Anesthesia for Children Undergoing Heart Surgery

15 Anesthesia for Children Undergoing Heart Surgery

Preoperative Evaluation

In the United States, 40,000 children are born each year with congenital heart disease (CHD),1 representing an incidence of 6 to 8 cases per 1000 live births.2 Children with chromosomal abnormalities such as trisomy 21 (i.e., Down syndrome) have a greater incidence of CHD. Having a sibling with CHD increases the risk of CHD, as does the presence of other congenital abnormalities.3 With improving diagnostic techniques, many children with CHD are diagnosed in the antenatal or early postnatal period. Associated with this improvement in diagnostic ability is a trend in most centers to undertake definitive repair earlier, with many patients undergoing corrective surgery in the neonatal period. Overall, about one half of all children with CHD undergo cardiac surgery in the first year of life, and about 25% undergo surgery in the first month of life.4,5

The perioperative management of children with complex cardiac defects requires a dedicated team of surgeons, cardiologists, anesthesiologists, intensivists, perfusionists, and nurses. Anesthesiologists caring for these children are challenged by some of the greatest physiologic aberrations encountered in clinical medicine. The anesthesiologists responsible for the care of these children require a comprehensive understanding of cardiac physiology and pathophysiology. The anesthesiologist must be able to adapt to each nuance of rapidly changing pathophysiology as it is encountered.

In addition to treating children with CHD, the pediatric cardiac anesthesiologist may also be responsible for the care of adults with CHD. This patient population with “grown-up CHD” is expanding as more children with CHD survive to adulthood. Along with CHD, they have some of the comorbidities of older age. The ideal approach for this group of patients is to be cared for in specialist units, but few are available. In the meantime, the care of these patients may fall to the most qualified physicians and the pediatric cardiac anesthesiologist. Some children have acquired rheumatic heart disease, some have cardiomyopathy, and others have undergone heart transplantation and require care from specialists.

For managing children with complex cardiac defects, there is increasing reliance on echocardiography and magnetic resonance imaging (MRI) to acquire diagnostic data. Although fewer children are being subjected to diagnostic angiography, more interventional cardiac catheterization procedures are being performed. Many conditions that would previously have been treated surgically are now treated in the angiography suite by interventional cardiologists, such as atrial septal defects (ASDs), patent ductus arteriosus (PDA), and ventricular septal defects (VSDs). Other interventions include dilating arteries with balloon catheters with and without stents and coiling of aberrant or excessive collateral vessels. The pulmonary artery is commonly balloon dilated and stented, and coarctation of the aorta is treated similarly by balloon dilation. Stenotic valves are also commonly dilated. These procedures have led to the risk being transferred from the angiography suite to the operating room.6 For the individual child, there has been a dramatic decrease in morbidity as increasing numbers of conditions are treated in the angiography suite, but the risks of complications that occur in the angiography suite have increased as more complex procedures are performed (see Chapter 20).

The Preoperative Visit and Evaluation

The preoperative visit is an important part of the overall management of anesthesia for children with CHD.7 The preoperative visit has several aims:

Medical Assessment

The anesthesiologist must have a clear and detailed understanding of the cardiac pathophysiology, the surgery to be undertaken, and associated congenital abnormalities or medical conditions. The medical assessment includes collation of information from the history, physical examination, and review of imaging and laboratory data. Most diagnostic information is obtained from the medical record. Particular attention should be paid to the echocardiographic, MRI, angiographic, and other imaging data; the chest radiograph; and the electrocardiogram. Many centers have joint cardiac conferences where decisions about treatment are discussed in a multidisciplinary forum. Reports from these meetings are valuable in the preoperative assessment.

In addition to gathering this specific diagnostic information, a directed history and physical examination should be performed to assess the overall condition of the child. Attention should be directed toward assessing the degree of cardiac failure, cyanosis, or risks of pulmonary hypertension and the prior surgical procedures and how this information may alter access to the central circulation and placement of invasive monitors. The general nutritional state of the child should be assessed; poor growth and development may be a sign of severe CHD. Other information should be sought that may have a bearing on the anesthetic plan. For example, is this repeat surgery, and does it require a repeat sternotomy? This has a bearing on line placement because a femoral bypass may be required, and it should be avoided for line placement. The previous use of aprotinin is important because the risk of anaphylaxis is increased in a second exposure and particularly if it has been given in the previous 6 months.8

The type of surgery to be performed is important. For example, if a Blalock-Taussig shunt is placed on the left, the arterial line should not be placed in the left arm because the trace will be lost or distorted during subclavian cross-clamping. If a Glenn shunt is planned, a short internal jugular line can be useful to monitor pulmonary artery pressure, but it should be removed early in the postoperative period so as not to risk the formation of thrombosis in the superior vena cava (SVC).

Good veins should be sought and marked for the application of local anesthetic cream. This is useful in sick children even if an inhalational induction is planned because it allows placement of a venous cannula during a very light plane of anesthesia and avoids myocardial depression from high concentrations of inhalation anesthetics.

Prescribing Premedication

The use of sedative premedication can be useful in cardiac anesthesia, but this practice varies widely. Numerous medications may be used, and numerous recommendations exist. Premedication for infants younger than 6 months of age is usually unnecessary. Premedication for older, healthy children who show little anxiety and with whom good preoperative rapport can be established is often unnecessary. However, older children, particularly those who have undergone previous surgery, have fears about anesthesia and surgery. It is important to address the fears of these children. Sedative premedication may play a pivotal role in achieving adequate anxiolysis for separation from the parents and induction of anesthesia. I prefer to avoid premedication in children with severe congestive heart failure. Cyanotic children such as those with tetralogy of Fallot (TOF) often benefit from sedative premedication because crying and struggling during induction of anesthesia may worsen their cyanosis. However, it is important that these cyanotic children are well supervised after premedication because they have a blunted response to hypoxia.9 In the United States, supplemental premedication is sometimes administered under the direct supervision of the anesthesiologist in the preoperative facility, providing for a calm child and gentle separation from the parents. In the United Kingdom, where induction of anesthesia takes place in a dedicated anesthesia room, parents are present until after the induction, often making additional premedication unnecessary.

