Anesthetic management of a patient with mitral stenosis undergoing mitral valve repair/replacement

Praveen Kumar Neema; Mukul Chandra Kapoor

doi:10.4103/2394-7438.182720

TUTORIAL

Year : 2016 | Volume : 2 | Issue : 2 | Page : 94-98

Anesthetic management of a patient with mitral stenosis undergoing mitral valve repair/replacement

Praveen Kumar Neema¹, Mukul Chandra Kapoor²
¹ Department of Anaesthesiology, AIIMS, Raipur, Chhattisgarh, India
² Department of Anaesthesiology, Max Smart Super Specialty Hospital, New Delhi, India

Date of Web Publication

19-May-2016

Correspondence Address:
Praveen Kumar Neema
Department of Anaesthesiology, AIIMS, Raipur - 492 099, Chhattisgarh
India

Source of Support: None, Conflict of Interest: None

Check

DOI: 10.4103/2394-7438.182720

Abstract

The results of mitral valve surgery have improved steadily. The current operative mortality rates for mitral valve surgery are reported to be in the region of 1.5% for mitral valve repair and 5.5% for mitral valve replacement. To ensure good perioperative patient outcome, it is imperative to follow management techniques based on sound scientific principles. In this review article, the authors describe anesthetic management, complexities of cardiopulmonary bypass and weaning from cardiopulmonary bypass in patients of varying severity of mitral stenosis.

Keywords: Anesthetic management, cardiopulmonary bypass, mitral stenosis

How to cite this article:
Neema PK, Kapoor MC. Anesthetic management of a patient with mitral stenosis undergoing mitral valve repair/replacement. MAMC J Med Sci 2016;2:94-8

How to cite this URL:
Neema PK, Kapoor MC. Anesthetic management of a patient with mitral stenosis undergoing mitral valve repair/replacement. MAMC J Med Sci [serial online] 2016 [cited 2023 Jun 4];2:94-8. Available from: https://www.mamcjms.in/text.asp?2016/2/2/94/182720

Introduction

Patients with mitral stenosis (MS) presenting for mitral valve (MV) repair or replacement show a wide range of clinical features varying from asymptomatic ones to the ones who are severely symptomatic even at rest. The asymptomatic ones may have severe MS, but demonstrate clinical features only on exercise testing. However, the majority of patients present with New-York Heart Association (NYHA) class III or IV status. Apparently, each patient is unique with reference to the pathophysiological effects of the MS. The patients who are asymptomatic and develop clinical features of MS on exercise are the least diseased ones whereas the ones who are symptomatic (dyspneic) at rest are severely diseased and may have hepatic and renal dysfunction and subnormal cardiac output (CO).^[1] In patients who are in NYHA class III or IV, pulmonary artery (PA) pressure is usually severely raised, the right ventricle (RV) is hypertrophied; consequently, its output is preload-dependent. The systemic arterial pressure in these patients is maintained by an increase in systemic vascular resistance (SVR).^[2] The cardiovascular stability in these patients is precariously balanced, and a mild decrease in RV preload, SVR, myocardial contractility, or RV perfusion can result in severe hypotension and cardiovascular collapse.

The role of preoperative evaluation of a patient with MS is to ascertain the NYHA status, to diagnose the severity of MS, to determine the heart rate and rhythm, to define the severity of PA hypertension (PAH), to define the severity of dysfunction of other organs secondary to pathophysiology of MS, to identify patients having precariously balanced hemodynamics, to identify associated comorbidities, and to optimize patients in congestive heart failure before surgery by anti-failure measures. The patients in atrial fibrillation are treated with oral anticoagulants to prevent arterial thromboembolic event;^[3] however, oral anticoagulants are discontinued preoperatively and replaced by subcutaneous heparin. The last dose of heparin should be administered at least 6–8 h before scheduled surgery. For the convenience of intraoperative anesthetic management, the authors sub-divide the patients with severe MS in “Three Groups” on the basis of absence or presence of the following pathophysiologic effects of MS, namely severity of PAH, RV hypertrophy, severity of tricuspid regurgitation, and hepatic and renal dysfunction [Table 1]. The severity of PAH, RV hypertrophy, and tricuspid regurgitation is assessed by Doppler echocardiography, and the dysfunction of the liver and kidney is assessed by laboratory investigation of serum creatinine and liver functions including prothrombin time.

