Late Complications of Cardiac Surgery

	INTRODUCTION

Top Next References

Changes in technology have produced a myriad of new devices and exposure techniques, and have redefined our concepts of the role for cardiopulmonary bypass, the very technology around which much of our specialty has been built. Parallel to these new procedures and advanced technologies has been an expansion of the knowledge of the early and late complications following these procedures.

The development and maintenance of large institutional, regional, national, and international databases have allowed surgeons to better understand the outcomes of many of the commonly performed procedures and at the same time to identify the risk factors which are predictive of improved or unsatisfactory outcomes. Both preoperative patient-specific risk factors as well as perioperative process-related factors have been studied extensively, profiling the spectrum of morbidity and mortality of patients undergoing cardiac surgery.

Postoperative complications can be classified in many ways, but the timing of onset or recognition is among the most useful classification systems. Short-term complications are those that are recognized almost immediately following the procedure, occurring within hours to up to thirty days following the procedure. Many data sets categorize short-term complications as those recognized during the initial or primary hospitalization or within the first thirty days following the procedure. Long-term complications, those occurring after the short-term period has elapsed, have also been well studied for many of the procedures that cardiac surgeons commonly employ.

Although typically framed in terms of six months, one year, many years, and extending up to decades after the procedure originally performed, any related cardiac event after the initial period of recovery can be considered within this important category of late surgical sequelae. The patient-related risk factors and the procedural or process-related factors predicting improved or deteriorated outcomes may be similar to those predicting the short-term sequelae for many of the procedures, but in most cases there is an additional set of factors predicting long-term outcomes. The levels of significance of the short-term predictive factors and their associated odds ratios have been different in all instances when long- and short-term outcomes have been compared.

This chapter addresses the long-term outcomes, also known as the late complications, of cardiac surgical procedures. In each instance, the differences and similarities when compared to the immediate or short-term complications will be explored in the context of an understanding of the predictive risk factors attributable to the patient, to the techniques involved with the procedures, and to the postoperative care whenever possible.

Given the long-term nature of the development of these late complications, the completeness of the follow-up, the nature of the follow-up system, and the ability to track large cohorts of patients are extremely important. It is clearly important for surgeons to have access to precise and complete long-term procedure-related data to fully understand current and evolving procedures, as well as to continually modify indications and surgical technique. By design, the long-term specific procedure-related outcomes and complications for valvular heart surgery, coronary heart disease, transplantation, and ventricular replacement devices are extensively covered in other chapters, and as such are not addressed in this chapter.

The recognition of long-term complications may depend upon the development of clinical symptoms and signs or upon routine scheduled provocative testing. Objective tests of cardiac function at rest and with exercise enhance the quantification of long-term outcomes of virtually all procedures we now commonly offer. Stratifying the outcomes of this objective noninvasive or invasive testing forms the basis of much of what we now use to determine indications for almost all aspects of preprocedural selection, intraoperative technique, and postoperative care. Medicine, based upon the pedestal of solid evidence, continues to support much of the decision making within our specialty.

This chapter will attempt to define the scientific basis of the late complications of cardiac surgery procedures. It is organized by the organ systems involved, and whenever possible the discussion is related directly to technical and perioperative considerations. Special attention is focused on neurologic, cognitive, and psychiatric complications, since those have been the focus of much interest and the basis of much procedurally driven decision making. Quality of life, months and years following all cardiac interventions, whatever the basis, underlies the ultimate goal of all of medicine. The closer we can come to understanding the relationships between patient and procedural factors and long-term complications that relate to quality of life, the better care our patients will receive.

	MYOCARDIAL ISCHEMIC COMPLICATIONS

Top Previous Next References

 
Myocardial infarction and ischemia have been reported to complicate most types of surgical procedures, but are particularly important in the setting of major cardiac, vascular, and pulmonary procedures. Recognition and early diagnosis are essential. With improvements in myocardial protection and better perfusion during cardiopulmonary bypass, the incidence of postoperative myocardial infarction has steadily fallen over the course of the past two decades.

