Cardiopulmonary Bypass: Principles and Practice

Previous Chapter | Next Chapter >




Quick Links to Sections in this Chapter








M. Kurusz: Department of Surgery, Division of Cardiothoracic Surgery, University of Texas Medical Branch, Galveston, Texas 77555–0528.

N. L. Mills: Department of Surgery, Division of Cardiothoracic Surgery, Tulane University School of Medicine, New Orleans, Louisiana 70112.

Some preexisting conditions or problems with the operation or function of the cardiopulmonary bypass (CPB) circuit can jeopardize the patient during CPB. Often the experience of a single team member or even several team members may be inadequate to form a plan of management for these problems. This chapter examines several such conditions or incidents that may be encountered infrequently and suggests management strategies derived from the published literature or the authors' experience.


Back to Quick Links

Atherosclerotic aorta

The atherosclerotic aorta can present problems during cannulation for CPB, application of clamps, delivery of cardioplegia, construction of proximal anastomoses for coronary artery bypass grafts, or valve replacement or repair. A large multicenter study identified the presence of proximal aortic atherosclerosis as the strongest predictor of stroke (1). This lends support to the theory that atherosclerotic emboli liberated by surgical manipulation of the aorta cause most strokes after CPB. Risk factors for ascending aortic atherosclerosis include: significant carotid, abdominal aortic, and left main coronary artery atherosclerosis; aortic wall irregularity on ascending aortic angiogram; adhesions between the ascending aorta and its adventitia; pale appearance of the ascending aorta; and minimal bleeding of an aortic stab wound (2).

In most cases, the surgeon can palpate the aorta at the time of surgery to select a nondiseased section as the site for insertion of the arterial perfusion cannula or proximal anastomoses. Such palpation may be performed before CPB by briefly clamping the superior and inferior venae cavae with straight Satinsky clamps to effect temporary (10- to 12-second) inflow occlusion that will rapidly decrease cardiac ejection and arterial pressure. With lower pressure in the aorta, gentle palpation can be used to assess the quality of the aortic wall to determine the location of disease-free sites for cannulation and placement of the aortic cross-clamp. Preoperative angiogram should also be performed during cardiac catheterization to assess the smoothness of the aortic lumen and detect wall irregularities (2). Transesophageal echocardiography and/or a sterile epiaortic probe placed directly on the aorta during surgery are used increasingly to visualize the ascending aortic lumen, with the latter ap- proach being more sensitive because of the interposition of the tracheal carina between the esophagus and distal ascending aorta. Culliford et al. (3) suggested routine use of transesophageal echocardiography in all patients over 65 years old to assess the degree of aortic atherosclerosis.

In severely diseased aortas, application of an aortic cross-clamp may not be deemed feasible because of increased risk of dislodgment of atherosclerotic debris or aortic dissection (Fig. 28.1). The so-called no-touch technique has evolved to better manage these patients (2). It relies on the internal mammary artery or gastroepiploic artery as conduits for coronary bypass grafts. In some patients requiring extensive revascularization, saphenous vein grafts can be anastomosed to the internal mammary artery. In no-touch cases, a femoral artery, axillary artery, or the underside of the distal aortic arch is used as the arterial cannulation site (2,4,5). Alternatively, either a long arch cannula can be inserted distal to the left subclavian artery (3) or a diffusion-tipped cannula can be used (6) to better disperse the flow jet at the CPB cannula tip. However, these cannulation techniques still present some risk of stroke hazard from insertion of the cannula in the severely diseased aorta. Cannulating from the left ventricular apex with the cannula advanced through the aortic valve has also been described (7).

FIG 28.1. Type 1, type II, and type III ascending aortic atherosclerosis. No clamp is safe on these types of ascending aortic disease. Type I: Circumferential ascending aortic calcification, which may be easily diagnosed preoperatively on the angiogram. Palpation of the ascending aorta at operation reveals firm calcification. Embolization or aortic injury that may be difficult to repair may result if the aorta is clamped. Type II: This pattern may be diagnosed preoperatively by noting an irregularity of the normally smooth lining of the ascending aorta on the left ventricular angiogram or aortic root injection. Visualization of the ascending aorta is now considered a mandatory part of workup before coronary artery bypass graft. Type III: Intraluminal liquid debris is the most elusive of the three patterns to diagnose before clamping the aorta. A pale appearance of the aorta or adherence of the adventitia to the ascending aorta may be the only diagnostic clues. Operative echocardiography wil reveal a thickened ascending aorta that will liberate liquid debris if a cross-clamp or partial occlusion clamp is applied. (From Mills NL, Everson CT. Atherosclerosis of the ascending aorta and coronary artery bypass: pathology, clinical correlates, and operative management. J Thorac Cardiovasc Surg 1991;102:546–553, with permission.)

We strongly recommend that the anesthesiologist perform simultaneous bilateral proximal (not bifurcation) carotid compression during any surgical manipulation of the severely diseased aorta to minimize atheromatous cerebral embolization. Rather than use conventional approaches to cardioplegia delivery, the surgeon may instead rely on electrical fibrillation or administration of a -adrenergic blocker such as esmolol to slow the heart during distal anastomoses (4). In these cases, left ventricular venting may be avoided. However, if significant aortic insufficiency is present, conventional venting methods should be used, which then requires needle vent placement in the ascending aorta for de-airing before weaning from CPB and which carries some degree of hazard in the diseased aorta. A case report recom mended use of an endoaortic clamp, as designed for minimally invasive cardiac surgery, to isolate the heart from CPB systemic flow (8). A central lumen allows delivery of cardioplegic solution or venting the aortic root. However, the prevailing wisdom is that this approach is unwise in the presence of severe atherosclerosis because of the risk of dislodging atherosclerotic debris. An aortic filtration system (9) that uses an umbrella screen inserted through a modified 24 French arterial cannula before aortic cross-clamp removal has been shown to capture particulate debris that presumably would have otherwise embolized systemically.

In summary, cannulation of the severely diseased ascending aorta may be associated with significant morbidity after CPB. Modification of conventional CPB cannulation, cardioplegia, and venting techniques can reduce the incidence of stroke in this challenging patient population. Newer diagnostic measurements such as release of S-100 , a marker of cerebral ischemia (10), and novel perfusion cannulation techniques under development may lead to further reductions in neurologic injury in the patient with diffuse atherosclerosis needing cardiac surgery.

Hematologic problems


Cold agglutinins are serum antibodies that become active at decreased blood temperature and produce agglutination or hemolysis of red blood cells. These antibodies are classically directed against antigens on the red blood cells but can also be nonspecific. The most clinically relevant characteristic of cold agglutinins is thermal amplitude, the temperature below which the antibodies become activated. As temperature drops below this threshold, antibody activity increases exponentially. In general, this activity reverses as rewarming occurs (11). Higher cold agglutinin titers (concentrations) are more clinically significant than low titers. There is no widely accepted definition for high versus low titer; however, Lee et al. (12) suggested titers less than 1:32 as being low and those greater than 1:128 as being high.

Rewarming activates the complement system to induce hemolysis in patients with cold agglutinins. For hemolysis to occur, the cold agglutinin and complement activities must overlap. That is, the temperature must be low enough for the cold agglutinins to activate but warm enough for a complement fixation to occur.

Aside from during hypothermic bypass, cold agglutinins seldom produce symptoms because activation most often occurs at temperatures well below the usual range of body temperature. With cold exposure, clinical signs may include acrocyanosis of digits, tip of the nose, or ears from agglutination-induced ischemia. Most commonly, immediate warming of the affected areas reverses the agglutination and thus the ischemia. A patient with prolonged hypothermic CPB would be at risk for multiorgan damage from prolonged vascular occlusion (13). This is an uncommon but poten tially catastrophic consequence of failure to recognize and treat the presence of cold agglutinins.

When patients undergo screening for cold agglutinins, a diagnosis can be easily made based on laboratory test results. If screening at 4°C is negative, no further screening is needed. If the screen is positive at 4°C, the thermal amplitude should be determined and the titer determined for each temperature at which the screen was positive. This will give more precise information for dealing with the cold agglutinin antibodies.

In patients who are not initially screened for cold agglutinins, a diagnosis may be made by astute observation. During hypothermic CPB, agglutination within the vessels may be noted, particularly if the surgeon is wearing magnifying loops. Hemolysis manifested by hemoglobinuria is most often recognized by pink or red-tinged urine. The latter occurrence, however, is relatively common even in the absence of cryoagglutination. If blood cardioplegia is used, the perfusionist may note agglutination in the cardioplegia delivery system as the blood is cooled (14–16). In addition, immediate agglutination of blood in a syringe during phlebotomy may indicate the presence of cold agglutinins. Agglutination also can be confirmed visually by immersing a test tube of blood into an ice slush solution and observing cell clumping on the side of the test tube that often disappears when the tube is warmed. Many cold agglutinins will present during a routine crossmatch done at room temperature. Any of the above findings suggest the presence of clinically significant cold agglutinins, and steps should be taken to prevent adverse reactions during CPB (14,17,18).

Both monoclonal and polyclonal cold agglutinins exist. The monoclonal types usually associated with lymphoreticular neoplasms are generally irreversible. The polyclonal antibodies are often associated with acute infectious diseases such as mycoplasma, infectious mononucleosis, or cytomegalovirus (13). Production of polyclonal cold agglutinins is typically transient and may remit spontaneously in weeks, but when present it may be associated with acute life-threatening intravascular hemolysis (13). Leach et al. (17) developed a comparison of clinically significant and insignificant cold agglutinins (Table 28.1). An important point in assessing the need to screen for cold agglutinins is that a failure to screen may lead to an adverse outcome, such as myocardial infarction, stroke, or acute renal failure. Without knowledge of the presence of cold agglutinins, these adverse outcomes may be attributed to another cause. For example, hemolysis may be attributed to mechanical trauma to blood from CPB (19). One report suggests that mechanical trauma to blood may be more important than cold-reacting autoantibodies when the thermal amplitude is less than 22°C (20).

Treatment of cold agglutinin disease during CPB essentially consists of prevention of complement activation and ultimately of agglutination or hemolysis. Treatment is based on the etiology and the severity of the problem. In a mild case of cold agglutinins (i.e., a case in which there is a very low thermal amplitude, such as 4°C) and/or a low titer of antibodies, minor or no changes in surgical or CPB techniques and, in some cases, treatment with corticosteroids to avoid hemolysis have been advised (21). For patients with low-titer nonspecific antibodies (and only these patients), Moore et al. (11) concluded that hypothermic CPB can be performed on these patients without increased risks of hemolytic or agglutination crises; further, minor degrees of hemolysis occur in all patients during hypothermic bypass, especially during the rewarming phase, whether or not cold agglutinins are present.

In a case with clinically significant cold agglutinins (i.e., high thermal amplitude, high titer, or clinical symptoms), a number of changes in surgical technique have resulted in successful surgery using CPB. In the case of cold agglutinins caused by acute infection (e.g., a recent viral illness), elective cardiac surgery should be postponed for several weeks, by which time the antibody may have disappeared (19). If the urgency of surgery precludes that approach, the most sensible approach is to use either normothermia or mild hypothermia using blood temperatures continuously maintained above the thermal amplitude to avoid the active temperature range of agglutination (19,22–28). Hence, the presence of cryoagglutinins with high titer or high thermal amplitude may represent a reasonable indication for the use of warm cardioplegia myocardial protection techniques while maintaining normothermic systemic temperatures.

When selecting cold cardioplegia, blood should be initially flushed out of the coronary circulation with warm crystalloid cardioplegic solution followed by cold cardioplegic solution. Just before removal of the aortic cross-clamp, warm cardioplegic solution should be used to prevent agglutination from blood exposed to a cold heart. Refinements of this latter technique consist of bicaval cannulation with tightening of caval tapes to avoid cooling large amounts of blood in the heart. A sump catheter may be placed in the right atrium to retrieve cardioplegic solution until the coronary sinus effluent is clear. In addition, lower than normal CPB systemic flows may be used to decrease the noncoronary collateral flow and subsequent cooling of this blood.

If it is uncertain whether this noncoronary collateral flow will cause problems, the heart may be maintained at a temperature above the thermal amplitude (29). Venting the left ventricle will avoid cooling and stagnation of blood in the left ventricular cavity. Crystalloid cardioplegia has been used rather than blood cardioplegia to avoid agglutination of the cells in the solution when delivered at low temperature (30). Adjuncts may include a myocardial insulation pad to prevent cooling of blood in structures adjacent to the heart. Using a septal temperature probe in the myocardium to keep the temperature greater than the thermal amplitude may prevent significant activation of cold agglutinins. However, once the red blood cells are flushed out of the coronary circulation, the heart can be cooled to provide myocardial protection provided the above adjuncts are used.

In addition, all fluids, blood, plasma, inspired gases, and bolus injections should be warmed (19) in the periods before and after bypass, especially if the cryoagglutinin has a high thermal amplitude. Antibody dilution by the extracorporeal circuit priming solution probably may reduce the tendency for activation of high titer cold agglutinins (W. Rock, personal communication, 1992). It seems reasonable to assume that hemodilution during CPB will reduce the concentration of the cryoagglutinin 30% to 50%; thus, little protection would appear likely unless the cold agglutinin titer is very low (1:2 or less).

The literature contains descriptions of successful cardiac surgery after plasmapheresis (28,31) or total exchange transfusions in patients with high titer high thermal amplitude cold agglutinins. There is some evidence that the patient's own red cells may be protected from hemolysis, and if transfusions are required, autologous packed red cells may be advantageous (19). In the case of unexpected agglutination encountered at the time of surgery, several techniques may be useful: verification by the blood bank that cold agglutination is present rather than an unrecognized alloantibody; the use of crystalloid cardioplegic solution to dilute the antibody in the coronary circulation; use of noncardioplegic techniques (e.g., electrical fibrillation); and maintenance of systemic temperatures greater than 28 to 30°C at which significant amounts of agglutination are unlikely to occur. In addition, the CPB circuit and blood cardioplegia delivery system should be monitored carefully throughout bypass for presence of cell aggregates, and CPB arterial line filters should be used in all cases.

Several cases of fibrin formation and clotting of membrane oxygenators during hypothermic CPB have been recently reported (32). Although cold agglutination was initially suspected, none of these patients exhibited agglutinin formation either during preoperative screening or postoperative hematologic workup. No abnormalities in blood coagulation factors VII and VIII or von Willebrand factor were demonstrated in blood samples tested postoperatively, nor were there significant differences between the affected patients' blood samples and control patients' blood. However, rapid CPB cooling in conjunction with use of efficient oxygenator heat exchangers having small blood pathways was associated with excessive premembrane CPB line pressure buildup that usually resolved with more moderate cooling strategies or by warming the perfusate.

In summary, all patients undergoing hypothermic CPB should be screened preoperatively for cold agglutinins (22–25,28,33). If an initial screen is positive, the cold agglutinins should be characterized as to thermal amplitude and titer. Clinical symptoms should also be sought. A patient with low titer, low thermal amplitude, and clinically asymptomatic can tolerate CPB with moderate hypothermia at very low risk with little or no alteration in technique. For clinically significant cold agglutinins with high thermal amplitude and/or titer, if due to a transient viral illness, elective surgery may be postponed several weeks in the hope that this will decrease cold agglutinins to insignificant levels. However, in severe cases in which the high titer high thermal amplitude cold agglutinins are not transient, precautions should be taken as outlined above to prevent an agglutination or hemolytic crisis (Fig. 28.2).

FIG 28.2. Algorithm for management of cold agglutinins in cardiopulmonary bypass.

Hemoglobinopathy and erythrocyte disorders

Sickle cell trait and disease

See Related Case Study from Yao & Artusio's Anesthesiology

In the normal adult, hemoglobin A comprises 96% to 97% of all hemoglobin. Hemoglobin A consists of two and two chains. Sickle cell hemoglobinopathy is a single-gene recessive abnormality that may be present in a heterozygous recessive form, which is termed sickle cell trait, or in a homozygous recessive form, which is expressed as sickle cell disease. The percentage of hemoglobin S and hemoglobin A varies with the individual, but sickle cell disease patients have predominantly hemoglobin S. This homozygous state found in 0.15% of African-Americans is associated with severe hemolytic anemia and vaso-occlusive phenomena, resulting from the increased blood viscosity that occurs when red cells aggregate and individually typically assume a sickle shape (34). Sickled cells have a very limited capacity to load and unload oxygen.

In contrast, patients with the sickle cell trait have a lower percentage of hemoglobin S, accounting for 20% to 45% of their total hemoglobin. Approximately 8% of African-Americans carry the heterozygous recessive trait. These individuals have few clinical problems, and except for severe or provoked conditions, they rarely experience sickle cell crises. Red blood cell sickling results from deoxyhemoglobin formation. The tendency toward sickling increases with hypoxemia, acidosis, increased concentrations of 2,3-diphospho- glyceric acid, infection, hypothermia, and capillary stagnation. A hypertonic environment that may lead to crenation of normal red blood cells also will lead to sickling. Hemoglobin S demonstrates increased osmotic and mechanical fragility, making hemolysis more likely. A hypotonic environment will lyse red blood cells with increased osmotic fragility.

In patients with sickle cell disease, some sickling begins to appear at 85% hemoglobin oxygen saturation, and sickling of red blood cells is complete at 38% hemoglobin oxygen saturation. In patients with sickle cell trait, sickling begins at hemoglobin oxygen saturations of approximately 40%. Sickling is reversible to a degree, but if it is repeated, the sickled cells become permanently damaged, resulting in markedly increased fragility and a shortened cell lifespan. In addition to increased blood viscosity potentially causing vascular occlusion, sickling can cause endothelial cell injury in the microvasculature. This may activate the intrinsic clotting system and exacerbate the vaso-occlusive phenomenon (34–37).

The operative strategy in sickle cell patients is to prevent sickling and thereby prevent hemolysis or vaso-occlusive phenomena intraoperatively and postoperatively. Because sickling results from decreased hemoglobin oxygen saturation, maintaining adequate arterial oxygen tension assumes paramount importance. Adequate capillary perfusion with short capillary transit times and avoidance of low output states (to prevent low mixed venous hemoglobin oxygen saturations) are also important (38–40). Continuous measurement of arterial and mixed venous hemoglobin oxygen saturations help to maintain adequate oxygen saturations. Because sickle crises are frequent when high concentrations of hemoglobin S are present (i.e., homozygous patients), marked reduction of sickling can be achieved by relative dilution of hemoglobin S with respect to hemoglobin A. This can be accomplished by using preoperative or intraoperative exchange transfusions.

Preoperative exchange transfusions, particularly applicable in patients who are already anemic, not only increase the prevalence of hemoglobin A relative to hemoglobin S but also suppress the production of hemoglobin S. This therapy improves the oxygen-carrying capacity of the blood by correcting the anemia and deficiency of hemoglobin A. In nonanemic patients, exchange transfusion may be accomplished intraoperatively. This type of transfusion is usually performed by sequestering the initial CPB venous drainage from the patient after priming the extracorporeal circuit with whole blood containing hemoglobin A (34,40). The goal of exchange transfusion is to achieve a hemoglobin A fraction of 60% to 70%, which is also the level sought when treating a major sickle cell crisis (39).

Acidosis shifts the oxyhemoglobin dissociation curve to the right, which increases the tendency toward sickling. This holds true particularly in venous blood, where sickling is most often initiated. Arterial and mixed venous blood gases should be measured frequently and any developing acidosis aggressively treated with sodium bicarbonate (39,40). Hypoperfusion may result from hypothermia, administration of cardioplegia, diminished intravascular volume, poor patient positioning, tourniquets, low CPB systemic flows, or low cardiac output states. It is important to avoid hypoperfusion because of the tendency of blood to desaturate during the capillary and venous phase of circulation, resulting in low hemoglobin oxygen saturation and red blood cell sickling. Hypoperfusion can usually be prevented by maintaining adequate systemic CPB flows and avoiding low cardiac output states both before and after bypass.

Localized sickling may occur in the heart during aortic cross-clamping because of the absence of coronary blood flow. This phenomenon may be avoided by flushing hemoglobin S out of the coronary arteries using either crystalloid cardioplegia or blood cardioplegia with a high fraction of hemoglobin A. Because mechanical prosthetic valves may predispose the patient to increased hemolysis, such valves are not recommended in these patients (40). Other means of avoiding mechanical blood trauma include minimizing the use of cardiotomy suction and venting. In patients with sickle cell disease, it appears advisable to minimize or avoid hypothermia during CPB. Despite the risks involved, numerous patients with homozygous sickle cell disease have successfully undergone CPB using the techniques described above (41–51). Successful cases have been described even using deep hypothermia with circulatory arrest (52).

