent, because the short expiratory time results in gas trapping that tends to normalize their FRC. They also use expiratory grunting as a method of expiratory braking to help maintain FRC. In infants with RDS treated in the pre-surfactant era, serial measurements of FRC and compliance have been shown to be sensitive indicators of illness severity.
Dynamic lung compliance has been shown to decrease as the clinical course worsens and to improve as the recovery phase begins. When mechanical ventilation is used in infants with noncompliant lungs resulting from surfactant deficiency, elevated distending pressures may be required initially to establish a reasonable FRC. Figure 2-4 shows the pressure–volume loop of a normal infant and that of an infant with RDS. A higher pressure is required to establish an appropriate lung volume in the infant with RDS than in the normal infant. However, this lung volume will be lost if the airway pressure is allowed to return to zero without the application of PEEP. Mechanical ventilation without PEEP leads to surfactant inactivation resulting in worsening lung compliance, and the repeated cycling of the terminal airways from below critical opening pressure leads to cellular injury and inflammation (atelectotrauma). This results in alveolar collapse, atelectasis, interstitial edema, and elaboration of inflammatory mediators.
Once atelectasis occurs, lung compliance deteriorates, surfactant turnover is increased, and ventilation/perfusion mismatch with increased intrapulmonary right-to-left shunting develops. A higher distending pressure and higher concentrations of inspired oxygen (FiO 2 ) will be required to maintain lung volume and adequate gas exchange, resulting in further injury. Early establishment of an appropriate FRC, administration of surfactant, use of CPAP or PEEP to avoid the repeated collapse and reopening of small airways (atelectotrauma), avoidance of overinflation caused by using supraphysiologic tidal volumes (volutrauma), and avoidance of use of more oxygen than is required (oxidative injury) all are important in achieving the best possible outcome and long-term health of patients.
The level of PEEP at which static lung compliance is maximized has been termed the best, or optimum, PEEP. This is the level of PEEP at which O 2 transport (cardiac output and O 2 content) is greatest. If the level of PEEP is raised above the optimal level, dynamic compliance decreases rather than increases. Additionally, venous return and cardiac output are compromised by excessive PEEP. One hypothesis for this reduction in dynamic lung compliance is that some alveoli become overexpanded because of the increase in pressure, which puts them on the “flat” part of the compliance curve (see Fig. 2-3, C ). Therefore, despite the additional pressure delivered, little additional volume is obtained. The contribution of this “population” of overexpanded alveoli may be sufficient to reduce the total lung compliance. It has been shown that dynamic lung compliance was reduced in patients with congenital diaphragmatic hernia (CDH) even though some of the infants had normal thoracic gas volumes. The reduction in dynamic lung compliance in patients with CDH is attributed to overdistention of the hypoplastic lung into the “empty” hemithorax after surgical repair of the defect. Because CDH infants have a reduced number of alveoli, they develop areas of pulmonary emphysema that persist at least into early childhood.
Based on available evidence, it seems prudent to avoid rapid reexpansion of the lungs in the treatment of CDH. Clinicians must be alert to any sudden improvement in lung compliance in infants receiving assisted ventilation (i.e., immediately after administration of surfactant or recruitment of lung volume). If inspiratory pressure is not reduced as compliance improves, cardiovascular compromise may develop because proportionately more pressure is transmitted to the mediastinal structures as lung compliance improves. The distending pressure that was appropriate prior to the compliance change may become excessive and lead to alveolar overexpansion and ultimately air leak. The use of volume-targeted ventilation would be ideal in these circumstances, because in this mode the ventilator will decrease the inspiratory pressure as lung compliance improves to maintain a set tidal volume.
Because the chest wall is compliant in the premature infant, use of paralytic agents to reduce chest wall impedance is rarely necessary. Little pressure is required to expand the chest wall of a premature infant (see chest wall curve in Fig. 2-2 ). In studies investigating the use of paralytic agents in premature infants at risk for pneumothoraces, no change in lung compliance or resistance was demonstrated after 24 or 48 hours of paralysis, and many of the infants studied required more rather than less ventilator support after paralysis.