Table 1: Grouping of the patients of mitral stenosis based on severity of PAP and secondary effects of mitral stenosis

Click here to view

The perioperative course of the patients belonging to Group I is usually similar to normal patients; however, it is important to control their heart rate. The patients belonging to Group II and III have moderate to severe PAH, varying degree of RV hypertrophy, and tricuspid regurgitation; in these patients, as described earlier in the pathophysiology of MS,^[4] the key to successful anesthetic management is to keep RV performance optimum. An increase in RV output in the presence of severe MS can produce pulmonary congestion and edema whereas the circulatory changes which can compromise RV performance and decrease RV output can result in hemodynamic collapse. Consequently, the anesthetic agents and the techniques selected for the conduct of anesthesia should take into account their circulatory effects, particularly their effects on RV performance. The determinants that compromise the RV performance include alterations in its preload, afterload, perfusion, and contractility, and heart rate and rhythm. Tachycardia and vasodilation are the two important mechanisms that compromise the perfusion and filling of the RV and its ability to generate adequate stroke volume against raised PA pressure.^[5] A precipitous increase in the PA pressure can also decrease the RV performance.^[6],[7] An inappropriately large tidal volume (TV) selected for mechanical ventilation is an important but a less recognized cause of hemodynamic compromise. A large TV interferes in two ways, it increases afterload to RV and decreases RV preload by decreasing venous return and by increasing juxtacardiac pleural pressure.^[8] Adequate RV preload is essential for optimal RV performance against severely raised PA pressure. Arguably, while managing these patients, one should prevent tachycardia and venodilation, preserve RV preload, systemic arterial pressure, and myocardial contractility, and avoid an increase in PA pressure. Although the mechanisms that can compromise RV performance are easy to comprehend, in practice, the objectives are difficult to achieve.

Various monitoring techniques are employed to detect hemodynamic effects of anesthetic agents and ventilatory techniques so that appropriate corrective measures can be taken. The transesophageal echocardiography (TEE) is employed to evaluate cardiac functions (myocardial contractility), ventricular preload, and adequacy of MV repair or evaluation of valve prosthesis after replacement. The PA catheter (PAC) is useful for evaluation of the left atrial (LA) pressure. Several other monitoring techniques are used to help safe conduct of cardiopulmonary bypass (CPB). In general, the patients undergoing MV repair or replacement are usually monitored with an electrocardiogram (ECG), pulse oximetry, noninvasive blood pressures, invasive blood pressures, blood gas analysis, central venous catheter, PAC, temperature, urine output, and TEE. An invasive arterial line is inserted before the induction of anesthesia, particularly in Group II and III patients for continuous monitoring of arterial pressure.

How to Achieve the Desired Goals?

In Group I patients, as stated earlier, a control of the heart rate to about 60–80 beats/min is adequate and keep the hemodynamics well within the acceptable limits. The control of the heart rate provides sufficient diastolic period for the emptying of LA which keeps the mean LA pressure low ^[9],[10] and avoids the complications of increased LA pressure such as pulmonary venous congestion; moreover, sufficient diastolic interval provides time for LV filling which helps maintaining CO. Premedication with any β-blocker (e.g., atenolol, 0.5 mg/kg) ensures adequate control of the heart rate.^[11] As stated earlier, the patients of Group II and III are precariously balanced hemodynamically, and great care is needed while anesthetizing them. In these patients, CO is usually subnormal, and the LV preload is dependent on the performance of the RV; whereas the systemic arterial pressure is maintained by sympathetic stimulation-mediated systemic vasoconstriction. All sedative premedicants decrease vascular tone by decreasing sympathetic activity; therefore, their administration can potentially decrease systemic arterial pressure as well as RV preload. In these patients, heavy premedication is poorly tolerated. The authors administer oral β-blocker (atenolol 0.5 mg/kg) routinely and intramuscular morphine 0.05 mg/kg selectively. Additional sedation may be considered before anesthesia induction, if the patient is inadequately sedated.