Hereditary spherocytosis

Hereditary spherocytosis, an autosomal dominant defect in red blood cell membranes, results in spherically shaped red blood cells that have increased osmotic and mechanical fragility. The usual treatment for hereditary spherocytosis resulting in hemolysis is splenectomy, which corrects the hemolysis and increases the shortened lifespan of the red blood cells to normal, although the cells retain their abnormal properties. Information describing CPB in these patients is limited. One case report describes a nonsplenectomized patient undergoing bypass with no apparent increase in blood destruction or in osmotic fragility over the baseline level (53). In addition, in a patient who had previously undergone splenectomy, no increase in hemolysis was noted during or after CPB, despite triple valve replacement using Bjork-Shiley mechanical valves (54). Other patients have had porcine valves inserted to minimize mechanically induced hemolysis. One study reported uneventful closure of an atrial septal defect in a 31-year old with hereditary spherocytosis and suggested that a short CPB time was important in avoiding complications (55). Hereditary elliptocytosis is a condition thought to be similar to hereditary spherocytosis; there is infrequent hemolysis and anemia, and no specific precautions are recommended in these patients (34,56).


Clinical CPB experience with other hemoglobinopathies is sparse. Another potential problem is that the hospital laboratory may neglect to notify the cardiac surgical team of rare but significant hematologic abnormalities before the patient's surgery.

Thalassemia minor patients exhibit no increase in red blood cell fragility, and therefore one would expect no hemolysis. Anemia in these patients is treated with transfusions. Hemolysis has been induced by a number of drugs in patients with glucose-6-phosphate dehydrogenase deficiency (56). This gender-linked inherited deficiency is present in 10% to 15% of African-American males. Susceptible individuals may develop explosive hemolysis when they receive drugs such as antimalarials, quinidine, phenacetin, or sulfonamides. These drugs should be avoided in susceptible patients undergoing surgery, including those undergoing CPB.

Acute methemoglobinemia may result from increased production of methemoglobin to levels far exceeding the usual amount, which is less than 1% of total circulating hemoglobin (57). Secondary or acquired methemoglobinemia is almost always caused by poisoning with chemicals or drugs classified either as direct oxidants such as nitrites or indirect oxidants such as benzocaine. Other such medications include high-dose methylene blue (58), nitroglycerin (59–61), nitroprusside, prilocaine, silver nitrate, sodium nitrate, flutamide (62), and sulfonamides.

The diagnosis of methemoglobinemia is made when cyanosis or oxygen desaturation occurs in the presence of an adequate arterial oxygen tension and is supported by a chocolate-brown color of blood rather than the usual dark blue of cyanosis (63). The diagnosis can be confirmed by spectrophotometry (64).

Treatment at times may need to precede definitive diagnosis. First, all probable offending drugs should be withdrawn. Next, maximal oxygen concentrations should be delivered to the oxygenator. If cyanosis or oxygen desaturation persists in the presence of high oxygen tension, pharmacologic treatment should begin. The drug of choice is methylene blue, 1 to 3 mg/kg administered in a 1% solution, which converts methemoglobin to active hemoglobin. The response to methylene blue is usually immediate and excellent, and because the treatment is relatively innocuous, its use should not be delayed. However, methylene blue may cause methemoglobinemia when a dose greater than 7 mg/kg is administered (65). If the patient fails to respond to methylene blue, the next line of treatment consists of high-dose vitamin C and, if necessary, exchange transfusion (59).

Polycythemia is defined as increased red cell mass. It occurs with cyanotic congenital cardiac defects as compensation for reduced oxygen delivery to tissues. The preoperative hematocrit in severe cases may exceed 70%, at which level blood viscosity is sufficiently high to compromise blood flow. Hemostatic abnormalities consistent with consumptive coagulopathy can occur (increased prothrombin and activated partial thromboplastin times, increased fibrin degradation products, thrombocytopenia, factor deficiencies, etc.) (66,67). Hemodilution beyond that normally used to prime the CPB circuit may better preserve the patient's coagulation status, so this may be an ideal situation for withdrawal of autologous blood before CPB. The degree of hemodilution needed to achieve a desired patient/pump hematocrit can be calculated using a formula shown in Figure 28.3 (66). Polycythemia vera is a hematologic disease that carries an increased risk of myocardial infarction.

FIG 28.3. Calculation of volume of crystalloid solution (cardiopulmonary bypass prime volume) necessary for hemodiluting to a desired hematocrit. EBV, estimated blood volume; CPB, cardiopulmonary bypass. *Factor (EBV/kg) is assumed to be 80 mL/kg for less than 10 kg body weight; 75 mL/kg for 10 to 20 kg of body weight; 70 mL/kg for more than 20 kg of body weight. The example addresses the polycythemic adult, but this condition is more prevalent in cyanotic pediatric patients with smaller required CPB circuit prime volumes. (From Milam JD, Austin SF, Nihill MR, et al. Use of sufficient hemodilution to prevent coagulopathies following surgical correction of cyanotic heart disease. J Thorac Cardiovasc Surg 1985;89:623–629, with permission.)

Other hematologic problems

Antithrombin III deficiency and heparin-induced thrombocytopenia are discussed thoroughly in Chapters 24 and 25. Other hypocoagulable disorders such as hemophilia A (factor VIII deficiency), hemophilia B (factor IX deficiency), or von Willebrand's disease generally require administration of fresh frozen plasma, cryoprecipitate, or factor VIII concentrates to counteract excessive bleeding tendencies after CPB (68,69). Despite the possible temptation to use less heparin for anticoagulation in these situations, standard full-heparin anticoagulation regimes should be used during CPB.

With the increased application of interventional cardiac procedures by cardiologists, patients may present for surgery with antiplatelet or thrombolytic therapy (e.g., aspirin, urokinase, streptokinase, tissue plasminogen activator, IIb/IIIc platelet receptor antagonists) that can cause alterations in results of conventional anticoagulation tests after heparin administration and severe bleeding after CPB. Standard treatment has been to administer platelet concentrates or fresh frozen plasma after CPB subsequent to protamine administration. Abciximab may be removed by hemoconcen tration during CPB (70). These patients will still likely require platelet transfusion, but the effect of the transfused platelets will be to provide hemostasis rather than act as receptors for circulating antibodies contained in abciximab.

Patients with a history of thrombophlebitis or other re current venous or arterial thromboembolism may be predisposed to hypercoagulability. These patients may require larger than expected and more frequent doses of heparin to maintain adequate anticoagulation during CPB. Cases of fatal thrombosis after CPB in patients with protein C deficiency have been reported (71). Lawson et al. (72) reported a successful CPB case in a patient with heterozygous protein C deficiency and outlined management recommendations that included continuous heparin anticoagulation therapy preoperatively (activated partial thromboplastin time 1.5 to 2.0 times baseline), normal heparin dosing (400 units/kg) just before bypass along with administration of fresh frozen plasma, reinstitution of heparin therapy after control of bleeding in the postoperative period, and discharge on oral warfarin.

Coagulation problems in pediatric patients undergoing CPB are usually more complex than those seen in the adult population (73). Neonates, in particular, have underdeveloped hemostatic systems when compared with older children and adults (74). This, coupled with the massive hemodilution and usual requirement for blood products in the CPB prime, may further obscure the hematologic picture with these patients (75). Use of the freshest bank blood obtainable (preferably less than 48 hours old) and fresh frozen plasma in the CPB circuit prime can minimize bleeding problems postoperatively. Administration of platelet concentrate or cryoprecipitate should improve coagulation parameters postoperatively; fresh frozen plasma administered after platelet transfusion may exacerbate bleeding (76).

In summary, patient history and preoperative testing can diagnose most hematologic disorders. In particular, patients with hepatic dysfunction may have serious bleeding postoperatively. In both elective cases and those presenting for CPB surgery emergently, consultation with an experienced hematologist is advised for safe management during the perioperative period. Dyke and Sobel (68) reviewed coagulation disorders and surgical management issues in greater detail.

Religious objections to blood transfusion

Avoidance of homologous blood transfusion is a desirable goal whenever CPB is used; with patients of the Jehovah's Witness faith it is mandatory because of their strict interpretation of the bible (77). Nearly 40 years ago, Cooley et al. (78) first reported the feasibility of open heart surgery in this patient population. In 1977, Ott and Cooley (79) reported additional results in a large series of Jehovah's Witness patients in whom no blood was transfused. However, there was a 10.7% mortality in those undergoing CPB (n = 39), with preoperative or postoperative anemia a contributing factor in 12 deaths.

Using current low prime membrane oxygenator circuits, intraoperative cell salvage, and reinfusion of shed mediastinal blood, CPB can be performed relatively safely, even in pediatric patients (77,80,81), reoperations, or those with complex anatomy (82). The lowest safe hematocrit on CPB is not known, but values of approximately 15% have been used successfully provided CPB systemic flows are maintained at levels to prevent development of metabolic acidosis.

To honor the patient's religious beliefs, it is necessary to maintain continuity between removed blood and the patient's vascular system when cell salvage is used (83). Figure 28.4 shows how this can be accomplished pre- and post-bypass. Some authors have further advocated use of heparin-bonded CPB circuits and lower levels of heparinization [activated clotting time (ACT) >280 seconds] with favorable results (84). Use of erythropoietin preoperatively to promote red cell growth and antifibrinolytics perioperatively also have been advocated (85,86). Von Son et al. (81) enumerated management strategies to minimize blood loss in these patients (Table 28.2). Lee and Martin (87) wrote an excellent review of CPB management in this patient population.

FIG 28.4. Schematic drawing of cardiopulmonary bypass (CPB) circuit for collection, processing, and reinfusion of blood after bypass while maintaining continuity with the patient's circulation. The reinfusion bag (top left) should initially be back-filled with patient's blood from an intravenous site to establish continuity with cell salvaged blood (from collection bag) before CPB. Cardiotomy suction can be used after bypass until protamine is administered with collected blood processed in the cell salvage system. Residual perfusate in the CPB circuit should be transferred to the cardiotomy reservoir and also processed by the cell-salvage system to minimize blood loss. A second cardiotomy reservoir (not shown) is used during CPB for conventional collection of suctioned and vent blood, which is drained into the venous reservoir. (Modified from Milan TP Jr, Whitmore J, Maddi R.Reoperative cardiac surgery in a Jehovah's Witness: role of continuous cell salvage and in-line reinfusion. J Cardiothorac Anesth 1989;3:211–214, with permission.)

Reoperative surgery

Patients undergoing repeat cardiac surgery often present problems because of adhesions that make a second sternotomy and vascular access for CPB cannulation technically difficult. There is also an increased risk of encountering major bleeding during dissection because cardiac structures or vessels may be adherent to the chest wall, particularly if the pericardium has not been closed during the initial operation (88). Normal anatomic landmarks are often obliterated, prolonging adequate surgical exposure and CPB times. Because of adhesions necessitating sharp dissection, these patients tend to bleed more and have higher transfusion rates than first time cases (89). The surgeon should dissect as little as possible during reoperative surgery to minimize large disrupted tissue surface areas that will bleed. Use of an argon beam coagulator in these cases may also minimize bleeding problems associated with extensive dissection.

In some cases, groin cannulation of the femoral or iliac artery and femoral vein must be used to establish CPB. Placing an adequately sized femoral arterial cannula can usually be accomplished easily, but placement of an adequately sized venous cannula may be more difficult. New long, thin-walled, kink-resistant femoral venous cannulas are now commercially available to permit positioning the cannula tip near the cavoatrial junction (90). However, because of their length, standard gravity siphonage may be inadequate to permit full CPB systemic flow. If maximizing the height differential between the patient's heart and the CPB venous reservoir or repositioning the cannula does not improve venous drainage, flow may be augmented with a centrifugal pump placed in the venous line (91,92). Activation of the centrifugal pump will exert additional negative pressure beyond that obtainable by height differential alone with a concomitant modest increase in venous line flow. Alternatively, a hard-shell venous reservoir may have regulated vacuum applied to its interior to effect additional venous line flow using the same principles (93).

Both methods of augmented venous drainage require careful monitoring of venous line pressure so that excessive levels of vacuum are not created (94). Excessive vacuum can collapse vascular walls into the venous cannula openings, thus impeding or stopping venous line flow (95). Excessive vacuum may be manifested by intermittent or staccato flow; in severe situations, the venous line may rhythmically jerk and relax as flow stops and is then reestablished with systemic venous return in the patient's cavae or right atrium. Levels of hemolysis will also quickly rise if excessive vacuum is exerted in the venous line (96).

Establishing CPB via the groin vessels will afford the surgeon more control if massive bleeding in the chest is encountered. Alternatively, if the arterial cannula has been placed, CPB may be established using the pump suckers and a vent as a source of venous return (so-called sucker bypass). This will allow surgical control of bleeding and provide adequate decompression of the heart and preserve patient hemodynamics until the surgeon can place a conventional venous cannula in the right atrium or right atrium/inferior vena cava. The perfusionist should be prepared with additional cannulas, connectors, and tubing if CPB must be established emergently during reoperations (97). It must be recognized that the risk of aortic dissection is much greater when femoral arterial cannulation is used, which is discussed later in this chapter.

In summary, the number of patients presenting for reoperation is increasing. Increased surgical experience, use of antifibrinolytic drugs, and newer CPB technology can reduce the risk of morbidity and mortality associated with these procedures to levels approaching primary cardiac operation (98,99). Augmented venous drainage techniques have also been applied during minimally invasive cardiac surgery (100) and are discussed in greater detail in Chapter 35.

Cardiopulmonary bypass after pneumonectomy

CPB techniques for patients who have undergone prior pneumonectomy have only recently been addressed in the literature (101,102). The technical aspects of conducting CPB in such patients are not significantly different from those in patients who have had lobectomy or no pulmonary resection.

Hemodilution for CPB has been associated with a decrease in postoperative pulmonary problems theoretically from dilution of noxious blood elements or from avoidance of noxious elements (e.g., microaggregates present in homologous blood). However, excessive hemodilution may predispose to pulmonary edema, which might be addressed by limiting hemodilution to a hematocrit fraction of more than 20% in postpneumonectomy patients.

Blood transfusions should be avoided not only because of potential infections or transfusion reactions but also because of possible pulmonary damage from such elements as platelets and white blood cells. Washed packed red blood cells appear more appropriate when transfusion is indicated; return of shed blood should be avoided. When platelet transfusion is indicated, steroids and diphenhydramine should be used as pretreatment and leukocyte-depleted platelet concentrate should be used. Leukocyte reduction of all transfused residual perfusate will lessen the potential for pulmonary injury (103).

Technically, the position of the heart may be distorted because of contracted fibrothorax and/or hyperinflation of the remaining lung. This may lead to technical difficulty in gaining exposure, especially after left pneumonectomy. Remote cannulation from the femoral vessels has been helpful in some patients. Monitoring of pulmonary artery or left atrial pressure, with possible left ventricular venting, is important because strict control of the level of pulmonary capillary hydrostatic pressure is critical both intraoperatively and postoperatively to prevent pulmonary edema. Air emboli in the pulmonary circuit would seemingly be less well tolerated, and this is addressed by standard de-airing techniques. CPB time directly correlates with postoperative lung water; therefore, at times, a less complete coronary revascularization may be preferable to incurring a long CPB time. For coronary bypass surgery, the proximal anastomoses may be performed off pump or with low-flow partial CPB.

Minimally invasive surgery

This topic is reviewed in detail in Chapter 35.


Back to Quick Links

The initiation of bypass is discussed in detail in Chapter 27. To briefly recapitulate a key issue, patency and continuity between the patient's vascular system (venous and arterial) and the CPB circuit must be assured for uncomplicated initiation of CPB. Cannula malposition, kinks in the CPB tubing, or clamps inadvertently left in place on cannulas or CPB tubing (venous or arterial) can prevent a smooth transition to full CPB flow. The absence of air bubbles in the CPB systemic line or arterial cannula must be visually verified to avoid systemic air embolism when CPB is started. Presence of abnormal anatomy (e.g., persistent left superior vena cava) or patient pathophysiology (e.g., aortic valve insufficiency) may compromise decompression of the heart during CPB. Presence of anatomic shunts may further obscure the surgical site, requiring placement of vents. Problems related to CPB cannulation and venting are discussed thoroughly in Chapters 5 and 6.

Acute dissection of the aorta is a serious often devastating problem that can occur with initiation of CPB. It is associated with a high mortality rate and, when recognized, inevitably leads to modification of the original surgical procedure. Fortunately, the incidence of acute dissection with ascending aortic cannulation is much less than when the femoral artery is used for the CPB systemic flow connection (104). Dissection may be delayed and not recognized until the postoperative period. This most often occurs in patients undergoing coronary artery bypass surgery who have a history of hypertension and severe atherosclerosis (105).

A triad of observations may aid recognition of acute aortic dissection. First and foremost is an unexpectedly increased CPB arterial line pressure as the systemic pump is activated. This often coincides with profoundly decreased systemic pressure measured by the indwelling arterial cathe ter. Third, the perfusionist may experience decreased venous drainage to the CPB circuit. At the surgical field, a hematoma may develop near the arterial cannulation site or there may be bleeding around the purse-string sutures.

If aortic dissection is suspected, the perfusionist should immediately stop CPB and notify the surgeon. The perfusionist and surgeon must confirm there are no kinks or clamps on the arterial line responsible for the increased CPB line pressure and/or decreased systemic pressure. The surgeon should palpate the aorta; if it is flaccid, this often indicates a major problem and acute dissection should be strongly suspected. It is important to clamp the venous line in this emergency situation to avoid exsanguinating the patient into the CPB reservoir. If dissection is confirmed, the arterial cannula is removed and reinserted in a different location so that CPB can be resumed. The dissection must be surgically repaired, which may involve replacement of the ascending aorta, quite possibly with deep hypothermia and circulatory arrest.

The risk of acute aortic dissection may be decreased by the surgeon observing a brisk backflow of blood at the time of insertion of the arterial cannula into the aorta. Once the CPB systemic flow line and cannula are joined, the perfusionist should observe and manually palpate pulsation in the CPB line that corresponds with the patient's pulse before starting bypass. Direct continuous measurement of arterial line pressure by aneroid gauge or electronic transducer allows observation of the CPB systemic line pressure during initiation and throughout the course of bypass.

Massive air embolism may occur during aortotomy for cannulation if an intraaortic balloon is in place (106,107). Briefly turning off the balloon pump whenever the aorta is opened, for instance, when placing cannulas (arterial or cardioplegia) or vents, will prevent aspirating room air into the aorta with intraaortic balloon deflation (108).


Back to Quick Links

CPB depends on two basic requirements: adequate blood volume to maintain appropriate blood flow and adequate gas flow (in the correct composition) to maintain appropriate gas exchange. Because of the invasive nature of cardiac surgery and CPB, a third requirement is avoidance of air embolism at all stages of the procedure. Myriad threats exist in accomplishing these objectives. Miscellaneous problems that may jeopardize patient well-being include errors in drug administration, contamination of the circuit or patient, and transfusion reactions or administration of mismatched blood products. Because of the patient's dependence on the CPB circuit and ancillary monitoring devices, equipment malfunction or failure is always a potential concern. Fortunately, the incidence of perfusion accidents is low and the incidence of adverse patient outcome (defined as injury requiring prolonged hospitalization, permanent injury, or death) is rare.

Publication of case reports and national surveys examining CPB experience has further heightened practitioner awareness of errors and malfunctions associated with CPB. Surveys by Stoney et al. (109), Kurusz et al. (110), and Mejak et al. (111) have elucidated incidence rates in the United States for a variety of perfusion mishaps occurring in the 1970s, 1980s, and 1990s, respectively. Other broad-based surveys by Wheeldon (112) and Jenkins et al. (113) have reported experiences in the United Kingdom and Australasia. The following discussion addresses the more common perfusion incidents based on these survey reports and the authors' experience (114,115). Although direct comparison of results among different respondents during different survey time frames may be challenged, some sense of trends in the major reported CPB incidents may be discerned in Figure 28.5, A and B.

FIG 28.5. A and B: Specific incidents are compared from the major cardiopulmonary bypass (CPB) perfusion accident surveys in the last three decades. The total caseload experience represented in each survey was as follows: Stoney, 373,819; Wheeldon, 33,000; Kurusz, 573,785; Jenkins, 27,048; Mejak, 653,621. Total numbers of incidents per 1,000 cases are plotted in solid bars with reported adverse patient outcomes (defined as permanent injury, significantly complicated patient recovery, prolonged hospital stay, or death) to the right in open bars; incident rates are shown numerically for all categories. aPatient outcome for reversed left ventricular vent incidents not reported on Stoney survey. bJenkins reported "coagulation problems" and Mejak reported "coagulation problems following bypass" instead of the more specific "disseminated intravascular coagulopathy" or "consumption coagulopathy" as was used by Stoney, Wheeldon, and Kurusz; these differences in survey question wording could have accounted for the higher incident rates reported by Jenkins and Mejak. cStoney asked about "failure of delivery of oxygen to pump," whereas Kurusz asked about "oxygenator failure"; Jenkins asked about several specific types of "oxygenator failures" and "gas supply failures," the most common being membrane leak; there were 35 cases reported in his survey, for which the oxygenator required replacement before or during CPB. dJenkins and Mejak reported 13 and 140 additional "electrical or mechanical failures," respectively, that were not distinguished.