The most common premedication is oral midazolam (0.5 to 1.0 mg/kg).10 However, the effect of midazolam may be unpredictable and may cause dysphoria. Numerous other medications have been recommended for this purpose, including ketamine, clonidine, temazepam, and chloral hydrate. The use of these drugs is often dictated by local preferences and is not always evidence based.

Upper Respiratory Tract Infection and Cardiac Surgery

Otherwise healthy children undergoing elective noncardiac surgery in the presence of an upper respiratory infection (URI) are more likely to suffer respiratory complications (Table 15-1). These complications typically are minor, are easily managed, and usually result in minimal morbidity.1113 The decision about whether to proceed with noncardiac surgery in a child with a URI is made on an individual basis.14

TABLE 15-1 Diagnosis of Upper Respiratory Tract Infection

Data from Schreiner MS, O’Hara I, Markakis DA, Politis GD. Do children who experience laryngospasm have an increased risk of upper respiratory tract infection? Anesthesiology 1996;85:475-80.

The decision to proceed with cardiac surgery in children is difficult. Although children with cardiac failure are prone to multiple URIs, they may also have signs that can mimic URIs. Surgery may be relatively urgent, and postponing surgery exposes the child to an increased risk. Cardiac surgery in children with URIs results in a prolonged stay in the ICU and prolonged ventilation times, although overall hospital stay is not prolonged. There is an increased incidence of pulmonary atelectasis and an increase in postoperative bacterial infections. There appears not to be any statistically significant increase in mortality rates (4.2% with URIs versus 1.6% without URIs) or long-term sequelae in children with URIs who undergo cardiac surgery. The URI group was significantly younger and smaller, which may account in part for the greater but statistically insignificant increased mortality rate.15 Although this increase in mortality was not statistically significant, it does raise concerns about the risks posed by URIs before cardiac surgery. Children who are scheduled for a Glenn shunt or completion of the Fontan circulation may be at particular risk because an increase in pulmonary vascular resistance (PVR) can adversely affect outcome. It is prudent to postpone surgery in a child with a URI who is scheduled for elective cardiac surgery. If the surgery is urgent, discussion with the surgical team is required to correctly assess the risks and benefits to the child.

Perioperative Challenges in Pediatric Cardiac Anesthesia

Intracardiac Shunting

In CHD, much of the pathophysiology involves communications between chambers or vessels that are normally separate, resulting in shunting of blood between ventricles, atria, the great arteries, or a combination of these, depending on the nature of the lesion. Management of shunting is a major consideration during anesthesia and requires an understanding of the factors that control shunting.

Impaired Hemostasis

Hemostasis is impaired after bypass in infants and children. This results from a combination of immature coagulation factor synthesis, hemodilution after bypass, and a complex interaction involving consumption of clotting factors and platelets. At birth, the levels of vitamin K–dependent coagulation factors in healthy, full-term neonates are only 40% to 66% of adult values. During the first month of life, these levels increase to 53% to 90% of adult values.20 However, in children with CHD, especially those with cyanosis or systemic hypoperfusion, coagulation factors often continue to be depressed due to impaired hepatic protein synthesis. Although antithrombin III levels are also low, true heparin resistance is rare in infants because of parallel decreases in coagulation factors.

At the onset of cardiopulmonary bypass (CPB), the introduction of the prime volume, which is two to three times greater than the child’s blood volume, dilutes the factor concentrations, particularly fibrinogen, to 50% of values before bypass and the platelet count to 30% of values before bypass. This degree of dilution occurs even when the pump circuit is primed with whole blood. Greater dilution may occur when packed red cells are used in the priming volume. At the conclusion of neonatal bypass, the activity of clotting factors is often extremely low, the fibrinogen concentration is frequently less than 100 mg/dL, and the platelet count has been reduced to 50,000 to 80,000/mm3.21,22 In addition to these quantitative changes, functional changes in the platelets occur during bypass. Extracorporeal circulation causes a loss of platelet adhesion receptors, activation of platelets, and formation of leukocyte-platelet conjugates. Platelet adhesion receptors are more depressed in children with cyanotic compared with acyanotic cardiac defects. Heparin also impairs platelet function independent of CPB.23

Cardiac surgery is associated with significant activation of the fibrinolytic system.24 Inadequate heparin levels during CPB may also contribute to postoperative bleeding because inadequate anticoagulation may allow activation of the hemostatic pathways. Activation causes the consumption of platelets and clotting factors. The standard measurement of anticoagulation, the activated clotting time (ACT), shows a poor correlation with heparin levels in children undergoing CPB.25 In one study, the use of heparin monitoring and heparin titration was associated with larger doses of heparin but smaller doses of protamine for antagonism.26 Activation of clotting cascades is also reduced, decreasing bleeding in the postoperative period.26 As a result of this multifactorial coagulopathy, blood loss is a greater problem in children than in adults and is a particular problem in neonates and small infants (see Chapter 17).27

Strategies to Reduce Bleeding after Bypass

In an effort to normalize factors and platelets to effective concentrations, some medical centers use fresh whole blood in the cardiopulmonary circuit prime. In adult patients and an in vitro aggregation study, transfusion of fresh whole blood provided equal or greater hemostatic and functional benefit when compared with transfusion of platelet concentrates. In children, transfusion of fresh whole blood less than 48 hours from harvest is associated with less blood loss compared with transfusion of reconstituted whole blood (e.g., packed erythrocytes, fresh frozen plasma [FFP], and platelets).28 However, fresh whole blood is often difficult to obtain. The units must be refrigerated for 24 to 48 hours while donor screening is performed, and storage causes significant platelet injury. Insistence on fresh whole blood places tremendous pressures on the transfusion service and donor center to coordinate the matching of donor types with recipient needs.

