Rudolph's Pediatrics, 22nd Ed.

CHAPTER 109. Life Support Systems

Julio Pérez Fontán

The past three decades have seen immense advances in the technologies used to replace the functions of failing organs or systems. During this time, mechanical ventilation has evolved from being used only as a last rite, when there was little hope of survival, to being a widely used and versatile technique that allows thousands of patients to recover from respiratory failure every year. Extracorporeal membrane oxygenation (ECMO), reintroduced into neonatal and pediatric intensive care medicine in the 1980s, is rapidly reaching similar status, as its indications and, perhaps more importantly, its limitations are better defined. More recently, external and implantable ventricular assist devices have undergone sufficient development to become viable alternatives for the continued support of patients with severe circulatory dysfunction, often as a bridge toward cardiac transplantation. While these technologies offer many possibilities, intensive care specialists must use them wisely by carefully selecting the patients who can benefit, by providing patients and families with realistic assessments of their potential, and by carefully evaluating the results to define better indications and to improve efficacy.


Today, any modern intensive care unit has at its disposal a wide array of ventilator models, many of them capable of effectively ventilating the lungs of adults as well as those of premature infants. Although these devices certainly make the task easier, their application to the treatment of infants and children requires more than a passing knowledge of respiratory physiology—it requires careful attention to monitoring the patient’s responses.


All commonly used methods of mechanical ventilation are based on the same principle: A mechanical device forces a gas mixture into the lungs during inspiration and then allows the passive recoil of lungs and chest wall to force gas out of the lungs during expiration. In most cases, the device uses a bellows or other type of pneumatic mechanism to inject a gas mixture into the airway lumen, either through a cannula inserted in the trachea or through a mask applied around the mouth and nose (positive-pressure ventilation). More rarely, the device lowers pressure around the patient’s chest and abdomen (negative-pressure ventilation), decreasing pleural and alveolar pressures relative to atmospheric pressure and drawing gas into the alveoli. Usually, a cannula must be placed in the trachea to prevent subatmospheric airway pressures from collapsing the pharynx and larynx during inspiration. Contrary to common misconception, positive- and negative-pressure ventilation apply similar strains to the lung tissue for a given level of lung inflation.3 Thus, it seems there would be no difference in the degree of lung injury caused by both methods, provided that tidal volume and functional residual capacity are kept equal. This consideration is, of course, limited by the practicality of its assumptions. Without entering a several-decade-long controversy on which method of ventilation is better,4 it seems clear that, by reducing rather than increasing pleural pressure relative to the atmosphere, negative-pressure ventilation may have a more beneficial effect on venous return to the heart than positive-pressure ventilation.5 However, the hemodynamic advantages do not make up for the impracticalities of negative-pressure ventilators, which include bulkiness, poor access to the patients, and inefficiency when high levels of support are needed.

The choice of an appropriate ventilatory strategy involves careful consideration of both the characteristics of the patient’s respiratory dysfunction and the ventilator’s capabilities. In general, the clinician must reconcile two fundamental and sometimes competing goals while designing such a strategy: (1) to maintain adequate gas exchange at the lowest possible inspired O2 concentration and (2) to minimize the accumulative effects of repeated stress and relaxation cycles on the lungs and airways.


The choice of optimal ventilatory pattern depends on the mechanical alterations produced by the disease in the lungs and airways, the homogeneous or heterogeneous distribution of these alterations, and the patient’s ability to contribute to the ventilatory effort. In practice, this choice materializes in a combination of several ventilatory settings, including ventilatory rate, tidal volume, inspiratory time, and inspiratory flow. These settings are organized into modes of ventilation, which are defined by the mechanisms that initiate, limit, and end (cycling) inspiration.

It is well established by both experimentation and clinical experience that no positive-pressure ventilator can duplicate the efficiency of the respiratory muscles.5 Even in individuals with healthy lungs, tidal volume, minute ventilation, and sometimes inspired O2 concentration must be raised above physiological levels to compensate for the increase in dead space and venous admixture that follows the institution of ventilatory support. The mechanisms of such increased requirements include the excessive distention of the conducting airways, the effacement of the vertical ventilation/perfusion gradients within the lungs, and the limitations imposed to movement and posture by the need to keep the endotracheal tube safely in place.6


Increasing the concentration of O2 in the inspired gas is the most predictable and rapid method to increase arterial PO2 in patients with increased venous admixture caused by ventilation/perfusion inequality, decreased ventilation, and diffusion abnormalities. Unfortunately, O2 administration has no direct beneficial effect on the gas-exchanging mechanism itself. Moreover, the use of O2 can compound the gas-exchange derangements. The presence of high concentrations of O2 in poorly ventilated alveolar-capillary units can cause reabsorption atelectasis as O2 in the alveolar gas is removed by pulmonary capillaries without being replaced by N2. Oxygen is in itself highly toxic to the lung tissues, even though its continued administration upregulates local antioxidant responses.7 Based on these considerations, it is reasonable, when ventilatory support is started, to increase inspired O2 concentration as much as needed to correct hypoxemia. Thereafter, it is advisable to adopt other strategies that minimize the continued need for O2.


