Current Review-

Care of the Adult Cardiac Surgery Patient

PART II- (Subsystem Approach Continued)

 Pezzella AT, Ferraris VA Lancey RA,
Care Of The Adult Cardiac Surgery Patient:
Part II Current Problems In Surgery Volume 41, Number 6, June 2004

Respiratory monitoring

Vasodilation with Normal/High Cardiac Output

Bleeding/Cardiac Tamponade

PART II- (Subsystem approach continued)

The major advances in the respiratory subsystem include preoperative assessment and preparation, better understanding of the anesthetic effect on lung function, the pathophysiological effects of CPB, the use of lower tidal volumes on mechanical ventilation, and fast track methodology to accelerate earlier extubation. Other respiratory considerations include pneumonia, pain control, atelectasis, bronchospasm, pulmonary aspiration associated with swallowing disorders, diaphragmatic dysfunction, the effect of pleurotomy, reintubation, prolonged mechanical ventilation, and pulmonary thromboemboli. The contemporary goal of pulmonary management is to effect a gradual but rapid transition from an anesthetized, intubated, ventilated patient to an awake, and extubated patient.

Patients with chronic obstructive lung disease (COLD), poor respiratory function, pneumonia, bronchospasm, smoking, and poor nutrition (serum albumin <2.5 g/dl) are at risk for respiratory dysfunction perioperatively. Preoperative cessation of smoking has not proven to be of benefit ⁽¹⁾. Pre-operative incentive spirometry, treatment of bronchospasm, and antibacterial treatment of pneumonia are highly recommended. Sivak ⁽²⁾ nicely discusses the need for ventilator dependency following cardiac surgery. In the Cleveland Clinic experience approximately 10% of adult cardiac surgery patients required mechanical ventilation beyond 48 hours. Additional pre-operative risk factors include emergency surgery, female gender, congestive heart failure, and decreased LV function. Operative factors include the effects of anesthesia or pulmonary function and the effects of CPB. CHF with pulmonary edema remains the most significant independent risk factor for prolonged mechanical ventilation ⁽³⁾.

Anesthesia decreases functional residual capacity (FRC) by 20%. This is related, in part, to changes in diaphragmatic motion, as well as decreasing hypoxic pulmonary vasoconstriction (HPVC) ⁽⁴⁾. The lung is a major target organ for dysfunction following cardiac surgery utilizing systematic hypothermia and CPB. Microemboli and activation of the coagulation system creates changes at the alveolar capillary level creating a widened A-a gradient, decreased lung compliance, and increased (FRC) ⁽³⁾. Membrane oxygenators have markedly decreased this response. The traditional approach to mechanical ventilation using tidal volumes (TV) of 10 – 15 cc/ Kg body weight has been challenged. The Acute Respiratory Distress Syndrome network ⁽⁵⁾, confirmed previous studies using lower TV’s (6cc/Kg), and increased positive end-expiratory pressures in order to decrease barotrauma, while maintaining adequate oxygenation and ventilation.

At the clinical level adequate drainage of the left chest and protection of the left phrenic nerve, have also contributed to decrease respiratory morbidity. Diuresis, limitation of fluid overload and optimization of LV function are essential to minimize pulmonary edema. Respiratory physiotherapy, especially incentive spirometry and early ambulation are important in the post extubation phase.

Advances in mechanical ventilation allow for more physiological weaning. Tobin ⁽⁶⁾ discusses the subject extensively. He describes four weaning methods: (1) the tube circuit at 5 – 10 minutes advanced to 1 – 2 hours; (2) intermittent mandatory ventilation (IMV) to 4/minute; (3) pressure support to 6 – 8 cm H₂O; and (4) a single daily T-tube trials for up to 2 hours. An example of criteria for extubation is summarized in (Figure 1). Clearly an awake patient, pain free, hemodynamically stable, no bleeding, and meets extubation criteria are candidates for fast track. A 4 – 8 hour window is the ideal goal. The role of steroids is controversial. Chaney et al ⁽⁷⁾ in a randomized study showed no value of methyl prednisolone (30mg /Kg during sternotomy and 30 mg/kg during initiation of CPB). Chaney et al ⁽⁸⁾ again confirmed the findings even with a decreased dose (15mg/kg). Yared et al ⁽⁹⁾ showed a decrease in extubation time but no overall change in morbidity or mortality utilizing dexamethasone 0.6mg/kg after induction of anesthesia. A recent general overview of variation in measurement of weaning parameters was done amongst respiratory therapists at nine Los Angeles hospitals ⁽¹⁰⁾. The four parameters noted in >90% were MIP, VT, VE, and RR (maximum inspiratory pressure, tidal volume, minute ventilation, and respiratory rate).

Pneumonia occurs in 1.5% to 9% of postoperative cardiac surgical patients. Pre-operative colonization is the predominant cause. Pre-operative monitoring of tracheal aspirates in high-risk patients may be of value in choosing more appropriate antibiotic coverage ^(11). Pain control is crucial to minimize complications and encourage aggressive pulmonary toilet. Aggressive treatment of atelectasis and bronchospasm is crucial to improved respiratory convalescence ^{(12, 13)}. Atelectasis secondary to poor respiratory effort or restrictive pleural effusions is a frequent source of early fever and pulmonary dysfunction, as well as serving as an entry for nosocomial pneumonia.

Pulmonary aspiration secondary to alterations in the swallowing reflex occurs more commonly than realized, not only in extubated patients but especially in the prolonged intubated group ⁽¹⁴⁾. Harrington et al ⁽¹⁵⁾ reported an incidence of 1.8% silent aspiration in CABG patients. Ferraris and co-workers ⁽¹⁶⁾ found 31 of 1042 (3.0%), patients having cardiac operations developed oropharyngeal dysphagia (OD) after operation. OD was more common in older patients with diabetes, renal insufficiency, hyperlipidemia, preoperative congestive heart failure, and in those having non-coronary artery bypass procedures. Postoperative neurologic complications are more common in patients with OD (10 of 31 patients with OD had some new neurologic complication compared to 36 of 1011 with a postoperative neurologic complication without OD, p<0.0001). Therapy consisting of modification of eating behavior and swallowing technique and in some severe cases enteral or parenteral feeding was successful in all surviving patients except one, but 4 patients required more than 4 months of supportive care before return to oral feeding was possible. Diaphragmatic dysfunction is discussed later. Pleurotomy is routine for Internal mammary Artery (IMA) harvesting. Major changes in pulmonary function postoperatively have not been noted after opening the pleura ⁽¹⁷⁾. Matsumoto et al, ⁽¹⁸⁾ showed a decrease in FVC in 79 patients undergoing CABG. In 2 of 3 groups having IMA’s, the group with skeletonized IMA’s had better FVC’s than with the musculo-fascial pedicle dissection. Patients with IMA’s have a larger incidence of pleural effusions. A variety of methods have been used to minimize the incidence of effusions with the attendant higher risk of pulmonary dysfunction and atelectasis. This includes prolonged duration of chest tube drainage, as well as a variety of drainage catheters and positioning to better drain the opened left chest cavity.

Reintubation is seen in two instances: self-extubation or failure to sustain adequate post-extubation ventilation ^(19-22). Underlying pulmonary disease, pulmonary edema from fluid overload, retained secretions, atelectasis/collapse, pleural effusion, and occasionally pulmonary emboli are the usual causes. The rate of reintubation in the postoperative cardiac surgery patient is less than 5% but is associated with prolonged hospital course, increased resource utilization and increased mortality. Prolonged mechanical ventilation has become more common e ^(23,24). Earlier standard or conventional tracheotomy, and especially the newer percutaneous methods, allow a more controlled and comfortable system for the patient and staff. The major risk of tracheotomy is mediastinitis. Curtis et al ⁽²⁵⁾ reported an incidence of 1.45% tracheotomy following CABG with an 8.6% mediastinitis rate, and overall 24.7% mortality. The time interval between tracheotomy and CABG was not predictive of mortality but a 1 to 2 week interval was recommended. Percutaneous dilatational tracheostomy (PDT) may become an attractive alternative in these cardiac surgery patients. Freeman et al ⁽²⁶⁾ reported a prospective randomized study of PDT on 80 patients. Performed in the ICU, the procedure was safe, quicker, and cost effective. However, PDT remains an elective procedure with standard tracheostomy for urgent, emergency situations.

Pulmonary embolism is common in the general population, affecting over 700,000 annually with a mortality of 20-40% ⁽²⁷⁾. The incidence following cardiac surgery is low, ranging from 0.6-1%. Since there is reluctance to use anticoagulant regimens routinely for fear of increased bleeding problems other strategies have been utilized. Early ambulation, graduated compression stocking, and the use intermittent pneumatic compression are utilized, the latter most commonly used in patients confined to bed ^(28,29).

The major considerations in the hematological subsystem include: preoperative hematological abnormalities, whether drug or non-drug related; the effect of CPB on blood components; postoperative surgical bleeding vs. coagulopathy; transfusion issues, i.e. blood and component therapy; and blood conservation strategies.

An increasing number of patients with pre-existing underlying blood dycrasias are undergoing successful cardiac surgery. A comprehensive review of this subject is not available. Most information is from anecdotal case reports and retrospective series (Figure 2). A planned pre-operative strategy in conjunction with the Hematology Service and the Blood Bank is prudent in order to insure proper success of the operation. Heparin induced thrombocytopenia can occur with or without thrombosis (HIT/HITT). It is a presumed IgG antibody-mediated reaction to the platelet factor 4 and heparin complex ⁽⁴⁹⁾. The incidence ranges from 1% – 3%. HIT/HITT usually develops 5 – 8 days after exposure to unfractionated heparin. A fall in platelets to 50% of the pre-heparin valve is diagnostic of HIT. The Toronto group summarizes nicely the non thrombotic (HIT) and thrombotic (HITT) types and the variety of substitutes to heparin ^(49). For patients requiring CPB, strategies to avoid heparin are crucial ⁽⁵⁰⁾. Follis ⁽⁵¹⁾ outlines a collective review of the literature. Aouifi et al ⁽⁵²⁾ presented 10 patients with HITT: 4 patients treated with danoproid sodium, and 6 patients treated with heparin sodium after pre-treatment with epoprostenol sodium. Results were mixed in this small group.