In practice, individual component therapy is used. In neonates and small infants with dilutional coagulopathy, platelets should be given in combination with cryoprecipitate to correct the defect in clotting. An initial dose of 10 mL of platelets/kg of body weight may need to be repeated. Platelets may be administered if bleeding persists and the platelet count is less than 100,000/mm3.29 Cryoprecipitate contains high concentrations of fibrinogen, factor VIII, von Willebrand factor, and factor XIII. Fibrinogen and von Willebrand factor are required for platelet adhesion and aggregation to occur. Platelet adhesion and aggregation are the fundamental first steps in primary hemostasis (see Chapter 10). The subsequent step of platelet degranulation switches on the entire coagulation cascade and cannot take place without adhesion and aggregation.30 Administration of FFP, for which there is no evidence of effectiveness in treating this type of coagulopathy, to the infant may excessively dilute the red cell mass and platelets.31

Transfusion guidelines have been described for adults and have been shown to reduce postoperative bleeding and transfusion requirements.32,33 However, similar guidelines have not been forthcoming for children in whom the practice appears to be more empirical. This is a less than ideal situation, and more work is urgently needed to produce well-validated guidelines. The thromboelastogram and the platelet count may be used to identify which children are likely to bleed after cardiac surgery.


The antifibrinolytics used in pediatric cardiac surgery include ε-aminocaproic acid (EACA) and tranexamic acid (TA). In many countries, aprotinin is no longer available, and in other countries, it has only very limited availability after its marketing license was withdrawn due to safety concerns. EACA and TA are lycine analogues that reduce bleeding after cardiac surgery in adults and children.34,35 They do not exert any antiinflammatory activity. Doses for pediatric cardiac surgery have not been clearly established.

Aprotinin is a serine protease inhibitor that has been studied thoroughly in adults. Early evidence demonstrated that it reduces bleeding, reduces the time taken to extubation, shortens ICU stay, and reduces overall mortality rates.36 However, subsequent studies have contradicted these earlier findings.37 The same volume of evidence has not been published for children, although several studies suggest that it is effective in reducing bleeding and that it reduces the time spent on the ventilator and in the ICU.3841 An increased risk of renal failure or stroke in adults undergoing revascularization surgery has been reported.42 The same investigators reported an increase in the 5-year mortality rate for adults after the use of aprotinin in revascularization surgery.43 Aprotinin has increased the 30-day mortality rate by as much as one third compared with TA or EACA.37 Comparable data for children have not been forthcoming (see Chapter 18). Aprotinin is available for use in New Zealand, Australia, and Canada. It appears that the early data regarding increased death rates have not been supported by subsequent studies and that the benefits may outweigh risks in specific populations.44

Topical Agents

The use of topical agents to promote clot formation and reduce bleeding in children after cardiac surgery is common. The most commonly used topical agents are fibrin sealants. Fibrin sealants mimic the stages of the blood coagulation process. Unlike the synthetic adhesives, they are biocompatible.45 Fibrin sealants are usually sourced from plasma components, and most contain virally inactivated human fibrinogen and thrombin with different quantities of factor XIII, antifibrinolytic agents, and calcium.45 When the fibrinogen and thrombin are mixed during the application process, the fibrinogen is converted to fibrin monomers. This results in the formation of a semirigid fibrin clot. By mimicking the later stages of the coagulation process, these sealants stop bleeding and assist in wound healing.45 They have significantly reduced bleeding in children.46


Desmopressin acts by increasing plasma concentrations of factor VIII and von Willebrand factor. It has been effective in reducing bleeding after CPB in adult cardiac surgery.49 Unfortunately, studies in children failed to demonstrate a similar effectiveness in reducing bleeding or transfusion requirements.50

Anesthesia Management for Surgery Requiring Cardiopulmonary Bypass


Noninvasive monitoring during pediatric cardiac surgery includes pulse oximetry, five-lead electrocardiography, an automated blood pressure cuff, a precordial or esophageal stethoscope, continuous airway manometry, inspired and expired capnography, anesthetic gas and oxygen analysis, multiple-site temperature measurement, and volumetric urine collection. The pulse oximeter is especially important when managing children with congenital cardiac disease. At least two probes should be placed on different limbs in the event that one fails during the procedure. In children with cyanotic heart disease, conventional pulse oximetry overestimates arterial oxygen saturation as saturation decreases51; this error tends to be exacerbated in the presence of severe hypoxemia.52 When monitoring children with a shunt across the ductus arteriosus, a probe should be placed on a right hand digit to measure preductal oxygenation, and a second probe should be placed on a toe to measure postductal oxygenation (children with a right-sided aortic arch may require the probe to be placed on a left-hand digit). Children undergoing repair of coarctation of the aorta should be monitored with a pulse oximeter on the right upper limb, because it may be the only reliable monitor during the repair, and blood pressure cuffs should be placed before and after the coarctation. These two cuffs may be cycled and the differential documented before and after surgical correction.

Monitoring end-tidal carbon dioxide tension (Petco2) is of value in most children. However, in children with cyanotic-shunting cardiac lesions, the Petco2 measurement may be less reflective of Paco2 because of ventilation-perfusion mismatching.53 Arterial blood gases are the most accurate measure of the adequacy of ventilation and oxygenation. To provide rapid decision making, it is helpful to have the blood gas analysis machine located in or near the cardiac operating room.54

Monitoring ionized calcium concentrations is essential during surgical procedures in which significant quantities of citrated blood are infused rapidly or when entire blood volumes are replaced. Neonates are particularly prone to disturbances in their ionized calcium concentration when citrated whole blood, FFP, or platelets are infused. Those with limited cardiac reserve tolerate ionized hypocalcemia poorly because of their greater sensitivity to the myocardial effects of citrate infusion (see Chapter 10).55 In isolation, the total serum calcium concentration is misleading.