In these patients, casual administration of commonly used anesthesia-inducing agents including thiopentone, propofol, midazolam, and fentanyl combination can potentially cause hemodynamic collapse; however, all these anesthetic agents are being used and have a reasonable safety record. Etomidate is considered the most cardio-stable induction agent for these patients. One should realize that in these patients, the CO is subnormal and preferentially distributed to well-perfused areas (vessel-rich group organs) such as heart, brain, liver, and kidney, and the circulation is slow;^[12] therefore, the volume of distribution of the administered drugs is decreased, and the drugs on administration are preferentially distributed to well-perfused areas. In other words, small total dose of anesthetic agents is required, and the effect of the drugs is delayed; therefore, one should titrate the effects of the intravenous induction agents over a relatively longer period. The skill of administration of these agents is most important, and rapid administration of an inappropriately large dose based on body weight is the most common cause of hemodynamic collapse. A thiopentone dose of 1–3 mg/kg is usually sufficient to induce sleep. The depth of anesthesia before endotracheal intubation is increased by a small dose of narcotic, fentanyl 2–5 μg/kg, and inhalation anesthetic delivery, while supporting face mask ventilation. Often in spite of slow and careful administration of a small dose of an anesthetic-inducing agent, hemodynamic collapse occurs. The causes of hemodynamic collapse in such a situation may include hypotension due to vasodilation due to suppression of sympathetic drive and decrease in RV preload and myocardial depression.

The hemodynamic stability can usually be restored by phenylephrine; however, one should remember that a large dose of phenylephrine can significantly increase the central blood volume and may lead to cardiac overload, pulmonary congestion, edema, and RV failure. Therefore, a small bolus dose of phenylephrine (10–25 μg) is initially administered which can be repeated if acceptable hemodynamics is not achieved; alternatively, one can start infusion of an inotrope with mild vasoconstrictor property such as epinephrine. Inotropes such as dopamine and dobutamine are also used, but can be counterproductive because of their chronotropic and vasodilating actions. Since these hemodynamic aberrations occur during induction of anesthesia, it is advisable to place invasive lines to detect and monitor these aberrations before induction of anesthesia; further; one should prepare the required drugs prior to the start of anesthesia. The authors recommend inserting an arterial pressure monitoring line in radial artery in Group I patients and femoral artery in Group II and III patients. The central venous line is usually placed after induction of anesthesia; however, a central venous access is secured by inserting a long line through basilica vein for administration of vasoactive drugs for patients belonging to Group II or III. As stated earlier, mechanical ventilation with a large TV (10 ml/kg) can compromise hemodynamics secondary to decrease in RV preload and increase in RV afterload and may decrease CO.^[8] Therefore, initially a low TV (5–6 mL/kg) is selected and titrated upward based on its effects on hemodynamics. The decrease in venous return during mechanical ventilation is exaggerated in the presence of hypovolemia and in sympathetic drive-dependent patients.

The period from the induction of anesthesia till beginning of CPB is critical particularly in patients belonging to Group III; therefore, it is advisable to shorten this period as much as possible. The common hemodynamic aberrations seen are hypotension after anesthesia induction until surgical incision and tachycardia and hypertension after surgical incision. The magnitude of these hemodynamic changes depends on the technique of anesthesia chosen for induction; in narcotic-based anesthesia technique, hypotension is common whereas in inhalational anesthesia technique, magnitude of hypotension is less and reverses easily on lowering inhaled concentration of the anesthetic agent. Hypertension and tachycardia are observed more frequently with inhalation anesthetic-based techniques at the time of surgical incision if the depth of anesthesia is not increased prior to surgical incision. The depth of anesthesia is maintained with narcotics and inhalation anesthetic agents; and can be titrated by BIS monitoring. Both the narcotic-based and the inhalation anesthetic-based or their combination-based techniques of anesthesia are in use and have a good safety record. At present, the combination technique including modest doses of a narcotic such as fentanyl up to 10 μg/kg and an inhalation anesthetic isoflurane is widely practiced. The inhalation anesthetic agents are easier to titrate, can be withdrawn easily if required, and provide myocardial protection.^[13],[14] Moreover, inhalation-based techniques provide an opportunity for early extubation.