In the past, paralysis was often used in larger infants who were “fighting the ventilator” or who were actively expiring against it despite the use of sedation and/or analgesia. It should be noted that poor gas exchange (inadequate support) is usually the cause rather than the result of the infant’s “fighting” the ventilator, and heavy sedation or paralysis masks this important clinical sign. The use of synchronized mechanical ventilation modes such as assist/control will obviate the need to paralyze or heavily sedate infants because they will then be breathing in synchrony with the ventilator.
During positive-pressure ventilation, the relative compliance of the chest wall and the lungs determines the amount of pressure transmitted to the pleural space. Increased intrapleural pressure leads to impedance of venous return and decreased cardiac output, a well-documented but largely ignored complication of positive-pressure ventilation. The relationship is described by the following equation:
P PL = P ¯ aw × ( C L / C L + C CW )
P ¯ aw
is mean airway pressure, C L is compliance of the lungs, and C CW is compliance of the chest wall.
Thus it can be seen that in situations of good lung compliance but poor chest wall compliance, transmission of pressure to the pleural space and hemodynamic impairment are increased. This situation commonly arises in cases of increased intra-abdominal pressure with upward pressure on the diaphragm, as may be seen in infants with necrotizing enterocolitis or after surgical reduction of viscera that had developed outside the abdominal cavity—for example, large omphalocele, gastroschisis, or CDH.
Resistance is the result of friction. Viscous resistance is the resistance generated by tissue elements moving past one another. Airway resistance is the resistance that occurs between moving molecules in the gas stream and between these moving molecules and the wall of the respiratory system (e.g., trachea, bronchi, bronchioles). The clinician must be aware of both types of resistance, as well as the resistance to flow as gas passes through the ventilator circuit and the endotracheal tube. In infants, viscous resistance may account for as much as 40% of total pulmonary resistance. The relatively high viscous resistance in the newborn is due to relatively high tissue density (i.e., a low ratio of lung volume to lung weight) and the higher amount of pulmonary interstitial fluid. This increase in pulmonary interstitial fluid is especially prevalent after cesarean section delivery and in conditions such as transient tachypnea of the newborn or delayed absorption of fetal lung fluid.
A reduction in tissue and airway resistance has been shown after administration of furosemide. Airway resistance ( R ) is defined as the pressure gradient ( P 1 − P 2) required to move gas through the airways at a constant flow rate ( <SPAN role=presentation tabIndex=0 id=MathJax-Element-5-Frame class=MathJax style="POSITION: relative" data-mathml='V˙’>V˙V˙
or volume per unit of time). The standard formula is as follows:
R = ( P 1 − P 2 ) / V ˙
Airway resistance is determined by flow velocity, length of the conducting airways, viscosity and density of the gases, and especially the inside diameter of the airways. This is true for both laminar and turbulent flow conditions.
Although in absolute terms airway resistance is elevated in the newborn infant, when corrected to lung volume (specific conductance, which is the reciprocal of resistance per unit lung volume), the relative resistance is lower than in adults. It is important to remember that because of the small diameter of the airways in the lungs of the newborn infant, even a modest narrowing will result in a marked increase in resistance.
Resistance to flow depends on whether the flow is laminar or turbulent. Turbulent flow results in inefficient use of energy, because the turbulence leads to flow in random directions, unlike with laminar flow, in which molecules move in an orderly fashion parallel to the wall of the tube. Therefore, the pressure gradient necessary to drive a given flow is always greater for turbulent flow but cannot be easily calculated. The Reynolds number is used as an index to determine whether flow is laminar or turbulent. It is a unitless number that is defined as follows:
Re = 2 r · v · d / η
When flow is laminar, resistance to flow of gas through a tube is described by Poiseuille’s law:
R ∝ L × η / r 4
Average values for airway resistance in normal, spontaneously breathing newborn infants are between 20 and 30 cm H 2 O/L/s, and these values can increase dramatically in disease states. Nasal airway resistance makes up approximately two-thirds of total upper airway resistance; the glottis and larynx contribute less than 10%; and the trachea and first four or five generations of bronchi account for the remainder. Average peak inspiratory and expiratory flow rates in spontaneously breathing term infants are approximately 2.9 and 2.2 L/min, respectively. Maximal peak inspiratory and expiratory flow rates average about 9.7 and 6.4 L/min, respectively. The range of flow rates generated by spontaneously breathing newborns (including term and premature infants) is approximately 0.6 to 9.9 L/min. Turbulent flow is produced in standard infant endotracheal tubes (ETTs) whenever flow rates exceed approximately 3 L/min through 2.5-mm internal diameter (ID) tubes or 7.5 L/min through 3.0-mm ID tubes. Flow rates that exceed these critical levels produce disproportionately large increases in airway resistance. For example, increasing the rate of flow through a 2.5-mm ID ETT from 5 to 10 L/min raises airway resistance from 32 to 84 cm H 2 O/L/s, more than twice its original value.