Inadequate blood flow

Acute aortic dissection can be a cause for inadequate blood flow during CPB and was discussed in the previous section because it most often is manifested at the time of initiation of CPB. Other causes of inadequate blood flow include low circulating volume that may result from a variety of causes. With the emphasis on reduced priming volumes for the CPB circuit and limited pre-CPB fluid administration by anesthesia personnel, there may be insufficient volume in the reservoir to maintain appropriate CPB systemic flow. Administration of vasodilators can drastically reduce CPB reservoir levels. Undetected blood leaks either from the CPB circuit or, more commonly, in the operative field or under surgical drapes will also reduce circulating blood volume. A particularly insidious blood leak may occur from Swan-Ganz catheter-induced pulmonary artery perforation, which carries a high risk of mortality (116,117). Hypothermia causes the catheter tip to stiffen and may predispose it to perforate the pulmonary artery (118). Deflating the balloon and withdrawing the catheter approximately 3 to 5 cm before or at the onset of CPB should lessen the risk of this complication (116), and a useful rule is to never inflate the balloon after systemic heparinization.

Inappropriate cannula size can reduce venous drainage, as will partial occlusion on the venous line. The venous or arterial cannulas may become malpositioned within the vasculature or inadvertently kinked or clamped, leading to reduced CPB blood flow (119–123). Rarely, the venous cannula may have a manufacturing defect that compromises the available cross-sectional area of its lumen, thus preventing normal venous drainage. Therefore, all cannulas should be inspected before use not only for proper size selection but also to confirm the absence of structural defects. A partially obstructed arterial cannula or oxygenator is manifested by elevated CPB systemic flow line pressures and lower than expected patient arterial pressures (124) or, if a centrifugal pump is being used, higher than expected revolutions per minute to maintain the desired flow (32). Another problem with centrifugal pumps is inaccurate flowmeter readings, which can arise when the flow probe is miscalibrated. With roller pumps, false displays of CPB pump output will result if the flow rate indicator on the CPB console is not set to match the tubing size (124).

Small quantities of entrained venous line air are rarely problematical in leading to a decrease in CPB systemic blood flow. However, if the entire venous line becomes filled with air, the gravity siphon effect will be lost (commonly referred to as air lock). This usually requires slowing or momentarily stopping CPB by clamping the venous line at the pump to allow personnel at the surgical field to refill the venous line with fluid and then "walking" any remaining air down the venous line into the venous reservoir to reestablish effective siphon drainage.

The CPB pump may fail, causing inadequate blood flow. Centrifugal pumps may decouple, meaning that the magnetic force to effectively rotate the pump has been lost between the pump console and base of the disposable pump head. This incident is often manifested by a high-pitched whining sound as the centrifugal pump motor accelerates to very high revolutions per minute; the in-line blood flowmeter also will show decreased flow that can progress to zero forward flow (and even retrograde flow) if not promptly corrected. An in-line valve in the CPB systemic flow line may prevent retrograde flow from the patient's aorta under these conditions (125). Underocclusion of a roller pump may lead to inadequate blood flow; this condition may be more difficult to detect than a centrifugal pump malfunction because in-line flowmeters are rarely used with roller pumps. Manifestations might include lower than expected CPB line or patient arterial pressures and decreased mixed venous hemoglobin oxygen saturation or PvO2with developing acidosis.

Inadequate gas exchange

Although there are fundamental physiologic issues related to whole body oxygen consumption and delivery during CPB (discussed in Chapter 27), mechanical problems related to inadequate gas exchange by the oxygenator include oxygenator failure and gas delivery system failures. Oxygenator failure is diagnosed by observing dark-colored blood exiting the oxygenator that cannot be corrected by increasing the concentration of oxygen in the ventilating gas. Blood gas analysis or in-line blood gas sensors can confirm visual assessment of inadequate oxygenation. Causes may include loss of gas supply to the CPB oxygen/air blender or flowmeter or, more rarely, failure of the blender. Leaks or obstructions in the gas delivery system have been reported (126–128). Use of an in-line oxygen analyzer, which is considered standard on anesthesia machines, should be incorporated into the oxygenator gas delivery line to monitor delivery of appropriate concentrations and warn of abnormally low oxygen flow. The optimal position of the sensor is to place it nearest the oxygenator (129). Kirson and Goldman (130) reviewed oxygenator gas delivery problems in detail and proposed a system consisting of an aspirating oxygen sensor and pneumotachography for more accurate detection of such problems. Problems with gas scavenging systems causing inadequate gas transfer in the oxygenator have been reported (131). The patient who is either too lightly anesthetized or is hypothermic with inadequate muscle relaxation may have higher oxygen consumption that can lead to decreased mixed venous hemoglobin oxygen saturation and decreased arterial PO2 if left uncorrected.

Inadequate anticoagulation leading to clotting in the oxygenator can cause inadequate gas transfer. The use of propofol anesthetic administered directly into the membrane oxygenator can potentially affect gas transfer, presumably by blocking pores or otherwise interfering with the membrane surface in the blood-contacting compartment of a microporous membrane. Therefore, propofol should be administered peripherally to permit deemulsification before it reaches the membrane oxygenator (132,133).

Fisher (134) recently provided data from the United Kingdom on the incidence of oxygenator failures requiring change-out. Two time periods were surveyed (1990 to 1992 and 1994 to 1996) during which the incidence was approximately 0.25 per thousand CPB procedures (1:4,000). Major causes for the change-outs differed between the two time periods. In the earlier survey, clotting in the oxygenator, particularly when aprotinin was being used, was the most frequently cited reason. In the latter survey, development of a high transoxygenator pressure gradient or a blood leak in the oxygenator was the major reason for change-out. Table 28.3 outlines steps for oxygenator change-out during CPB. Hart et al. (135) also described a technique for membrane oxygenator change-out that does not require stopping CPB. Earlier techniques with bubble oxygenator change-out required short periods (2 to 3 minutes) of circulatory arrest (136,137).

Electrical problems

Minor and major electrical problems are relatively common during CPB but only rarely cause adverse patient outcome. The entire CPB console or individual components can fail to operate as a result of wall power or electrical cord failures. Most hospitals have backup emergency generators to supply the operating room with electricity in the event of power outage. Sometimes these backup generators can fail to come on in a timely manner, making it prudent to have alternative electrical backup such as a uninterruptible power supply in the operating room and in-line with the CPB console for dealing with power failures. Some newer CPB consoles have battery power built in to provide continuous function in the event of electrical failure. All CPB systemic pumps should have provision for manual hand cranking readily available in the event backup electrical sources fail. The CPB console can be sabotaged, either by inappropriate conduct by untrained personnel or by malicious actions by those with more sinister intentions. In either situation, performance of a thorough pre-bypass checklist by the primary perfusionist at the time of setup should uncover potential malfunctions before placing the patient on bypass.

Electronic components within the console can fail, leading to erratic operation or cessation of function of individual components. Runaway pumps have been reported (138,139), as have electrical fires that can lead to loss of function of CPB components (140). Power surges or brownouts can lead to inaccurate readings on the CPB console or physiologic monitors. Piezo-electric or static charges created by roller pump rotation can interfere with the electrocardiogram (ECG) (141). This may be prevented by grounding the CPB console to a metal-jacketed temperature port in the oxygenator (142). ECG electrodes, lead wires, and cables should be intact to also avoid ECG "noise" when the patient is on CPB (143). Transducers may "drift" over the course of several hours, leading to inaccurate readings. It is therefore prudent to re-zero transducers before weaning from bypass to ensure accuracy of measured parameters that can influence management decisions (144). Temperature probes may become disconnected or malpositioned, leading to improper displays. It is therefore advisable to place at least two patient temperature probes for redundancy. In-line blood gas monitors or oximeters may display erroneous values unless calibrated against standard laboratory samples. The electrocautery can interfere with monitor displays; use of a shielded ECG cable will minimize such interference.

There are a variety of alarms used on anesthesia, perfusion, and ancillary equipment in the cardiac surgical operating room. Often, similar sounding high-pitched signals can make rapid identification of the source of the alarm difficult. Cases of misinterpreted alarms (e.g., intraaortic balloon and low-level alarm on CPB reservoir) have distracted the perfusionist with serious patient consequences (145). Alarms may be obtrusive, and there may be a tendency among some practitioners to disable alarms that are designed to give early warning or automatically respond if user-set thresholds are exceeded when potentially hazardous patient conditions are met (146,147).

Perfusionists, anesthesiologists, and nurses must know how to troubleshoot equipment used during CPB. Performing team drills for the unexpected perfusion crisis such as oxygenator change-out, hand cranking for electrical failure, or dealing with massive air embolism can be life saving during these rare events. Cooper et al. (148) cited lack of familiarity with equipment as one major factor in operating room mishaps, and Gaba (149) provided an excellent review of human error during conduct of anesthesia that has applicability to performance of CPB.

Air embolism

Concern over air embolism during CPB has been a constant focus of all involved in the perioperative care of the cardiac surgical patient. Air embolism (venous or arterial) may occur not only from CPB (pump air) but during a variety of surgical procedures (surgical air) or at the head of the table from actions or inaction by anesthesia personnel (anesthetic air) (150). The pathophysiology of air embolism has been extensively studied in laboratory experiments, many of which predated the clinical use of CPB. Case reports detailing mechanisms of air embolism have appeared periodically in the literature. The next sections review the more common etiologies, followed by a discussion of treatment strategies for this complication.

Surgical (operative) air

In 1914, the danger of air embolism during cardiac surgery was reported by Carrel (151), who wrote, "The opening of the ventricles or of the pulmonary artery and the aorta is always followed by entrance of air into the heart." Support for his statement was the observation of ventricular fibrillation and death in animals after coronary artery air embolism.

Air embolism on the left side of the heart was well known to thoracic surgeons before the advent of open heart surgery. Reyer and Kohl (152) reported 10 cases of venous or arterial air embolism, 5 of which resulted in the patients' deaths, during a variety of surgical or diagnostic procedures. Kent and Blades (153) further warned that the two major hazards of thoracic surgery were infection and embolic phenomena. In their animal experiments, air embolism was found to be well tolerated on the venous side, in the absence of a patent foramen ovale, but fatal with small injections of air into the pulmonary veins.

Geoghegan and Lam (154) reported that the mechanism of death due to air embolism in dogs (0.25 to 2.0 mL/kg) was either coronary (immediate death) or cerebral (severe brain damage). Benjamin et al. (155) further sought to define the mechanisms of air embolism by injecting varying amounts of air (0.5 to 8.0 mL/kg) into the left atrium, left ventricle, aortic root, common carotid, or descending aorta of dogs. Left atrial air embolism was fatal 100% of the time, whereas the same volumes of air injected into the left ventricle caused death in 83% of the animals. Aortic root and carotid air embolism were better tolerated, and large volumes of air (up to 10 mL/kg) were required to cause death when given into the descending aorta. They, like other investigators (156–164), noted the dangers of left heart air but concluded that small amounts of air in the systemic circulation were generally well tolerated during surgical procedures if appropriate resuscitative maneuvers were undertaken when required.

Many retrospective reviews of early clinical experience with CPB have been published. Callaghan et al. (165) analyzed 60 deaths in 250 CPB patients operated on between 1956 and 1961. A variety of causes described included seven cases of cerebral damage, four of which were from air embolism. Ehrenhaft et al. (166) reported 19 of 244 (7.7%) patients undergoing open heart surgery suffered cerebral damage. Systemic air embolism was the suspected etiology because many operations involved closure of septal defects. Like Carrel nearly 50 years earlier, they warned of air entrance to the left side of the heart or aorta with subsequent embolization when the normal circulation was restored. Allen (167) reported cerebral damage in 18 of 500 (3.6%) patients undergoing repair of valvular or congenital cardiac defects and warned of the propensity of air to collect in the left atrium near the right superior pulmonary vein. Sloan et al. (168) reported 78 of 600 (13%) patients died after CPB; air trapped in the left ventricle was identified as the source of the air and was believed responsible for 49 deaths. Nicks (169) reported systemic air embolism in 40 of 340 (11.7%) patients undergoing congenital or valvular procedures; 10 patients died. Fishman et al. (170) later confirmed the left atrium and pulmonary veins as locations of trapped air whenever the left heart was opened. Anderson et al. (171) and Lin (172) also warned of the risks of pulmonary hypertension due to right-sided air embolism.

The preference of cannulation of the ascending aorta for CPB, although much safer than the femoral artery site, increased the risk of cerebral air embolism. Gomes et al. (173) found that air embolism via the femoral artery was five times less likely to involve the cerebral vessels if the air originated from the CPB arterial line. Beckman et al. (174) determined the optimum method of placement of the ascending aortic cannula to lessen the risk of air entry. Direct insertion of the cannula without use of a side-biting clamp, which tended to trap a small amount of air, was found to be safest.

In the article by Mills and Ochsner (114) on mechanisms of air embolism, two additional surgical sources were described: unexpected resumption of the heart beat and inadequate steps to remove air after cardiotomy. Coronary air embolism with cardioplegia techniques was reported in 1981 (175) and 1986 (110). Although air embolism was generally thought not to occur during coronary bypass operations, Hughes (176) reported the possibility of intraventricular air with right superior pulmonary vein venting that could draw air in via coronary arteriotomy, especially when the left anterior descending coronary artery was opened. Robicsek and Duncan (177) and Lee (178) subsequently confirmed this mechanism. Air also can be drawn retrograde through an opened coronary artery if an aortic root vent is used and placed under significant negative pressure.

Cardiopulmonary bypass (pump) air

Air embolism originating from the CPB circuit may enter the patient's vascular system either from the arterial line or by other mechanisms, many of which are discussed in other chapters. Reed et al. (179) reviewed the many mechanisms for air embolism from the CPB circuit, some of which have been reported in the literature and others learned through direct or anecdotal experience.

Arterial line air embolism due to emptying of the CPB reservoir, as occurred during Dennis' early case (180), has been a common cause. The current popular use of membrane oxygenators in which the blood is drawn from a venous reservoir and then pumped through the oxygenator may have decreased the incidence of arterial line air embolism that was much more prevalent when bubble oxygenators were widely used (110). However, inattention to the reservoir level can still cause air to be transmitted to the CPB systemic flow line, regardless of oxygenator type. The importance of maintaining an adequate volume in the CPB reservoir was reported in 1958 (181) and has been one of the fundamental safety principles taught in perfusion training programs since their inception.

High pump flows in conjunction with vertical CPB reservoir outlets may cause vortexing of air into the systemic flow line. Newer model reservoirs with angled or horizontal outlets are less prone to this condition.

The arterial roller pump head tubing may rupture, causing arterial air embolism. Cases of electrical malfunction of the arterial pump (roller and centrifugal) were previously mentioned (138,139); in some circumstances the malfunctioning pump can spontaneously accelerate to a high speed, drawing air into the CPB systemic flow line. Less dramatic, but just as dangerous, the arterial roller pump may rotate slowly before or after CPB and may be unnoticed by the perfusionist, leading to emptying of the reservoir and subsequent air embolism. If the arterial pump is not operating, a second roller pump for blood cardioplegia can draw air into both the cardioplegia delivery circuit and the arterial line.

Accidental disconnects, punctures, cuts, or openings, such as stopcocks left open to atmosphere, in the arterial line can cause air embolism depending on flow conditions. Generally, such disruptions will cause blood loss from the circuit, but under low CPB systemic flow conditions air may enter the arterial line (182,183). Centrifugal pumps, when connected directly to the patient's venous system, can draw air into the CPB circuit if intravenous lines are open to atmosphere. Breaks in the integrity of the CPB arterial line are especially favorable for air entry if they occur on the negative (inlet) side of the pump. Strictures in the arterial line on the positive side from kinks, application of clamps, or excessive flows through small-diameter connectors or cannulas can produce air embolism from cavitation effects (184).

The oxygenator or venous or cardiotomy reservoir may become pressurized and transmit air into blood lines connected to the patient if ports designed to vent them to atmosphere become occluded (185,186). The current popularity of vacuum-assisted venous return has the potential for pressurizing the cardiotomy reservoir if vacuum is not maintained and the pump suction or vent pumps continue to operate. Vacuum-assisted venous drainage also has the potential to deprime the arterial line filter or oxygenator if vacuum is applied before establishing CPB. A pressurized cardiotomy reservoir may transmit air retrograde to the patient's heart via the vent line (114) or to the arterial filter if the purge line is connected to the reservoir and a one-way valve is not incorporated into the purge line. Retrograde venous air embolism from bubble oxygenators due to blockage of the gas scavenging line was described by Wells and Stiles (187) and confirmed by others (188,189). Alternatively, if the gas phase of either a silicone rubber or microporous membrane oxygenator becomes pressurized above blood phase pressure, ventilating gas can enter the arterial line through the membrane material.

Anesthetic air

The risk of air embolism in the course of anesthetic management of patients undergoing open heart surgery most commonly occurs with intravenous or monitoring lines. Despite controversy regarding the risk of small quantities of venous air, cases of fatal air embolism during neurosurgical procedures with patients in the sitting position were reported as early as 1902 (190). During animal experiments, Goodridge sought to dispel the earlier report by Hare (191) that venous air embolism was innocuous. He concluded (190) ". . .I would say that I believe the statement, "that large quantities of air may be introduced into the veins without unfavorable results' to be pernicious teaching and not supported by fact."

Inappropriate ventilation of the patient during insertion of CPB cannulas or a left atrial monitoring line can cause air embolism. Expanding the lungs fully to displace pulmonary venous air is an important adjunct to surgical de-airing maneuvers; if not performed properly, air may be retained and later embolize to the arterial circulation.

The risk of greater than 50% nitrous oxide ventilation in promoting bubble growth has been reported by several authors (192–195) and was reviewed by Munson (196). Wells et al. (197) found increased cerebrospinal fluid markers indicative of cerebral ischemia when nitrous oxide anesthesia was used in conjunction with bubble oxygenation. Conventional wisdom is that nitrous oxide should be avoided from the time of cannulation for CPB until emergence from CPB. Many would also advise its avoidance after CPB, because the likelihood of some intravascular air is relatively high, especially in procedures where the heart has been opened.

Treatment of air embolism

Peirce (198) wrote that treatment of iatrogenic air embolism has not kept pace with other medical knowledge, perhaps because it is a rare and frequently unrecognized complication. Intraoperative treatment for air embolism in open heart surgical patients is unique when compared with other medical procedures where air embolism may occur. If coronary air embolism results in myocardial dysfunction, the circulatory needs of the patient can be met by the CPB circuit. Systemic heparinization, a requisite for CPB, has been shown to be of benefit in reducing interactions of bubbles with blood but may be deleterious if brain infarction has occurred (199). Hemodilution to levels commonly used during CPB will reduce blood viscosity and improve tissue perfusion when air embolism occurs. With the chest open, the surgeon is able to aspirate air directly from the heart chambers or vessels. Venting, induction of hypothermia (200,201), retrograde coronary sinus (202,203) or cerebral perfusion (114), or direct cardiac massage are all available if air embolism occurs during the open heart operation. If right coronary artery air embolism is suspected (isolated acute ST segment elevation in the inferior ECG leads is highly suggestive), the surgeon can transiently raise the pressure in the proximal aorta by gently pinching the aorta distal to the arterial cannula with the left hand while using the index finger of the other hand to assess proximal aortic pressure. The transiently increased aortic pressure will push intracoronary air bubbles through the coronary circulation, which often can be seen if significant right coronary air embolism has occurred. This maneuver should be tried in all patients before implementing intraaortic balloon counterpulsation when there is difficulty encountered in weaning from CPB.

Drug or fluid therapy by the anesthesiologist also may be instituted immediately via the CPB circuit. Packing the patient's head in ice will decrease cerebral metabolism and may be beneficial (204). Ventilating the patient with 100% oxygen favors bubble resolution and can limit cerebral ischemia (205). If air embolism occurs before or after CPB, conventional means of treatment can immediately be used such as positioning the patient head-down and turning the patient into the left lateral recumbent position (206,207) or instituting cardiopulmonary resuscitation (208).