Myocardial Protection and Anesthesia during Cardiopulmonary Bypass

The standard preparations for CPB are made, and before commencing the CPB, adequacy of anticoagulation, anesthesia, and cannulation is ensured, drips are turned off, monitors are recalibrated, and pupils are checked. The anesthesia during CPB is maintained by administration of sevoflurane or isoflurane to the extracorporeal circuit as tolerated. The CPB management for the patients belonging to Group I is straight forward. However, patients of Group II and III often develop severe hypotension on commencement of CPB, which often needs treatment by small boluses of phenylephrine. The patients of Group III who are in congestive heart failure and have features of systemic venous congestion can have excessive venous return and flooding of venous reservoir. In such cases, the venous line can be primed retrograde and the excess volume in the venous reservoir can be siphoned out; the perfusionist should be informed about retrograde priming at the time of setting of CPB circuit, if the same is planned. Alternatively, if the venous reservoir is flooded and estimated hemoglobin on CPB is above 10 g%, the perfusionist can be advised to collect one or two units of blood from the arterial re-circulation line of the CPB circuit; if the venous reservoir is flooded and hemoglobin is <7 g%, the perfusionist can be advised to hemo-concentrate the prime by using hemofiltration ^[5] or by administration of furosemide. The collected blood can be added to the venous reservoir during CPB, if required.

Systemic hypothermia and cold crystalloid or blood cardioplegia are the standard practices adapted for myocardial protection. At the end of the surgical procedure, the patient is warmed to 35°C rectal temperature or 36–37°C nasopharyngeal temperature. Overwarming is avoided which is possibly associated with adverse neurological outcome and results in significant vasodilation and low systemic arterial pressure at the time of termination of CPB. The patients with RV hypertrophy and dilation (Group II and III patients) are susceptible to develop RV dysfunction due to inappropriate increases in its preload, afterload, and decreased myocardial perfusion; therefore, utmost care should be taken while separating these patients from CPB. The heart is meticulously de-aired before removal of aortic clamp. After removal of aortic clamp, the arterial inflow is maintained at approximately 1.5 L/min/m ² and the LA/LV vent is run at approximately 25 revolutions per min until the return of myocardial tone and effective contraction. It should be remembered that residual trapped air is often released after the return of effective myocardial contractions; therefore, the aortic root needle inserted for the administration of cardioplegia is usually left open or actively vented for letting out released air. Active LA/LV venting and aortic root venting can remove air from the left heart and may prevent systemic and coronary artery air embolism. It should be noted that residual air release often results in embolism of the right coronary artery (RCA) perhaps because of its anatomical position and the frequent use of the Trendelenburg (head down) position of the patient at the time of aortic clamp release. Air emboli in the RCA circulation result in ST-segment elevation in ECG leads II, III, and aVF (inferior leads). RCA air embolism is most damaging in the presence of RV hypertrophy and PAH. Coronary air emboli during CPB are managed by continuing CPB and administration of intravenous inotropic agents. Pulsatile blood flow, partial aortic clamp application, and/or increased blood pressure and force of myocardial contraction help facilitate transit of gaseous emboli through the coronary circulation.^[15] In case of RCA embolism, weaning from CPB should be delayed until trapped air is moved out.

Weaning from Cardiopulmonary Bypass

The process of weaning from CPB entails progressive transition from full mechanical circulatory support to spontaneous heart activity with the aim of providing sufficient blood flow through the pulmonary and systemic circulation at a suitable pressure. Preoperative status and the behavior of the patient during anesthesia induction and prebypass period provide vital clues to the expected course while weaning from CPB. Prior to weaning from CPB, serum potassium, blood gases, and hematocrit are optimized, the lungs are inflated manually, and mechanical ventilation is started with oxygen 100% at a TV of 6–8 mL/kg body weight, inhalation anesthetic agent, sevoflurane or isoflurane (0.5%) is added to oxygen, and a heart rate of about 80–90/min is ensured, and if spontaneous rate is <80/min, pacing is initiated. Before terminating CPB, all the monitors are re-calibrated and placed online. The authors electively start a modest dose of epinephrine infusion, 0.05 - μg/kg/min, about 10–15 min before weaning, which is titrated upward or downward after separation from CPB depending on the hemodynamics. In patients with PAH, milrinone due to its pulmonary vasodilatory effect can be a good choice.

The process of weaning from CPB is considered complicated because of CPB-induced inflammatory response and the possibility of myocardial dysfunction due to myocardial ischemia during aortic clamp, inadequate myocardial protection, myocardial edema, preexisting myocardial dysfunction, paravalvular leak, iatrogenic damage to circumflex artery,^[16] etc. Therefore, an uncertainty is involved in separation from CPB. The first few minutes after weaning from CPB are important when decisions regarding pharmacological support, ventricular assistance, and additional surgical interventions have to be taken quickly based on clinical information, hemodynamic parameters, ECG, and TEE examination to avoid myocardial damage and to ensure adequate oxygen transport.