Temperature monitoring during CPB is a critical guide to adequate brain cooling and to appropriate rewarming before separation from bypass. Because it is not practical to measure brain temperature directly, surrogate measuring sites are used. The tympanic membrane, nasopharyngeal, and rectum have been used; the nasopharyngeal site most closely matches true brain temperature and is the site at which temperature is most often monitored. The tympanic and rectal sites tend to overestimate the brain temperature.56,57 Measurement of skin temperature gives an indication about peripheral perfusion and provides information about adequate peripheral rewarming.

After induction of anesthesia, an arterial catheter should be placed in children who will undergo CPB. The radial artery may be percutaneously cannulated with relative ease, even in infants. In small infants, the femoral arteries are frequently used for arterial access, and the axillary arteries are commonly used. The brachial artery is avoided because it is an end artery. Catheters placed in the dorsalis pedis or posterior tibial artery often provide inaccurate hemodynamic data, especially after separation from bypass, and it may become difficult to sample blood for laboratory testing. In the rare circumstance that peripheral arterial cannulation cannot be accomplished, the surgeon may place a catheter in the internal mammary artery after sternotomy, and a sterile monitoring line may be passed over the drapes. Ultrasound guidance often is used for insertion of arterial catheters.

Central venous lines can be very useful. For cardiac surgical procedures, there are two commonly used methods of obtaining central access. The decision of which to use may be determined in part by institutional bias. In the first method, the cardiac surgeons expose the heart quickly and have it available for inspection and estimation of filling pressures. Central lines can be readily established from the field and handed off to the anesthesia team. These transthoracic central lines are useful but carry a small amount of risk.28 In the second method, percutaneous insertion of central venous lines is particularly indicated for long, complex procedures, especially when access to the infant is limited or the heart is not exposed. Percutaneous cannulation of the central circulation through the internal jugular approach or the subclavian approach has been demonstrated to be safe.29 However, it is important to appreciate that insertion of central venous lines through the internal jugular or subclavian route may fail or may be associated with pneumothorax, hemorrhage, and hematoma formation after puncture of major arteries.2931 Cannulation of the external jugular vein may avoid some of these serious complications when the catheter can be successfully threaded into the central circulation.32 Increasingly, ultrasound-guided techniques are being employed to establish central venous access (see Chapter 48). In the United Kingdom, the use of ultrasound for the placement of these lines is recommended by the National Institute of Clinical Excellence (NICE), and ultrasound is used routinely for the placement of central lines.

In children with unrestrictive VSDs or ASDs, including hearts with a single ventricle or single atrium, central venous pressure is equivalent to left ventricular filling pressure. Cannulation of vessels that drain into the SVC should be approached with caution in children with univentricular anatomy who may undergo the Fontan procedure, because thrombosis of the SVC can be a devastating complication. In such circumstances, the femoral veins may be the preferred sites for venous pressure monitoring.

Pulmonary arterial catheters in children with intracardiac defects usually provide little more information than a simple central line, are difficult to insert without fluoroscopy, and may not provide meaningful measurements of cardiac output. As a result, they are rarely used in pediatric cardiac patients.

Transesophageal Echocardiography

Use of perioperative echocardiography has become the standard of care in the United States.58,59 In adult practice, anesthesiologists usually perform the TEE, but in children, the TEE is more commonly performed by a pediatric cardiologist. This may reflect the increased complexity of congenital lesions and the difficulty in accurately assessing these lesions and their repairs. TEE has been cost-effective when used routinely during pediatric cardiac surgery.60 The use of TEE can have a significant impact on surgical and medical management. In one large study, a second bypass run was undertaken in 7.3% of cases based on the findings of the TEE. There was a surgical alteration in the management of 12.7% and medical alteration in 18.5% of cases. Pediatric cardiac anesthesiologists usually can perform TEE before and after bypass if they have received adequate training.61

The introduction of small probes with multiplane capability has greatly increased the use of TEE, even in infants and neonates.62,63 In 1999, a survey of centers in the United States indicated that 93% used intraoperative echocardiography and that all but one used TEE.64 The American Society of Echocardiography and the Society of Cardiovascular anesthesiologists have published guidelines for performing a comprehensive intraoperative TEE in adults65 and children.66

Although the use of TEE in children usually is safe, complications do occur and may be more common in small infants.67 Complications include damage to the mouth, oropharynx, esophagus, and stomach. Other complications include hemodynamic disturbance as a result of compression of the left atrium or other structures. Interference with the airway also occurs in a small number of cases. This includes inadvertent extubation, right main-stem intubation, and compression of the tracheal tube. However, the overall incidence is small, approximately 2%.68 Information gathered from the TEE examination takes place before and after bypass and may be divided broadly into two categories: hemodynamic assessment with monitoring and structural diagnostic information. Hemodynamic information includes information about ventricular function and filling.69 Diagnostic information relates to confirmation or otherwise of preoperative findings and assessment of the adequacy of the surgical repair.

Induction of Anesthesia

In the United Kingdom, induction of anesthesia takes place in an anesthesia room, which is immediately adjacent to the operating room, and the parents are almost always present for induction. Anesthesia is commonly induced while the child is sitting with or being held by a parent. It is possible to involve some parents to the extent that they can hold the mask for the child as he or she is anesthetized. After the child is asleep, he or she is transferred to the anesthetic trolley, where venous and arterial access is secured and the trachea is intubated. This practice differs from most centers in North America, where induction of anesthesia occurs in the operating room.

Induction can be achieved using an intravenous or inhalational technique. Ketamine may be given intramuscularly or orally when intravenous access is difficult and mask induction is refused. The type of induction should be tailored to the child and the cardiac defect. If intravenous access has already been established, an intravenous induction is preferred. If intravenous access has not been achieved, a decision is made about the optimal induction technique. In severely ill children, it is advisable to obtain intravenous access before induction of anesthesia. The use of local anesthetic cream such as EMLA or Ametop Gel helps to reduce the pain of injection. Use of local anesthetic cream is also helpful if an inhalational induction is to be used because it allows insertion of an intravenous cannula during a much lighter plane of anesthesia. This requires that the appropriate veins are selected during the preoperative visit and that clear instructions are given to nursing staff about where the cream should be applied.