The use of platelet inhibitors has become an integral part of the medial armamentarium in the treatment of acute coronary syndromes. Many of these patients are coming to elective, urgent, and emergency cardiac surgery. Aspirin (ASA) clearly reduces mortality following myocardial infarction. It works by inhibiting cycloxygenase, thus reducing the synthesis of thromboxane A₂ (TXA₂) and thereby decreasing stimulated platelet aggregation ⁽⁵³⁾ (Figure 3). The half-life of platelets is 7 – 10 days, but the effective circulating pool of platelets is probably reconstituted in 3-4 days after aspirin ingestion. The use of ASA clearly increases perioperative blood loss ^(54,55). Ferraris and co-workers ⁽⁵⁵⁾ reviewed the records of 2,606 consecutive patients undergoing CABG to identify patients with a history of ASA ingestion up until the time of operation. ASA ingestion was correlated with postoperative blood transfusion using multivariate analysis. Stratification of the results according to the frequency of ASA use showed that ASA was an independent multivariate predictor of postoperative blood transfusion only in high-risk patients. In addition to identifying ASA as a significant predictor of increased postoperative blood transfusion, these authors suggest that a high-risk cohort of ASA users exists. A platelet function testing algorithm that combines preoperative risk factor assessment, template bleeding times, and flow cytometry may allow the identification of this high-risk cohort (ASA hyper-responders) who are at increased risk for bleeding. The current strategy is to continue ASA up to and through the perioperative period in most patients but to delay operation, if possible, in certain high-risk patients. Despite this, there is current evidence to suggest that ASA is safe and early use of ASA postoperatively i.e. within six hours post-up is beneficial to outcome and survival ⁽⁵⁶⁾.

Other drugs, including clopidogrel (Plavix) and the IIb / IIa inhibitors [ Abciximab (Rheo-pro), eptifibitatide (Integrilin), and tirofiban (Aggrastat)] inhibit platelet function ⁽⁵³⁾ (Figure 3). Cessation of drugs and platelet administration usually reverse the effect within twelve hours. Abciximab usually requires a 24 – 48 hour delay of surgery, if feasible, to decrease bleeding risks ⁽⁵⁷⁾. Pre-operative use of thrombolytic agents, warfarin, and fractionated heparin require individualized approaches ⁽⁵⁸⁾.

CPB creates changes in the coagulation mechanism. Hemodilution, hypothermia, flow through non-organic surfaces, heparin, protamine, anesthesia, and other drugs effects, all contribute to a coagulopathic state. Platelet activation of the complement pathway result from these inciting events ⁽⁵⁹⁾. Hemodilution causes thrombocytopenia and decreased coagulation factors. Hypothermia impairs platelet thromboxane A₂ synthesis ⁽⁶⁰⁾.

The major concern in the hematological subsystem is post-operative bleeding. The incidence of bleeding requiring re-operation within the first 24-48 hours continues to decrease. The STS database reports an incidence of 2.32% in CABG patients, and 3.10% in redo CABG patients (http:/www.ctsnet.org/doc/2988), and the Northern New England Study group has shown a decline from 3.6% to 2% over a 4-year period ⁽⁶¹⁾.

Resternotomy for bleeding is a definite risk factor for adverse outcomes. Unsworth-White et al ⁽⁶²⁾ in a study of 2,221 patients showed a reoperation rate of 3.8%. By multivariate analysis there was increased ICU stay, IAB use, and mortality. Likewise, the Barnes Hospital group ⁽⁶³⁾ had a 4.2% rexxploration rate with subsequent increase in renal failure, prolonged mechanical ventilation, ARDS, sepsis, atrial arrhythmias, and mortality. Criteria for reexploration are not well established. General guidelines include the following amounts of chest table drainage: 500cc/hr for one hour, 400cc/hr for two hours, and 300cc/hr for three hours ⁽⁶⁴⁾. Increasing PEEP has proven to be of no value in limiting postoperative bleeding ⁽⁶⁵⁾. Reexploration is usually done in the operative room although reexploration for bleeding in the Critical Care Unit has been advocated with surprisingly good results and no increase in wound infection rates ^(66,67). Hartstein et al, ⁽⁶⁸⁾ presented a useful algorithm for excessive mediastinal bleeding (Figure 4).

An appreciation of the coagulopathic challenges following CPB has sparked a number of drug regimens to ameliorate the effect: Aprotinin (Trasylol) is a serine protease inhibitor. Its hemostatic effect is related to antifibrinolytic properties and protection of platelets. It activates factor X 11 (Hageman factor). A multicenter, double blind study of 287 patients undergoing redo CABG showed that both high dose (2 x 10⁶ KIU loading dose, 2 x 10⁶ KIU in CPB circuit, 5 x 106 KIU/hr during surgery) and low dose aprotinin (1 x 10⁶/ 1 x 10⁶/ 2.5 x 10⁵ ) reduces transfusion requirements without a higher risk for perioperative MI ⁽⁶⁹⁾. Different dose regimes have been advocated ⁽⁷⁰⁾. Concerns regarding graft occlusion and stroke remain but have largely been addressed by several randomized studies (Image Trial) that compare postoperative graft patency in aprotinin and control groups ⁽⁷¹⁾. The conclusions confirm that early vein graft occlusion was increased by aprotinin. Presently, aprotinin is used primarily for patients who are at high risk for postoperative bleeding. An exciting new postulate of aprotinin action shows that aprotinin, by two disparate properties, selectively blocks the proteolytically activated thrombin receptor on platelets, the protease – activated receptor 1 (PAR 1). This gives both a hemostatic and antithrombotic action ⁽⁷²⁾_. An exciting area has been recent evidence that aprotinin, as well as ASA, and phosphodiesterase inhibitors, may ameliorate the inflammatory effects of CPB ⁽⁷³⁾.

Desmopressin (DDAVP) is a synthetic vasopressin analogue, which acts by increasing the concentration of Von Willibrand factor, an important mediator of platelet adhesion ⁽⁷⁴⁾. It is an expensive drug with some effect on CABG patients receiving ASA and who have preoperative renal failure ⁽⁷⁵⁾.

Epsilon – aminocaproic acid (Amicar) limits fibrinolysis and prevents the plasmin-mediated degradation of platelet glycoprotein IB, thus preserving and perhaps enhancing platelet aggregate function ⁽⁷⁶⁾. Vander Salm et al ⁽⁷⁷⁾ showed that Amicar limits fibrinolysis and causes decreased postoperative bleeding.

Tranexamic acid competively inhibits plasminogen activation, and at higher concentrations non-competively inhibits plasmin with the overall effect of reducing or eliminating fibrinolysis ⁽⁷⁸⁾. Karski et al ⁽⁷⁹⁾ showed decreased post-operative blood loss but no difference in coagulation profile in double blind matched groups.

Even with recognizing the effects of pre-operative drugs and CPB, aggressive monitoring of blood loss with early reoperation, and use of ameliorating drugs, bleeding still occurs. A variety of strategies have evolved to create bloodless surgery, or at best better blood and blood product conservation. More that 22 million blood components are transfused per year in the USA with the attendant risks of hepatitis, HIV, CIV, and transfusion reactions and an overall cost of 5 – 7 billion dollars ⁽⁸⁰⁾. 10% of blood product use is attributed to cardiac surgery. Guidelines are emerging to suggest a more restrictive blood transfusion policy. The Canadian Critical Care Trials group ⁽⁸¹⁾ restricted transfusion for hemoglobin <7.0 g/ deciliter versus 10.8g/deciliter with no change in mortality in a matched series of 838 patients. An NIH consensus conference concluded that a hemoglobin >10.9g/ deciliter does not require blood whereas <7.0g/deciliter benefited from blood ⁽⁸²⁾.

The requirement of blood transfusions for open-heart surgery has decreased. In 1971, the average was eight units of blood per case. In 1977, the average for CABG was 2 – 6 units, with all other open-heart procedures averaging 4 – 7 units per patient. Magovern et al ⁽⁸³⁾ in 1996 found a mean of 2.5 units of packed cells transfused per patient in a prospective group and 2.9 in a retrospective group. Intraoperative strategies to decrease blood usage include heparin-coated circuits, decreased priming volumes of the CPB circuit (1400cc Plasmalyte versus 600-800cc), hemofiltration, and cell saver ^(84,85). A difficult area has been the transfusion threshold for the hemodiluted low hemoglobin on bypass or immediately post bypass. Goodnough et al, ⁽⁸⁶⁾ proposed that on bypass hemoglobin/ hematocrit be maintained at ³ 8g/ d</ ³ 24% for high risk patients and ³ 7g/ d</ ³ 21% for low risk patients. Paone and Silverman ⁽⁸⁷⁾ ignore arbitrary "transfusion triggers." Helm and Isom ⁽⁸⁸⁾ present practical guidelines for red blood cell transfusion. (Figure 5)

Jehovah Witnesses refuse, on religious grounds, both homologous blood transfusion, and autologous blood not in continuity with the body. This includes pre-donation of autologous blood. Intraoperative autologous blood donation is feasible as long as there is continuity with the body. Cell saver autotransfusion through a continuous circuit, as well as continuous shed mediastinal auto transfusion system, are both permitted. Rosengartt combined the NYU experience with Jehovah Witnesses to develop a "bloodless" surgery protocol for all patients ^(89,90) (Figure 6). A few areas of this protocol are highlighted. Introduced in 1987, recombinant human Erythropoietin (Epoetin alfa) stimulates red cell production in the bone marrow and increase hemoglobin, as well as inducing reticulocytosis ⁽⁹¹⁾. However, D' Ambra et al ⁽⁹²⁾ showed no difference in perioperative transfusion requirements in three groups (Epoetin alfa 300 IU/Kg; 150 IU/Kg; placebo). The 5-day preoperative course was challenged by Goodnough ⁽⁹³⁾. A longer course of treatment (14 days, beginning 9 days preoperatively) was recommended. Predonated autologous blood has not proven practical in cardiac surgery patients, especially with the increasing number of non-elective cases ⁽⁹⁴⁾. Intraoperative sequestration is more practical yet is limited ⁽⁹⁵⁾. Yet, despite preservation of red cell mass there is no decrease in the incidence of post-operatively bleeding ⁽⁹⁶⁾.