A very simple technique of separation from CPB practiced by the authors is described. Once anesthesiologist, surgeon, and perfusionist indicate their preparedness, the venous line is clamped, the arterial pressure, the myocardial contractility, the preload of the left ventricle in transgastric mid-papillary view, and the heart are carefully observed, and as soon as any one of the following is achieved or noticed - a systolic arterial pressure of 60–80 mmHg, an adequately filled left ventricle as indicated by TEE, a CVP of 5 mmHg, a visually just full RV, or slowing of the heart, the arterial inflow from the extracorporeal circuit is discontinued. A higher systolic arterial pressure (~80 mmHg) is targeted in Group I patients while in sick patients of Group II and III, the arterial inflow is terminated early at ~60–70 mmHg. Thereafter, the hemodynamic parameters are carefully observed and depending on the hemodynamic parameters and TEE findings, preload, heart rate, and inotropic support are further increased. If preload is augmented, LV filling and contractility are assessed continuously by TEE during transfusion and care is taken to prevent LV distention. Satisfactory hemodynamic parameters are usually achieved within 10–15 min; thereafter, the MV is examined for adequacy of repair, or the valve prosthesis is examined for paravalvular leaks and for the satisfactory movement of the prosthesis leaflets or the disc.

After ensuring satisfactory hemodynamics and valve repair or prosthesis functioning, anticoagulation is reversed by slow administration of the calculated dose of protamine. The remaining postbypass period is an exercise in adjustment of preload and vasoactive drugs. For volume replacement, the collected blood and the residual perfusate in the extracorporeal circuit are administered slowly or autologous packed red blood cells are given. Intravenous furosemide is administered to take care of hemodilution. The strategy of weaning from CPB with gradual augmentation of preload and inotropic support is aimed to avoid distension of LV as well as the RV and to facilitate movement of accumulated tissue fluid in the vascular compartment and is based on the following evidences available in the literature (1) hypovolemia and decreased intravascular pressure lead to fluid movement from tissues to blood,^[17] (2) short periods of controlled hypotension are well tolerated by patients under anesthesia and by trauma victims,^[18] (3) the sick patients with valvular heart disease have their CO in low normal range,^[19] (5) ultrafiltration improves outcome in sick patients undergoing cardiac surgery with CPB,^[20] (6) administration of inotropic drugs after CPB and cardioplegic cardiac arrest may hasten resolution of myocardial edema,^[21] (7) required CO is less than normal in well sedated and mechanically ventilated patients,^[22] and in early postbypass period if core temperature is <35.5°C.^[23] Moreover, the patients with severe MS generally have their CO in subnormal range and they are well adapted to it.

However, in case patient shows slowing of the heart or distention of the LV or RV, the inotropic support is increased and the heart is either decompressed, or the CPB is re-established. While CPB is being re-established, attempt is made to find out the cause of failure, and appropriate measures are initiated which may include rest on CPB, the second run of CPB to repair residual lesion if any, and/or addition of more inotropic support or circulatory support by intra-aortic balloon pump. In patients with RV dysfunction, pulmonary vasodilators (nitroglycerin and sodium nitroprusside) along with milrinone or levosimendan can be considered.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