The most common inhalational induction agent is sevoflurane. Sevoflurane is very rapid acting and should be used with care in the child with CHD because high concentrations can produce bradycardia, hypotension, and apnea if not titrated carefully. Concentrations should be rapidly reduced after an adequate level of anesthesia is achieved. In children who are cyanotic with a right-to-left shunt and reduced pulmonary blood flow, inhalational inductions are slow. The addition of nitrous oxide can aid an inhalational induction in two ways. First, because it is odorless, it can be started before the introduction of the sevoflurane, allowing the child to be somewhat sedated before the stronger smelling agent is started. Second, it allows a smoother and more rapid induction compared with sevoflurane alone. Concentrations of up to 70% nitrous oxide can be used to smooth induction of anesthesia even in cyanotic children, but the nitrous oxide should be replaced with air and oxygen or 100% oxygen as soon as intravenous access is obtained and a muscle relaxant is given. It is not always necessary to use a mask for an inhalational induction because cupped hands are often more acceptable to the child, particularly one who is frightened of the mask. It is important to tell the child about each event before it happens and to demonstrate the action on yourself, a parent, or toy animal. You should also offer the child the opportunity to hold the mask, or if the child is accompanied by a parent, offer the child the choice of the parent holding the mask. Good premedication often aids this process (see Chapter 4).

For sick children in whom it may be preferable to use an intravenous induction, various options are available. For example, in neonates with coarctation of the aorta or with hypoplastic left heart syndrome who are not ventilated before coming to the operating room, one approach is to administer fentanyl in a dose of 2 to 3 µg/kg, followed by pancuronium and then by a very low dose (i.e., sedative dose) of sevoflurane or isoflurane. Fentanyl obtunds the hypertensive response to intubation, and the pancuronium maintains cardiac output by maintaining the heart rate. The very-low-dose volatile agent provides the sedation or anesthesia. In older children, etomidate is a very good induction agent, providing stable hemodynamics, although it does cause pain on injection. Ketamine is also widely used for intravenous induction in neonates and older children. Ketamine maintains or increases blood pressure, heart rate, and cardiac output. The exact mechanism of these effects of ketamine is unknown; ketamine may stimulate the release of endogenous stores of catecholamines, although it is a negative inotrope in the denervated heart.72 This negative inotropic effect may make ketamine a poor choice in children in whom catecholamine stimulation may already be maximal, such as in severe cardiomyopathy. It may also be a poor choice if tachycardia is undesirable, such as in the case of aortic stenosis.

Monitoring should ideally be applied before induction begins, but applying monitoring can upset the child, which can be detrimental (e.g., the child with TOF who begins to cry and precipitates a “tet spell”). A pulse oximeter probe may be the only monitor that is applied before induction of anesthesia. Sevoflurane or isoflurane may provide another advantage by offering a degree of ischemic preconditioning to the heart and to other organs, particularly the brain and kidney. In a double-blind study of adult patients undergoing coronary bypass grafting, exposure to 4% sevoflurane for 10 minutes before cross-clamping reduced the degree of myocardial dysfunction and renal damage postoperatively.73 It is thought that the same effect is observed in children.74

Institution and Separation from Bypass

Before initiation of CPB, the surgeon requests heparin to be given. After administration of heparin (preferably flushed through a central venous catheter) but before the initiation of bypass, the ACT should be determined. The ACT measurement should be at least three times greater than the baseline value. When bypass is started, any additional anesthetic drugs should be administered, and ventilation should cease. Both hypertension and hypotension may complicate bypass. Blood pressure may be controlled within the normal range using α-adrenergic blockers or agonists such as phenylephrine and phentolamine. The child is usually cooled at this stage, using the nasopharyngeal temperature as a guide. If the heart is to be stopped, cardioplegia is given by the perfusionist after the aorta is cross-clamped to provide myocardial protection during the period of ischemia. Cardioplegia is usually repeated every 20 to 30 minutes, although it is not required if the surgery is performed while the heart is beating. Myocardial damage is related to the duration of the aortic cross-clamping and the effectiveness of the myocardial protection.

At an appropriate time during the surgery, the cross-clamp is removed, and perfusion to the heart is restored. The heart usually starts to beat in normal sinus rhythm, although this is not always the case. In the early phase of reperfusion, it is possible for various degrees of heart block to occur. However, these effects are usually short-lived, and as the effects of cardioplegia wear off, normal sinus rhythm is usually restored. In addition, heart block may result from damage to the conducting system during surgery.

After release of the cross-clamp, any inotropes or vasodilators that are required are usually started. Rewarming may have begun before release of the cross-clamp, but more commonly, the child is rewarmed after release of the clamp.

When the child has adequately rewarmed, as reflected by a normal core and minimal core-peripheral temperature difference, good heart function has returned, the child’s lungs are adequately ventilated, and any inotropes required have been started, the child is ready to be separated from bypass. If a TEE probe is in place, the heart should be scanned for the presence of air. If air is present, further de-airing should occur before attempting to come off bypass. In the initial stages after separating from bypass, additional volume can be administered by the perfusionist through the aortic cannula, usually under the direction of the surgeon or anesthesiologist. Many centers institute modified ultrafiltration at this point. This involves taking arterial blood from the aortic cannula and passing it through the ultra-fine filter. This blood, which is oxygenated and warm, is then reinfused into the right atrium. When this process is complete, a thorough TEE examination can be undertaken.

When the team is satisfied with the TEE result, the surgeon asks for protamine to be administered. Before this is done, the perfusionist and the surgical team should be informed that protamine is about to be administered. The surgeons should remove any pump suckers from the field, and the perfusionist should stop all pump suction. This is done to ensure that no protamine enters the bypass circuit in case it is necessary to go back on bypass for any reason. The ACT can be checked along with the blood gas analysis. The ACT should return to levels before bypass. Any blood products required are usually given after the administration of protamine, usually while the surgeons are achieving hemostasis. As soon as the chest is closed, the child can then be transferred to the ICU.