Autologous platelet-rich plasma (PRP) initially showed a decrease in homologous plasma and platelet usage when used in conjunction with intraoperative auto transfusion ⁽⁹⁷⁾. A subsequent meta-analysis failed to confirm an advantage ⁽⁹⁸⁾ Khuri et al ⁽⁹⁹⁾ showed improved hemostasis when cryopreserved homologous platelet transfusion were used compared to liquid storage of platelets. An exciting area of use of PRP is the creation of autologous platelet-rich plasma from whole blood and combining it with thrombin and calcium to form a coagulum. It is then used as glue in a variety of clinical situation. www.plateletgel.net. An exciting area has been the development of the PFA-100 ® platelet analyzer to assess the quality of platelets ⁽¹⁰⁰⁾.

Autotransfusion of shed mediastinal blood remains controversial ⁽¹⁰¹⁾. The results of decreased homologous transfusion, cost-effectiveness, and safety have been raised. Recent studies from Alabama, United Kingdom (UK), and Montreal show mixed results, regarding decreased costs, decreased homologous transfusion, and risk ^(101-104).

Retrograde autologous priming of the CPB circuit has been shown to decrease transfusion after coronary artery surgery. This involves displacing circuit prime at the initiation of CPB with the patient’s blood draining both antegrade through the venous cannula and retrograde through the arterial cannula ⁽⁸⁹⁾. A UK study ⁽¹⁰⁵⁾ randomized 104 patients to autologous priming and control. 49% of the control group required blood transfusion versus 17% in the autologous priming group.

Perioperative renal dysfunction and insufficiency remains a serious source of morbidity and mortality following cardiac surgery. The incidence of renal failure in CABG patients in the 1997 STS database was 3.14%, with 0.87% of patients requiring dialysis (http://www.ctsnet.org/doc2988). In general, patients with pre-operative renal dysfunction (serum creatine >1.5mg d/L) have a higher incidence of stroke, bleeding complications, dialysis, prolonged mechanical ventilation, LOS, and death ⁽¹⁰⁶⁾. Mangano and co-workers ⁽¹⁰⁷⁾, reviewed 2,222 patients undergoing CABG and found that 7.7% had postoperative renal dysfunction and 1.4% had oliguric renal failure requiring dialysis. Mortality was 0.9% for those with no renal failure, 19% for those with renal dysfunction, and 63% for oliguric renal failure requiring dialysis. Chertow ⁽¹⁰⁸⁾ studied 43,642 VA patients undergoing CABG or valve surgery. The overall risk of acute renal failure requiring dialysis was 1.1%. The mortality in this group was 63.7%, as compared to 4.3% mortality in patients without ARF. Decreased myocardial function and advanced atherosclerosis were independent risk factors for the development of dialysis dependent renal failure. Preoperative renal function correlated with post-operative renal failure. The risk of ARF was 0.5%, 0.8%, 1.8%, and 4.9% with baseline serum creatinine concentrations of < 1mg/dL, 1.0 – 1.4 mg/dL, 1.5 – 1.9 mg/dL, and 2.0 – 2.9 mg/dL, respectively.

The focus of renal management includes preoperative assessment, intra-operative modifications, and postoperative management. Patients with recent or long- standing hypertension should undergo renal angiography at time of catheterization to assess renal artery stenosis. Critical renal artery stenosis can be treated preoperatively in hopes of improving postoperative renal function and allowing better control of postoperative hypertension. To optimize preoperative renal function contrast loads are minimized and safer contrast agents are employed. Operative considerations include limiting the duration of CPB and non-pulsatile CPB with mean arterial pressures >60 mmHg ^(109). Additional effects of CPB include trauma to the blood elements, especially erythrocytes, with increased free hemoglobin levels and micro-particle embolic insults to the kidneys. The increased release of catecholomines and inflammatory cytokines also affect renal function adversely. The resultant pathophysiological effect of these adverse stimuli causes decreased renal blood flow with an ensuing decrease in glomerular filtration rate and increase in renal vascular resistance. Hypotension and pressor agents accentuate this response. Again, during CPB mean arterial pressures are usually maintained in the 50 – 70 mmHg range. Ultrafiltration is utilized in long pump runs primarily to decrease volume overload in patients with renal dysfunction

The post-operative goal of these interventions is to maintain adequate renal perfusion pressure and a urine output in the 0.5cc / kg/ hour range. Following CPB there is usually a brisk diuresis (>200 – 300 cc/hr). Volume replacement is required to maintain a MAP > 60. There is no value in routine use of dopamine as renal protection in patients with normal pre-operative renal function ⁽¹¹⁰⁾. In fact, there is data to suggest that dopamine may exacerbate renal tubular injury, as evidenced by an increase in urinary excretion of retinal binding protein (RBP), a marker of tubular dysfunction ⁽¹¹¹⁾. Persistence of oliguria with a rising creatinine requires a more aggressive approach beyond cessation of pressors, maintenance of adequate filling pressures, and intermittent loop diuretics. Continuous furosemide (Lasix) or bumetanide infusion (BUMEX) in both stable and unstable hemodynamic patients may be useful ⁽¹¹²⁾. The continuous infusion of diuretics, including mannitol and dopamine may minimize the potential toxic effects of bolus doses of diuretics in a compromised kidney ⁽¹¹³⁾. Unusually high urine outputs following operation may be due to spontaneous diuresis following CPB, accentuated response to diuretics, improved filtration from low dose dopamine, or, rarely, SIADH. Volume replacement and the occasional use of desmopressin (4mg IV) is the usual approach.

An increasing number of patients with underlying chronic renal failure on dialysis are undergoing open-heart surgery. Several recent series show improved results in this group of patients but outcomes are still significantly worse than patients without renal dysfunction. Franga et al ⁽¹¹⁴⁾, reported on 44 dialysis patients undergoing CABG with an 11.4% operative mortality, but a 73% complication rate, and 32% 5-year actuarial survival rate. Kaplon et al, ⁽¹¹⁵⁾ studied 42 patients undergoing valve replacement with dialysis dependent renal failure. There was no perioperative mortality, but 5-year survival was only 33% for mechanical prosthesis and 27% for bio-prosthetic valve replacement. The Northern New England Cardiovascular Disease study group reported an annual death rate of 3.8% for non-renal failure CABG patients, 16.9% for all renal failure patients, and interestingly 7.7% for renal failure patients without diabetes or PVD, and 23.0% for those with diabetes and PVD ⁽¹¹⁶⁾. Cardiac surgery following renal transplantation has been reported with an operative mortality of 8.8% ⁽¹¹⁷⁾.

Patients undergoing cardiac surgery may present with weight gain and decreased oncotic pressure. The endocrine changes related to surgical stress, anesthesia, and CPB also influence volume changes. Exogenous losses and replacement completes the changes associated with cardiac surgery. The BSA (body surface area) is obtained pre-operatively on all patients and is the reference point for determination of operative fluid management as well as CPB perfusion flows (usually 2.0 liters/min/m² or 50cc/min/kg at normothermia).

The overall effects on water balance, electrolytes, and subsystem fluid balance are important. Following CPB the extra cellular fluid volume (ECFV) increases by 20-30% with an increase in both sodium retention and potassium excretion. Mean fluid accumulation ranges around 800cc/m² per hour of CPB ⁽¹¹⁸⁾. Colloid oncotic pressure (COP) is reduced by as much as 50% and returns to normal over a 2-week period ⁽¹¹⁹⁾. A perioperative weight gain of 4-6 kg is normal. Spontaneous and forced diuresis restores pre-operative weight in the average patient by the 3rd to 5th post-operative day. Weight gain, and decreased oncotic pressure following CPB has brought more emphasis on intraoperative and post-operative strategies to limit extracellular fluid accumulation, especially in compromised patients. Decreasing volume primes for CPB along with use of colloid solutions to maintain COP >20 mmHg is a reasonable goal ⁽¹²⁰⁾. Wilkes et al ⁽¹²¹⁾ showed a lower cumulative blood loss with albumin compared to hydroxyethyl starch (Hetastarch). Post-operatively, accumulation of extravascular pulmonary water has not been shown to be affected by the prime type nor the type of fluid administered post-operatively ⁽¹²²⁾. The use of colloid or crystalloid administration in fluid resuscitation provides equivalent results with colloid limiting extracellular water accumulation but at increased cost ⁽¹²³⁾. In the USA a variety of colloids for priming, resuscitation, or volume expansion are available. These include fresh frozen plasma, human serum albumin, human plasma protein fraction (containing albumin; alpha, beta, and gamma globulins), dextran (a non-protein branched polysaccharide containing D-glucose units, with preparations designated according to molecular weight e.g. dextran 40 or 70), and hydroxethyl starch.