Flow conditions are likely to be at least partially turbulent (“transitional”) when ventilator flow rates exceed 5 L/min in infants intubated with 2.5-mm ID ETTs or when rates exceed 10 L/min in infants with 3.0-mm ID ETTs. With turbulent flow, resistance increases exponentially. The resistance produced by infant ETTs is equal to or higher than that in the upper airway of a normal newborn infant breathing spontaneously. The increased resistance caused by the ETT poses little problem as long as the infant receives appropriate pressure support from the ventilator, because the machine can generate the additional pressure needed to overcome the resistance of the ETT. However, when the infant is being weaned from the ventilator or if the infant is disconnected from the ventilator with the ETT still in place, the infant may not be capable of generating sufficient effort to overcome the increase in upper airway resistance created by the ETT. LeSouef et al. measured a significant reduction in respiratory system expiratory resistance after extubation in premature newborn infants recovering from a variety of respiratory illnesses, including RDS, pneumonia, and transient tachypnea of the newborn.
Airway or Tube Length
Resistance is linearly proportional to tube length. The shorter the tube, the lower the resistance; therefore it is good practice to cut ETTs to the shortest practical length. Shortening a 2.5-mm ID ETT from 14.8 cm (full length) to half its length is feasible, because the depth of insertion in a small preemie is usually about 6 cm. This would reduce the resistance of the tube to half. Cutting the tube to 4.8 cm reduces the flow resistance in vitro to essentially that of a full-length tube of the next size (3.0-mm ID ETT). These relationships are consistent for the range of flows generated by spontaneously breathing newborns.
Airway or Tube Diameter
In a single-tube system, the radius of the tube is the most significant determinant of resistance. As previously described, Poiseuille’s law states that resistance is inversely proportional to the fourth power of the radius. Therefore, a reduction in the radius by half results in a 16-fold increase in resistance and thus the pressure drop required to maintain a given flow. It is important to fully appreciate that resistance to flow increases exponentially as ETT diameter decreases. This is one of the reasons extremely low birth-weight infants are difficult to wean from mechanical ventilation. In a multiple tube system, like the human lung, resistance depends on the total cross-sectional area of all of the tubes. Although the individual bronchi decrease in diameter as they extend toward the periphery, the total cross-sectional area of the airway increases exponentially.
Because resistance increases to the fourth power as the airway is narrowed, even mild airway constriction can cause significant increases in resistance to flow. This effect is exaggerated in newborn infants compared to adults because of the narrowness of the infant’s airways. Resistance during inspiration is less than resistance during expiration because the airways dilate upon inspiration ( Fig. 2-8 ). This is true even though gas flow during inspiration usually is greater than that during expiration, because as we saw above, the relationship between resistance and flow is linear, whereas that to radius is geometric. There is an inverse, nonlinear relationship between airway resistance and lung volume, because airway size increases as FRC increases. Lung volume recruitment therefore reduces resistance to airflow. Any process that causes a reduction in lung volume, such as atelectasis or restriction of expansion, results in increased airway resistance. At extremely low volumes, resistance approaches infinity because the airways begin to close as residual volume is approached (see Fig. 2-2 ).