The source of air must be determined and promptly interrupted to prevent further transmission of air into the patient. If the source of air is the CPB systemic flow line, then CPB should be stopped immediately (114,209). The arterial and venous lines should be clamped to prevent additional embolization or exsanguination. After confirming sufficient volume in the CPB reservoir, air in the arterial line should be purged out by aspirating with a large syringe or refilling the line using the systemic pump. If a pressurized reservoir or CPB component is the source, the pressure should be relieved before releasing clamps on lines directly connected to the patient.

If air has filled the aorta, the arterial line can be disconnected from the arterial cannula. After clamping the venous cannula(s) and disconnecting the venous line, the arterial line and venous line are joined so perfusate can be quickly recirculated back to the CPB reservoir to remove the air. Once cleared of air, the lines are reconnected to the appropriate cannulas and CPB may be resumed. If air has not reached the arterial line filter, it can be vented out the purge line by slowly advancing flow with the systemic pump. Air can be removed from a centrifugal pump by detaching it from the console and positioning the pump head outlet uppermost to take advantage of buoyancy effects; with the pump head de-aired, recirculation or resumption of CPB can then be accomplished. Air may be removed from the membrane oxygenator by recirculating at high flow via the membrane recirculation line. Tapping and inverting components, as is performed during initial priming, may be required to effectively remove CPB air. If the arterial filter has been filled with air, it may be necessary to clamp the filter out of the circuit and use an arterial filter bypass line to quickly reestablish CPB blood flow. Alarms to detect air embolism often must be turned off to reestablish CPB systemic flow but then should be reengaged after the patient is safely back on bypass.

Placing the patient in the Trendelenburg position after air embolism is common practice but may not aid in bubble removal from cerebral vessels deep within the brain. Butler et al. (210) studied the distribution of carotid artery injections of air (0.5 to 1.0 mL) in dogs placed in 0 to 30-degree Trendelenburg position. They concluded that regardless of the position, if the heart was ejecting blood, bubbles would not be retarded in their distribution into the brain. In additional in vitro studies in the same report using a simulated carotid artery, the authors verified that the buoyant properties of the bubbles were not sufficient to prevent the blood from carrying bubbles in the direction of flow. They did suggest, however, that in conditions of circulatory arrest or extremely low flow, bubbles would have a propensity to rise in stagnated blood.

More recently, Mehlhorn et al. (211) studied the effects of body repositioning on the hemodynamic response to large infusions of venous air (2.5 mL/kg at a rate of 5 mL/sec) in dogs. Testing the supine, left lateral recumbent (Durant's maneuver), left lateral recumbent with 10-degree head-down, or right lateral recumbent positions, they found no significant differences among the various body positions in terms of heart rate, blood pressure, pulmonary artery, central venous, or left ventricular end-diastolic pressures, or cardiac output. There also were no significant differences in recovery times regardless of body position.

If large quantities of air are suspected of having entered the cerebral vessels, retrograde cerebral perfusion, as first described by Mills and Ochsner (114), may be used with remarkable results. There are a number of case reports now in the literature (212–220) confirming the benefit of this technique, as well as anecdotal information from survey respondents (110). The fact that retrograde cerebral perfusion is being used electively more frequently during aortic arch surgery may facilitate implementation of this method under emergency condition. Table 28.4 outlines steps to perform retrograde cerebral perfusion for massive air embolism.

Drug therapy for arterial air embolism is aimed first at raising the arterial blood pressure to force bubbles through tissue vasculature to the venous side of the circulation (221). Corticosteroids, diuretics, antiplatelet agents, anticonvulsants, and barbiturates have been advocated to decrease cerebral manifestations of ischemic injury due to air embolism. Lidocaine pretreatment has been shown to be beneficial in an experimental setting (222). Surface tension-reducing agents have been proposed (223–225). More recently, the use of perfluorocarbons to enhance oxygen delivery has been suggested for treatment of air embolism (226); however, it apparently has not been used clinically for treatment of air embolism (227). Such agents may also reduce the surface tension at the bubble–blood interface and would aid in absorption and dissolution of the bubbles (228,229). Speiss et al. (230,231) studied the protective effects of prophylactically administered perfluorocarbon solutions in experimental coronary and cerebral air embolism in the laboratory. Treated animals had significantly fewer dysrhythmias and less decrease in myocardial function when pretreated.

The patient with known or suspected air embolism should be ventilated with 100% oxygen (232). Ventilation with nitrous oxide, if it is being used, should be discontinued to decrease the possibility for bubble expansion (233,234). Hlastala and Van Liew (235) showed that a 2-mm bubble will disappear in approximately 1 hour if the patient is breathing oxygen; if the patient is not being ventilated and is breathing room air, the same size bubble will persist for 9 hours.

Regardless of when air embolism occurs during the intraoperative period, most authors have recommended that the surgical procedure should be completed (114,209). An air embolism incident in the course of an open heart surgical procedure can create confusion on the part of various team members. Having a predetermined treatment plan (209,236) for dealing with this event can be life saving. Tovar et al. (237) recently published an excellent review of management of air embolism. A proposed algorithm from their article is shown in Figure 28.6.

FIG 28.6. Algorithm for postoperative management of air embolism. (From Tovar EA, DelCampo C, Borsari A, et al. Postoperative management of cerebral air embolism: gas physiology for surgeons. Ann Thorac Surg 1995;60:1138–1142, with permission.)

The most effective postoperative treatment for air embolism is compression in a hyperbaric chamber and ventilation with 100% oxygen (238–248). Although the logistics of moving a critically ill patient on a ventilator and with invasive monitoring and multiple intravenous lines can be daunting, treatment in such a facility has been credited for many complete recoveries. Recoveries have been reported even if such treatment is delayed (249,250). However, it must be recognized that access to a multiplace hyperbaric chamber may be limited. There are only approximately 100 such facilities in the United States, most of which are located on the coastlines or near large bodies of water and therefore are at great distances from many hospitals performing open heart surgery. The patient with suspected air embolism should not be transported by air because reduced air pressure during flight will cause intravascular bubbles to expand. Brenner (209) suggested that treatment regimens for air embolism should be continued unless the patient expires or is diagnosed as brain dead. Table 28.5 outlines management strategies for treatment of air embolism.

Miscellaneous problems

The CPB circuit or its components may become contaminated by inattention to detail during setup or operation. The obvious risk of bacterial contamination via the patient's bloodstream necessitates observing sterile technique at all stages of use of CPB. Infection after cardiac surgery is an uncommon but serious complication leading to prolonged duration of mechanical ventilation and intensive care unit stay. One third of these patients may die of causes related to infection (251).

The practice of dry preassembly of the CPB circuit is not without risk. The Centers for Disease Control and Prevention recently reported an investigation into an outbreak of gram-negative bacteremia in nine cardiac surgery patients at one hospital (252). The source of the infections was found to be caused by housekeeping personnel inadvertently contaminating uncovered preassembled disposable pressure transducers on the CPB console with sprayed cleaning water. This complication can be prevented by ensuring all blood-contacting surfaces and devices on the CPB circuit are kept covered until time of use. Alternatively, the entire CPB circuit can be set up just before use to minimize risk of contamination. Hospital-prepared cardioplegic solution may be the source of contamination by Enterobacter cloacae (253,254) due to defective equipment used by manufacturing pharmacy personnel. Thus, meticulous attention to quality control, including batch sterility testing of the cardioplegic solution, will minimize this risk. Alternatively, commercially prepared cardioplegic solutions are usually subject to more stringent quality controls.

Drug errors by any cardiac surgical team member are relatively common and can lead to adverse outcomes. The most commonly cited drug-related problem in one survey (110) was overdosage of either a vasodilator or vasoconstric tor. Although these types of errors may be short-lived and of no clinical consequence, drug errors related to anticoagulation can be fatal. Protamine administration during CPB can effectively render the oxygenator, reservoir, and arterial line filter unusable due to development of gross clot. A case of inadvertent protamine administration instead of heparin has been reported (255). The patient survived, but a massive clot was found in the oxygenator, a smaller clot in the arterial filter, and fibrin strands in the cardiotomy suction tubing. No activated coagulation time or heparin assay testing was performed. The hospital has now changed their policy to no longer store protamine in the operating room; it is now reportedly not drawn up until the surgeon requests that protamine be administered.

Another serious complication is transfusion of ABO-incompatible bank blood. The manifestation is an acute hemolytic reaction that can damage the kidneys. Prevention lies in adherence to strict double checking of blood units against the patient's identification wristband. Maintenance of urine output with diuretics, intravenous fluid support, and administering corticosteroids and cardiotonic drugs has been used with success (256).


Back to Quick Links


Although the rate of cardiovascular disease in pregnancy has steadily declined over the past several decades, it remains a topic of interest because of the high rate of fetal mortality when an operation requiring CPB is needed. The incidence of all maternal cardiovascular disease during pregnancy is currently about 1.5% (257). The most frequently performed procedures include closed mitral commissurotomy, open mitral commissurotomy, mitral valve replacement, and aortic valve replacement. Credit for the first case using CPB in a pregnant patient has been attributed to Leyse et al. (258) in 1958, with three more cases reported subsequently (259,260). Other cardiac problems associated with pregnancy in which operation may be necessary are atrial septal defect, patent ductus arteriosus, ventricular septal defect, and idiopathic hypertrophic subaortic stenosis (261). Because of intravenous drug abuse, infectious endocarditis assumes an increasing role among cardiovascular diseases in pregnancy.

A number of fundamentals in the physiology of pregnancy affect the timing and use of CPB in the pregnant patient. The first trimester is the trimester of organogenesis. Therefore, any injury to the embryo during this period may result in teratogenesis. This is particularly true of drugs. An important example is warfarin, a well-known teratogen. Because of the small size of the warfarin molecule, it can cross the placenta. The neonatal morbidity and mortality for patients taking warfarin throughout their pregnancy may be 40% or greater. For this reason, patients undergoing warfarin therapy are often changed to heparin when pregnancy is confirmed. The large heparin molecule cannot cross the uteroplacental barrier (262). However, the rate of spontaneous abortion increases with heparin.

Other factors associated with CPB that may lead to teratogenesis include hypoxemia and decreased perfusion. During the second trimester of pregnancy, the period of organogenesis is finished; therefore, teratogenesis is not seen. The risk of premature labor is much less than in the third trimester, when the risk rises significantly. The hypervolemia and anemia of pregnancy also are less in the second trimester than in the third trimester, as are the hemodynamic demands because of the smaller size of the fetus and lesser uterine blood flow.

Numerous physiologic changes occur during pregnancy. Because there is no autoregulation of uterine blood flow, nonpulsatile flow during CPB may compromise fetal blood supply. After the 28th to 32nd week of gestation, the patient's blood volume has increased by 30% to 50%. Heart rate increases 10% to 15% and stroke volume increases 30% to 35% with an associated 15% decrease in systemic vascular resistance. During pregnancy, uterine blood flow represents 10% to 15% of the cardiac output, as compared with only 1% in a nonpregnant female (261).

The lowest risk for CPB surgery is believed to occur during the second trimester, with reported maternal mortality risk during this time varying from 1.5% to 5% (263–266). In particular, a survey (257) regarding bypass surgery on pregnant women revealed only 1 maternal death in 68 cases. In general, the maternal risk is probably not increased over that in a nonpregnant patient. Fetal risk, however, is quite high, ranging from 10% to 50% (261,264–272). The previously cited survey (257) revealed a 20% fetal mortality in the same 68 CPB cases.

A closer examination of the causes of fetal risk reveals multiple factors. There are numerous possible causes for fetal hypoxia. Low CPB systemic flows or hypotension during bypass can result in fetal hypoxia because of the lack of autoregulation of the uterine blood flow. Low hemoglobin oxygen saturation and possibly the lack of pulsatile flow, uterine arteriovenous shunts, or uterine arterial spasm may contribute to fetal hypoxia. Hypoxia may result from particulate or bubble embolization to the uteroplacental bed (273). Uterine blood flow also may be compromised by venous obstruction of the inferior vena cava resulting from improper cannula placement (274).

Hypoxia is best diagnosed by noting fetal heart rate (FHR) decelerations. The normal FHR is between 120 and 160 beats/min. The FHR may decrease to 80 to 100 beats/min with hypothermia, but an FHR less than 60 beats/min suggests a high probability of life-threatening fetal distress. Acidosis may contribute to fetal bradycardia. Fetal bradycardia, defined as FHR less than 120 beats/min at normothermia and FHR less than 80 beats/min at hypothermia, may be treated by increasing the CPB systemic flow rate, which usually increases the heart rate toward normal (270,275,276). However, one problem that may occur is a temporary increase in the heart rate followed by a heart rate reduction that does not respond to increased flow. For this reason, time on CPB is a significant factor and should be minimized (265).

The initiation of uterine contractions most frequently causes fetal death (264,265). The dilution of progesterone during CPB secondary to hemodilution may predispose to the onset of uterine contractions (270), which can be detected by a tachodynamometer. When the FHR is monitored simultaneously, the FHR pattern may show late decelerations with each uterine contraction, indicating hypoxia. This usually indicates that blood flow across the myometrium stops as uterine contraction pressure increases to exceed the uteroplacental arteriolar pressure. Most uterine contractions resolve after the termination of CPB. Risk of initiation of uterine contractions does not end, however, with termination of CPB; therefore, it is recommended that uterus be monitored for contractions for 72 hours postoperatively (269). The rewarming cycle during bypass is associated with the onset of uterine contractions (269). Uterine contractions are treated with tocolytic agents (see below). Progesterone may also be useful.

Perfusion problems constitute a potential source of fetal complications. These include nonpulsatile perfusion, inadequate perfusion pressure, inadequate systemic blood flow, embolic phenomena to uteroplacental circulation, or alterations in placental blood flow secondary to cannulation. Meffert and Stansel (262) noted an increase in fetal mortality when mitral valve replacement was performed, as compared with that resulting from mitral commissurotomy. Those authors believed that this difference resulted from the longer CPB times required for mitral valve replacement. Fetal mortality rate has been noted to increase with an increase in CPB time in some reviews (265), yet Weiss et al. (266) could find no correlation between time on CPB and neonatal outcome in their recent survey of the literature.

A number of strategies have evolved to avoid complications, and techniques have been devised to reduce the rate of fetal complications. Because of the marked increase in cardiac output during pregnancy, normal cardiac output during the third trimester of pregnancy may be 6 L/min; therefore, a perfusion flow of 4 L/min represents only two thirds that of the normal flow and could result in fetal hypoperfusion. Using a membrane oxygenator and an arterial line filter in the CPB circuit minimizes particulate and gaseous emboli. FHR can be adequately maintained by perfusion at either normothermia or mild hypothermia. In pregnant patients, most authors recommend systemic pressures ranging from 60 to 75 mm Hg, preferably accomplished with high perfusion flows rather than with -adrenergic agonists. FHR monitoring should be used to detect the decreases in heart rate associated with poor fetal perfusion. Because FHR is correlated with CPB systemic flow, fetal bradycardia should be treated with an increase in CPB flow rate. Both FHR and uterine activity monitoring should be maintained for 48 to 72 hours postoperatively. Table 28.6 outlines management strategies for the pregnant patient.

In the third trimester of pregnancy, particularly because of the increased rate of premature labor, expectant management of the patient allowing the fetus to stay in utero as long as possible plays an increasing role. Options may include expectant management of the mother with delivery or cesarean section followed by maternal cardiac surgery, maternal surgery with the fetus remaining in utero, or simultaneous cesarean section and maternal cardiac surgery. The approach selected should balance the maturity of the fetus with the severity of maternal cardiac disease and the risk of labor and delivery with uncorrected cardiac disease. If uteroplacental insufficiency is present (suggested by fetal bradycardia), delivery of the fetus before CPB should be considered. Although the exact role is not known, avoidance of CPB altogether by using a closed procedure such as closed mitral commissurotomy or balloon valvuloplasty may be an option in patients at particularly high risk for undergoing CPB.

Fetal complications could also be theoretically attributed to nonpulsatile perfusion, hyperoxygenation, or anticoagulation with heparin. At present, none of these factors has been clearly identified as the cause of fetal problems. Hypothermia also causes FHR decelerations; therefore, normothermia or mild hypothermia (32°C) is generally used. However, Buffolo et al. (277) reported use of deep hypothermia with circulatory arrest for aortic arch surgery with survival of the mother and fetus. Others (271,278) also reported successful cases in which hypothermia with a period of circulatory arrest was required. Most sources believe that a hematocrit exceeding 22% should be maintained as well as a CPB systemic flow index of 3.0 L/min/m2. To avert the potential effects of aortocaval compression by the gravid uterus, central cannulation is preferable to femoral arterial or venous cannulation. Late in pregnancy, the right flank should be elevated to prevent aortocaval compression (262).

When necessary, tocolytic agents are used. At term, the uterus is more responsive to tocolytic agents than it is earlier in pregnancy (269). A number of tocolytic agents are available. The agent should be chosen because of its effectiveness and its lack of side effects. A number of tocolytic agents may have cardiovascular side effects, particularly -adrenergic agonists such as terbutaline (261,270,279). Ritodrine and magnesium sulfate have proven most effective in clinical trials (268). Other approaches include ethanol or a combination of a -adrenergic agonist and progesterone (270). Because of limited side effects, magnesium sulfate may be the best tocolytic agent for pregnant patients undergoing CPB associated with diabetes mellitus or hypertension.

If measures to avoid or treat fetal complications fail and fetal distress occurs, measures must be taken to correct this problem. If fetal distress can be prevented, fetal mortality will decline. Metabolic acidosis, when present, is treated with sodium bicarbonate. Glucose is given to replenish decreased fetal glycogen stores, and reduced arterial or venous hemoglobin oxygen saturations are promptly corrected by increasing oxygenator gas flow, CPB systemic flow, or hemoglobin concentration as indicated (275). Inotropic or vasopressor agents should be avoided when other treatments may be substituted. For example, if the patient has lost blood and needs volume replacement, transfusion would be preferred over vasopressor agents. Epinephrine has been recommended because of its rapid onset, brief duration, and minimal unwanted side effects at moderate doses. At high doses, epinephrine predominantly manifests -adrenergic effects. Because blood flow to the gravid uterus is primarily under -adrenergic control, -adrenergic agonists will decrease uterine blood flow in normotensive pregnant patients. Drugs that exhibit a combination of - and -adrenergic activity are most useful because they increase maternal blood pressure without reducing uterine blood flow (273,280). Examples of such drugs are ephedrine and moderate-dose epinephrine. A reduction in uterine blood flow has been seen in animal studies in which hypotension was treated with dopamine or hypertension was treated with nitroprusside. In addition, nitroprusside crosses the placenta in animals and may liberate free cyanide ions and cause metabolic acidosis. Large doses of nitroprusside in gravid ewes uniformly result in maternal and fetal death (263,281).

In summary, CPB during pregnancy is associated with a low rate of maternal complications but a high risk to the fetus. If possible, bypass surgery should be avoided during pregnancy. If the disease process dictates that surgery is necessary, the first and third trimester are best avoided to minimize teratogenesis or premature initiation of uterine contractions and labor, respectively. It appears likely that fetal distress and complications can be avoided by maintaining a high mean arterial pressure, high CPB systemic flows, normothermia or mild hypothermia, by the use of fetal monitoring, and by early treatment of FHR decelerations and uterine contractions. If fetal distress does become apparent, corrective measures must be taken.

Renal failure

The effects of CPB on renal function is discussed fully in Chapter 19. However, in patients with preexisting renal disease, cardiac surgery and CPB present difficulties primarily related to their inability to excrete potassium and tendency to become fluid overloaded during the perioperative period. Secondary problems involve anemia, acidosis, platelet dysfunction, and hypofibrinogenemia, all of which may be exacerbated by conventional CPB. The patient on chronic hemodialysis may also present with sepsis, bleeding tendencies, malnutrition, and glucose intolerance.

Hemodialysis can be performed concurrently with CPB and offers benefits over either scavenging hyperkalemic cardioplegic solution (282) or using conventional hemoconcentration, which is not as effective in removing potassium. The first case of hemodialysis during CPB has been attributed to Soffer et al. (283) in 1979. The patient was a 55 year old in congestive heart failure from aortic insufficiency secondary to infective endocarditis. The patient had been on maintenance hemodialysis for 10 months before admission for aortic valve replacement. The CPB circuit was primed with two units of packed red blood cells, fresh frozen plasma (900 mL), and balanced electrolyte solution (300 mL), which resulted in an initial hematocrit of 30% on CPB. A conventional hemodialysis machine was connected to the CPB circuit, drawing blood from the venous line and returning it to the cardiotomy reservoir (flow rate 180 mL/min); the dialysate flow rate was 500 mL/min. Serum potassium values ranged from 3.2 to 4.3 mEq/L, and the patient did not require hemodialysis again until postoperative day 3.