Control of Systemic and Pulmonary Vascular Resistance during Anesthesia

In some children with hypoplastic left heart syndrome (HLHS) who present for a Norwood procedure, excessive blood flow to the lungs resulting from a relatively low PVR and a relatively high SVR steals blood from the systemic circulation, leading to hypotension, myocardial ischemia, and progressive acidosis. However, when the reverse occurs and the PVR is greater than the SVR, the child develops progressive desaturation. Similar pathophysiology exists with other duct-dependent circulations and to some extent with other shunting lesions. It may prove difficult to manipulate the SVR and PVR predictably because control of PVR is poorly understood, vasoactive drugs usually are distributed on both sides of the circulation, and pharmacologic attempts to modify shunting have produced unpredictable results.75 Despite these problems, several techniques have proved useful in manipulating the relative PVR and SVR. Potent inhalational anesthetics appear to reduce SVR more than PVR. PVR is decreased in children by increasing inspired oxygen to 100% and by hyperventilation to a pH of 7.6 or greater. Positive end-expiratory pressure, acidosis, hypothermia, and the use of 30% or less inspired oxygen can increase PVR. Because vasoconstrictors such as phenylephrine increase SVR more than PVR, they are effective acutely in reducing right-to-left shunting and increasing left-to-right shunting in the operating room.

During cardiac surgical procedures, a direct method of selectively increasing PVR or SVR is to have the surgeon place partially obstructing tourniquets around pulmonary arteries or the aorta to increase resistance so that flow to the opposite side of the circulation increases. Although these are only temporary measures, they may reestablish a better relative balance of resistances and a more normal physiology in a deteriorating clinical situation.

Anesthetic Drugs Used in Pediatric Cardiac Anesthesia

Inhalational Agents


Sevoflurane is the induction agent of choice in pediatric anesthesia.76,77 It is associated with little myocardial depression or dysrhythmias.7880 It has specific advantages over halothane when used in children with CHD, particularly in children younger than 1 year of age and in cyanotic children.81 In contrast to halothane, sevoflurane causes no reduction in heart rate at 1.0 and 1.5 minimal alveolar concentrations (MACs) in healthy children compared with awake values.82 However, at greater concentrations, it can slow the heart rate and cause respiratory depression. Both features are important in children with CHD because a slow heart rate reduces cardiac output and hypoventilation leads to hypercarbia and hypoxia, which can increase PVR. In the absence of nitrous oxide, sevoflurane causes less depression of myocardial contractility than halothane during induction of anesthesia. Sevoflurane does cause a mild decrease in SVR, but in common with halothane and isoflurane, it does not perturb the shunt between the right and left sides of the heart through an ASD or VSD when it is given in anesthetic concentrations of about 1 MAC in 100% oxygen.83 Sevoflurane has caused conduction abnormalities in susceptible patients.84 It should also be used with great caution in children with severe ventricular outflow tract obstruction (see Chapter 6).85


At equipotent concentrations, isoflurane causes similar hemodynamic depression in neonates and infants compared with halothane. Isoflurane typically is not used for induction of anesthesia because of the high frequency of laryngospasm (greater than 20%).86 Inadequate ventilation because of laryngospasm or other causes quickly leads to large increases in PVR due to hypoxemia and hypercarbia. This increase in PVR and the resulting pulmonary hypertension is poorly tolerated in small children with heart disease, especially in the presence of right-to-left shunting (see Chapter 6).


In the United States and the United Kingdom, the use of halothane has all but ceased, but it is still widely used in other parts of the world. It is included here for completeness. Uptake of halothane in infants younger than 3 months of age is more rapid than it is in adults. This also is the case for the uptake of halothane by the myocardium.87 Although the effects of halothane on the human neonatal myocardium are unknown, young rodents have a reduced cardiovascular tolerance for halothane but require greater amounts for anesthesia.88 Studies have shown a significant incidence of hypotension with bradycardia in infants with normal cardiovascular systems during induction with halothane.89 During induction of anesthesia in normal infants, halothane decreases the cardiac index to 73% of awake values at 1.0 MAC and to 59% at 1.5 MAC.90 The MAC for halothane in infants 1 to 6 months of age is the greatest of any age group.91 This increased anesthetic requirement in infants, combined with the immaturity of their cardiovascular system, explains in part the relative cardiovascular intolerance of halothane by infants. Atropine has been used intramuscularly before induction to partially compensate for the myocardial depression of halothane by reducing bradycardia and hypotension. Although halothane may produce some degree of hypotension, an increase in arterial saturation in children with cyanotic CHD may occur.92

A careful induction with sevoflurane is usually well tolerated in children with mild to moderate heart disease. However, large concentrations of potent inhalational agents may be an unwise choice for induction in young infants with severe cardiac disease. In children of any age with marginal cardiovascular reserve and in those with severe desaturation of systemic arterial blood due to right-to-left shunting, inhalational anesthetic-induced myocardial depression and systemic hypotension are poorly tolerated. A more appropriate use of these anesthetic agents in children with severe heart disease is the addition of low concentrations of the inhalational agent to control hypertensive responses after an intravenous induction (see Chapter 6).

Nitrous Oxide

Nitrous oxide should be avoided for maintenance of anesthesia in children with CHD because of the risk of enlarging intravascular air emboli and the potential to increase the PVR. Nitrous oxide may expand microbubbles and macrobubbles, increasing obstruction to blood flow in arteries and capillaries. In all children with right-to-left shunts, there is a potential for these bubbles to be shunted directly into the systemic circulation and coronaries. Care must be taken to ensure that no air bubbles are accidentally injected into the veins. Adverse outcomes after coronary air embolism are exacerbated by nitrous oxide.93 The hemodynamic effects of venous air embolism are increased by nitrous oxide, even without paradoxical embolization.94 In children with preexisting right-to-left shunts, paradoxical air embolism is clearly a potential problem; but even those with large left-to-right shunts can transiently reverse their shunts. This is particularly true during coughing or a Valsalva maneuver, when the normal transatrial pressure gradient is reversed. Several studies have demonstrated right-to-left shunting of microbubbles of air after injection of saline into the right atrium during these maneuvers.9597 Because coughing and Valsalva maneuvers may occur during anesthesia induction, even the most rigorous attention to avoiding air bubbles in intravenous lines may not prevent small amounts of air from reaching the systemic circulation. Microbubbles have also been observed after CPB.98