The product of ventilatory rate by the tidal volume determines the ventilator’s contribution to the total minute ventilation, the remainder of it being supplied by the patient’s spontaneous effort (if any). Tidal volume is usually preset by adjusting either the volume delivered into the ventilator’s circuit (volume-controlled ventilators) or, less frequently, the pressure generated at the airway opening (pressure-controlled ventilators). Some infant ventilators are time-cycled. When the safety pressure limit in these ventilators is set higher than the peak airway pressure, tidal volume is determined by the inspiratory flow circulating constantly through the circuit and by the duration of the inspiratory time (when the airway pressure exceeds the safety pressure, the ventilator behaves as if it were in the pressure-controlled mode). Many ventilators combine options for volume-, pressure-, and time-cycled control of tidal volume.

When instituting ventilatory support in a volume-controlled mode, the tidal volume generated by the ventilator is always greater than the actual tidal volume delivered into the lungs. The difference usually results from leaks and compression of the gas in the ventilator tubing. Leaks may be apparent as a discrepancy between the inhaled and exhaled volume. Although their exact magnitude is difficult to estimate, they can be quite large with the uncuffed endotracheal tubes commonly used in infants and small children and often vary with the position of the head. Because of gas leaks and compression (and frequent inaccuracies in volume measurement by the ventilator), the selection of an appropriate tidal volume should always be corroborated by the size of the chest wall excursions, independent of the information provided by the ventilator.

During pressure-controlled ventilation, tidal volume is determined by the magnitude and timing of the airway pressure increase during inspiration. In adjusting these two variables, it is important to remember that it is alveolar and not airway opening pressure that drives the change in lung volume. How airway and alveolar pressures relate to each other depends on the inspiratory flow and the combined resistance of the endotracheal tube and airways (Fig. 109-1). When inspiratory flow is constant (as is the case with volume-controlled modes of ventilation and with infant constant-flow time-cycled ventilators), airway opening pressure is greater than alveolar pressure throughout inspiration. Thus, the peak inspiratory pressure at the airway opening underestimates the peak alveolar pressure by a magnitude proportional to the inspiratory flow and the resistance of the tube and airways. When a decelerating inspiratory flow is applied (as is usually the case with pressure-controlled modes), airway opening and alveolar pressures equilibrate after gas flow ceases at the end of inspiration. Accordingly, peak airway opening and alveolar pressures are similar.

The choices of tidal volume and ventilatory rate are clearly interdependent. It is a good starting rule that their combination should mimic the patient’s own ventilatory pattern. Small tidal volumes at a relatively high ventilatory rate offer advantages in infants and children with severe restrictive disease. In contrast, slightly larger tidal volumes and a slower ventilatory rate are appropriate in patients with airway obstruction. Especially when the obstruction is intrathoracic, enough time must be allowed for expiration to be completed (Fig. 109-2). Otherwise, the subsequent inspiration interrupts expiratory flow, causing pressure in the alveoli to exceed that in the proximal airway. The resultant increase in lung volume is difficult to detect. If uncorrected, it may flatten the combined volume-pressure relationship of the lungs and chest wall, decrease tidal volume, and interfere with venous return to the heart.