An interesting aspect of fluid administration is the determination of fluid responsiveness. Michard et al ⁽¹²⁴⁾ reviewed 12 studies and concluded that dynamic versus static parameters are more predictive of fluid response. The stated parameters including right atrial pressure (RAP) and pulmonary artery occlusion pressure (PAOP) are not as accurate as the dynamic parameters, including the respiratory changes in RAP, arterial pressure, and aortic blood velocity. These reflect changes in preload as a reflection of changes in pleural pressure.

Calcium

Calcium ion is necessary for the excitation – contraction-coupling mechanism for myocardial muscle contraction. Changes in calcium ion concentration can occur during cessation of CPB, low cardiac output, transfusion of citrated blood, and sepsis. The concentration of calcium ion is greatest in the intracellular space with small amounts in the extracellular fluid (ECF). The level of ionized or active form of calcium is the true determinant of calcium function. Calcium levels bound to albumin change with the degree of serum albumin whereas ionized levels remain unchanged ^(125,126).

The major effect of potassium imbalance is on cardiac electrical activity. There are major shifts of potassium during cardiac surgery with CPB. Cardioplegia coupled with decreased urine output, decreased insulin levels, and RBC hemolysis all accentuates hyperkalemia ⁽¹²⁷⁾, whereas an accentuated diuresis, insulin for hyperglycemia, and alkalosis postoperatively cause hypokalemia ⁽¹²⁸⁾. Aggressive treatment of hypokalemia decreases the incidence of perioperative arrhythmias, particularly ventricular irritability. Serum potassium chloride levels and replacement protocols are an integral part of the early post-operative routine orders. In general, the serum potassium rises approximately 0.1meq/L for each 2meq kel given.

Magnesium is the second most common intracellular cation after potassium. It is involved in cardiac excitability and muscle contraction. This is accomplished secondary to its role as an ATP cofactor and calcium antagonist. In the clinical setting, especially following hemodilution and CPB, hypomagnesaemia is common (> 70%) and is associated with enhanced atrial and ventricular arrhythmias. The anti-arrhythmic effects of magnesium are related to its activity as a calcium-channel blocker, activation of ATP, and regulation of intracellular potassium. Again aggressive treatment of hypomagnesemia is recommended, especially to decrease the incidence of arrhythmias ^(129-132).

Hyponatremia in the absence of hyperlycemia is uncommon following cardiac surgery. Hypotonic or dilutional hyponatremia is a reflection of excess free water with relatively normal sodium stores ⁽¹³³⁾. Hypernatremia (Na⁺ > 150) may complicate regulation of acid-base balance and further use of Na HCO₃ to correct metabolic acidosis is contraindicated because of concerns about the CNS effects of extreme hypernatremia. In this setting, THAM is usually indicated to correct acidosis.

Acid-Base Disorders ⁽¹³⁴⁾

Respiratory acidosis is common in the sedated patient and precludes aggressive fast track weaning. Metabolic acidosis with lactate accumulation is an ominous sign in the early perioperative period. A dilutional hyperchloremic acidosis is sometimes seen in crystalloid volume replacement, especially in patients with RV failure. Low cardiac output renal dysfunction or splanchnic ischemia is the usual causes. Respiratory alkalosis and metabolic alkalosis are usually seen later as compensatory mechanisms with mixed effects, especially in response to diuretic therapy. Severe alkalemia (blood pH > 7.60) can cause arteriolar constriction at the cerebral and myocardial level.

The neurological sequelae of open-heart surgery continue to receive attention in the lay literature as well as medical journals. Barbat and Caplan have summarized nicely the neurological complications of cardiac surgery ⁽¹³⁵⁾. The incidence of neurological injury approaches 6% following open-heart surgery. Included in the STS database for 1997 are permanent focal stroke <2%, transient stroke, 1%, delirium <3%, and coma >24hours <0.5% (http://www.ctsnet.org/doc/2998). These events are usually apparent within the first few hours after operation. Later deficits are seen, especially with hemodynamic instability or atrial fibrillation. Ricotta et al ⁽¹³⁶⁾ associated carotid stenosis >50%, redo heart surgery, valve surgery, and prior stroke with increased postoperative neurological risk. Newman et al ⁽¹³⁷⁾ developed a stroke risk index. Key predictor variables were age, previous neurological disease, diabetes, vascular disease, redo CABG, unstable angina, and pulmonary disease. John et al ⁽¹³⁸⁾ reviewed, 19,224 patients in New York State. The stroke rate was 1.4% following CABG with 24.8% mortality in that group. Multivariable logistic regression identified the following predictor variables: calcified aorta, prior stroke, age, carotid artery disease, duration of CPB, renal failure, PVD, smoking, and diabetes. Intraoperative factors that may cause postoperative neurologic deficits include particulate macroembolization of air, debris, or thrombus ⁽¹³⁹⁾; microembolization of wbc’s, platelets, fibrin ⁽¹⁴⁰⁾; duration of CPB ⁽¹⁴¹⁾; cerebral hypoperfusion during non-pulsatile CPB; and hypothermic circulatory arrest ^(142). Preventive measures to limit postoperative neurological injury have evolved. The approach to concomitant carotid artery disease is controversial with no unified consensus. Akins et al ⁽¹⁴³⁾ reported 200 consecutive patients with combined carotid endarterectomy and CABG, The mortality was 3.5% and permanent stroke 3%. However, Borger et al ⁽¹⁴⁴⁾ reported a meta analysis of 844 combined and 920-staged operations. The crude mortality rate was 4.7% versus 2.9%, for the staged procedures, and the stroke rate was 6.0% versus 3.2%. The authors concluded that a staged approach to the combination of coronary artery disease and carotid occlusive disease was preferred if clinical circumstances do not mandate a combined approach (e.g. TIA’s in the presence of unstable angina). The selective use of carotid ultrasonography is crucial in identifying patients at risk for postoperative neurological deficit ⁽¹⁴⁵⁾. Kouchoukos and his group are advocates of assessing atherosclerosis of the ascending aorta utilizing intraoperative epiaortic ultrasonography scanning ⁽¹⁴⁶⁾. The single clamp technique to avoid partial occlusion of the aorta for proximal coronary anastamosis has not proven an absolute effective method of avoiding neurological injury ⁽¹⁴⁷⁾. However, a recent study utilizing the brain marker S-100 showed improved protection with the single aortic cross-clamp technique ⁽¹⁴⁸⁾. Operations following a recent stroke represent a unique challenge. Zisbrod et al ⁽¹⁴⁹⁾ reported on 15 patients undergoing surgery 2 to 28 days, (mean 12.7 ± 7.9 days) following recent neurological injury with 14/15 survivors. This can present a difficult decision, especially in patients with an embolic cerebral deficit undergoing urgent or emergency surgery for infectious valvular endocarditis ⁽¹⁵⁰⁾. Non-pulsatile CPB, mean perfusion pressures, and pH control on CPB have all been evaluated ^(151,152). The debate of decreased neurological injury by utilizing OPCAB procedures continues ⁽¹⁵³⁾. Intraoperative monitoring of neurological status has been mentioned. There has not been widespread use of transcranial doppler ultrasound to assess high-pitched signals as microemboli ⁽¹⁵⁴⁾. TEE is routinely used in valve operations to assess intracardiac air. Post-operative assessment of neurological injury is difficult. Floyd et al ⁽¹⁵⁵⁾ has nicely summarized a rationale approach (Figure 7). An exciting area is the use of biochemical markers. S100 b is a calcium-binding protein in the brain. It has been suggested as a marker of brain damage following CPB ^(156-158) . Neuropsychiatric deficits are very common, occurring in up to 50% - 70% of patients following operation ⁽¹⁵⁹⁾. The cause of these neuro-psychiatric disorders is uncertain and remains controversial. Newman and the Duke group ⁽¹⁶⁰⁾, reported a series of 261 patients with cognitive decline of 53% at discharge, 36% at six weeks, 24% at six months, and 42% at five years. Roach et al, ⁽¹⁶¹⁾ associated a higher risk of neuro-psychiatric decline in older patients with systolic hypertension, pulmonary disease, and excessive consumption of alcohol. van Dijk et al ⁽¹⁶²⁾ reviewed twelve cohort studies. A pooled analysis of six comparable studies yielded 22.5% of patients with a cognitive deficit at 2 months following surgery. The debate will continue regarding neurological sequellae following CABG utilizing on pump (ONCAB) or off pump (OPCAB) techniques.

Seizures are unusual following cardiac surgery. Pre-operative seizure disorders are usually under therapeutic control. Anesthesia is also an anti-seizure protective mechanism. However, perioperatively residual muscle relaxants causing tremors, and chills or shivering can mimic seizures. Electrolyte and glucose imbalance are occasional causes. An ischemic or hemorrhagic CNS event is a rare cause of seizures and, if postoperative seizures are due to these, this is grave prognostic event. On the other hand, alcohol withdrawal or full-blown delirium tremens is a serious but unusual perioperative event. The mortality, historically, in all surgical patients is as high as 20% ⁽¹⁶³⁾. The usual approach is a high index of suspicion in the high-risk patient, followed by perioperative sedation, thiamine, and multivitamins. Rarely is IV alcohol employed. Paraplegia is discussed with aortic surgery.