Nitrous oxide can increase PVR in adults.99,100 However, in a 50% inspired concentration, it does not appear to affect PVR or pulmonary artery pressure in infants.101 Nitrous oxide mildly decreases cardiac output at this concentration.102 Avoidance of the use of nitrous oxide has been suggested in children with limited pulmonary blood flow, pulmonary hypertension, or depressed myocardial function. In the well-compensated child who does not require 100% inspired oxygen, nitrous oxide (usually at concentrations of 50%) may be used during induction of anesthesia but discontinued before tracheal intubation. If a reduced inspired oxygen concentration is indicated to maintain an appropriate balance between PVR and SVR after tracheal intubation, air may be added to the inspired gas mixture (see Chapter 6).

Intravenous Induction Agents


Ketamine is a dissociative anesthetic agent that is a good analgesic. It increases blood pressure, heart rate, and cardiac output. Although the mechanism of the stimulation of blood pressure and heart rate has not been established, it is thought to stimulate the release of endogenous stores of catecholamines. Ketamine exerts a negative inotropic effect on the denervated heart.103 I think that this combination of effects makes it a poor choice for children in whom sympathetic stimulation may already be maximal, such as in those with severe cardiomyopathy. It is also a poor choice if tachycardia is undesirable, such as in a child with aortic stenosis. Ketamine is thought to have minimal effect on PVR in children with CHD as long as the airway and ventilation are well preserved,104,105 although it has occasionally increased PVR in children undergoing cardiac catheterization. Ketamine is quite a versatile anesthetic that may be administered intramuscularly and orally when intravenous access is difficult or an inhalational induction is contraindicated. The usual intravenous dose of 2 mg/kg produces a very predictable response, and an intramuscular dose of 8 to 10 mg/kg (combined with 0.1 mg/kg of intramuscular midazolam) is less predictable. The oral dose of ketamine is 5 to 6 mg/kg. The use of ketamine varies greatly from one institution to another, with some units using it extensively and others using it rarely (see Chapter 6).


Etomidate is a very safe drug, with an LD50 to ED50 ratio of 26 in animals models.106 This ratio indicates that the lethal dose (LD) is 26 times greater than the effective dose (ED). It is a short-acting anesthetic with little effect on systemic blood pressure, heart rate, and cardiac output after a single dose in healthy children.107 Etomidate has a favorable hemodynamic profile even when used in shocked children and appears to have a low risk of clinically important myoclonus or status epilepticus, pain on intravenous injection, and nausea and vomiting.108,109 The major concern about etomidate is the increased mortality rates reported when it is administered as a continuous infusion. This grave side effect has been attributed to adrenal suppression.110112 The inhibition of steroid synthesis occurs after a prolonged infusion and after a single dose of etomidate, and it has created a controversy about its use as an anesthetic agent, particularly in the ICUs in some jurisdictions.113 However, newer analogues of etomidate have addressed these deficiencies and may lead to a surge in its use in the future (see Chapter 6).


Propofol is a rapidly acting intravenous hypnotic agent that may be administered as a single dose or by continuous infusion. It has no analgesic properties. Propofol has mild antiemetic properties.114 Its short duration of action is the result of rapid redistribution and metabolism, which also allows the drug to be given by continuous infusion without accumulation. Induction doses decrease SVR, blood pressure, and cardiac output; the effect on heart rate varies. The ED50 for propofol in infants and small children is greater than it is in adults.110117 If propofol is given very slowly, smaller doses are required to achieve the anesthetic state, although the induction time increases. A slower infusion also results in more stable hemodynamics.118 Pain on injection and involuntary movement after intravenous propofol have been concerns that have been overcome (see Chapter 6). Although propofol can be used safely in children with CHD, it is typically avoided as an induction agent in those with severe CHD because of its effects on SVR and blood pressure. It should be avoided in those with a fixed cardiac output such as severe aortic or mitral stenosis because it may cause severe hypotension. It can be used by infusion during CPB to reduce awareness and may be particularly useful if an early extubation is planned (see Chapter 6).



As in adults with severe cardiac disease, intravenous fentanyl combined with pancuronium and 100% oxygen or air and oxygen provides an excellent induction technique in very sick children with CHD, although it is not an amnestic. Inclusion of intravenous midazolam or another amnestic agent is strongly urged to avoid awareness. In neonates and infants, the use of high-dose opioid anesthesia provides excellent hemodynamic stability, with suppression of the hormonal and metabolic stress response.119,120 When fentanyl or other opioids are combined with nitrous oxide, the negative inotropic effects of nitrous oxide may be evident, particularly in sicker children.121 The high-dose fentanyl technique is effective in preterm neonates undergoing ligation of a PDA.122 In high-risk, full-term neonates and in older infants with severe CHD, the high-dose fentanyl technique in doses of up to 75 µg/kg, combined with pancuronium maintains stable hemodynamics during induction, tracheal intubation, and surgical incision.123 Oxygen saturation is well maintained and often improves during induction, even in cyanotic children.124 The cardiac index, SVR, and PVR in infants given 25 µg/kg of fentanyl do not change significantly.125 Combining pancuronium with fentanyl is desirable because the vagolytic effects of pancuronium offset the potential vagotonic effects of fentanyl. The hemodynamic stability reported in infants with the combination of high-dose fentanyl and pancuronium may not be replicated when other muscle relaxants are used (see Chapter 6).126


Sufentanil (5 to 20 µg/kg), an alternative to fentanyl, is 5 to 10 times more potent than fentanyl but has a large margin of safety.127 It is highly lipophilic and is rapidly distributed to all tissues. It is infrequently used in infants and children with CHD.