When lung volume decreases as a result of disease (eg, surfactant deficiency, pulmonary edema), alveoli lose their mechanical stability and collapse easily. Under such conditions, the number and volume of the alveoli that remain open is very sensitive to the volume of the preceding inflations (a property that physiologists call volume history dependence). This is why early ventilators were designed to provide periodic intermittent sighs or breaths larger than the prevalent tidal volume (a practice that has, incidentally, come back into vogue with some modifications and the new name of recruitment maneuvers). It soon became obvious that raising the functional residual capacity (FRC) of the lungs by increasing the airway pressure at the end of expiration was a much more effective alternative. Ostensibly, the objectives of positive end-expiratory pressure (PEEP) are to recruit collapsed alveoli and to prevent the closure of those that are open. When these goals are achieved, the ventilation/perfusion profile of the lungs improves and venous admixture decreases8 (see Fig. 102-5). As an added benefit, lung compliance and alveolar ventilation increase, improving the efficiency of mechanical ventilation while reducing the workload of the respiratory muscles in patients who also breathe spontaneously. However, when PEEP fails to recruit a predominance of collapsed or underventilated alveoli, well-ventilated alveoli become overdistended. Blood flow may then be forced through alveolar-capillary units with a low ventilation/perfusion ratio, overcoming their hypoxic vasoconstriction. As a result, venous admixture increases and lung compliance and alveolar ventilation decrease. The unpredictable nature of these changes emphasizes the importance of careful clinical assessment after changing ventilatory strategy.

FIGURE 109-1. Relationship between airway and alveolar pressure during volume- and pressure-controlled ventilation. Most volume-controlled modes generate a constant flow during inspiration (A). Under these circumstances, the combined resistance of the endotracheal tube and airways (see diagram on the left) establish a constant pressure gradient between the airway opening and the alveoli (ΔP). Once the gradient is established, both airway opening and alveolar pressure form parallel ramps separated by a difference, ΔP. Because the peak airway pressure is reached while gas is still flowing into the lungs, the peak pressure recorded by the ventilator underestimates peak alveolar pressure. In contrast, most pressure-controlled modes use a decelerating flow during inspiration (B). This results in a large ΔP at the beginning of inspiration, when flow is high. ΔP fades quickly and disappears completely when gas flow ceases at the end of inspiration. From that point on, airway opening and alveolar pressures are similar, and lung volume is held constant until expiration starts. Volume and pressure modes generate the same peak alveolar pressure for a given tidal volume. However, the decelerating flow pattern does it with a lower peak airway opening pressure, a feature that sometimes is incorrectly claimed to offer advantages in terms of lung strain.


The amplitude and time profile of the inspiratory flow greatly influences the distribution of tidal volume in the lungs, especially when lung disease has a heterogeneous distribution.9 When inspiratory time is shortened and the tidal volume is delivered at a high flow rate, gas tends to flow preferentially toward areas of the lung that have a low resistance or a low compliance (or a short time constant, as the product of compliance by resistance is known). Consequently, areas served by obstructed airways receive a small proportion of the tidal volume, which in some situations may be an advantage, because these areas take longer to empty during expiration and are therefore prone to trap gas. Areas where the alveoli are prone to collapse because of pulmonary edema or surfactant deficiency, for example, receive a disproportionate amount of the tidal volume. In contrast, when inspiratory time is lengthened, there is more time for pressures to equilibrate across high-resistance airway segments, especially if flow ceases at the end of inspiration (see Fig. 109-1). Under such circumstances, tidal volume is distributed primarily to areas in which alveolar compliance is higher, regardless of whether the airways that serve these areas are obstructed or not. The resultant pattern of ventilation may be beneficial if the less-ventilated areas are also less perfused, but the preferential flow to areas with a long time constant renders the duration of the expiratory time critical to avoid gas trapping.


In patients with homogeneous restrictive disease, mean airway pressure (the integral of the pressure at the airway opening divided by the duration of the respiratory cycle) correlates well with arterial PO2. Although this observation cannot be safely extrapolated to situations in which the disease is not homogeneous or, much less, when it is obstructive, it has become common practice to characterize the level of ventilatory support based on the mean airway pressure value needed to achieve satisfactory arterial oxygenation. The value of mean airway pressure as a predictor of arterial PO2 originates from the direct link that exists between this pressure and alveolar recruitment. Increases in tidal volume or peak inspiratory pressure, ventilatory rate, inspiratory time, inspiratory flow, and positive end-expiratory pressure all have the effect of raising mean airway pressure and mean airway volume.

FIGURE 109-2. Increase in functional residual capacity (FRC) produced by shortening expiratory time during mechanical ventilation. When expiratory flow becomes interrupted by the following inspiration, alveolar pressure remains higher than the positive end-expiratory pressure (PEEP) measured at the airway opening. The difference (ΔPEEP), often referred to as auto-PEEP or inadvertent PEEP, increases lung volume at end-expiration relative to the previous breath. FRC continues to increase breath after breath until the increased recoil of the lungs prevents further increases. The increased lung volume may be apparent only as an increased peak airway pressure.