Phrenic Nerve Injury (PNI)

The major effect of phrenic nerve injury is diaphragmatic paralysis or dysfunction ⁽¹⁶⁴⁾. The causes of injury include local cold injury from pericardial ice saline/slush, cautery damage to the nerve during harvest of the internal mammary artery (IMA), and mobilizations of the pericardium during redo operations. With a decrease of local cold application the incidence of PNI has decreased from 24-73% to 2-17%. IMA harvesting remains a major cause, secondary to retraction, transection, severance, cautery injury, or devascularization of the nerve ⁽¹⁶⁵⁾. The sequellae of PNI is unpredictable. In patients with COPD, the morbidity is higher in terms of longer mechanical ventilation and hospital stay ⁽¹⁶⁶⁾. In many patients the diaphragm will normalize on chest roentgenogram at 1 year ⁽¹⁶⁷⁾.

Brachial Plexus Injury/Peripheral Nerve Injury

Brachial plexus injury is seen less due to the understanding that excessive retraction of the sternum during median sternotomy is the usual etiology, secondary to first rib impingement on the lower brachial plexus ^{(168, 169)}. Again IMA harvesting can cause this injury as well ⁽¹⁷⁰⁾. Ulnar nerve injury results from compression due to malpositioning of the upper extremity ⁽¹⁷¹⁾. Injury to the common peroneal nerve is related to stretch or compression at the level of the fibula head of the lower extremity resulting in palsy or plegia of dorsiflexion and eversion of the foot ⁽¹⁷²⁾.

Muscular weakness and atrophy, though uncommon in cardiac surgery patients, is seen in the chronic critical care patient ^(173,174). The major problem in this group is delayed mechanical ventilatory weaning and the causes are multifactorial with malnutrition, electrolyte disturbances, and cardiac cachexia/CHF being the leading contributing factors.

The incidence of infection in cardiac surgery is low (Figure 8). The major considerations in the infectious subsystem include: prophylaxis, nosocomial infection, the spectrum of wound infections, evaluation of fever, and sepsis/septic shock state. Prophylactic antibiotics in cardiac surgery are an established concept. Starr’s group ⁽¹⁷⁵⁾ reported a 20% incidence of staphylococcal prosthetic valve endocarditis in the pre-prophylactic era. Kreter et al ⁽¹⁷⁶⁾ performed a 30-year meta analysis of prophylactic antibiotics. There was consistent benefit with a significant decrease in wound infection rate associated with use of prophylactic antibiotics. Yet administration beyond 48 hours was of no benefit. Classen et al, ⁽¹⁷⁷⁾ confirmed that the pre-operative administration within 2 hours of the surgical incision resulted in significant reduction in wound infections. The Medical Letter summarizes the likely pathogens as staphylococcus, epidermidis, staphylococcus, Corynebacterium, and enteric gram-negative bacilli ⁽¹⁷⁸⁾. Recommended drugs include cefazolin or cefuroxime 1-2 grams IV. This dose is usually given within 30 minutes of the incision and continued for 24 hours at 8-hour intervals. Operations lasting linger than 4 hours require re-dosing of the antibiotic dose. Vancomycin, 1 gm IV (or 20 mg, per Kg) is recommended for valve procedures and in situations in which methicillin-resistant Staph·aureus (MRSA) and Staph·epidermidis are a frequent cause of post-operative wounds infection, or for patients allergic to penicillins or cephalosporins. Consultation with hospital infection control may be appropriate where resistant organisms are of concern. Vancomycin has been occasionally associated with vasodilatation in the perioperative period ⁽¹⁷⁹⁾. An interesting finding in nasal carriers of pathogens has prompted a recommendation for nasal topical treatment. Perl et al ⁽¹⁸⁰⁾ found intranasal mupirocin prevents staphylococcus aureus infections in patients with nasal carriage of this pathogen.

Cardiac surgery in the HIV patient is controversial. There is no conclusive data on the effect of CPB in the immunocompromised HIV positive, or AIDS patient ⁽¹⁸¹⁾. In asymptomatic HIV positive patients with CD4 counts >200, cardiac surgery should not be denied ⁽¹⁸²⁾ (The Center for Disease Control [CDC] stages HIV by CD4 cell counts: Stage I > 500, Stage II 200 to 499, and Stage III < 200 with CD4 counts < 200 defining AIDS regardless of presence of infection or malignancy) ⁽¹⁸²⁾.

Acute-phase proteins as systemic responses to inflammation have received recent attention especially C₃, C₄, and C-reactive protein ⁽¹⁸³⁾. C-reactive protein (CRP) is a nonspecific marker of inflammation. There is evidence to suggest that atherosclerosis and degenerative aortic valvular stenosis are associated with chronic inflammatory processes ⁽¹⁸⁴⁾. Fransen et al, ⁽¹⁸⁵⁾ reported an interesting group of 593 cardiac surgery patients, eighty-seven of whom had a perioperative infection. In the group with a perioperative infection CRP levels were significantly elevated compared to those without any source of infection.

The risks of toxic or allergic reaction, emergence of resistant bacteria (especially methicillin-resistant staphylococcus aureus [MRSA] and Vancomycin resistant enterococcus [VRE]), and super-infection (especially with clostridium difficile) are still low in the cardiac surgery population. Yet MRSA is becoming more of a problem in some cardiac surgery units. The Montreal Heart Institute reported a series of 39 MRSA cardiac surgery infections in 13,199 patients from 1992 to 2000 ⁽¹⁸⁶⁾. Implementation of preventive measures, including nasal screening, preventive isolation, application of mupirocin, and prophylaxis with Vancomycin and alcohol gels, decreased the incidence to 0 % in 2000. However, in a setting of high prevalence of MRSA, Vancomycin prophylaxis does not guarantee total control of surgical infections. The Israel group ⁽¹⁸⁷⁾ showed a higher prevalence of surgical site infections secondary to methicillin – susceptible staphylococcus organisms with Vancomycin prophylaxis, as opposed to more frequent infections with β lactom-resistant organisms in those receiving cefazolin prophylaxis. In addition, patients undergoing cardiac surgery are still susceptible to infection secondary to poor nutrition or debilitation, diabetes, invasive devices (urinary catheters, indwelling vascular catheters, chest tubes, endotracheal tubes), immunological suppression from anesthesia, cardiopulmonary bypass, and immunosuppressive drugs, especially in the transplant group. Recently, Chelemer et al ⁽¹⁸⁸⁾ reported an association of bacterial infection and red blood cell transfusion. A prospective cohort study of 533 CABG patients over a 7-month period yielded a 14.1% bacterial infection rate. The adjusted rates were 4.8% for no transfusion, 15.2% for one to two units, 22.1% for three to five units, and 29.0% for greater than six units.

Hospital acquired infections (nosocomial) can come from the surgical wound, the lung, urinary tract, device related infections, or the GI tract. The overuse or inappropriate use of antibiotics has added to the emergence and increase of nosocomial infections, as well as the emergence of antibiotic resistant organisms, and further super infection with less pathogenic organisms ⁽¹⁸⁹⁾. The impact of nosocomial infection is enormous with resulting prolonged LOS, multiorgan dysfunction (MOD), and increased hospital mortality. Kollef et al ⁽¹⁹⁰⁾, studied 605 consecutive cardiac surgery patients. 21.7% acquired at least one nosocomial infection. Risk factors included prolonged mechanical ventilation, prolonged antibiotics, urinary catheter, and female gender. Mortality was increased in patients with nosocomial infections, 11.5% versus 3.2%. MOD was the most significant determinant for hospital mortality, with pneumonia, sepsis, female gender, CPB and severity of illness (Apache III Criteria) as the major risk factors for MOD.

The incidence of wound infection in cardiac surgery is low. Cardiac surgery is classified as non-contaminated clean surgery. The STS database reports an incidence of 0.02% to 2.45% in the 1997 survey of 161,018 surviving CABG patients (Figure 8). The incidence in redo patients was not significantly higher. Most cardiac centers have an infection surveillance committee to systematically track and document infections. Guidelines are published by the Center for Disease Control (CDC) for nosocomial infections for postoperative pneumonia, and for surgical wound infections ⁽¹⁹¹⁾. Surgical site infections (SSI) include incisional SSI’s and organ/space SSI’s (ie. postoperative infections of body cavities or organs manipulated by surgeons) ⁽¹⁹²⁾. A number of schemes have been published for predicting wound infections. The National Nosocomial Infection Survey (NNIS) risk index is a popular one ⁽¹⁹³⁾. Flaherty et al ⁽¹⁹⁴⁾ has nicely summarized the topic of infection surveillance and control measures, ICU design, and infection control policies and procedures.

Surgical wounds are discussed further in with the wound subsystem section. Pneumonia is primarily seen in the prolonged mechanical ventilator patient. Hospital acquired pneumonia (HAP) is second only to UTI as a cause of nosocomial infection but carries the highest mortality rate ⁽¹⁹⁵⁾. However, an increasing number of patients have microorganism colonization prior to surgery. This is especially true for smokers, and patients with COPD. Carrel et al ⁽¹⁹⁶⁾, assessed tracheal aspirates prior to intubation in 500 patients undergoing CABG. Ninety-one had a positive gram stain. The incidence of pneumonia was 15.3% in that group versus 3.6% in the group with a negative gram stain. Prolonged prophylaxis did not decrease the incidence in either group.