In 1984, Geronemus and Schneider (284) reported continuous arteriovenous hemodialysis using a hollow-fiber hemoconcentrator to treat acute renal failure in 10 critically ill patients. Peritoneal dialysis fluid was slowly pumped (15 to 20 mL/min) countercurrent to blood flow that relied on the patient's arterial blood pressure as the driving force. Because of these low flow rates, the duration of continuous arteriovenous hemodialysis was maintained for between 23 and 108 hours. They found that advantages included simplicity of the extracorporeal circuit, hemodynamic stability during treatment, and excellent urea and creatinine clearance. The technique was less effective in treatment of volume overload with fluid removal rates between 25 and 100 mL/hr.

In 1985, Hakim et al. (285) reported a series of 26 patients undergoing cardiac surgery, including cardiac transplantation, who had impaired or absent renal function preoperatively. In five cases a conventional hemodialysis machine was used in the operating room by accessing the patient's circulation via the CPB circuit. The other 21 had hemodialysis performed using continuous arteriovenous hemodialysis via a shunt in parallel with the CPB circuit (blood flow 200 to 300 mL/min). Benefits described were no fluid retention after CPB, no increase in the serum potassium (average 4.5 mEq/L) or blood urea nitrogen, unlimited ability to use potassium cardioplegia, and no need for increased heparin dosages to maintain anticoagulation. Murkin et al. (286) also reported intraoperative hemodialysis in 12 patients with renal failure. All patients underwent conventional hemodialysis the evening before surgery and had a portable dialysis machine connected to the CPB circuit. Blood flow was between 150 and 300 mL/min and dialysate flow was 500 mL/min to maintain the serum potassium at 4.0 mEq/L. Benefits included simplicity and intraoperative control of metabolic changes when other methods would have been ineffective.

Wheeldon and Bethune (287) published an excellent review of the principles of hemofiltration (hemoconcentration) and hemodialysis during CPB, including available equipment and suggested circuitry. Besides providing safe intraoperative management of patients in renal failure, they concluded that hemodialysis or hemoconcentration was a simple, efficient, and inexpensive method for control of patient blood volume and blood conservation during CPB. Sutton (288) also recently reviewed CPB management for patients with renal failure and provided a useful table listing sieving coefficients of drugs and ions (Table 28.7).

Other authors (289–292) reported the benefits of intraoperative hemodialysis when compared with routine hemodialysis before and after CPB (usually postoperative day 1) to control serum potassium and reverse fluid overload. A primary benefit appears to be delay of reinstitution of conventional hemodialysis with systemic heparinization until postoperative day 2 or 3. Hamilton et al. (290) described use of a hollow-fiber hemoconcentrator to perform hemodialysis during CPB. They relied on regulated gravity flow of peritoneal dialysis fluid matched to blood flow (300 to 500 mL/min) through the hemoconcentrator. For increased removal of solutes, the dialysate flow can be increased. However, if the serum potassium is low, dialysate flow can be stopped while maintaining hemoconcentration. If the serum glucose becomes elevated, they recommended using normal saline as the dialysate. They also noted the importance of careful monitoring of the blood pH and HCO3values, with sodium bicarbonate administration as necessary. A diagram of hemodialysis using two roller pumps to control dialysate flow in and out of the hemoconcentrator is shown in Figure 28.7.

FIG 28.7. Circuit for performing hemodialysis with a hollow-fiber hemoconcentrator during cardiopulmonary bypass. Blood and fluid flow is in direction of arrows, and bold Xs represent placement of tubing clamps. The arterial filter purge line (top right) is used as a source of blood (estimated flow 150 to 300 mL/min) through the hemoconcentrator and is returned to the venous reservoir. Dialysate fluid is drawn from large bag by roller pump #1 and pumped through the outer chamber of the hemoconcentrator while simultaneously being drawn (roller pump #2) from the hemoconcentrator effluent port where it is discarded into a collection cannister (lower left). The speed of pump #1 must always be less than the speed of pump #2 to prevent dialysate from crossing hollow fibers in the hemoconcentrator and entering the bloodstream. For most effective solute removal, blood flow and dialysate flow should be countercurrent. Increasing the flow of dialysate will increase the rate of removal of solutes. Frequent laboratory measurements of serum electrolytes, acid-base status, and whole blood activated clotting time should be performed when using intraoperative hemodialysis. If hemodialysis is not required, hemoconcentration alone can be accomplished by stopping pump #1 and using pump #2 to exert a slight negative pressure in the hemoconcentrator. The volume of fluid collected in excess of that removed by pump #1 from the dialysate bag represents plasma water removed from the patient's circulation. The volume of removed fluid or collected dialysate should be monitored continuously when using this technique. If both hemodialysis and hemoconcentration are not required, both pumps are turned off and blood is simply allowed to shunt through the hemoconcentrator to avoid stasis.

In summary, the leading cause of death in patients undergoing chronic renal dialysis is cardiovascular disease (291), and there are in excess of 180,000 patients currently on hemodialysis for end-stage renal disease in the United States (293). The risk of morbidity or mortality after coronary artery bypass surgery is only slightly increased in this patient population (294), and it is likely that more patients with renal failure will require open heart surgery in the future. Hemodialysis concurrent with CPB can be performed safely using equipment and techniques familiar to perfusionists. Attention to maintenance of adequate hematocrit, electrolyte and fluid balance, and control of hemodynamics during hemodialysis on CPB are important for successful patient outcome.

Acquired immunodeficiency syndrome

The AIDS epidemic is unlike previously known epidemics in that the disease may lie clinically silent for 10 years or more after the organism is introduced. In June 1979, a 32-year-old man in New York City, infected with HIV, appeared at first to be a medical curiosity, but the syndrome was rapidly recognized, and by June 1983, over 1,600 cases had been reported in the United States. Health care workers have always lived with the risks associated with contracting potentially fatal diseases; however, rarely in the course of history has a disease provoked such an emotional response, intense debate, and such a plethora of research efforts. The apparently uniform fatal outcome of clinical AIDS mandates introspection and close scrutiny of any invasive procedures performed by health care workers. This section briefly discusses the management of patients harboring an HIV infection who undergo cardiac surgery.

Homologous blood might best be considered a toxic substance. For cardiac surgical team members, exposure to blood is a daily risk. Gerberding et al. (295) studied the risks of exposure in 1,307 consecutive surgical procedures and reported that accidental exposure to blood occurred in 84 instances. The incidence of exposure increased when procedures lasted more than 3 hours and when blood loss exceeded 300 mL. Their data supported the practice of double gloving and the use of waterproof garments and face shields. Blood contact events were reported in 28% of 684 operations in a university medical center setting (296). The authors concluded that most contacts were preventable and reflected indifference to universal precautions that were part of their standard operating room policy.

The so-called universal precautions advised by the Centers for Disease Control and Prevention did not address the unique problems that occur in the operating room (297). Cardiothoracic cases (298) revealed a 58% contact (p < 0.001) with blood by personnel and multiple contacts in 12 cases. Although no case of operating room transmission of HIV had yet been documented, the authors concluded it had no doubt occurred and would certainly be documented in the future.

AIDS has clearly become a national obsession. The National Commission on AIDS concluded that mass screening programs would interfere with doctor–patient relationships and would encourage a false sense of security. This is because a 6- to 12-week window period normally occurs between infection with the HIV virus and seropositivity with traditional HIV screening tests. The medicolegal ramifications of testing patients present a significant problem, although, in part, this may be obviated by complying with local statutes for informed consent. No solution has been developed for the emergency patient who must rapidly undergo CPB when there is not sufficient time to obtain test results; thus implementation of universal precautions necessitate that all patients be treated as potentially infected.

The extent to which blood should be considered a toxic substance is realized when one studies the history of a nurse health care worker infected in Iowa. The worker was infected during a resuscitative effort on an HIV-positive patient when blood seeped from an intravenous line onto her ungloved left index finger, which had been previously cut during gardening. She became HIV-positive 3 months later. If a patient develops clinical AIDS after open heart surgery, he or she will likely blame the surgery or its personnel, when in fact a positive HIV test may have predated the patient's surgery.

The ELISA screening test, which may be completed in a few hours, has few false-negative results but a significant number of false positives. The confirmatory test is the Western blot, which has a low false-positive rate. Patients confirmed positive with this test should have adequate counseling, which is available in most hospital settings at this time.

The American Medical Association and the American College of Physicians have endorsed statements that doctors may not ethically refuse to treat a patient solely because the patient has AIDS or is seropositive for HIV infection. The subcommittee on AIDS of The American College of Surgeons Governors Committee does not currently recommend discontinuation of practice for HIV-positive surgeons (299). Patient-to-health care worker transmission has been reported and appears to be much more frequent than the reverse.

The Society of Thoracic Surgeons Committee on AIDS has made the following recommendations to thoracic surgeons. First, before elective surgery, make every attempt to identify patients who are positive for infection with the HIV by voluntary testing. In doing so, be careful to comply with any applicable confidentiality statutes and informed consent statutes. Second, use universal precautions against exposure to blood and other body fluids in the operating room and throughout the hospital, regardless of whether the status of the HIV serology is known. Third, take a leadership role to encourage easing the current restrictions on HIV testing so that we may learn about the risk of transmission of HIV, the role that positive HIV status may play in affecting the results of treatment of other disease processes, and the possibility that surgery and bypass may accelerate the development of AIDS in the HIV-positive patient. Finally, encourage hospital laboratories to develop the capability of providing rapid screening of patients for HIV.

In 1989 Condit and Frater (300) surveyed cardiothoracic surgeons' attitudes regarding whether AIDS patients should be operated on. Two thirds of respondents indicated they would operate on this patient population, but the decision would be based on medical indications. However, there was a concern expressed that CPB might accelerate the progression of the disease. With current treatment regimens, HIV patients are living much longer than they did more than a decade ago at the time of the survey. No association between CPB and progression to AIDS in the HIV-positive patient have been established (301–303), and HIV patients are now operated on with less concern than was expressed in the previous decade.

In summary, the AIDS epidemic has positively and negatively affected the use of CPB. Health care workers have become much more cautious since the identification of its etiology. The modes of HIV transmission have been established with a high degree of certainty, and in infected patients, viral presence has been established in virtually all body fluids (semen, vaginal secretions, blood, urine, saliva, cerebrospinal fluid, mother's milk, and almost certainly feces). All hospitals should have the drug azidothymidine immediately available for health care workers who experience an injury from an HIV-positive patient. There is laboratory evidence that azidothymidine given within 15 minutes of such injury can prevent HIV viral migration into the injured worker's cells (304).

A practical technique to minimize accidental needle punctures or cuts during cardiac surgery is to establish a policy whereby no surgical intruments or needles are handed off or passed directly between the scrub nurse and surgeon. Instead, a white towel on the sterile field can be used to delineate a space for placing and retrieving instruments by only one surgical team member at a time. Full-blown AIDS is still 100% fatal despite intense research efforts and newer drug regimens to control the disease. Health care professionals working in cardiac surgery have an obviously increased risk of exposure, and universal precautions to provide barriers cannot be overemphasized.

Malignant hyperthermia

See Related Case Study from Yao & Artusio's Anesthesiology

Malignant hyperthermia is a syndrome of acute hyperthermia (core temperatures may exceed 42°C) and/or myotonic reactions initiated by a hypermetabolic state of skeletal muscle. This syndrome can be triggered by administration of potent inhalation anesthetics (e.g., halothane, sevoflurane, and isoflurane) and succinylcholine. The syndrome was originally described in 1962 (305). The incidence is approximately 1 in 15,000 children and 1 in 50,000 adults. Early on, management of patients with malignant hyperthermia consisted of copious amounts of chilled intravenous Ringer's lactate solution, lavage of stomach and bladder with iced solutions, and large doses of procainamide or procaine (306). Currently, dantrolene sodium from 1 to 10 mg/kg is the treatment of choice, along with some of the active cooling measures listed above. Unlike previous pharmacologic measures (e.g., procainamide), dantrolene specifically addresses the causative mechanism, which is impaired reuptake of ionized calcium from the cytosol into storage sites located in the sarcoplasmic reticulum of skeletal muscle my ocytes. Before the introduction of dantrolene, mortality was greater than 60% (307). Before dantrolene was available, CPB had been used unsuccessfully and later successfully as a method of controlled cooling in the treatment of malignant hyperthermia (308).

Byrick et al. (309) described the anesthetic management of a patient with biopsy-proven malignant hyperthermia who underwent coronary artery bypass grafting. Measures included pretreatment with dantrolene, removal of the halothane vaporizer from the oxygenation-inspired gas pathway, and the use of high-dose fentanyl for anesthesia and pancuronium for muscle relaxation. Cold potassium (10 mEq/L) cardioplegia was used; inotropic agents and calcium (possible trigger agent) were avoided. The patient continued to receive dantrolene every 6 hours for 24 hours after the operation. No evidence of malignant hyperthermia was encountered. Other authors (310) used similar management strategies with success, the common theme of which is avoidance of trigger agents. Pretreatment with dantrolene is no longer recommended.

The diagnosis of malignant hyperthermia may be complicated or delayed by the coincidental use of CPB and may require an increased index of suspicion. Cases in which malignant hyperthermia occurred while on CPB have been reported (311,312). Both patients were successfully treated with a single dose of dantrolene (1 mg/kg). Early recognition and treatment of malignant hyperthermia is important because the marked hypermetabolism may exceed the oxygen delivery capacity of the oxygenator, especially at normothermia. During CPB, unexpected hyperthermia or temperature rises and unexplained metabolic and respiratory acidosis would provide the most likely clues to the diagnosis. When suspected, it appears sensible to induce or maintain CPB hypothermia while awaiting dantrolene-induced resolution of the clinical syndrome. Repeated doses of dantrolene may be required. Dantrolene is known to cause skeletal muscle weakness and causes myocardial depression in animals; caution is advised in its use in cardiac patients (309,310). Schwartz and Hensley (313) presented two cases and reviewed the physiologic basis and clinical diagnosis of this rare complication.

Miscellaneous pathophysiology

There is no specific contraindication for CPB in the elderly patient, and several reports demonstrated successful outcomes for coronary artery bypass surgery in octogenarians (314–316). Coronary bypass surgery has even been performed in a 100-year-old patient for unstable angina with favorable outcome for 3 years after surgery (C.D. Williams, personal communication, 1999). Others reported a low in cidence of cardiac-related mortality in patients after coronary bypass or valve replacement surgery (317). However, the risk of morbidity or mortality is increased in the elderly patient population (318).

Rady et al. (319) statistically determined perioperative predictors of morbidity and mortality in elderly patients (defined as age older than 75 years) having surgery with CPB. They reviewed the records of 1,157 patients who had cardiac surgery within a 30-month period at one hospital (14% of total caseload during the same time frame) and found that predictors of postoperative morbidity were: preoperative intraaortic balloon; preoperative serum bilirubin of more than 1.0 mg/dL; blood transfusion with more than 10 units of packed red blood cells; CPB time more than 120 minutes (aortic cross-clamp time greater than 80 minutes); return to operating room for surgical exploration; heart rate more than 120 beats/min; use of inotropes or vasopressors after CPB and upon admission to the intensive care unit; and anemia beyond postoperative day 2. Predictors of mortality (n = 90/1,157 or 8%) were similar and included: preoperative cardiogenic shock; serum albumin less than 4.0 g/dL; systemic oxygen delivery less than 320 mL/min/m2before surgery; blood transfusion more than 10 units; CPB time more than 140 minutes (aortic cross-clamp time more than 120 minutes; return to operating room for surgical exploration; mean arterial pressure less than 60 mm Hg; heart rate more than 120 beats/min; central venous pressure more than 15 mm Hg; stroke volume index less than 30 mL/min/m2; requirement for inotropes; arterial bicarbonate less than 20 mmol/L; plasma glucose more than 300 mg/dL after surgery; and anemia beyond postoperative day 2.

The decision to operate on the elderly patient should consider these findings. Maintenance of higher than normal CPB perfusion pressure (e.g., minimum 70 mm Hg), having a lower target threshold for administration of blood products, and minimizing CPB time may address several of the risk factors outlined above. In addition, the carotid arteries of all patients over 75 years of age should be checked preoperatively by noninvasive examination. For the cachectic patient, an immunologist can perform an anergy skin test battery (tetanus toxoid, diphtheria toxoid, steptococcus, old tuberculin, candida, trichophyton, and proteus), which can be used to screen those patients who would benefit from nutritional therapy before their surgery (320). Those patients determined to be anergic will have a high incidence of morbidity and mortality associated with open heart surgery.

The obese patient may be at a slightly increased risk of superficial wound complications or atrial dysrhythmias, but obesity per se is not a risk factor for adverse outcome after cardiac surgery (321). The standard criteria for CPB systemic flow indices based on body surface area may be safely lowered to 1.8 to 2.0 L/min/m2to avoid very high CPB flows (322), provided acceptable metabolic parameters are maintained. Occasionally, there may be difficulties in achiev ing adequate venous drainage in the obese patient (323); these problems can be overcome by use of assisted venous drainage methods (see Chapter 27). In the occasional morbidly obese patient, perhaps especially one who is also quite tall, the gas exchange capacity of the oxygenator may prove marginal at normothermia. If this problem develops, possible approaches include deepening anesthesia, cooling the patient, and using two oxygenators in parallel.

Klemperer et al. (324) showed that patients with noncardiac liver cirrhosis (Child class A) tolerate CPB and cardiac surgery with only minor morbidity. However, with moderate to advanced cirrhosis, morbidity and mortality are markedly high (80% mortality). All patients in their series (n = 13) had chest tube drainage and requirement for blood transfusion at a rate three times higher than patients without liver disease. Because the liver produces most coagulation factors but also clears activated factors and fibrinolytic components from the circulation (325), bleeding after CPB is the major concern in this patient population. Patients with liver disease also frequently have reduced platelet numbers and function (326) and increased fibrinolysis secondary to low-grade disseminated intravascular coagulopathy (327), which may be aggravated by the hematologic derangement inherent with CPB. Because of the high risk of morbidity in the patient presenting for cardiac surgery with chronic active hepatitis, a liver biopsy should be performed. If white cells are found, elective surgery should be postponed until the infection is controlled.

Patients with gastrointestinal disease may require excessive volume on CPB due to an increased tendency to third-space fluids. Buzzelli and Trittipoe (328) reported a case in which a patient with end-stage renal disease and an acutely perforated Meckel's diverticulum required over 4 L of fluid and blood administration plus 1,500 mL of crystalloid cardioplegic solution during 96 minutes of CPB. After bypass, the patient's abdomen was markedly distended, which required laparotomy and closure of the perforated bowel.


Back to Quick Links

An unanticipated perfusion mishap or the unusual or unappreciated patient condition can turn a routine CPB case into a challenge that will tax even the most experienced practitioners. Although statistically the current risk of patient serious injury or death directly related to CPB has been reduced to levels of a fraction of 1% and are certainly less than the risk of untreated heart disease in most cases, perfusion safety should continue to be a guiding principle whenever CPB is used (329). For the team that experiences an adverse patient outcome, reexamination of practices can yield rewards for future patients. A systematic approach to how CPB is performed may eliminate inherent potential errors that will occur no matter how diligent the team.

Newer CPB technology will undoubtedly continue to evolve. If device or equipment failure or fluids administered to the patient on CPB are suspected as contributory to an adverse patient outcome, they should be sequestered in a secure area for later examination. Like other endeavors that rely on complex tightly coupled technologies (330), CPB accidents more often result from operator error than from device failure. This chapter was written in the hope that awareness and communication of past failures will educate dedicated clinicians in avoiding their repetition.


Back to Quick Links

  • To reduce the incidence of stroke after CPB in the patients with severe atherosclerosis of the ascending aorta, transesophageal echocardiographic visualization of the aortic lumen, alternative cannulation sites and techniques, bilateral proximal carotid compression during surgical manipulation of the aorta, electrical fibrillation or -adrenergic blockers for myocardial protection, and no-touch techniques should be considered.

  • All patients undergoing hypothermic CPB should be screened for cold agglutinins.

  • Patients found to exhibit low titer (e.g., less than 1:32) low thermal amplitude for cold agglutinins and who are clinically asymptomatic can tolerate CPB with moderate hypothermia.

  • Patients with high thermal amplitude and/or titer (e.g., more than 1:128) with clinical symptoms of cold agglutinins require that the temperature be maintained above the thermal amplitude; additionally, warm crystalloid cardioplegia (followed by cold with insulation of the heart from adjacent structures) may prevent activation of cold agglutinins.