The idea that ventilation-induced lung injury is caused by excessive stretch of the lung tissues rather than by a direct effect of the pressures needed to ventilate the lungs has reached axiomatic value in the past few years.10 However, the intimate mechanisms that lead to the disruption of the delicate structures of the lung and the inflammatory and fibroproliferative response that follows11 are far from clear. It is reasonable to assume, however, that the interplay between lung volume and the modifications produced by disease in the elastic properties of the lung determines how much deformation and damage occurs. For instance, the absence of lung surfactant in the newborn’s respiratory distress syndrome reduces lung volume by promoting alveolar collapse. When positive pressure is applied to the airway in order to restore alveolar patency, the stresses generated within the lung tissues can become very high, even if the lung volume remains lower than it would be if the lungs were healthy. The reason is that the increased surface tension of the alveolar gas-liquid interface is transmitted through the alveolar septa to airways, blood vessels, and neighboring alveolar walls. The resultant strain in these tissue elements can disrupt their integrity and can stimulate the release of inflammatory cytokines and growth factors that alone can initiate a response of inflammation and repair.

Thus, reducing stress (or, more correctly, the strain or deformation caused by the stress) within the lungs is an important priority in planning ventilatory strategy. To do so, the clinician is often faced with a basic conflict between the need to preserve oxygenation and ventilation and the clear long-term advantage of keeping tidal volume, ventilatory rate, and positive end-expiratory pressure to a minimum. The recognition of this dilemma gave rise to ventilatory strategies that allow arterial PCO2 to rise above its physiological values (permissive hypercapnia) and arterial PO2 to decrease to the lowest levels compatible with tissue oxygenation, all the while keeping tidal volumes to a minimum.


As mechanical ventilators increase in complexity, the clinician faces new choices regarding the ventilatory rate and tidal volume that are most appropriate for a given patient, and they must determine the mechanism (pressure or flow) and interval in which the breaths are triggered and the extent to which the machine complements the patient’s own breathing effort (Fig. 109-3). Many of the modes of mechanical ventilation defined by these options are designed to help the patient adapt to the ventilator or to discontinue ventilatory support (a process often referred to as weaning from mechanical ventilation).

Many ventilator modes are brand-specific adaptations purported to offer increased control, safety, or breathing autonomy for the patient. Most modern ventilators contain mechanisms that detect changes in pressure or flow created by spontaneous breaths. Refinements in these mechanisms have led to mechanical ventilation modalities that adjust the delivery of ventilator breaths to the patient’s rhythm and to the patient’s effort and breathing efficiency on a continuous basis. At an increasing rate, computer-based technologies are being used to ensure that the patient receives a target tidal volume and a target minute ventilation or to provide continuous adjustment of flow delivery based on the changing pressure-volume relationships of the respiratory system.

The old-fashioned controlled mechanical ventilation, a volume- or pressure-targeted mode in which all breaths are initiated and completed by the ventilator, has been replaced by a variety of synchronized modes, whereby breaths can be initiated by the patient and completed by the ventilator. For instance, assist-control ventilation allows the patient to trigger an unlimited number of ventilator breaths, with a backup rate activated when the patient’s breathing rate decreases below a certain threshold. In synchronized intermittent mandatory ventilation (SIMV), ventilator breaths are delivered at a preestablished rate but with a variable interval, allowing the patient an opportunity to initiate some of these breaths and to breathe freely from the ventilator circuit in the interval between ventilator breaths; this is often used to wean the patient from the ventilator. Pressure support is also a weaning mode in which the ventilator complements the early phase of the patient’s inspiratory effort at a preestablished inspiratory pressure. The patient’s breath triggers the ventilator’s contribution (by a decrease in airway pressure or the initiation of inspiratory flow). The pressure is released when the inspiratory flow decreases below a preset level, signaling that the breath is completed and expiration may start. The advantages of pressure support are that it is initiated by the patient and allows the tidal volume to vary depending on the individual’s respiratory drive. A potential disadvantage in children is that a large leak around an endotracheal tube may prevent the ventilator from initiating expiration in a timely manner. This will result in a prolonged inspiratory phase and patient discomfort.