Consistent with the above physiologic principles, the preponderance of evidence indicates that the application of PEEP and CPAP decreases airway resistance. ETT resistance is of considerable clinical importance. It has been shown that successful extubation is accomplished more often in infants coming directly off intermittent mandatory ventilation than after a 6-hour preextubation trial of endotracheal CPAP. Nasal CPAP circuit design, specifically its resistance, and the means by which nasal CPAP is attached to the patient are the most important determinants of CPAP success or failure.
Viscosity and Density
Gas viscosity is negligible relative to the viscosity of fluids. However, gas density can be of clinical significance. The relationship between airway resistance and the density of the gas in turbulent flow is directly proportional and linear. Decreasing the density of the gas by two-thirds, such as occurs when heliox, a mixture of 80% helium and 20% O 2 , is administered, reduces airway resistance to one-third compared to that when room air is breathed. Heliox can be useful for reducing upper airway resistance (and work of breathing) in patients with obstructive disorders such as laryngeal edema, tracheal stenosis, and BPD. Gas density is influenced by barometric pressure, so airway resistance is slightly decreased at high altitudes, although this has little clinical significance.
Work of Breathing
Breathing requires the expenditure of energy. For gas to be moved into the lungs, force must be exerted to overcome the elastic and resistive forces of the respiratory system. This is mathematically expressed by the following equation:
Work of breathing = Pressure ( force ) × Volume ( displacement )
Work of breathing is the force generated to overcome the frictional resistance and static elastic forces that oppose lung expansion and gas flow into and out of the lungs. The workload depends on the elastic properties of the lung and chest wall, airway resistance, tidal volume (V T ), and respiratory rate. Approximately two-thirds of the work of spontaneous breathing is the effort to overcome the static elastic forces of the lungs and thorax (tissue elasticity and compliance). Approximately one-third of the total work is applied to overcoming the frictional resistance produced by the movement of gas and tissue components (airflow and viscous).
In healthy infants exhalation is passive. A portion of the energy generated by the inspiratory muscles is stored (as potential energy) in the lungs’ elastic components; this energy is returned during exhalation, hence it is also referred to as nondissipative work , in contrast to the frictional forces that are lost or dissipated as heat. If the energy required to overcome resistance to flow during expiration exceeds the amount of elastic energy stored during the previous inspiration, work must be done not only during inspiration but also during expiration; thus exhalation is no longer entirely passive.
In infants, energy expenditure correlates with oxygen consumption. Resting oxygen consumption is elevated in infants with RDS and BPD. Mechanical ventilation reduces oxygen consumption by decreasing the infant’s work of breathing. Work of breathing is illustrated in a dynamic pressure–volume loop (see Fig. 2-7 ). Pressure changes during breathing can be measured with an intraesophageal catheter or balloon, and volume changes can be measured simultaneously with a pneumotachograph. During inspiration (ascending limb of the loop) and expiration (descending limb of the loop), both elastic and frictional resistance must be overcome by work. If only elastic resistance needed to be overcome, the breathing pattern would follow the compliance line; however, because airway resistance and tissue viscous resistance must also be overcome, a loop is formed (hysteresis). The areas ABCA and ACDA in Figure 2-7 represent the inspiratory work and the expiratory work, respectively, performed to overcome frictional resistance. The area ABCEA represents the total work of breathing during a single breath.
The diaphragm is responsible for the majority of the workload of respiration. The most important determinant of the diaphragm’s ability to generate force is its initial position and the length of its muscle fibers at the beginning of a contraction. The longer and more curved the muscle fibers of the diaphragm, the greater the force the diaphragm can generate. In situations in which the lung is hyperinflated (overdistended), the diaphragm is flattened and thus at a mechanical disadvantage.
The application of PEEP or CPAP (continuous distending pressure [CDP]) may reduce the work of breathing for an infant whose breathing is on the initial flat part of the compliance curve secondary to atelectasis (see Fig. 2-3, A ). In this situation, CDP should reduce the work of breathing by increasing FRC and bringing breathing to a higher level on the pressure–volume curve where the compliance is higher (see Fig. 2-3, B ). Reductions in respiratory work with the application of CDP have been shown in newborns recovering from RDS and in babies after surgery for congenital heart disease.