  • Sickle cell disease is a homozygous gene recessive abnormality characterized by a predominance of hemoglobin S; sickled cells have a limited capacity to load and unload oxygen, exhibit increased osmotic and mechanical fragility, increase blood viscosity, and can cause vascular occlusion.

  • Sickle cell trait is a heterozygous recessive abnormality in which hemoglobin S comprises 20% to 40% of the total hemoglobin.

  • Tendency for sickling occurs with hypoperfusion, hypoxemia, acidosis, increased concentrations of 2,3-diphosphoglyceric acid, infection, hypothermia, and capillary stagnation.

  • Exchange transfusion before or with the initiation of CPB should raise the hemoglobin A fraction more than 60% to decrease the risk of sickling, which will minimize hemolysis and prevent vaso-occlusive phenomenon.

  • Acute methemoglobinemia is most often induced by chemicals or drugs and is diagnosed when cyanosis or oxygen desaturation (chocolate-brown–colored blood) occurs despite an adequate arterial oxygen tension.

  • The most effective treatment of methemoglobinemia consists of methylene blue administration (1 to 2.5 mg/kg); additionally, the oxygenator should be ventilated with 100% oxygen and high CPB systemic flows should be used.

  • Patients with polycythemia may be hemodiluted to normal CPB levels by either removal of blood before bypass or using larger volumes of crystalloid solution in the CPB prime.

  • Jehovah's Witness patients may safely undergo CPB by minimizing circuit prime and fluid administration, use of cell salvage and hemoconcentration, and return of residual perfusate after bypass while maintaining continuity between removed blood and the patient to accommodate the patient's religious beliefs.

  • Reoperative patients may require alternative cannulation sites and augmented venous drainage techniques to adequately establish CPB.

  • Acute aortic dissection should be suspected if the CPB arterial line pressure unexpectedly increases and there is a simultaneous decrease in systemic pressure and/or venous drainage.

  • Venous or arterial air embolism may occur from improper operation of CPB, surgical technique, or through intravenous lines.

  • Systemic air embolism carries a greater risk than venous air embolism because of the potential for cerebral involvement.

  • Air embolism originating from CPB may enter the patient's systemic circulation from the arterial line or by other mechanisms such as improperly operated vents, cardioplegia delivery, or vacuum-assisted venous drainage or from pressurized CPB components.

  • Improperly de-aired intravenous lines or inappropriate ventilation of the patient during insertion of cannulas or vents or during surgical de-airing maneuvers can result in air embolism.

  • If massive air embolism from the CPB circuit occurs, bypass should be stopped immediately, the source of air determined, and efforts made to remove the air from the circuit and patient's vasculature.

  • Retrograde cerebral perfusion may be an effective treatment if significant air is suspected to have entered the patient's cerebral circulation.

  • The sterility of preassembled CPB circuits must be maintained by ensuring all blood-contacting surfaces and components are kept secure and covered until time of use.

  • CPB in the pregnant patient is associated with a low risk of maternal complications but a high risk to the fetus, particularly during the first and third trimesters when teratogenesis or premature initiation of uterine contractions may occur.

  • Maintenance of mean arterial pressures more than 60 mm Hg, CPB flow indices of 2.5 to 3.0 L/min/m2, normothermia or mild hypothermia, avoidance of potassium-based cardioplegia, use of FHR monitoring, and prompt treatment of uterine contractions with tocolytic agents can minimize complications in the mother and fetus.

  • Hemodialysis during CPB for patients in renal failure can be performed using a hemoconcentrator to lower elevated serum potassium, remove excess fluid, and delay reinstitution of maintenance hemodialysis until postoperative day 2 or 3.

  • CPB can be performed safely on HIV-positive patients by using universal precautions and modifiying the way surgical instruments are exchanged to prevent blood exposure and accidental needlesticks.

  • Malignant hyperthermia may be suspected during CPB in the patient who exhibits unexpected temperature rises and unexplained metabolic and respiratory acidosis; it is treated with dantrolene (1 to 2.5 mg/kg).


Back to Quick Links

    1. Roach GW, Kanchuger M, Mangano CM, et al. Adverse cerebral outcomes after coronary bypass surgery. N Engl J Med 1996;335:1857–1863.

    2. Mills NL, Everson CT. Atherosclerosis of the ascending aorta and coronary artery bypass: pathology, clinical correlates, and operative management. J Thorac Cardiovasc Surg 1991;102:546–553.

    3. Culliford AT, Colvin SB, Rohrer K, et al. The atherosclerotic ascending aorta and transverse arch: a new technique to prevent cerebral injury during bypass: experience with 13 patients. Updated in 1994. Ann Thorac Surg 1994;57:1051–1052.

    4. Olearchyk AS. Case report: calcified ascending aorta and coronary artery disease. Ann Thorac Surg 1994;59:1013–1015.

    5. Sabik JF, Lytle BW, McCarthy PM, et al. Axillary artery: an alternative site of arterial cannulation for patients with extensive aortic and peripheral vascular disease. J Thorac Cardiovasc Surg 1995;109:885–891.

    6. Groom RC, Hill AG, Kuban B, et al. Aortic cannula velocimetry. Perfusion 1995;10:183–188.

    7. Tanaka T, Kawamura T, Ohara K, et al. Transapical aortic perfusion with a double-barreled cannula. Ann Thorac Surg 1978;25:209–214.

    8. Liddicoat JR, Doty JR, Stuart RS. Case report: management of the atherosclerotic ascending aorta with endoaortic occlusion. Ann Thorac Surg 1998;65:1133–1135.

    9. Reichenspurner HH, Navia J, Berry G, et al. Particulate embolic capture by an intra-aortic filter device during cardiac surgery. J Thorac Cardiovasc Surg 2000;119:233–241.

    10. Kumar P, Dhital K, Hossein-Nia M, et al. S-100 protein release in a range of cardiothoracic surgical procedures. J Thorac Cardiovasc Surg 1997;113:953–954.

    11. Moore RA, Geller EA, Mathews ES, et al. The effect of hypothermic cardiopulmonary bypass on patients with low-titer, nonspecific cold agglutinins. Ann Thorac Surg 1984;37:233–238.

    12. Lee M-C, Chang C-H, Hsieh M-J. Use of a total wash-out method in an open-heart operation. Ann Thorac Surg 1989;47:57–58.

    13. Pruzanski W, Shumak KH. Biologic activity of cold-reacting autoantibodies. Part 2. N Engl J Med 1977;297:583–589.

    14. Dake SB, Johnston MFM, Brueggeman P, et al. Detection of cold hemagglutination in a blood cardioplegia unit before systemic cooling of a patient with unsuspected cold agglutinin disease. Ann Thorac Surg 1989;47:314–315.

    15. Bracken CA, Gurkowski MA, Naples JJ, et al. Case Conference: Case 6-1993, Cardiopulmonary bypass in two patients with previously undetected cold agglutinins. J Cardiothorac Vasc Anesth 1993;7:743–749.

    16. Fischer GD, Claypoole V, Collard CD. Increased line pressures in the retrograde blood cardioplegia line: an unusual presentation of cold agglutinins during cardiopulmonary bypass. Anesth Analg 1997;84:454–456.

    17. Leach AB, Van Hasselt GL, Edwards JC. Cold agglutinins and deep hypothermia. Anaesthesia 1983;38:140–143.

    18. Pruzanski W, Shumak KH. Biologic activity of cold-reacting autoantibodies. Part 1. N Engl J Med 1977;297:538–542.

    19. Shahian DM, Wallach SR, Bern MM. Open-heart surgery in patients with cold-reactive proteins. Surg Clin North Am 1985;65:315–322.

    20. AuBuchon JP, Stofan BA, Davey RJ. Hemolysis during extracorporeal circulation: significance of cold-reactive auto-antibodies and mechanical trauma [Abstract]. Blood 1983;62[Suppl 1]:42a.

    21. Schreiber AD, Herskovitz BS, Goldwein M. Low-titer cold-hemagglutinin disease: mechanism of hemolysis and response to corticosteroids. N Engl J Med 1977;296:1490–1494.

    22. Berreklouw E, Moulijin AC, Pegels JG, et al. Myocardial protection with cold cardioplegia in a patient with cold autoagglutinins and hemolysins. Ann Thorac Surg 1982;33:521–522.

    23. Klein HG, Faltz LL, McIntosh CL, et al. Surgical hypothermia in a patient with a cold agglutinin. Transfusion 1980;20:354–357.

    24. Wertlake PT, McGinniss MH, Schmidt PJ. Cold antibody and persistent intravascular hemolysis after surgery under hypothermia. Transfusion 1969;9:70–73.

    25. Williams AC. Cold agglutinins: cause for concern? Anaesthesia 1980;35:887–889.

    26. Diaz JH, Cooper ES, Ochsner JL. Cardiac surgery in patients with cold autoimmune diseases. Anesth Analg 1984;63:349–352.

    27. Blumberg N, Hicks G, Woll J, et al. Successful cardiac bypass surgery in the presence of a potent cold agglutinin without plasma exchange [Letter]. Transfusion 1983;23:363–364.

    28. Agarwal SK, Ghosh PK, Gupta D. Collective review: cardiac surgery and cold-reactive proteins. Ann Thorac Surg 1995;60:1143–1150.

    29. Landymore R, Isom W, Barlam B. Management of patients with cold agglutinins who require open-heart surgery. Can J Surg 1983;26:79–80.

    30. Holman WL, Smith SH, Edwards R, et al. Agglutination of blood cardioplegia by cold-reacting autoantibodies. Ann Thorac Surg 1991;51:833–836.

    31. Paccagnella A, Simini G, Nieri A, et al. Cardiopulmonary bypass and cold agglutinin [Letter]. J Thorac Cardiovasc Surg 1988;95:543.

    32. Blomback M, Kronlund P, Aberg B, et al. Pathologic fibrin formation and cold-induced clotting of membrane oxygenators during cardiopulmonary bypass. J Cardiothorac Vasc Anesth 1995;9:34–43.

    33. Dove SK, Raskin SA, Lawson DS. Cold agglutinins and hypothermic cardiopulmonary bypass. Perfusion 1992;7:3–6.

    34. Szentpetery S, Robertson L, Lower RR. Complete repair of tetralogy associated with sickle cell anemia and G-6-PD deficiency. J Thorac Cardiovasc Surg 1976;72:276–279.

    35. Hebbel RP. Beyond hemoglobin polymerization: the red blood cell membrane and sickle disease pathophysiology [Review]. Blood 1991;77:214–237.

    36. Ballas SK, Mahandas N. Pathophysiology of vaso-occlusion. Hematol Oncol Clin North Am 1996;10:1221–1239.

    37. Setty BNY, Stuart MJ. Vascular cell adhesion molecule-1 is involved in mediating hypoxia-induced sickle red blood cell adherence to endothelium: potential role in sickle cell disease. Blood 1996;88:2311–2320.

    38. Chun PKC, Flannery EP, Bowen TE. Open-heart surgery in patients with hematologic disorders. Am Heart J 1983;105:835–842.

    39. Fox MA, Abbott TR. Hypothermic cardiopulmonary bypass in a patient with sickle-cell trait. Anaesthesia 1984;39:1121–1123.

    40. Ingram CT, Floyd JB, Santora AH. Aortic and mitral valve replacement after sickle cell crisis [Letter]. Anesth Analg 1982;61:802–803.

    41. Harris LC, Haggard ME, Travis LB. The coexistence of sickle cell disease and congenital heart disease: a report of three cases with repair under cardiopulmonary by-pass in two patients. Pediatrics 1964;33:562–570.

    42. Yacoub MH, Baron J, Et-Etr A, et al. Aortic homograft replacement of the mitral valve in sickle cell trait. J Thorac Cardiovasc Surg 1970;59:568–573.

    43. Craenen J, Kilman J, Hosier DM, et al. Mitral valve replacement in a child with sickle cell anemia. J Thorac Cardiovasc Surg 1972;63:797–799.

    44. Heiner M, Teasdale SJ, David T, et al. Aortocoronary bypass in a patient with sickle cell trait. Can Anaesth Soc J 1979;26:428–434.

    45. Riethmuller R, Grundy EM, Radley-Smith R. Open heart surgery in a patient with homozygous sickle cell disease. Anaesthesia 1982;37:324–327.

    46. Baxter MRN, Bevan JC, Esseltine DW, et al. The management of two pediatric patients with sickle cell trait and sickle cell disease during cardiopulmonary bypass. J Cardiothorac Anesth 1989;3:477–480.

    47. Balasundaram MS, Duran CG, Al-Halees Z, et al. Cardiopulmonary bypass in sickle cell anaemia: report of five cases. J Cardiovasc Surg 1991;32:271–274.

    48. Pagani FD, Polito RJ, Bolling SF. Case report: mitral valve reconstruction in sickle cell disease. Ann Thorac Surg 1996;61:1841–1843.

    49. Kingsley CP, Chronister T, Cohen DJ, et al. Case conference. Case 2-1996, anesthetic management of a patient with hemoglobin SS disease and mitral insufficiency for mitral valve repair. J Cardiothorac Vasc Anesth 1996;10:419–424.

    50. Shulman G, McQuitty C, Vertrees RA, et al. Case report: acute normovolemic red cell exchange for cardiopulmonary bypass in sickle cell disease. Ann Thorac Surg 1998;65:1444–1446.

    51. Yung GL, Channick RN, Fedullo PF, et al. Successful pulmonary thromboendarterectomy in two patients with sickle cell disease. Am J Respir Crit Care Med 1998;157:1690–1693.

    52. Longenecker VW, Hartley MB, Dingmann M, et al. Cardiopulmonary bypass in the sickle cell anemia patient using profound hypothermia and circulatory arrest: a case report. J Extracorp Technol 1998;30:135–139.

    53. Moyes DG, Holloway AM, Hutton WS. Correction of Fallot's tetralogy in a patient suffering from hereditary spherocytosis. S Afr Med J 1974;48:1535–1536.

    54. Gayyed NL, Bouboulis N, Holden MP. Open heart operation in patients suffering from hereditary spherocytocis. Ann Thorac Surg 1993;55:1497–1500.

    55. Dal A, Kumar RS. Open heart surgery in presence of hereditary spherocytosis. J Cardiovasc Surg 1995;36:447–448.

    56. deLeval MR, Taswell HF, Bowie EJW, et al. Open heart surgery in patients with inherited hemoglobinopathies, red cell dyscrasias and coagulopathies. Arch Surg 1974;109:618–622.

    57. Mansouri A. Methemoglobinemia [Review]. Am J Med Sci 1985;289:200–209.

    58. Hedlund KD, Sanford DM, Coyne DP. Methemoglobinemia during cardiopulmonary bypass? A case report and review. Proc Am Acad Cardiovasc Perfus 1987;8:232–235.

    59. Robicsek F. Acute methemoglobinemia during cardiopulmonary bypass caused by intravenous nitroglycerin infusion. J Thorac Cardiovasc Surg 1985;90:931–933.

    60. Bojar RM, Rastegar H, Payne DD, et al. Methemoglobinemia from intravenous nitroglycerin: a word of caution. Ann Thorac Surg 1987;43:332–334.

    61. Zurick AM, Wagner RH, Starr NJ, et al. Intravenous nitroglycerin, methemoglobinemia, and respiratory distress in a postoperative cardiac surgical patient. Anesthesiology 1984;61:464–466.

    62. Jackson SH, Barker SJ. Case reports: methemoglobinemia in a patient receiving flutamide. Anesthesiology 1995;82:1065–1067.

    63. Cooper JR, Keats AS. Methemoglobinemia diagnosed as a consequence of cardiopulmonary bypass. Tex Heart Inst J 1985;12:103–106.

    64. Armstrong D, Schmalfuss V, Reed CC, et al. A case study: methemoglobinemia during cardiopulmonary bypass. Proc Am Acad Cardiovasc Perfus 1985;6:179–181.

    65. Kastrup EK. Antidotes. In: Kastrup EK , ed. Drug facts and comparisons . St. Louis: J.B. Lippincott, 1986:2012–2013.

    66. Milam JD, Austin SF, Nihill MR, et al. Use of sufficient hemodilution to prevent coagulopathies following surgical correction of cyanotic heart disease. J Thorac Cardiovasc Surg 1985;89:623–629.

    67. Ware JA, Reaves WH, Horak JK, et al. Defective platelet aggregation in patients undergoing surgical repair of cyanotic congenital heart disease. Ann Thorac Surg 1983;36:289–294.

    68. Dyke C, Sobel M. The management of coagulation problems in the surgical patient. Adv Surg 1991;24:229–257.

    69. Manley NJ, Williams DR, Pierce WS. Double valve replacement in a hemophiliac. AmSECT Proc 1976;4:46–50.

    70. Poullis M, Manning R, Haskard D, et al. Brief communication: Reopro removal during cardiopulmonary bypass using a hemoconcentrator. J Thorac Cardiovasc Surg 1999;117:1032–1034.

    71. Ridley PD, Ledingham SJM, Lennox SC, et al. Case report: protein C deficiency associated with massive cerebral thrombosis following open heart surgery. J Cardiovasc Surg 1990;31:249–251.

    72. Lawson DS, Darling EM, Ware RE, et al. Case report: management considerations for a heterozygous protein C deficient patient undergoing open heart surgery with cardiopulmonary bypass. J Extra-Corp Technol 1995;27:172–176.

    73. Petaja J, Peltola K, Sairanen H, et al. Fibrinolysis, antithrombin III, and protein C in neonates during cardiac operations. J Thorac Cardiovasc Surg 1996;112:665–671.

    74. Andrew M, Paes B, Johnston M. Development of the hemostatic system in the neonate and young infant. Am J Pediatr Hematol Oncol 1990;12:95–104.

    75. Kern FH, Morana NJ, Sears JJ, et al. Coagulation defects in neonates during cardiopulmonary bypass. Ann Thorac Surg 1992;54:541–546.

    76. Miller BE, Mochizuki T, Levy JH, et al. Predicting and treating coagulopathies after cardiopulmonary bypass in children. Anesth Analg 1997;85:1196–1202.

    77. Stein JI, Gombotz H, Rigler B, et al. Open heart surgery in children of Jehovah's Witness: extreme hemodilution on cardiopulmonary bypass. Pediatr Cardiol 1991;12:170–174.

    78. Cooley DA, Crawford ES, Howell JF, et al. Open heart surgery in Jehovah's Witnesses. Am J Cardiol 1964;13:779–781.

    79. Ott DA, Cooley DA. Cardiovascular surgery in Jehovah's Witnesses: report of 542 operations without blood transfusion. JAMA 1977;238:1256–1258.

    80. Tsang VT, Mullaly RJ, Ragg PG, et al. Bloodless open-heart surgery in infants and children. Perfusion 1994;9:257–263.

    81. Van Son JAM, Hovaguimian H, Rao IM, et al. Strategies for repair of congenital heart defects in infants without the use of blood. Ann Thorac Surg 1995;59:384–388.

    82. St. Rammos K , Bakas AJ, Panagopoulos FG. Mitral valve replacement in a Jehovah's Witness with dextrocardia and situs solitus. J Heart Valve Dis 1996;5:673–674.

    83. Malan TP Jr, Whitmore J, Maddi R. Reoperative cardiac surgery in a Jehovah's Witness: role of continuous cell salvage and in-line reinfusion. J Cardiothorac Anesth 1989;3:211–214.

    84. Aldea GS, Shapira OM, Treanor PR, et al. Effective use of heparin-bonded circuits and lower anticoagulation for coronary artery bypass grafting in Jehovah's Witnesses. J Card Surg 1996;11:12–17.

    85. Rosengart TK, Helm RE, Klemperer J, et al. Combined aprotinin and erythropoietin use for blood conservation: results with Jehovah's Witnesses. Ann Thorac Surg 1994;58:1397–1403.

    86. Chikada M, Furuse A, Kotsuka Y, et al. Open-heart surgery in Jehovah's Witness patients. Cardiovasc Surg 1996;4:311–314.

    87. Lee RB, Martin TD. Religious objections to blood transfusion. In: Mora CT , ed. Cardiopulmonary bypass, principles and techniques of extracorporeal circulation . New York: Springer-Verlag, 1995:473–480.

    88. Dobell ARC, Jain AK. Catastrophic hemorrhage during redo sternotomy. Ann Thorac Surg 1984;37:273–278.

    89. Bracey AW, Radovancevic R, Radovancevic B, et al. Blood use in patients undergoing repeat coronary artery bypass graft procedures: multivariate analysis. Transfusion 1995;35:850–854.

    90. Jones RE, Fitzgeral D, Cohn LH. Reoperative cardiac surgery using a new femoral venous right atrial cannula. J Card Surg 1990;5:170–173.