References

1.	Hugenholtz PG, Ryan TJ, Stein SW, Abelmann WH. The spectrum of pure mitral stenosis. Hemodynamic studies in relation to clinical disability. Am J Cardiol 1962;10:773-84.
2.	Sukernik MR, Martin DE. Anesthetic management for the surgical treatment of valvular heart diseases. In: Hensley FA, Martin DE, Gravlee GP, eds. A Practical Approach to Cardiac Anesthesia. 4^th ed. Philadelphia: Lippincott Williams & Wilkins, 2008:316-47.
3.	Kottkamp H, Hindricks G, Breithardt G. Role of anticoagulant therapy in atrial fibrillation. J Cardiovasc Electrophysiol 1998;9 8 Suppl: S86-96.
4.	Neema PK. Pathophysiology of mitral valve stenosis. MAMC J Med Sci 2015;1:25-7.
5.	Neema PK, Singha SK, Manikandan S, Muralikrishna T, Rathod RC, Dhawan R, et al. Case 6-2011: Aortic valve replacement in a patient with aortic stenosis, dilated cardiomyopathy, and renal dysfunction. J Cardiothorac Vasc Anesth 2011;25:1193-9.
6.	Belik J, Light RB. Effect of increased afterload on right ventricular function in newborn pigs. J Appl Physiol 1989;66:863-9.
7.	Vlahakes GJ, Turley K, Hoffman JI. The pathophysiology of failure in acute right ventricular hypertension: Hemodynamic and biochemical correlations. Circulation 1981;63:87-95.
8.	Pinsky MR. Heart-lung interactions. In: Grenvik A, Ayres SM, Holbrook PR, Shoemaker WC, editors. Text Book of Critical Care. Noida, India: Saunders; 2000. p. 1204-21.
9.	Ashcom TL, Johns JP, Bailey SR, Rubal BJ. Effects of chronic beta-blockade on rest and exercise hemodynamics in mitral stenosis. Cathet Cardiovasc Diagn 1995;35:110-5.
10.	Klein HO, Sareli P, Schamroth CL, Carim Y, Epstein M, Marcus B. Effects of atenolol on exercise capacity in patients with mitral stenosis with sinus rhythm. Am J Cardiol 1985;56:598-601.
11.	Neema PK, Waiker HD. Atenolol premedication in patients undergoing closed mitral commissurotomy. Ann Card Anaesth 2003;6:42-6. [PUBMED]
12.	Borden CW, Ebert RV, Wilson RH, Wells HS. Studies of the pulmonary circulation. II. The circulation time from the pulmonary artery to the femoral artery and the quantity of blood in the lungs in patients with mitral stenosis and in patients with left ventricular failure. J Clin Invest 1949;28 (5 Pt 2):1138-43.
13.	Sepac A, Sedlic F, Si-Tayeb K, Lough J, Duncan SA, Bienengraeber M, et al. Isoflurane preconditioning elicits competent endogenous mechanisms of protection from oxidative stress in cardiomyocytes derived from human embryonic stem cells. Anesthesiology 2010;113:906-16.
14.	Kato R, Foëx P. Myocardial protection by anesthetic agents against ischemia-reperfusion injury: An update for anesthesiologists. Can J Anaesth 2002;49:777-91.
15.	Utley JR, Stephens DB. Air embolus during cardiopulmonary bypass. In: Utley JR, editor. Pathophysiology and Techniques of Cardiopulmonary Bypass. Baltimore, MD: Williams and Wilkins; 1983. p. 78-100.
16.	Hiltrop N, Bennett J, Desmet W. Circumflex coronary artery injury after mitral valve surgery: A report of four cases and comprehensive review of the literature. Catheter Cardiovasc Interv 2016; doi: 10.1002/ccd.26449. [Epub ahead of print].
17.	Lundvall J, Länne T. Large capacity in man for effective plasma volume control in hypovolaemia via fluid transfer from tissue to blood. Acta Physiol Scand 1989;137:513-20.
18.	Bickell WH, Wall MJ Jr., Pepe PE, Martin RR, Ginger VF, Allen MK, et al. Immediate versus delayed fluid resuscitation for hypotensive patients with penetrating torso injuries. N Engl J Med 1994;331:1105-9.
19.	Lake CL. Chronic treatment of congestive heart failure. In: Kaplan JA, editor. Cardiac Anesthesia. Philadelphia: WB Saunders; 1993. p. 681-751.
20.	Montenegro LM, Greeley WJ. Pro: The use of modified ultrafiltration during pediatric cardiac surgery is a benefit. J Cardiothorac Vasc Anesth 1998;12:480-2.
21.	Allen SJ, Geissler HJ, Davis KL, Gogola GR, Warters RD, de Vivie ER, et al. Augmenting cardiac contractility hastens myocardial edema resolution after cardiopulmonary bypass and cardioplegic arrest. Anesth Analg 1997;85:987-92.
22.	Ramsay J. How much cardiac output is enough? J Cardiothorac Vasc Anesth 2002;16:1-3.
23.	Corso PJ, Hockstein MJ. New techniques in management of the cardiac surgery patient. In: Grenvik A, Ayres SM, Holbrook PR, Shoemaker WC, editors. Text Book of Critical Care. Noida, India: WB Saunders; 2000. p. 1130-55.