In the last few years, there has also been a resurgence of noninvasive modalities designed to obviate the need for endotracheal intubation or to provide an earlier transition to spontaneous breathing. BiPAP(biphasic positive airway pressure) is a mode named after a commercial ventilator that provides ventilatory support through a nasal mask. A pressure-controlling valve switches between inspiration and expiration by alternating two levels of positive airway pressure, expiratory positive airway pressure (EPAP), and inspiratory positive airway pressure (IPAP), depending on whether the patient’s inspiratory flow exceeds a preset limit (EPAP to IPAP) or whether inspiratory flow decreases below a preset threshold or ceases altogether for a preset time (IPAP to EPAP). Continuous positive airway pressure (CPAP) allows the patient to breathe spontaneously from a pressurized constant flow circuit or by activating a demand valve at a constant expiratory pressure. It is often applied to newborns and small infants through a facial or nasal mask, nasal prongs, or a pharyngeal cannula. Care must be exercised when a demand valve system is used, because the high resistance of the valve can impose an intolerable load on the diaphragm, especially in infants and small children.


The idea of ventilating the lungs at frequencies well above the physiological range began as a creative method to support gas exchange while suppressing spontaneous respirations in experimental animals. It soon became clear that ventilatory rates higher than 2 Hz (breath/second) and as high as 12 Hz provided adequate oxygenation and ventilation, with tidal volumes that appeared lower than the estimated dead space.12 This observation gave rise to extensive speculation that alveolar ventilation during high-frequency ventilation involved singular mechanisms of gas transport. According to the same speculation, these mechanisms made it possible to ventilate the lungs with airway pressures lower than those needed to produce the bulk movement of gas characteristic of spontaneous and conventional mechanical ventilation. Today’s consensus is that gas exchange during high-frequency ventilation follows primarily the same principles as during conventional mechanical ventilation. Improvements in alveolar stability and regional gas-flow distribution may account for the gas-exchange advantages reported by some studies comparing high-frequency ventilation to other methods. The development of guidelines identifying prospectively the circumstances when the use of high frequencies would offer advantages is very much needed.

FIGURE 109-3. Diagrams of pressure and lung volume changes with several modes of ventilatory support. Arrows indicate triggered breaths as denoted by the small negative airway pressure inflection preceding the breath. See text for explanation and abbreviations.

Two types of devices are used to provide high-frequency ventilation: high-frequency oscillation and jet ventilation (Fig. 109-4). High-frequency oscillators generate a sinusoidal pressure waveform, which in modern ventilators is rectified to lengthen the expiratory phase and to minimize the gas trapping that was common in early prototypes. In addition to the ventilatory frequency, the operator can adjust the duration of the inspiratory time, the amplitude of the pressure oscillations, and the estimated mean airway pressure. Expiration is facilitated, a feature that may under some conditions limit expiratory flow. High-frequency jet ventilators use a cannula that is built into the wall of a special endotracheal tube or that is inserted directly into the airway lumen to deliver a rapid burst of inspiratory gas. Expiration is passive, a feature that limits the ventilatory frequencies achievable by this method to levels considerably lower than those used during high-frequency oscillatory ventilation.


Infants and children usually require sedation to alleviate the hardship and discomfort that are inevitable during ventilatory support, to facilitate adaptation to the ventilator, and to prevent movements that may dislodge the endotracheal tube when one is used. A combination of benzodiazepines and opioids is commonly used, taking advantage of the anxiolytic properties of the former and the analgesic and cough-suppressant effects of the latter. Both types of medications have the advantage of decreasing ventilatory drive when high levels of support are needed. However, this advantage turns into a disadvantage at the time of weaning, when the continued need for sedation must be balanced against the imperative that the patient sustains an appropriate ventilatory effort. Nondepolarizing neuromuscular blockers are given on rare occasions to suppress all respiratory muscular activity. The use of these medications must be reduced to a minimum because of the risks incurred if the endotracheal tube becomes dislodged and because the immobility that they cause prevents continuous assessment of the patient’s state of sedation and neurological function.

FIGURE 109-4. Diagrams comparing the inspiratory and expiratory phase (as indicated by the arrow direction) and gas-flow profile during high-frequency oscillation and jet ventilation. A: High-frequency oscillation uses a piston or a vibrating membrane to generate a quasisinusoidal flow in the airways. Expiration is facilitated by the vacuum action of the piston or membrane, resulting sometimes in airway collapse (as shown by the airway narrowing in the drawing). Although modern oscillators minimize this problem by lengthening expiration, mean alveolar pressure may be underestimated by mean proximal airway pressure. B: High-frequency jet ventilators use a cannula to deliver a burst or jet of inspiratory gas into the trachea. Additional gas is entrained from an indwelling endotracheal tube. Expiration is passive, a feature that poses a lower limit to the maximal frequency achievable by this method.