    91. Praeger PI, Pooley RW, Moggio RA, et al. Simplified method for reoperation on the mitral valve. Ann Thorac Surg 1989;48:835–837.

    92. Solomon L, Sutter FP, Goldman SM, et al. Augmented femoral venous return. Ann Thorac Surg 1993;55:1262–1263.

    93. Darling E, Kaemer D, Lawson S, et al. Experimental use of an ultra-low prime neonatal cardiopulmonary bypass circuit utilizing vacuum-assisted venous drainage. J Extra-Corp Technol 1998;30:184–189.

    94. Toomasian JM, McCarthy JP. Total extrathoracic cardiopulmonary support with kinetic assisted venous drainage: experience in 50 patients. Perfusion 1998;13:137–143.

    95. Kurusz M, Deyo DJ, Sholar AD, et al. Laboratory testing of femoral venous cannulae: effect of size, position and negative pressure on flow. Perfusion 1999;14:379–387.

    96. Pedersen TH, Videm V, Svennevig JL, et al. Extracorporeal membrane oxygenation using a centrifugal pump and a servo regulator to prevent negative inlet pressure. Ann Thorac Surg 1997;63:1333–1339.

    97. Reed CC, Kurusz M, Lawrence AE Jr. Effects of surgical technique on cardiopulmonary bypass. Safety and techniques in perfusion . Stafford, TX: Quali-Med, 1988:187–189.

    98. Antunes MJ. Techniques of valvular reoperation. Eur J Cardiothorac Surg 1992;6[Suppl I]:S54–S58.

    99. Watanabe G, Haverich A, Speier R. Third-time coronary artery revascularization. Thorac Cardiovasc Surg 1993;41:163–166.

    100. Toomasian JM, Peters WS, Siegal LC. Extracorporeal circulation for port-access cardiac surgery. Perfusion 1997;12:83–91.

    101. Berrizbeitia LD, Anderson WA, Laub GW, et al. Case report: coronary artery bypass grafting after pneumonectomy. Ann Thorac Surg 1994;58:1538–1540.

    102. Soltanian H, Sanders JH Jr, Robb JC, et al. Case report: hybrid myocardial revascularization after previous left pneumonectomy. Ann Thorac Surg 1998;65:259–260.

    103. Gu YJ, Obster R, Haan J, et al. Biocompatibility of leukocyte removal filters during leukocyte filtration of cardiopulmonary bypass perfusate. Artif Organ 1993;17:660–665.

    104. Soorae AS, Cleland J, O'Kane H. Delayed non-mycotic false aneurysm of ascending aortic cannulation site. Thorax 1977;32:743–748.

    105. Murphy DA, Craver JM, Jones EL, et al. Recognition and management of ascending aortic dissection complicating cardiac surgical operations. J Thorac Cardiovasc Surg 1983;85:247–256.

    106. Mills NL, Morris JM. Air embolism associated with cardiopulmonary bypass. In: Waldhausen JA , Orringer MB , eds. Complications in cardiothoracic surgery . St. Louis: Mosby, 1991:60–67.

    107. Coughlan DWS, Diehl JT, Rice TW, et al. Ascending aortic insertion of the intra-aortic balloon catheter: the potential for air embolism. Proc Am Acad Cardiovasc Perfus 1986;7:161–162.

    108. Kurusz M, Lick SD, Conti VR. Air embolism with intraaortic balloon counterpulsation during cardiopulmonary bypass [Letter]. J Thorac Cardiovasc Surg 1998;115:1393.

    109. Stoney WS, Alford WC Jr, Burrus GR, et al. Air embolism and other accidents using pump oxygenators. Ann Thorac Surg 1980;29:336–340.

    110. Kurusz M, Conti VR, Arens JF, et al. Perfusion accident survey. Proc Am Acad Cardiovasc Perfus 1986;7:57–65.

    111. Mejak BL, Stammers A, Rauch E, et al. A retrospective study on perfusion accidents and safety devices. Perfusion 2000;15:51–61.

    112. Wheeldon DR. Can cardiopulmonary bypass be a safe procedure? In: Longmore DB , ed. Towards safer cardiac surgery . Lancaster, UK: MTP, 1981:427–446.

    113. Jenkins OF, Morris R, Simpson JM. Australasian perfusion incident survey. Perfusion 1997;12:279–288.

    114. Mills NL, Ochsner JL. Massive air embolism during cardiopulmonary bypass-causes, prevention, and management. J Thorac Cardiovasc Surg 1980;80:708–717.

    115. Kurusz M, Wheeldon DR. Risk containment during cardiopulmonary bypass. Semin Thorac Cardiovasc Surg 1990;2:400–409.

    116. Fleisher AG, Tyers GFO, Manning GT, et al. Management of massive hemoptysis secondary to catheter-induced perforation of the pulmonary artery during cardiopulmonary bypass. Chest 1989;95:1340–1341.

    117. Thrush DN, Jeffries D. Pulmonary hemorrhage during cardiac surgery. J Cardiothorac Vasc Anesth 1991;5:377–378.

    118. Cohen JA, Blackshear RH, Gravenstein N, et al. Increased pulmonary artery perforating potential of pulmonary artery catheters during hypothermia. J Cardiothorac Vasc Anesth 1991;5:234–236.

    119. Kulkarni MG. A complication of aortic cannulation. J Cardiovasc Surg 1968;9:207–208.

    120. Krous HF, Mansfield PB, Sauvage LR. Carotid artery hyperperfusion during open-heart surgery. J Thorac Cardiovasc Surg 1973;66:118–120.

    121. Ross WT, Lake CL, Wellons HA. Cardiopulmonary bypass complicated by inadvertent carotid cannulation. Anesthesiology 1981;54:85–86.

    122. McLeskey CH, Cheney FW. A correctable complication of cardiopulmonary bypass. Anesthesiology 1982;56:214–216.

    123. Sudhaman DA. Accidental hyperperfusion of the left carotid artery during CPB [Letter]. J Cardiothorac Vasc Anesth 1991;5:100–101.

    124. Tempe D, Khanna SK. Accidental hyperperfusion during cardiopulmonary bypass: suggested safety features [Letter]. Ann Thorac Surg 1998;65:306.

    125. Springer MA, Korth DC, Guiterrez PJ, et al. An in vitro analysis of a one-way arterial check valve. J Extra-Corp Technol 1995;27:29–33.

    126. Gravlee GP, Wong AB, Charles DJ. Hypoxemia during cardiopulmonary bypass from leaks in the gas supply system [Letter]. Anesth Analg 1985;64:646–653.

    127. Roth JV, Fried DW. Catastrophic cardiopulmonary bypass accident [Letter]. J Cardiothorac Vasc Anesth 1993;7:253–254.

    128. Robblee JA, Crosby E, Keon WJ. Hypoxemia after intraluminal oxygen line obstruction during cardiopulmonary bypass. Ann Thorac Surg 1989;48:575–576.

    129. Kurusz M, Andrews JJ, Arens JF, et al. Monitoring oxygen concentration prevents potential adverse patient outcome caused by a scavenging malfunction: case report. Proc Am Acad Cardiovasc Perfus 1991;12:162–165.

    130. Kirson LE, Goldman JM. A system for monitoring the delivery of ventilating gas to the oxygenator during cardiopulmonary bypass. J Cardiothorac Vasc Anesth 1994;8:51–57.

    131. Jerabek CF, Walton HG, Doerfler S. The effect of gas scavenging on hollow fiber membrane oxygenator performance. Proceedings 27th International Conference American Society of Extracorpeal Technology , 10–14 March 1989:24–29.

    132. Nader-Djalal , Khadra WZ, Spaulding W, et al. Correspondence: Does propofol alter the gas exchange in membrane oxygenators? Ann Thorac Surg 1998;66:298–299.

    133. Tarr TJ, Kent AP. Sequestration of propofol in an extracorporeal circuit. J Cardiothorac Anesth 1989;3[Suppl 1]:75.

    134. Fisher AR. The incidence and cause of emergency oxygenator changeovers. Perfusion 1999;14:207–212.

    135. Hart MA, Abshier DA, Pacheco SL, et al. A technique for the change-out of a malfunctioning membrane oxygenator without terminating cardiopulmonary bypass. Proc Am Acad Cardiovasc Perfus 1988;9:105–113.

    136. McMillan J, Smith BF, Cooper E, et al. Case report: emergency changing of an oxygenator during total cardiopulmonary bypass. Anaesth Intens Care 1979;7:271–272.

    137. Reed CC, Stafford TB. Cardiopulmonary bypass , 2nd ed. Houston: Texas Medical Press, 1985:408.

    138. Kurusz M, Shaffer CW, Christman EW, et al. Runaway pump head: new cause of gas embolism during cardiopulmonary bypass. J Thorac Cardiovasc Surg 1979;77:792–795.

    139. Ballard S, Biggan DE. Runaway pumphead. In: Stafford TB , Toomasian JM , Kurusz M , eds. Case reports. I. Clinical studies in extracorporeal circulation. Houston: PREF Press, 1994:94–99.

    140. Kurusz M. Fire in the heart-lung machine during mitral valve replacement. In: Stafford TB , Toomasian JM , Kurusz M , eds. Case reports. I. Clinical studies in extracorporeal circulation . Houston: PREF Press, 1994:85–93.

    141. Khambatta HJ, Stone JG, Wald A, et al. Electrocardiographic artifacts during cardiopulmonary bypass. Anesth Analg 1990;71:88–91.

    142. Metz S. ECG artifacts during cardiopulmonary bypass and alternative method [Letter]. Anesth Analg 1991;72:715–716.

    143. Emergency Care Research Institute. ECG artifact in the OR. Health Dev 1991;20:140–141.

    144. Hug CC Jr. The anesthesiologist's response to a low-output state after cardiopulmonary bypass: etiologies and remedies. J Card Surg 1990;5:259–262.

    145. Kurusz M, Faulkner SC. Anecdotes and case reports from the perfusion accident survey. Proc Am Acad Cardiovasc Perfus 1987;8:261–264.

    146. Quinn ML. Semipractical alarms: a parable. J Clin Monit 1989;5:196–200.

    147. Beneken JEW, van der Aa JJ. Alarms and their limits in monitoring. J Clin Monit 1989;5:205–210.

    148. Cooper JB, Newbower RS, Kitz RJ. An analysis of major errors and equipment failures in anesthesia practice=mconsiderations for prevention and detection. Anesthesiology 1984;60:34–42.

    149. Gaba DM. Human error in anesthetic mishaps. Int Anesthesiol Clin 1989;27:137–147.

    150. Kurusz M, Butler BD, Katz J, Conti VR. Air embolism during cardiopulmonary bypass. Perfusion 1995;10:361–391.

    151. Carrel A. Experimental operations on the orifices of the heart. Ann Surg 1914;60:1–6.

    152. Reyer GW, Kohl HW. Air embolism complicating thoracic surgery. JAMA 1926;87:1626–1630.

    153. Kent EM, Blades B. Experimental observations upon certain intracranial complications of particular interest to the thoracic surgeon. J Thorac Cardiovasc Surg 1942;11:434–445.

    154. Geoghegan T, Lam CR. The mechanism of death from intracardiac air and its reversibility. Ann Surg 1953;139:351–359.

    155. Benjamin RB, Turbak CE, Lewis FJ. The effects of air embolism in the systemic circulation and its prevention during open cardiac surgery. J Thorac Cardiovasc Surg 1957;34:548–552.

    156. Clowes GHA Jr. Experimental procedures for entry into the left heart to expose the mitral valve. Ann Surg 1951;134:957–968.

    157. Potts WJ, Riker WL, de Bord R, et al. Maintenance of life by homologous lungs and mechanical circulation. Surgery 1952;31:161–166.

    158. Helmsworth JA, Clark LC, Kaplan S, et al. Artificial oxygenation and circulation during complete by-pass of the heart. J Thorac Cardiovasc Surg 1952;24:117–133.

    159. Lewis FJ, Taufic M. Closure of atrial septal defects with the aid of hypothermia. Surgery 1953;33:52–59.

    160. Miller BJ, Gibbon JH Jr, Greco VF, et al. The production and repair of interatrial septal defects under direct vision with the assistance of an extracorporeal pump-oxygenator circuit. J Thorac Surg 1953;26:598–616.

    161. Eguchi S, Bosher LH Jr. Myocardial dysfunction resulting from coronary air embolism. Surgery 1962;51:103–111.

    162. Goldfarb D, Bahnson HT. Early and late effects on the heart of small amounts of air in the coronary circulation. J Thorac Cardiovasc Surg 1963;46:368–378.

    163. Taber RE, Maraan BM, Tomatis L. Prevention of air embolism during open-heart surgery: a study of the role of trapped air in the left ventricle. Surgery 1970;68:685–691.

    164. Justice C, Leach J, Edwards WS. The harmful effects and treatment of coronary air embolism during open-heart surgery. Ann Thorac Surg 1972;14:47–53.

    165. Callaghan JC, Despres JP, Benvenuto R. A study of the causes of 60 deaths following total cardiopulmonary bypass. J Thorac Cardiovasc Surg 1961;42:489–496.

    166. Ehrenhaft JL, Claman MA, Layton JM, et al. Cerebral complications of open-heart surgery: further observations. J Thorac Cardiovasc Surg 1961;42:514–526.

    167. Allen P. Central nervous system emboli in open heart surgery. Can J Surg 1963;6:332–337.

    168. Sloan H, Morris JD, Mackenzie J, et al. Open heart surgery: results in 600 cases. Thorax 1962;17:128–138.

    169. Nicks R. Arterial air embolism. Thorax 1967;22:320–326.

    170. Fishman NH, Carlsson E, Roe BB. The importance of the pulmonary veins in systemic air embolism following open-heart surgery. Surgery 1969;66:655–662.

    171. Anderson RM, Fritz JM, O'Hae JE.Pulmonary air emboli during cardiac surgery. J Thorac Cardiovasc Surg 1965;49:440–449.

    172. Lin C-Y. Pulmonary air embolism: reappraisal of the importance in open-heart surgery. Nagoya J Med Sci 1967;30:365–372.

    173. Gomes OM, Pereira SN, Castagna RC, et al. The importance of the different sites of air injection in the tolerance of arterial air embolism. J Thorac Cardiovasc Surg 1973;65:563–568.

    174. Beckman CR, Hurley F, Mammana R, et al. Risk factors for air embolization during cannulation of the ascending aorta. J Thorac Cardiovasc Surg 1980;80:302–307.

    175. Sturm J. Air embolism during mitral valve replacement [Letter]. J Thorac Cardiovasc Surg 1981;81:804–805.

    176. Hughes D. Air embolism during cardiopulmonary bypass [Letter]. J Thorac Cardiovasc Surg 1981;82:639.

    177. Robicsek F, Duncan GD. Retrograde air embolization in coronary operations. J Thorac Cardiovasc Surg 1987;94:110–114.

    178. Lee ME. Air embolism in coronary bypass operations [Letter]. J Thorac Cardiovasc Surg 1988;95:543.

    179. Reed CC, Kurusz M, Lawrence AE Jr. Air embolism. Safety and techniques in perfusion . Stafford, TX: Quali-Med, 1988:239–246.

    180. Dennis C. Perspective in review: one group's struggle with development of a pump-oxygenator. Trans Am Soc Artif Intern Organ 1985;31:1–11.

    181. Milnes RF, vanderWoude R, Morris JD, et al. Problems related to a bubble oxygenator system. Surgery 1958;42:986–992.

    182. Baird RJ, Miyagishima RT. The danger of air embolism through a pressure-perfusion cannula. J Thorac Cardiovasc Surg 1963;46:212–219.

    183. Kumar AS, Jayalakshmi TS, Kale SC, et al. Management of massive air embolism during open heart surgery. Int J Cardiol 1985;9:413–416.

    184. Tanasawa I, Wotton DR, Yang W-J, et al. Experimental study of air bubbles in a simulated cardiopulmonary bypass system with flow constriction. J Biomech 1970;3:417–424.

    185. Comer TP, Saxena NC, Hamel NC. Full recovery of a patient after oxygenator replacement during open-heart surgery. Cardiovasc Dis Bull Tex Heart Inst 1980;7:165–168.

    186. Pfefferkorn RO, Rose MW, Pfefferkorn SP. An unreported pathway for the introduction of air embolism from and unvented hardshell cardiotomy reservoir: a case report. Proceedings First World Congress on Open Heart Technology, Brighton UK 1981 . London: Franklin Scientific Projects, 1982:78–79.

    187. Wells WJ, Stiles QR. Massive venous air embolism during cardiopulmonary bypass. Ann Thorac Surg 1981;31:86–89.

    188. Pickard LR. Venous air embolism [Letter]. Ann Thorac Surg 1982;33:102–103.

    189. Emergency Care Research Institute. Hazard: scavenging gas from membrane oxygenators. Health Dev 1987;16:343–344.

    190. Goodridge M. Entrance of air into the veins, and its treatment. Am J Med Sci 1902;124:461–476.

    191. Hare HA. The effect of the entrance of air into the circulation. Therap Gazette 1889;5:606–610.

    192. Tisovec L, Hamilton WK. Newer considerations in air embolism during operation. JAMA 1967;201:376–377.

    193. Munson ES. Transfer of nitrous oxide into body air cavities. Br J Anaesth 1974;46:202–209.

    194. Garcia C, Albin MS, Bunegin L. Effect of nitrous oxide on air bubble volume in arterial air embolism [Letter]. Crit Care Med 1989;17:1236.

    195. Losasso TJ, Black S, Muzzi DA, et al. Detection and hemodynamic consequences of venous air embolism. Does nitrous oxide make a difference? Anesthesiology 1992;77:148–152.

    196. Munson ES. Pathophysiology and treatment of venous air embolism: a review. Middle East J Anesthesiol 1988;9:315–325.

    197. Wells DG, Podolakin W, Mohr M, et al. Nitrous oxide and cerebrospinal fluid markers of ischemia following cardiopulmonary bypass. Anaesth Intens Care 1987;15:431–435.

    198. Peirce EC II. Cerebral gas embolism (arterial) with special reference to iatrogenic accidents. HBO Rev 1980;1:161–188.

    199. Dutka AJ. A review of the pathophysiology and potential application of experimental therapies for cerebral ischemia to the treatment of cerebral arterial gas embolism. Undersea Biomed Res 1985;12:403–421.

    200. Spampinato N, Stassano P, Gagliardi C, et al. Massive air embolism during cardiopulmonary bypass: successful treatment with immediate hypothermia and circulatory support. Ann Thorac Surg 1981;32:602–603.

    201. Steward D, Williams WG, Freedom R. Hypothermia in conjunction with hyperbaric oxygenation in the treatment of massive air embolism during cardiopulmonary bypass. Ann Thorac Surg 1977;24:591–593.

    202. Fundaro P, Santoil C. Massive coronary gas embolism managed by retrograde coronary sinus perfusion. Tex Heart Inst J 1984;11:172–174.

    203. Sandhu AA, Spotnitz HM, Dickstein ML, et al. Retrograde cardioplegia preserves myocardial function after induced coronary air embolism. J Thorac Cardiovasc Surg 1997;113:917–922.

    204. Bauer RO, Campbell M, Goodman R, et al. Aeroembolism treated by hypothermia: report of a case. Aerospace Med 1965;36:671–675.

    205. Fine J, Fischmann J. An experimental study of the treatment of air embolism. N Engl J Med 1940;223:1054–1057.

    206. Alvaran SB, Toung JK, Graff TE, et al. Venous air embolism: comparative merits of external cardiac massage, intracardiac aspiration, and left lateral decubitus position. Anesth Analg 1978;57:166–170.

    207. Musgrove JE, MacQuigg RE. Successful treatment of air embolism. JAMA 1952;150:28.

    208. Ericsson JA, Gottlieb JD, Sweet RB. Closed-chest cardiac massage in the treatment of venous air embolism. N Engl J Med 1964;270:1353–1354.

    209. Brenner WI. A battle plan in the event of massive air embolism during open heart surgery. J Extra-Corp Technol 1985;17:133–137.

    210. Butler BD, Laine GA, Leiman BC, et al. Effect of Trendelenburg position on the distribution of arterial air emboli in dogs. Ann Thorac Surg 1988;45:198–202.

    211. Mehlhorn U, Burke EJ, Butler BD, et al. Body position does not affect the hemodynamic response to venous air embolism in dogs. Anesth Analg 1994;79:734–739.

    212. Watanabe T, Shimasaki T, Kuraoka S, et al. Retrograde cerebral perfusion against massive air embolism during cardiopulmonary bypass [Letter]. J Thorac Cardiovasc Surg 1992;104;532–533.

    213. Brown JW, Dierdorf SF, Moorthy SS, et al. Venoarterial cerebral perfusion for treatment of massive arterial air embolism. Anesth Analg 1987;66:673–674.