If the lung already is overinflated, increasing CDP will not result in a decrease in the work of breathing (see Fig. 2-3, C ). The one exception here is when lung overinflation is the result of airway collapse, as can be seen in infants with BPD. In this unique situation, higher CDP will maintain airway patency and relieve air trapping, reducing lung volume to a more normal level. Alveolar overdistention caused by any reason is often accompanied by an increase in PaCO 2 (indicating decreased alveolar ventilation) and a decrease in PaO 2 , despite an increase in FRC.
The time constant of a patient’s respiratory system is a measure of how quickly his or her lungs can inflate or deflate—that is, how long it takes for alveolar and proximal airway pressures to equilibrate. Passive exhalation depends on the elastic recoil of the lungs and chest wall. Because the major force opposing exhalation is airway resistance, the expiratory time constant ( K t ) of the respiratory system is directly related to both lung compliance ( C L ), which is the inverse of elastic recoil, and airway resistance ( R aw ):
K t = C L × R aw
The time constants of the respiratory system are analogous to those of electrical circuits. One time constant of the respiratory system is defined as the time it takes the alveoli (capacitor) to discharge 63% of its V T (electrical charge) through the airways (resistor) to the mouth or ventilator (electrical) circuit. By the end of three time constants, 95% of the V T is discharged. When this model is applied to a normal newborn with a compliance of 0.005 L/cm H 2 O and a resistance of 30 cm H 2 O/L/s, one time constant = 0.15 second and three time constants = 0.45 second. In other words, 95% of the last V T should be emptied from the lung within 0.45 second of when exhalation begins in a spontaneously breathing infant. In a newborn infant receiving assisted ventilation, the exhalation valve of the ventilator would have to be open for at least that length of time to avoid air trapping. Inspiratory time constants are roughly half as long as expiratory, largely because airway diameter increases during inspiration. This relationship between inspiratory and expiratory time constants accounts for the normal 1:2 inspiratory/expiratory (I:E) ratio with spontaneous breathing.
The concept of time constants is key to understanding the interactions between the elastic and the resistive forces and how the mechanical properties of the respiratory system work together to modulate the volume and distribution of ventilation. A working knowledge of time constants is essential for choosing the safest and most effective ventilator settings for an individual patient at a particular point in the course of a specific disease process that necessitates the use of assisted ventilation. It must be recognized that compliance and resistance change over time and, therefore, the optimal settings need to be reevaluated frequently.
Patients are at risk of incomplete emptying of a previously inspired breath when their lung condition involves an increase in airway resistance with no or only a modest reduction in lung compliance. They also are at risk when the pattern of assisted ventilation does not allow sufficient time for exhalation—that is, the lungs have an abnormally long time constant—or there is a mismatch between the time constant of the respiratory system (time constant of the patient + that of the ETT + that of the ventilator circuit) and the expiratory time setting on the ventilator. In these situations, the end result is gas trapping. This gas trapping is accompanied by an increase in lung volume and a buildup of pressure in the alveoli and distal airways referred to as inadvertent PEEP or auto-PEEP .
Important clinical and radiographic signs of gas trapping and inadvertent PEEP include (1) radiographic evidence of overexpansion (e.g., increased anteroposterior diameter of the thorax, flattened diaphragm below the ninth posterior ribs, intercostal pleural bulging), (2) decreased chest wall movement during assisted ventilation, (3) hypercarbia that does not respond to an increase in ventilator rate (or even worsens), and (4) signs of cardiovascular compromise, such as mottled skin color, a decrease in arterial blood pressure, an increase in central venous pressure, or the development of a metabolic acidosis. Such late signs of air trapping should never occur today, because all modern ventilators give us the ability to monitor flow waveforms, which allow us to graphically see whether expiration has been completed before the next breath begins.
Time constants are also a function of patient size, because total compliance is proportional to size. The much shorter time constants of an infant are reflected in the more rapid normal respiratory rate, compared to adults. To keep the concept simple, remember that whales and elephants have very large lungs and very long time constants; hence they breathe very slowly. Mice and hummingbirds have tiny lungs with extremely short time constants and have a very rapid respiratory rate to match. Everything else being equal, large infants have longer time constants than “micropreemies.” Any decrease in compliance makes the time constant shorter, and therefore tachypnea is the usual clinical sign of any condition leading to decreased compliance.