Advances in the biological inertness of the materials used to build membrane oxygenators and advances in the reliability and efficiency of the blood pumps made it possible in the 1970s to begin expanding the benefits of cardiopulmonary bypass from the operating room to the intensive care unit. Through the tenacity of a few pioneers like Robert Bartlett,13 ECMO became a viable method for respiratory and circulatory support in pediatric patients even after initial controlled studies failed to demonstrate differences in outcome compared to conventional mechanical ventilation in adults.14

Almost 30 years later, ECMO offers a potentially salvaging therapy to infants and children with hypoxemic respiratory failure and, perhaps as a more natural extension of the origins of the therapy, to infants and children with cardiogenic shock after surgical correction of complex congenital heart lesions or cardiomyopathy. Paradoxically, the rates of ECMO referral have decreased throughout the world, in great part because of new therapeutic options and generally improved care by more conventional or less invasive means. ECMO is now used much more rarely in newborns with respiratory failure, the primary beneficiaries of the early efforts, and is used more selectively as a means of support in infants and children with severe circulatory impairment. The reversibility of lung and myocardial injury is the most significant determinant of survival and therefore should be the most important consideration in establishing indications. Yet, refinements in technique and the ability to quickly institute support have led to new practical applications of ECMO, such as in patients suffering from unexpected circulatory arrest while in the hospital.15

FIGURE 109-5. Diagram of blood flow during venoarterial (left) and venovenous (right) extracorporeal membrane oxygenation (ECMO). During venoarterial ECMO, a portion of the venous return to the heart () is diverted into the ECMO circuit, where it undergoes O2 and CO2 exchange in the membrane oxygenator and is then returned under pressure to the aorta, where it mixes with blood from ventilated () and unventilated () areas of the lung. Systemic blood flow () is the sum of , , and . Arterial O2 content reflects the flow-weighted sum of these three components. During venovenous ECMO,  is diverted from the central venous circulation and then returned at a more proximal point in the same venous circulation. In this case,  and  have already undergone partial gas exchange in the membrane oxygenator.

ECMO is usually provided using two alternative circuit designs: venoarterial ECMO and venovenous ECMO (Fig. 109-5). At least in the United States, venoarterial ECMO is used most frequently in pediatric patients, because it is familiar and because it can be used simultaneously to support gas exchange and cardiac output. In the typical arrangement, a portion of the patient’s venous return is redirected via a cannula placed in the right atrium (usually through the internal jugular vein) to a venous reservoir in the ECMO circuit, where a rotary or centrifugal pump forces it sequentially through a membrane oxygenator and a countercurrent heat exchanger before returning it to the arterial circulation via another cannula positioned in the ascending aorta (usually through the carotid artery). In the oxygenator, blood flows in contact with a permeable membrane that separates it from a flow of oxygen and carbon dioxide. The composition and gas-flow rate are adjusted to optimize the oxygen and carbon dioxide contents of the blood exiting the oxygenator.

The ECMO pump provides the driving force for the return of blood to the aorta. Because blood is removed from the right atrium by gravity, it is the flow of venous blood that limits ECMO flow. Lowering the venous reservoir relative to the right atrium increases the flow of blood into the circuit. However, the increase becomes limited by the collapse of the right atrium or the systemic veins. Increases in intrathoracic pressure (eg, PEEP), decreases in the circulating blood volume, or venous dilation reduce the maximal flow that can be achieved for a given right-atrial pressure.

In venovenous ECMO, blood is removed and returned to the venous system. Consequently, this method supports only gas exchange and is not suitable for patients requiring cardiac support. Unlike venoarterial ECMO, systemic and pulmonary blood flow are the same. Extracorporeal carbon dioxide removal has been used successfully in adults with hypercapnic respiratory failure. It capitalizes on the extreme diffusibility of carbon dioxide across synthetic membranes to remove large volumes of this gas without the need for a substantial extracorporeal diversion of venous blood. Extra-corporeal carbon dioxide removal has no advantages in patients with refractory hypoxemia and is rarely applied in pediatrics patients.


Infants and children with uni- or biventricular failure often have no intrinsic abnormalities in lung function, and therefore they might not benefit from an external oxygenator and carbon dioxide exchanger. In recent years, the medical industry has produced several increasingly smaller ventricular assist devices that can be used to support the function of one or both ventricles. Many of these devices are now practical for use in small children and infants. Potential indications include cardiogenic shock after bypass surgery and cardiomyopathies. The availability of implantable devices has considerably improved the autonomy and quality of life of many pediatric heart transplantation candidates, often allowing them to be discharged from the hospital while they await their new organs.