Catheter related infections (i.e., bladder and vascular related) are not infrequent in the critical care unit. A consensus group has recently published guidelines for management of intra-vascular catheter-related infection ⁽¹⁹⁷⁾. Gordon et al, ⁽¹⁹⁸⁾, at the Cleveland Clinic reported 40,207 patients over 10 years admitted to the cardiothoracic ICU. The mean crude infection rate was 0.01/ 1,000-patient-care days and rose during the study period. Staphylococcus aureus (12%), Coagulase-negative staphylococci (11%), Candida albicans (11%), Pseudomonas aeruginosa (10%), and Enterococci (9%) were the major pathogens. Sources included primary bloodstream infection (BSI) (33%), intravascular devices (27%), lower respiratory tract infections (17%), and surgical wound infections (12%). The crude mortality rate among 246 patients with BSIs from January 1993 to September 1995 was 37%.

Fever is a common problem in the critical care unit and following cardiac surgery. All fever is not infection. Clark et al, ⁽¹⁹⁹⁾ outlined nicely the pathogenesis of fever. Stimulated leukocytes release endogenous pyrogen, which stimulates the hypothalamus, with local prostaglandin (PGE₂) release and increased c-AMP. The net effect is to increase the thermoregulatory set point. This mechanism is the most common cause of postoperative fever. This in turn causes posturing, shivering and vasoconstriction. Noninfectious causes of fever relative to cardiac surgery include myocardial infarction, post-pericardiotomy syndrome, atelectasis, and drug fever. Infectious causes include wound infection, UTI, pneumonia, catheter sepsis, and loculated areas of contaminated blood accumulation (e.g. pericardial, pleural, retroperitoneal, leg wound spaces). O’Grady et al, ⁽²⁰⁰⁾, collated the opinions of 13 experts in critical care regarding the cause of postoperative fever. Utilizing evidence based data a series of recommendations were made regarding diagnosis and treatment. Acceptable means of assessment of temperature included intravascular, bladder, oral cavity, rectum, external auditory canal, but not axillary. A temperature > 38.3° C (100.9º F) was a reasonable trigger for clinical assessment, but not laboratory or radiologic evaluation. Additionally, clinical assessment was recommended for the euthermic or hypothermic patient where there was a suspicion of infection. Appropriate treatment is based on the accurate determination of fever and its causes, but assuming that fever is caused by a non-infectious etiology is potentially hazardous in cardiac surgical patients because of the dire consequences of missing a treatable infection. Fever is a poor marker of postoperative bacteremia. Kohman et al ⁽²⁰¹⁾ noted a 3.2% incidence of fever in a retrospective study of 835 febrile CABG patients. Despite this, an aggressive approach is warranted. The role of blood cultures in the evaluation of fever is not standardized. Badillo et al ⁽²⁰²⁾ outlines the role of blood cultures in the febrile postoperative patient. The yield of true positive bacteremia ranges from 4.0% to 5% with a contamination rate ranging from 32% to 46.9%.

Sepsis/Septic Shock

The incidence of sepsis and septic shock in cardiac surgery is low and difficult to document in the literature. Early recognition and treatment is critical to improve outcome in the septic patient, particularly following cardiac surgery. The basic pathophysiology of sepsis includes systemic inflammation, coagulation changes, impaired fibrinolysis, and subsequent target organ failure with overall multiorgan failure, irreversible shock, and death ⁽²⁰³⁾. The sepsis syndrome of altered organ function in response to infection occurs in over 500,000 cases/year in the USA with a mortality of 20-50% ⁽²⁰⁴⁾. The death rate from sepsis is approximately 7.7 per 100,000 population ⁽²⁰⁴⁾. Activation of the coagulation system and inhibition of the fibrinolytic system are central to the sepsis process. Disseminated intravascular coagulation (DIC) is the ultimate result in the unrecognized, untreated patient. Holmes et al, ⁽²⁰⁵⁾ reviewed the role of vasopressin, both a vasopressor and antidiurectic hormone, to support the vasodilated patient with a suggestion that this treatment may be helpful in certain situations. Bernard et al ⁽²⁰⁶⁾ for the PROW/ESS study group showed a distinct survival advantage in treatment of severe sepsis utilizing drotrecogin alfa (activated) or recombinant human activated protein C. The mechanism of action is a modulation of the systemic inflammatory, procoagulant, and fibrinolytic reaction to infection. In a randomized study of 1690 patients the mortality rate was 30.8% in the placebo group, versus 24.7% in the treatment group.

Cardiac surgery is associated with 3 traumatic events: the stress of surgical trauma, the effect of anesthesia, and the pathophysiology of cardiopulmonary bypass. The metabolic response to surgery and critical illness has been recently reviewed ⁽²⁰⁷⁾. The stress of the operation illicits an increase in circulatory levels of norepinephrine and epinephrine with no change in the urinary clearance of these catecholomines. Other hormones with increased expression following CPB include glucagon, ACTH cortisol, ADH, renin, aldosterone, and growth hormone. Insulin and testosterone levels decrease. The increased cortisol causes increased blood glucose, and protein breakdown, (catabolism), salt and water retention, and hypokalemia. Clearly, the hormonal response is directly proportioned to the degree of stress ⁽²⁰⁸⁾. Anesthetic drugs modulate the stress response, whether by epidural, IV or inhalation methods. The increased binding of drugs changes the pharmacokinetics of drugs (i.e. absorption, distribution, clearance, and excretion), and alters the immune response after CPB.

Obesity is an uncertain risk factor for cardiac surgery, despite the fact that the overall risk of death in the general population from morbid obesity is high ⁽²⁰⁹⁾. Body mass index (BMI) (the weight in kilograms divided by the square of the height in meters) is a useful method of measuring obesity. Patients at the extremes of BMI are at increased risk in certain studies of outcome from cardiac operations. Engelman et al, ⁽²¹⁰⁾ studied 5168 patients looking at BMI and albumin levels. A BMI < 20 and albumin levels < 2.5 g / d L were independent risk factors for mortality. A BMI > 30 was associated with increased sternal wound and saphenous vein harvest site infection. Recently, Kuduvalli et al ⁽²¹¹⁾ showed an increase in mortality at four years for obese patients (BMI > 30) following CABG. The association of age and obesity in a general population cohort revealed higher mortality with increased BMI up to age 75 ⁽²¹²⁾. Yet, Brandt et al ⁽²¹³⁾, have shown no increased morbidity and mortality in a series of 500 consecutive patients with a BMI ≥ 30.0 Kg/m².

Diabetes Mellitus ^(214-216)

Type I diabetes is a result of diminished insulin production, and is all insulin dependent. Type II diabetes is a result of insulin resistance. Type II diabetes affects 8.4% of US men and 7.7% of US women. Over 10 million Americans are diagnosed with diabetes, with another 5 million estimated to have undiagnosed diabetes. 90% of all diabetics are type 2. Over 65% of type 2 diabetics die of heart disease. Insulin-dependent diabetes mellitus is associated with long-term complications, especially retinopathy, nephropathy, neuropathy, and cardiovascular disease. Diabetics tend to have silent ischemia, smaller coronary vessels, diffuse coronary and peripheral vascular disease and decreased LV function. Complications, including wound infection are higher in diabetics. Interestingly, the results of CABG are superior to PTCA in the diabetic population, despite increased surgical morbidity ⁽²¹⁷⁾.

The incidence of diabetes ranges from 10 to 20% in the cardiac surgery population. During CPB the increase in growth hormone, catecholomines, and cortisol cause increased glucose mobilization. There is a decrease in insulin secretion as well. The subsequent hyperglycemia usually persists for 24 hours following surgery. A sliding scale insulin regimen, and in, many cases, an insulin drip at 3-4 units/hour suffices to keep the blood glucose level in the 120-180-mg/dL level. There is recent data to suggest that maintaining blood glucose under 200 is associated with a decreased wound infection rate and improved outcome following cardiac procedures ⁽²¹⁸⁾. Diabetics are at a higher risk for other perioperative nosocomial infections, particularly pneumonia and urinary tract infection. Recently, a more aggressive approach to maintain blood glucose at or below 110 mg/dL in critically ill patients resulted in a reduction of mortality from 8.0% to 4.6% at 12 months ⁽²¹⁹⁾.

The thyroid hormone axis has received increased attention in cardiac surgery. T₃, T₄, and TSH levels decline following CPB ^(220,221). Complicated patients show persistent decline following surgery. There is evidence to suggest that pre-operative thyroid function should be determined in patients on thyroxin therapy in order to better manage problems that may develop in the postoperative period ⁽²²²⁾. The role of thyroid hormone replacement in post-operative low cardiac output patients remains unclear ⁽²²³⁾. The Duke T₃ study group ⁽²²⁴⁾ studied 211 CABG patients. Administration of IV T₃ following CPB had no major effect on hemodynamic performance. Mullis-Jansson et al, ⁽²²⁵⁾, however showed an improvement in LV performance, in a similar study of 170 patients. Yet, there is no present clear-cut data to support routine thyroid hormone supplementation in the postoperative cardiac surgical patient.

Surgical stress activates the hypothalamic-pituitary-adrenal (HPA) axis and increases plasma ACTH and cortisol levels. Acute adrenal insufficiency following cardiac surgery is a rare but potentially catastrophic event ⁽²²⁶⁾. Adrenal hemorrhage is the most common cause. This is a rare event, occurring in of 5 of 4,364 adult CABG patients in one older retrospective study ⁽²²⁷⁾. Other adrenal considerations include glucocorticoid coverage in patients on exogenous steroid treatment, with suppressed HPA ⁽²²⁸⁾.

Free radicals play an important role in the response to stress of all kinds especially ischemia and the changes caused by CPB. The systemic oxidative stress response, mediated by free radicals and ameliorated by free radical scavengers, is important in the genesis and manifestation of critical illness ⁽²²⁹⁾. This is a state where toxic reactive oxygen intermediates overcome endogenous antioxidant defenses. Antioxidant therapies have been advocated to reduce the harmful effects of the oxidative stress response. These include N-acetyl/cysteine, selenium, vitamins E and C, superoxide desmutase, catalase, lazaroids, and allopurinol.