Extremely low birth-weight infants with RDS have decreased compliance but initially normal airway resistance. This means that the time constants are extremely short. Equilibration of the airway and alveolar pressures occurs very quickly (i.e., early in the inspiratory cycle). Reynolds estimated that the time constant in RDS may be as short as 0.05 second. This means that 95% of the pressure applied to the airway is delivered to the alveoli within 0.15 second, a value consistent with clinical observation. Short time constants make rapid-rate conventional ventilation feasible in these infants and makes them ideal candidates for high-frequency ventilation.
Term infants with meconium aspiration or older growing preterm infants with BPD have elevated airway resistance and correspondingly longer time constants; therefore they are most at risk of inadvertent PEEP. They should be ventilated with slower respiratory rates and longer inspiratory and, especially, expiratory times. Evidence of air trapping should be actively sought by examining ventilator waveforms, before clinical signs of CO 2 retention and hemodynamic impairment develop. It should be noted that the proximal airway PEEP level does not indicate the level of alveolar PEEP nor does it demonstrate the occurrence of alveolar gas trapping. Even under conditions of zero proximal airway PEEP, alveolar PEEP levels and the degree of gas trapping may be dangerously high if the baby has compliant lungs, increased airway resistance, or both (i.e., a prolonged time constant).
Although it is useful clinically to think of the infant’s respiratory system as having a single compliance and a single resistance, we know this is not really the case. The resistance and compliance values we obtain from pulmonary function measurements are essentially weighted averages for the respiratory system. There are populations of respiratory subunits with a range of discrete compliance and resistance values, whereas what we measure at the airway are averaged values for those populations of subunits.
Mechanisms of Gas Transport
Ventilation or gas transport involves the movement of gas by convection or bulk flow through the conducting airways and then by molecular diffusion into the alveoli and pulmonary capillaries. This makes possible gas exchange (O 2 uptake and CO 2 elimination) that matches the minute-by-minute metabolic needs of the patient. The driving force for gas flow is the difference in pressure at the origin and destination of the gases; for diffusion, it is the difference in the concentrations between gases in contiguous spaces. Gas flows down a pressure gradient and diffuses down a concentration gradient. The predominant mechanism of gas transport by convection is bulk flow, whereas the predominant mechanism of gas transport by diffusion is Brownian motion.
Ventilation of the alveoli is an intermittent process that occurs only during inspiration, whereas gas exchange between alveoli and pulmonary capillaries occurs throughout the respiratory cycle. This is possible because a portion of gas remains in the lungs at the end of exhalation (FRC); the remaining gas provides a source for ongoing gas exchange and maintains approximately equal O 2 and CO 2 tensions in both the alveoli and the blood returning from the lungs.
During spontaneous breathing, inspiration is achieved through active contraction of the respiratory muscles. A negative pressure is produced in the interpleural space, a portion of which is transmitted via the parietal and visceral pleura through the pulmonary interstitial space to the lower airways and alveoli. A pressure gradient between the outside atmospheric pressure and the airway and alveolar pressures results in gas flowing down the pressure gradient into the lungs ( Fig. 2-9 ). Interpleural pressure is more negative than alveolar pressure, which is more negative than mouth and atmospheric pressures.
When an infant receives negative-pressure ventilation, pressure is decreased around the infant’s chest and abdomen to supplement the negative-pressure gradient used to move gas into the lungs, mimicking the normal physiologic function. During positive-pressure ventilation, the upper airway of the infant ( Fig. 2-10 ) is connected to a device that generates a positive-pressure gradient down which gas can flow during inspiration. The pressure in the ventilator circuit and in the upper airway is greater than the alveolar pressure, which is greater than the interpleural pressure, which is greater than the atmospheric pressure. The negative intrathoracic pressure during spontaneous or negative pressure respiration facilitates venous return to the heart. Positive-pressure ventilation alters this physiology and inevitably leads to some degree of impedance of venous return, adversely affecting cardiac output.