Remifentanil is an ultra-short-acting opioid that is rapidly metabolized in the plasma and tissue by nonspecific esterases to an inactive metabolite. It has a very brief elimination half-life, with a context-sensitive half-life of only 3 minutes, independent of the duration of infusion. In pediatric cardiac surgery, it is an attractive alternative to fentanyl that provides intense analgesia during the most stimulating parts of surgery but facilitates rapid awakening and weaning from mechanical ventilation without residual opioid effect. Its pharmacodynamics are unaffected by CPB.128 It provides stable hemodynamic conditions in children, although there is a tendency toward bradycardia and systemic hypotension.129131 It has no negative inotropic effect, even in the failing heart.132

A significant concern is the development of acute tolerance with increasing analgesic requirements after discontinuing remifentanil.133135 Some studies have suggested that this is not clinically important.136 Strategies to prevent tolerance to remifentanil have included intravenous magnesium infusions as well as nitrous oxide.137,138 Remifentanil is also used for prolonged sedation of children in the ICU. Many units have moved toward early extubation and discharge from the ICU after cardiac surgery (i.e., fast tracking), and remifentanil is a useful drug in this setting (see Chapter 6). Consideration must be given to transitioning to a longer-acting opioid before discontinuation of remifentanil.

Neuromuscular Blocking Drugs

Pancuronium has been studied in depth in children with CHD. When administered over a 60- to 90-second interval, pancuronium maintains heart rate and blood pressure.140 An intubating bolus dose of pancuronium may produce tachycardia and increase cardiac output. This bolus dose effect is sometimes desirable to support cardiac output in infants in congestive heart failure because their stroke volume is fixed. Pancuronium may be the muscle relaxant of choice when high-dose opioid techniques are used to offset the vagotonic effects of opioids such as fentanyl. Other muscle relaxants are also widely used, particularly if they are to be extubated in the operating room or early in the ICU.

Long-Term Neurocognitive-Developmental Outcomes Associated with Anesthesia

Concerns have been raised about the possibility that many of the anesthetic agents such as inhalational anesthetics, propofol, ketamine, and midazolam may cause long-term neurocognitive-developmental problems in neonates and young infants.139 This effect is thought to result from the neuronal apoptosis caused by these agents. The opioids have not been implicated in these changes, but this may change in time. There is no evidence to directly link anesthetic exposure in infancy to long-term neurocognitive defects. There is much ongoing research in this area (see Chapter 23).

Regional Anesthesia

The use of regional anesthesia to provide pain relief during and after cardiac surgery in adults also reduces the stress response to surgery and may reduce morbidity and mortality. In adults undergoing cardiac surgery, the benefits of regional anesthesia include earlier extubation, fewer respiratory complications, a reduction in renal failure, fewer strokes, and less myocardial damage after CPB.141143 In animals, thoracic epidural anesthesia reduces myocardial damage after coronary occlusion.144 The same benefits may be achieved by using intrathecal (spinal) analgesia. High spinal anesthesia using bupivacaine reduces the stress response to CPB and β-adrenergic dysfunction and improves cardiac performance after cardiac surgery in adults.145

Good research into regional anesthesia and analgesia in pediatric cardiac surgery is limited. Caudal morphine has been used to provide postoperative analgesia and has produced good analgesia for about 6 hours while reducing analgesic requirements for up to 24 hours.146 Two retrospective studies in children147,148 included a variety of regional anesthetic techniques. Most children were extubated in the operating room, although ∼4% of them required reintubation within 24 hours. Adverse effects included emesis (39%), pruritus (10%), urinary retention (7%), postoperative transient paresthesia (3%), and respiratory depression (1.8%). The rate of adverse effects was less with a thoracic catheter epidural approach compared with various caudal, lumbar epidural, and spinal approaches. Hospital duration of stay was unaffected by the presence of regional anesthesia complications. Although this study appears to indicate that regional analgesia is safe, the numbers in the study are too small to conclude that regional analgesia is safe for pediatric cardiac surgery.

The use of regional anesthesia in cardiac surgery for children remains controversial.149,150 The main concern is the risk of bleeding and the potential for disastrous neurologic complications. The risks may be greater in children than in adults because of the presence of collateral vessels, increased venous pressure, coagulopathy related to cyanosis, and the use of aspirin. There remain many unanswered questions regarding neuraxial block in children, such as the true incidence of epidural hematoma, the time delay required between placement of the epidural catheter and full anticoagulation, and the correct management of a bloody tap. The estimated risk of epidural hematoma during cardiac surgery in adults is 1 case per 1000 patients and 1 case per 2400 patients for spinal and epidural block, respectively.151 Whether the risks are similar or greater in children cannot be determined because, the numbers of children involved in studies are too small. A large, randomized, prospective study to evaluate a true risk-benefit ratio without bias is needed; until such data are available, various commentators have advised great caution with the use of regional analgesia for cardiac surgery, and some have suggested that it may not be possible to perform the study required because of ethical considerations.152

Fast Tracking

Fast tracking refers to abbreviating the perioperative period of children undergoing cardiac surgery. It should include every phase of the child’s journey from referral and preoperative evaluation to less invasive surgery, early weaning from respiratory support, extubation, and discharge from the ICU and hospital.

Early extubation of pediatric patients after cardiac surgery offers advantages in terms of cost and reduced morbidity associated with longer ICU stays.153157 The success of this approach depends on the close teamwork of a multidisciplinary team, with every member of the team working toward the same goal. Successful fast tracking usually requires the development of care pathways to ensure that the quality of patient care is not compromised.158 Early extubation and discharge from the ICU requires preplanning and the adoption of a technique that facilitates this goal. The use of very large doses of fentanyl is not appropriate; alternative techniques have been used, including smaller doses of fentanyl in combination with inhalational agents159,160

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

May 25, 2016 | Posted by in ANESTHESIA | Comments Off on Anesthesia for Children Undergoing Heart Surgery

Full access? Get Clinical Tree

Get Clinical Tree app for offline access