    214. Diethrich EB, Koopot R, Maze A, et al. Successful reversal of brain damage from iatrogenic air embolism. Surg Gynecol Obstet 1982;154:572–575.

    215. Ghosh PK, Kaplan O, Barak J, et al. Massive arterial air embolism during cardiopulmonary bypass. J Cardiovasc Surg 1985;26:248–250.

    216. Hendricks FFA, Bogers AJJC, de la Riviere AB, et al. The effectiveness of venoarterial perfusion in treatment of arterial air embolism during cardiopulmonary bypass. Ann Thorac Surg 1983;36:433–436.

    217. Noritake S, Kitayama H, Matsuno S, et al. Massive air embolism during cardiopulmonary bypass: a case report of successful management by temporary retrograde perfusion through the superior vena cava. Kyobu Geka 1985;38:274–277.

    218. Stark J, Hough J. Air in the aorta: treatment by reversed perfusion. Ann Thorac Surg 1986;41:337–338.

    219. Toscano M, Chiavarelli R, Ruvolo G, et al. Management of massive air embolism during open-heart surgery with retrograde perfusion of the cerebral vessels and hyperbaric oxygenation. Thorac Cardiovasc Surg 1983;31:183–184.

    220. Rozanski J, Szufladowicz M. Successful treatment of massive air embolism. J Card Surg 1994;9:430–432.

    221. Goldstone J, Towan HJ, Ellis RJ. Rationale for use of vasopressors in treatment of coronary air embolism. Surg Forum 1978;29:237–239.

    222. Evans DE, Kobrine AI, LeGrys DC, et al. Protective effect of lidocaine in acute cerebral ischemia induced by air embolism. J Neurosurg 1984;60:257–263.

    223. Padula RT, Eisenstat TE, Bronstein MH, et al. Intracardiac air following cardiotomy: location, causative factors, and a method for removal. J Thorac Cardiovasc Surg 1971;62:736–742.

    224. Eiseman B, Baxter BJ, Prachuabmoh K. Surface tension reducing substances in the management of coronary air embolism. Ann Surg 1959;149:374–380.

    225. Malette WG, Fitzgerald JB, Eiseman B. Aeroembolus: a protective substance. Surg Forum 1960;11:155–156.

    226. Menasche P, Pinard E, Desroches A-M, et al. Fluorocarbons: a potential treatment of cerebral air embolism in open-heart surgery. Ann Thorac Surg 1985;40:494–497.

    227. Menasche P, Fleury J-P, Piwnica A. Update: fluorocarbons: a potential treatment of cerebral air embolism in open-heart surgery. Ann Thorac Surg 1992;54:392–393.

    228. Butler BD, Kurusz M. Fluorocarbon treatment for cerebral air embolism [Letter]. Ann Thorac Surg 1986;42:350–351.

    229. Speiss BD, McCarthy RJ, Tuman KJ, et al. Perfluorocarbon emulsions and air embolism [Letter]. Ann Thorac Surg 1987;44:223.

    230. Speiss BD, McCarthy RJ, Ivankovich AD. Protection from coronary air embolism by a perfluorocarbon emulsion (FC-43). J Cardiothorac Anesth 1987;1:210–215.

    231. Speiss BD, Braverman B, Woronowicz AW, et al. Protection from cerebral air emboli with perfluorocarbons in rabbits. Stroke 1986;17:1146–1149.

    232. Annane D, Troche G, Delisle F, et al. Effects of mechanical ventilation with normobaric oxygen therapy on the rate of air removal from cerebral arteries. Crit Care Med 1994;22:851–857.

    233. Presson RG Jr, Kirk KR, Haselby KA, et al. Effect of ventilation with soluble and diffusible gases on the size of air emboli. J Appl Physiol 1991;70:1068–1074.

    234. Tuman KJ, McCarthy RJ, Speiss BD, et al. Effects of nitrous oxide on coronary perfusion after coronary air embolism. Anesthesiology 1987;67:952–959.

    235. Hlastala MP, Van Liew HD. Absorption of in vivo inert gas bubbles. Respir Physiol 1975;24:147–158.

    236. Schabel RK, Berryessa RG, Justison GA, et al. Ten common perfusion problems: prevention and treatment protocols. J Extra-Corp Technol 1987;19:392–398.

    237. Tovar EA, Del Campo C, Bosari A, et al. Postoperative management of cerebral air embolism: gas physiology for surgeons. Ann Thorac Surg 1995;60:1138–1142.

    238. Armon C, Deschamps C, Adkinison C, et al. Hyperbaric treatment of cerebral air embolism sustained during an open-heart surgical procedure. Mayo Clin Proc 1991;66:565–570.

    239. Bove AA, Clark JM, Simon AJ, et al. Successful therapy of cerebral air embolism with hyperbaric oxygen at 2.8 ATA. Undersea Biomed Res 1982;9:75–79.

    240. Halliday P, Anderson DN, Davidson AI, et al. Management of cerebral air embolism secondary to a disconnected central venous catheter. Br J Surg 1994;81:71.

    241. Hart GB. Treatment of decompression illness and air embolism with hyperbaric oxygen. Aerospace Med 1974;45:1190–1193.

    242. Kindwall EP. Massive surgical air embolism treated with brief recompression to six atmospheres followed by hyperbaric oxygen. Aerospace Med 1973;44:663–666.

    243. Lar LW, Lai LC, Ren LW. Massive arterial air embolism during cardiac operation: successful treatment in hyperbaric chamber under 3 ATA [Letter]. J Thorac Cardiovasc Surg 1990;100:928–930.

    244. Meijne NG, Schoemaker G, Bulterijs AB. The treatment of cerebral gas embolism in a high pressure chamber: and experimental study. J Cardiovasc Surg 1963;4:757–763.

    245. Menkin M, Schwartzman RJ. Cerebral air embolism: report of five cases and review of the literature. Arch Neurol 1977;34:168–170.

    246. Murphy BP, Harford FJ, Cramer FS. Cerebral air embolism resulting form invasive medical procedures: treatment with hyperbaric oxygen. Ann Surg 1985;201:242–245.

    247. Peirce EC II. Specific therapy for arterial air embolism [Editorial]. Ann Thorac Surg 1980;29:300–303.

    248. Takita H, Olszewski W, Schimert G, et al. Hyperbaric treatment of cerebral air embolism as a result of open-heart surgery: report of a case. J Thorac Cardiovasc Surg 1968;55:682–85.

    249. Mader JT, Hulet WH. Delayed hyperbaric treatment of cerebral air embolism: report of a case. Arch Neurol 1979;36:504–505.

    250. Bitterman H, Melamed Y. Delayed hyperbaric treatment of cerebral air embolism. Israel J Med Sci 1993;29:22–26.

    251. Ryan T, McCarthy JF, Rady MY, et al. Early bloodstream infection after cardiopulmonary bypass: frequency rate, risk factors, and implications. Crit Care Med 1997;25:2009–2014.

    252. Rudnick JR, Beck-Sague CM, Anderson RL, et al. Gram-negative bacteremia in open-heart-surgery patients traced to probable tap-water contamination of pressure-monitoring equipment. Infect Control Hosp Epidemiol 1996;17:281–285.

    253. Hughes CF, Grant AF, Leckie BD, et al. Cardioplegic solution: a contamination crisis. J Thorac Cardiovasc Surg 1986;91:296–302.

    254. Talbot GH. Contamination of cardioplegic solution [Letter]. J Thorac Cardiovasc Surg 1986;92:966.

    255. Metz S. Administration of protamine rather than heparin in a patient undergoing normothermic cardiopulmonary bypass. Anesthesiology 1994;80:691–694.

    256. Ti LK. Survival after large ABO-incompatible blood transfusion. Can J Anaesth 1998;45:916.

    257. Becker RM. Intracardiac surgery in pregnant women. Ann Thorac Surg 1983;36:453–458.

    258. Leyse R, Ofstun M, Dillard DH, et al. Congenital aortic stenosis in pregnancy, corrected by extracorporeal circulation: offering a viable male infant at term but with anomalies eventuating in his death at four months of age-report of a case. JAMA 1961;176:1009–1012.

    259. Dubourg G, Broustet P, Bricaud H, et al. Correction complete d'une triade de Fallot en circulation extra-corporelle chez une femme enceinte. Arch Mal Coeur 1959;52:1389–1391.

    260. Kay CF, Smith K. Surgery in the pregnant cardiac patient. Am J Cardiol 1963;12:293–295.

    261. Werch A, Lambert HM, Cooley D, et al. Fetal monitoring and maternal open heart surgery [Letter]. South Med J 1977;70:1024.

    262. Meffert WG, Stansel HC Jr. Open heart surgery during pregnancy. Am J Obstet Gynecol 1968;102:1116–1120.

    263. Pedersen H, Finster M. Anesthetic risk in the pregnant surgical patient. Anesthesiology 1979;51:439–451.

    264. Parry AJ, Westaby S. Cardiopulmonary bypass during pregnancy. Ann Thorac Surg 1996;61:1865–1869.

    265. Pomini F, Mercogliano D, Cavalletti C, et al. Cardiopulmonary bypass in pregnancy. Ann Thorac Surg 1996;61:259–268.

    266. Weiss BM, von Segesser LK, Seifert B, et al. Outcome of cardiovascular surgery and pregnancy: a systematic review of the period 1984–1996. Am J Obstet Gynecol 1998;179:1643–1653.

    267. Izquierdo LA, Kushnir O, Knieriem K, et al. Effect of mitral valve prosthetic surgery on the outcome of a growth-retarded fetus. A case report. Am J Obstet Gynecol 1990;163:584–586.

    268. Caritis SN, Edelstone DI, Mueller-Heubach E. Pharmacologic inhibition of premature labor. Am J Obstet Gynecol 1979;133:557–578.

    269. Lamb MP, Ross K, Johnstone AM, et al. Fetal heart rate monitoring during open heart surgery. Br J Obstet Gynecol 1983;88:669–674.

    270. Korsten HHM, Van Zundert AAJ, Mooij PNM, et al. Emergency aortic valve replacement in the 24th week of pregnancy. Acta Anaesth Belg 1989;40:201–205.

    271. Strickland RA, Oliver WC Jr, Chantigian RC, et al. Anesthesia, cardiopulmonary bypass, and the pregnant patient. Mayo Clin Proc 1991;66:411–429.

    272. Chambers CE, Clark SL. Cardiac surgery during pregnancy. Clin Obstet Gynecol 1994;37:316–323.

    273. Conroy JM, Bailey MK, Hollon MF, et al. Anesthesia for open heart surgery in the pregnant patient. South Med J 1989;82:492–495.

    274. Harrison EC, Roschke EJ. Pregnancy in patients with cardiac valve prostheses. Clin Obstet Gynecol 1975;18:107–123.

    275. Koh KS, Friesen RM, Livingstone RA, et al. Fetal monitoring during maternal cardiac surgery with cardiopulmonary bypass. Can Med Assoc J 1975;112:1102–1104.

    276. Reisner LS. Cardiac dysfunction: special considerations during pregnancy. In: Utley JR , ed. Pathophysiology and techniques of cardiopulmonary bypass . Vol. III. Baltimore: Williams & Wilkins, 1985:15–29.

    277. Buffolo E, Palma JH, Gomes WJ, et al. Case report: successful use of deep hypothermic circulatory arrest in pregnancy. Ann Thorac Surg 1994;58:1532–1534.

    278. Daley R, Harrison GK, McMillan IKR. Direct-vision pulmonary valvotomy during pregnancy. Lancet 1957;273:875–876.

    279. Ravindran R, Viegas OJ, Padilla LM, et al. Anesthetic considerations in pregnant patients receiving terbutylline treatment. Anesth Analg 1980;59:391–392.

    280. Rolbin SH, Levinson G, Shnider SM, et al. Dopamine treatment of spinal hypotension decreases uterine blood flow in the pregnant ewe. Anesthesiology 1979;51:37–40.

    281. Naulty JS, Cefalo RC, Lewis P. Fetal toxicity of nitroprusside in the pregnant ewe. Am J Obstet Gynecol 1980;139:708–711.

    282. Kopman EA. Scavenging of potassium cardioplegic solution to prevent hyperkalemia in hemodialysis-dependent patients. Anesth Analg 1983;62:780–782.

    283. Soffer O, MacDonell C Jr, Finlayson DC, et al. Intraoperative hemodialysis during cardiopulmonary bypass in chronic renal failure. J Thorac Cardiovasc Surg 1979;77:789–791.

    284. Geronemus R, Schneider N. Continuous arteriovenous hemodialysis: a new modality for treatment of acute renal failure. Trans Am Soc Artif Intern Organs 1984;30:610–613.

    285. Hakim M, Wheeldon D, Bethune DW, et al. Haemodialysis and haemofiltration on cardiopulmonary bypass. Thorax 1985;40:101–106.

    286. Murkin JM, Murphy DA, Finlayson DC, et al. Hemodialysis during cardiopulmonary bypass: report of twelve cases. Anesth Analg 1987;66:899–901.

    287. Wheeldon D, Bethune D. Haemofiltration during cardiopulmonary bypass. Perfusion 1990;5[Suppl]:39–51.

    288. Sutton RG. Renal considerations, dialysis, and ultrafiltration during cardiopulmonary bypass. Int Anesthesiol Clin 1996;34:165–176.

    289. Ilson BE, Bland PS, Jorkasky DK, et al. Intraoperative versus routine hemodialysis in end-stage renal disease patients undergoing open-heart surgery. Nephron 1992;61:170–175.

    290. Hamilton CC, Harwood SJ, Deemar KA, et al. Haemodialysis during cardiopulmonary bypass using a haemofilter. Perfusion 1994;9:135–139.

    291. Koyangi T, Nishida , Endo M, et al. Coronary artery bypass grafting in chronic renal dialysis patients: intensive perioperative dialysis and extensive usage of arterial grafts. Eur J Cardiothorac Surg 1994;8:505–507.

    292. Garrido P, Bobadilla JF, Albertos J, et al. Cardiac surgery in patients under chronic hemodialysis. Eur J Cardiothorac Surg 1995;9:36–39.

    293. United States Health Care Financing Administration. End stage renal disease program highlights [Fact sheet] . August 1997.

    294. Marshall WG Jr, Rossi NP, Meng RL, et al. Coronary artery bypass grafting in dialysis patients. Ann Thorac Surg 1986;42[Suppl]:S12–S15.

    295. Gerberding JL, Littell C, Tarkington A, et al. Risks of exposure of surgical personnel to patient's blood during surgery at San Francisco General Hospital. N Engl J Med 1990;322:1788–1793.

    296. Popejoy SL, Fry DE. Blood contact and exposure in the operating room. Surg Gynecol Obstet 1991;172:480–483.

    297. Centers for Disease Control. Recommendation for preventing transmission of infection with human T-lymphotropic virus type III/lymphadenopathy-associated virus in the workplace. MMWR Morb Mortal Wkly Rep 1985;35:681–686, 691–695.

    298. Centers for Disease Control. Recommendation for preventing transmission of infection with human T-lymphotropic virus Type III/lymphadenopath-associated virus in the workplace. MMWR Morb Mortal Wkly Rep 1986;35:21–23.

    299. Subcommittee on AIDS of The American College of Surgeons Governors Committee. 1991 Clinical Congress highlights: statement on the surgeon and HIV infection. ACS Bull 1991;76:28–31.

    300. Condit D, Frater RWM. Human immunodeficiency virus and the cardiac surgeon: a survey of attitudes. Ann Thorac Surg 1989;47:182–186.

    301. Lemma M, Vanelli P, Beretta L, et al. Cardiac surgery in HIV-positive intravenous drug addicts: influence of cardiopulmonary bypass on the progression to AIDS. Thorac Cardiovasc Surg 1992;40:279–282.

    302. Aris A, Pomar JL, Saura E. Cardiopulmonary bypass in HIV-positive patients. Ann Thorac Surg 1993;55:1104–1108.

    303. Uva MS, Jebara VA, Fabiani JN, et al. Cardiac surgery in patients with human immunodeficiency virus infection: indications and results. J Card Surg 1992;7:240–244.

    304. Robicsek F, Duncan GD, Masters TN, et al. Can AIDS be prevented after injury with contaminated instruments? Ann Thorac Surg 1990;49:984–986.

    305. Denborough MA, Foster VFA, Lovell RRH, et al. Anaesthetic deaths in a family. Br J Anaesth 1962;34:395–396.

    306. Ryan JF, Kerr WS Jr. Malignant hyperthermia: a catastrophic complication. J Urol 1973;109:879–883.

    307. Britt BA, Kalow W. Malignant hyperthermia, a statistical review. Can Anaesth Soc J 1970;17:293–301.

    308. Ryan JF, Donlon JV, Malt RA, et al. Cardiopulmonary bypass in the treatment of malignant hyperthermia. N Engl J Med 1974;290:1121–1122.

    309. Byrick RJ, Rose DK, Ranganathan N. Management of a malignant hyperthermia patient during cardiopulmonary bypass. Can Anaesth Soc J 1982;29:50–54.

    310. Bahret PM, Larach DR, Williams DR, et al. Malignant hyperthermia: a case report. Proc Am Acad Cardiovasc Perfus 1986;7:177–180.

    311. MacGillivray RG, Jann H, Vanker E, et al. Development of malignant hyperthermia obscured by cardiopulmonary bypass. Can Anaesth Soc J 1986;33:509–514.

    312. Quinn RD, Pae WE Jr, McGary SA, et al. Development of malignant hyperthermia during mitral valve replacement. Ann Thorac Surg 1992;53:1114–1116.

    313. Schwartz AJ, Hensley FA Jr. Case conference 3, 1990. J Cardiothorac Anesth 1990;4:385–399.

    314. Utley JR, Leyland SA. Coronary artery bypass grafting in the octogenarian. J Thorac Cardiovasc Surg 1991;101:866–870.

    315. Tsai T-P, Nessim S, Kass RM, et al. Morbidity and mortality after coronary artery bypass in octogenarians. Ann Thorac Surg 1991;51:983–986.

    316. Ko W, Kreiger KH, Lazenby D, et al. Isolated coronary artery bypass grafting in one hundred consecutive octogenarian patients. J Thorac Cardiovasc Surg 1991;102:532–538.

    317. Ranger WR, Glover JL, Shannon FL, et al. Coronary artery bypass and valve replacement in octogenarians. Am Surg 1996;62:941–946.

    318. Cane ME, Chen C, Bailery BM, et al. CABG in octogenarians: early and late events and actuarial survival in comparison with a matched population. Ann Thorac Surg 1995;60:1033–1037.

    319. Rady MY, Ryan T, Starr NJ. Perioperative determinants of morbidity and mortality in elderly patients undergoing cardiac surgery. Crit Care Med 1998;26:225–235.

    320. Johnson WC, Lurich F, Meguid MM, et al. Role of delayed hypersensitivity in predicting postoperative morbidity and mortality. Am J Surg 1979;137:536–542.

    321. Moulton MJ, Creswell LL, Mackey ME, et al. Obesity is not a risk factor for significant adverse outcome after cardiac surgery. Circulation 1996;94[Suppl II]:II-87–II-92.

    322. Kirklin JW, Barratt-Boyes BG. Cardiac surgery , 2nd ed. New York: Churchill Livingstone, 1993:80.

    323. Babka RM, Caviness ML, Howell J, et al. The effects of obesity on cardiopulmonary bypass using the Bentley BOS-10 oxygenator. Proc Am Acad Cardiovasc Perfus 1982;3:60–63.

    324. Klemperer JD, Ko W, Kreiger K, et al. Cardiac operations in patients with cirrhosis. Ann Thorac Surg 1998;65:85–87.

    325. Humphries JE. Transfusion therapy in acquired coagulopathies. Hematol Oncol Clin North Am 1994;8:1181–1201.

    326. Joist JH. Hemostatic abnormalities in liver disease. In: Colman RW , Hirsh J , Marder VJ , eds. Hemostasis and thrombosis: basic principles and clinical practice. Philadelphia: J.B. Lippincott, 1994:906.

    327. Rastogi P, Gupta SC, Bisht D. Coagulation studies in patients with cirrhosis of liver: bleeders vs. non-bleeders. Indian J Pathol Microbiol 1990;33:323–327.

    328. Buzzelli TJ, Trittipoe RE. Massive volume loss during cardiopulmonary bypass and its association with Meckel's diverticulum. J Extra-Corp Technol 1997;29:45–48.

    329. Kurusz M, Wheeldon DR. Perfusion safety. Perfusion 1988;3:97–112.

    330. Perrow C. Normal accidents, living with high-risk technologies . New York: Basic Books, 1984.