Natriuretic Peptides

Atrial natriuretic peptide (ANP) and B-type natriuretic peptide (BNP) are naturally occurring diuretic peptides that act locally as hormones and markers of stress. These peptides are secreted from the heart and brain in varying quantities. They maintain homeostasis of fluid balances and blood pressure through natriuresis, vasodilatation, and inhibition of renin and aldosterone secretion ^(230,231). Since ANP is a peptide, it can be synthesized using pharmaco-genetic engineering. Commercially prepared ANP has been used to induce diuresis in patients who are recalcitrant to the usual diuretic regimens. This is a costly, mostly unconfirmed, therapy that may prove useful with further controlled studies. BNP is an excellent marker of endstage cardiac disease and may prove to be a most useful indicator of the presence and exacerbations of CHF ⁽²³²⁾.

The incidence of GI complications following cardiac surgery range from 0.6% to 2% with an overall mortality ranging from 14.8% to 72% (Figure 9). Non-pulsatile CPB causes transient decrease in hepatic, gastric, and splanchnic blood flow. There is a subsequent decrease in gastric pH and increased GI permeability. In the usual setting the GI consequences are minimal. Yet the combination of decreased GI organ perfusion with prolonged periods of CPB, low cardiac output, excessive blood loss and atheroembolization all accentuate GI damage. Acute mesenteric ischemia with subsequent intestinal infarction is a devastating complication that accounts for as much as 25% of all GI complications following cardiac surgical procedures. The resultant mortality rate for acute mesenteric ischemia is between 60% and 100% ^(241,242). Early surgical intervention (<6 hours) is associated with 48% mortality, and this rises to 99% with delays in surgical intervention (> 6 hours). Clearly gut mucosol hypoperfusion and bacterial translocation is the common etiology. Additionally, Ghosh et al ⁽²⁴³⁾, identified by logistic multivariate analysis in 39 of 5349 patients undergoing cardiac surgery utilizing CPB, six predictors of intestinal ischemia: duration of cross clamp, use of inotropic support, IAB, blood transfusions, triple vessel CAD, and peripheral vascular disease.

Given the high incidence of UGI bleeding in the series shown in Figure 9, prophylaxis for gastrointestinal bleeding from stress ulcers has been advocated in the critical care setting. This is especially true for patients requiring mechanical ventilation. H₂ receptor antagonists are the most common agents used. However, a higher gastric pH has been associated with an increased incidence of nosocomial pneumonia ⁽²⁴⁴⁾. The cytoprotective agent sucralfate does not alter gastric pH and is associated with a lower incidence of nosocomial pneumonia. In a multicenter randomized study of 1200 patients, H₂ receptor antagonist ranitidine (Zantac) was compared with sucralfate ⁽²⁴⁵⁾. The incidence of GI bleeding was 1.7% for ranitidine and 3.8% for sucralfate. There was no statistical difference in nosocomial pneumonia (19.1% vs. 16.2%). The association between upper GI bleeding, helicobacter pylori infection and anticoagulant is very high in heart valve surgery patients ^(246,247). Empirical treatment for H·pylori may be indicated in certain heart valve patients following operation, especially when warfarin is indicated because of cardiac valve considerations. There are at least four treatment regimens over 10-14 day courses that are available ⁽²⁴⁸⁾.

The etiology of pancreatitis following CPB remains unknown. The postulated causes include ischemia, and drugs implicated in the induction of pancreatitis include steroids, phenylephrine, norepinephrine, narcotics, and calcium chloride. In a prospective study of 300 patients a multivariate analysis showed that pancreatic cellular injury was associated with preoperative renal insufficiency, valve surgery, post-operative hypotension, and the administration of more than 800mg of calcium chloride per square meter of body-surface area ⁽²⁴⁹⁾. Again in a retrospective study of 5621 patients undergoing CPB the overall mortality rate was 44% in a group of 15 (0.44%) sustaining pancreatic complications ⁽²⁵⁰⁾. No specific etiological factor was noted.

Acute acalculous accounts for 2-15% of acute cholecystitis. Ultrasound, CT, and HIDA scanning are of significant value in the diagnosis. An aggressive surgical approach is advocated ⁽²⁵¹⁾.

In general, pre-operative liver dysfunction or failure is a significant risk factor for cardiac surgery and advanced cirrhosis may preclude cardiac surgical interventions. Klemperer et al, ⁽²⁵²⁾ , reported on 13 patients with non-cardiac cirrhosis undergoing cardiac surgery. Two of 8 patients with Child Class A cirrhosis had major morbidity and none had mortality, whereas 5 of 5 (100%) with child class B cirrhosis had morbidity and 4 of 5 (80%) died. Bizouain et al, ⁽²⁵³⁾, reported 10 patients with Child-Pugh Class A and 2 with Class B of which 3 of 12 died.

Over 15% of hospitalized patients experience feeding problems and potential or actual pulmonary aspiration (swallowing dysfunction (SD) or oropharyngeal dysphasia) with Ferraris et al ⁽¹⁶⁾, reporting an incidence of 3% of 1,042 patients over a two-year period (see Respiratory Subsystem). This is particularly true in long-term orotracheal intubation patients. Hogue et al, ⁽²⁵⁴⁾, reported an incidence of 4% of 869 patients undergoing cardiac surgery over a one-year period. The duration of tracheal intubation (>5 days) and the perioperative use of TEE were associated risk factors for SD.

A not uncommon problem is paralytic leus. The differentiation from partial and total small or large bowel obstruction may be difficult. Acute colonic psuedo-obstruction (Ogilvie’s syndrome) is a rare event, usually seen with long-term mechanical ventilation ⁽²⁵⁵⁾. Diarrhea and Clostriduim difficile are a difficult problem in the immunosuppressed patient. The incidence on a general surgery service approaches 2% ⁽²⁵⁶⁾. Evaluation including stool toxin assay is followed with oral or enteral metronidazole. Dallal et al ⁽²⁵⁷⁾ evaluated 2,334 patients with C. difficile colitis over an 11-year period. The incidence increased from 0.68% to 1.2% with life threatening symptoms rising from 1.6% to 3.2%. 64 patients died. Of 44 patients requiring colectomy, 57% died.

The major vascular concerns relative to cardiac surgery are the added risk related to the presence or adequacy of the greater saphenous veins, presence of peripheral vascular disease, cerebrovascular disease, and distal extremity viability, especially during vein harvesting, and the use of IAB’s. The documentation of ABI’s (Ankle Brachial Index= ankle Doppler systolic/brachial Doppler systolic pressures) is essential. An ABI <0.8 indicates the presence of lower extremity occlusive arterial disease.

The traditional incision for cardiac surgery is the median sternotomy. Originally described by Milton in 1897, it was reintroduced by Julian in 1957 ⁽²⁵⁸⁾. It gained favor over the right or left thoracotomy approaches, as well as the bilateral anterior thoracotomy. Recent minimally invasive procedures include; hemisternotomy, parasternotomy, and limited anterior thoracotomy approaches.

The median sternotomy provides ready access to the entire heart, and ascending aorta. It is the most common approach for open-heart surgery. It is well tolerated, relatively easy to perform, and has low incidence of infection, or technical problems. However high risk, complicated operations with persistent bleeding and hemodynamic instability may preclude primary sternal closure. Hemodynamic comprise secondary to increased tissue edema is the usual reason. Delayed sternal closure allows hemodynamic stabilization, and diuresis ^(259,260).

Delayed wound closure or delayed primary closure (DPC) is not new concept. It was initially applied to dirty abdominal wounds where the skin and subcutaneous tissue were left open and closed 3-6 days following initial fascial closure. Delayed closure of the entire abdominal cavity was a more recent application of this concept. The two factors involved include; (1) prevention or reduced incidence of deep and superficial inflection; and (2) the inability to close the abdominal cavity secondary to hemodynamic instability with massive tissue and organ edema. Anderson et al ⁽²⁶¹⁾ outlined nicely the recent Brigham and Women’s Hospital experience. Open Chest Management occurred in 1.7% (87 of 5,177) of cardiac procedures. Hospital survival was 76%. Major complications included deep sternal infection (n=4); CVA (n=8); and dialysis (n=13). Multivariate analysis revealed ventricular assistance and reoperation for bleeding as independent predictors of in-hospital mortality.

Superficial and deep sternal wound surgical site (SSI) infections are significant complications of cardiac surgery. Deep sternal infection with associated mediastinitis occurs in 1-2% of cardiac operations with a resultant mortality approaching 10% ⁽²⁶²⁾. There is a large body of literature related to recognition and treatment, particularly complex wound reconstruction techniques. Borger et al, ⁽²⁶³⁾ reported the Toronto experience of deep sternal wound infections. In 12,267 patients from 1990 to 1995 there was an incidence of 0.75%. Risk factors included male sex and diabetes. In those undergoing CABG, bilateral IMA grafting, diabetes and male sex were the major risk factors. Baskett et al, ⁽²⁶⁴⁾ , reported an incidence of 0.25% in a series of 9,771 patients from 1987 to 1997. Mortality was 8.3%. Trick et al, ⁽²⁶⁵⁾ noted antibiotic prophylaxis > 2 hours prior to surgery, preoperative blood glucose > 200 mg/dL, and sternal wound staple closure with a normal body mass index were risk factors for deep sternal site infection. Braxton et al, ⁽²⁶⁶⁾ noted a one-year survival of 78% with mediastinitis and 95% without. Olsen et al ⁽²⁶⁷⁾ reviewed the Barnes-Jewish Hospital, St. Louis experience. An attempt was made to develop predictive models for deep and superficial chest surgical-site infections following CABG operations. The incidence of deep chest infection was 1.9% and 2.3% for superficial infection. Multivariate analysis yielded obesity, diabetes transfusion of 4 or more units of blood, and use of IAB as independent risk factors for deep infection. The predictors for superficial infection included obesity, current smoking, and use of pre-operative anti-platelet drugs. Francel and Kouchoukos ^(268,269) reports a practical and rational approach to wound problems after sternotomy. One hundred fifty one patients were treated over a 6-year period (a 3% infection rate). A practical algorithm for the total spectrum of wound difficulties is outlined. The recent use of vacuum-assisted closure (VAC) therapy has shown promise for deep sternal wound infections ^(270,271). The rational is that vacuum accelerates micro circulation and angiogenesis to the wound bed.

A major source of morbidity following CABG has been leg problems related to donor saphenous vein harvesting ⁽²⁷²⁾. The continuous or interrupted skin incision technique remains the usual method of vein procurement. Wound infection, phlebitis, leg edema, cutaneous nerve damage, and early decreased leg mobility remain the major complications, and cause for early hospital readmission. Utley et al, ⁽²⁷³⁾ found impaired wound healing in 24.3% of patients. Wound healing was defined as inflammation, separation, cellulitis, lymphangitis, drainage, necrosis, or abscess requiring surgical control. A variety of ingenious endoscopic vein harvesting techniques utilizing small incisions and CO₂ insufflation have emerged with encouraging early results ^(274,275). Endothelial and smooth muscle function of harvested veins has shown to be similar in both the open and endoscopic techniques ⁽²⁷⁶⁾.

Preoperative malnutrition i.e. debilitation or cachexia (i.e. > 10% weight loss over 6 months) is a risk factor for post-operative complications. However, in many situations, pre-operative hyper alimentation is not feasible in many patients undergoing cardiac surgery, particularly urgent and emergency operations. The role of preoperative nutritional support of cardiac surgical patients is unclear ⁽²⁷⁷⁾. In general, the majority of cardiac surgical patients do not require extensive nutritional support perioperatively. In elderly patients an albumin level of < 3.5 g/dL is associated with increased risk of infection and poor outcome ⁽²⁷⁸⁾. An optimal nutritional indicator is body mass index (BMI)(body weight in kilograms divided by the square of the height in meters). In general, a BMI <15Kg/m² is associated with an increase in morbidity ⁽²⁷⁹⁾. Post-operative patients have accelerated catabolic protein loss, usually requiring 25-40 Kcal.kg/day. In those patients requiring prolonged post-operative mechanical ventilation the trend is toward earlier tracheostomy (i.e. at 10-14 days) and enteral hyper-alimentation, preferably introduced distal to the pylorus via a duo tube or percutaneous endoscopic gastrostomy (PEG). Little data is available on the outcome of these patients, since many are transferred to independent chronic care facilities with mechanical ventilator capability. The role of the sophisticated area of nutritional pharmacology in complex post-operative cardiac surgery patients is unknown. The role of arginine, glutamine, and n-3 fatty acids in nutritional support offers exciting expectations ⁽²⁸⁰⁾.

Figure One

Parameters Monitored Core temperature, Paco_2, patient alertness

• O₂ SAT ³ 92% or Pao₂ > with FIO₂ < 0.5 and PEEP < 5 cm H₂O
• Core temperature ³ 36.0° C
• End-tidal CO₂ < 45

Therapeutic Response

• When patient is alert, hemodynamically stable, normothermic with critical respiratory values mean with T-piece adapter assessing arterial blood gases q 30-60 min
• Stop weaning if any of following appear:

1. Pao₂ < 60 mm Hg
2. Paco₂ > 55 mm Hg
3. pH < 7.30
4. Respiratory rate > 30
5. 20-beat/min rise in heart rate
6. 20-mm Hg rise in MAP

________

* Adapted from Morris, DC, St. Clair, D. Management of Patients After Cardiac Surgery. Current Problems in Cardiology April 1999; 208-209

Reproduced with permission

• Cold agglutinins ⁽³⁰⁾ Plasma exchange;

• Liver dysfunction Vitamin K; FFP
• Sickle Cell disease ^{(31)(32)(33)(34))} Decrease hemoglobins (Hbs) to < 30%; intraoperative exchange transfusion
• Hemophilia B (Christmas disease) ⁽³⁵⁾ Factor IX concentrate
• Hemophilia A ⁽³⁶⁾ Factor VIII concentrate
• Protein C deficiency ⁽³⁷⁾ ACT > 300; avoid Amicar
• Pernicious Anemia ⁽³⁸⁾ Erythropoietin
• Leukemia ⁽³⁹⁾ No supplemental Therapy
• Factor XII deficiency (Hageman factor) ⁽⁴⁰⁾ Exogenous Factor XII administration
• AT III deficiency ⁽⁴¹⁾ AT III supplement
• Methemoglobin ⁽⁴²⁾ Methylene Blue
• ITP ⁽⁴³⁾ Immunoglobulin; platelet transfusion
• Von Willebrand’s disease (¯ Factor VIII) ⁽⁴⁴⁾ Admin. Factor VIII to 50- 100%
• Acquired Angioedema (¯ C₁-INH) ⁽⁴⁵⁾ Tranexamic Acid
• Werlhof’s disease ⁽⁴⁶⁾ (thrombocytopenia) Pre-op immunoglobulin
• Factor VII deficiency ⁽⁴⁷⁾ Admin Factor VII to 10-15%
• Polycythemia vera ⁽⁴⁸⁾ Phelebotomy; hydroxyurea

Figure Three

* From Frangos SG, Chen AH, Sumpio B. Vascular Drugs in the New Millennium. J Am

Coll Surg 2000; 191: 76-92

Reproduced with permission

­ aPTT, PT, ACT Heparin excess Protamine

­ PT (>1.2 control), ¯ Fibrinogen Factory deficiency Fresh frozen plasma

­ FDP, ­ D-dimers, TEG evidence of acid, aprotinin (antifibrinolytic

________

*From Hartstein G, Janssen M. Treatment of Excessive Mediastinal Bleeding After

Cardiopulmonary Bypass. Ann Thorac Surg 1996; 62: 1951-1954

Reproduced with permission

Figure Five

Guidelines for red blood cell transfusion ⁽⁸⁸⁾*

Perioperative Factors Recommended Adjustment of Adjustment of

time point affecting tolerance transfusion trigger (hematocrit %) trigger upward trigger downward

____________

* Adapted from Helm RE, Isom OW. Chapter-16Indications for red cell transfusion. In: Kreiger KH, Isom WO, ed. Blood Conservation in Cardiac Surgery. Springer, New York, 1998; p. 427

Figure Six

Figure Six (con’t)

Technical Blood Conservation Measures Incorporated in Comprehensive Multimodality Blood Conservation Program ^(90)*

Principle Technique

Minimize autologous Intraoperative Autologous donation

blood loss Cell salvage device used "skin to skin"

Reclamation of residual pump-circuit

Minimize Minimize crystalloid administration

Hemodilution Retrograde Autologous priming

Small-volume oxygenator and circuitry

Optimize coagulation status Full and sustained rewarming

Minimize unnecessary transfusions Adherence to strict transfusion guidelines

Early return to operating room for excessive

bleeding

* From Helm RE, Rosengart TK, Gomez M, et al. Comprehensive Multimodality Blood Conservation: 100 Consecutive CABG Operations Without Transfusion. Ann Thorac Surg 1998; 65: 125-136

Reproduced with permission

* From Floyd TF, Cheung AT, Stecker MM, et al. Postoperative Neurological

Assessment and Management of the Cardiac Surgical Patient. Seminars in Thor Cardio

Vasc Surg 2000; 119: 138-147

Reproduced with permission

Figure Eight

STS Database *

1997 – CABG

161,108 Patients

Infection

Sternum - Superficial 0.73%

Sternum - Deep 0.63%

Leg 1.26%

IAB site 0.02%

Septicemia 0.92%

UTI 1.52%

Pneumonia 2.45%

Prosthetic Valve SBE 0.00%

Catheter device related not recorded

*From http://www.ctsnet.org/doc/2988

Patients Incidence Mortality Most Common

Huddy, 1991 ⁽²³³⁾ 4,473 0.78% 63% GI bleeding

Ohri, 1991 ⁽²³⁴⁾ 4,629 0.58% 14.8% GI bleeding

Tsiotos, 1994 ⁽²³⁵⁾ 19,246 0.60% 26% Complicated Peptic

Ott, 1995 ⁽²³⁶⁾ 8,448 0.65% 72% -------

Spotnitz, 1995 ⁽²³⁷⁾ 1,831 2.0% 57% UGI bleeding

Perugini, 1997 ⁽²³⁸⁾ 1,477 2.1% 19.4% GI bleeding

Zacharias, 2000⁽²³⁹⁾ 4,463 1.9% 17%/50% UGI bleeding

Akpinor, 2000⁽²⁴⁰⁾ 4,401 0.5% 41.7% GI bleeding

Figure One Weaning from Mechanical Ventilation

Figure Two Spectrum of Anecdotal Cases of Blood Dycrasias Undergoing

Cardiac Surgery

Figure Three Action of Antiplatelet Drugs

Figure Four Treatment Protocol for excessive Mediastinal Bleeding after CPB

Figure Five Guidelines for red blood cell transfusion

Figure Six Pharmacologic Blood Conservation measures incorporated

Figure Seven Postoperative Neurological Evaluation

Figure Eight STS database- CABG Infections

Figure Nine Gastrointestinal (GI) Complications