The diagnosis of postoperative myocardial infarction has traditionally been based on electrocardiography (EKG). Presence of Q waves has been deemed sufficient to diagnose a perioperative infarction. However, Svedjeholm et al compared accuracy of Q-wave–based postoperative myocardial infarction diagnosis with biochemical markers in 302 consecutive patients. EKG-positive criteria for myocardial infarction were present in 8.1%, but when correlated with CPK-MB and troponin-T levels, only 1% of these cases qualified as positive for myocardial infarction. More than 25% of Q waves were associated with plasma levels of troponin-T below the reference level.¹ Troponin-T is the most reliable marker for postoperative myocardial infarction according to Hake et al, who studied EKG, CPK-MB, and troponin-T in patients suffering myocardial infarctions following coronary surgery.²

Iyer studied 12,003 coronary artery bypass graft patients treated over a 12-year period and calculated an overall mortality rate of 0.99% with a postoperative myocardial infarction rate of 1.34%. Bypass time of greater than 100 minutes and presence of unstable angina were the two most important predictors for postoperative myocardial infarction.³

	COMPLICATIONS OF ARRHYTHMIA AND PACING PROCEDURES

Top Previous Next References

 
Atrial Arrhythmias

With the advent of various techniques for approach (standard or minimally invasive, epicardial or endocardial) and methods of energy delivery for ablation procedures (cryo, radiofrequency, or microwave), the resultant increase in the number of these operations performed has produced a reasonable database of outcomes.

Guang et al compared freedom from atrial fibrillation at 3 years in two groups of patients requiring mitral valve surgery. Ninety-six patients with mitral valve replacement and radiofrequency ablation (modified Maze III) of the left atrium were contrasted with 87 undergoing mitral valve replacement alone. At three years, 77% of the first cohort were free of atrial fibrillation versus 25% of group II.⁴

Mohr et al evaluated a series of 234 patients undergoing radiofrequency ablation with 31.6% having radiofrequency ablation alone, 24.4% with concomitant mitral repair, 16.2% with mitral valve replacement, 5.1% with aortic valve replacement, and 5.0% with coronary artery bypass grafting. Median sternotomy was used in 43.2% and right video-assisted limited thoracotomy in 56.8%. Of the total group, 83.9% were discharged in sinus rhythm, 7.6% had atrial fibrillation, and 8.5% had atypical flutter. Pacemakers were needed in 9.8%, with 1.3% developing an aortoesophageal fistula with a 4.2% hospital and 2.1% 30-day mortality. At 6 months, 81% of patients were in sinus rhythm; at 12 months, the proportion was 72.5%.⁵ Combined endocardial and epicardial approaches have also been tried with similar good success.⁶ These studies have documented the efficacy of these procedures in alleviating atrial arrhythmias and in maintaining sinus rhythm at reasonable durations of follow-up.

Circumferential radiofrequency ablation around the pulmonary vein ostia alone, producing atrial electroanatomic remodeling, has been described by Pappone et al. In that study, 251 patients (179 with paroxysmal atrial fibrillation and 72 in permanent atrial fibrillation) were treated with radiofrequency for lesions surrounding the pulmonary veins with intraoperative evidence of pulmonary vein electrical isolation. At 10.4 months follow-up, 85% of the patients with paroxysmal atrial fibrillation and 68% of those with permanent atrial fibrillation were in sinus rhythm. Age, atrial fibrillation duration, and ejection fraction were not predictors for failure to convert. Left atrial diameter, however, was noted to be significantly higher in permanent atrial fibrillation patients with recurrence.⁷ Chen et al performed atrial size reduction with ablation in a group of 119 patients, and noted that in patients who converted to sinus rhythm, the right and left atrial sizes were smaller than in those who did not convert.⁸ Lim et al studied the influence of atrial fibrillation on outcome following mitral valve repair in 400 patients, and noted that atrial fibrillation did not affect early outcome or durability of mitral repair, but patients in atrial fibrillation had lower survival at 3 years of 82% (vs. 93% in controls) and 73% versus 88% at 5 years. Age older than 70 years, poor left ventricular function, and atrial fibrillation were negative predictors in univariate analysis.⁹

Repetitive atrial flutter following left atrial Maze procedures was noted by Usui et al to be associated with a right atrial isthmus. In their series of 41 patients, 9.8% developed atrial flutter, and it was inducible in all of them through the right atrial isthmus, which was treated subsequently by linear ablation.¹⁰

Albage et al studied hormonal activation in the perioperative phase and noted that weight gain from fluid retention was higher after the Maze procedure due to elevated antidiuretic hormone and aldosterone levels.¹¹ Bauer et al, in their study of 72 patients, found that atrial transport function (ATF) return to normal in 87% (echocardiography) and 78% (magnetic resonance imaging) at 12 months. Independent predictors of return of ATF were smaller preoperative left atrial diameter and better preoperative left ventricular ejection fraction.¹²

Ventricular Arrhythmias

Preoperative risk factor analysis for new onset tachyarrhythmias after cardiac surgery by Mayr et al revealed older age, history of congestive heart failure, sepsis, systemic inflammatory response, and multiorgan dysfunction as highly predictive.¹³ In a study of 382 coronary artery bypass grafting patients by Steinberg et al, 3.1% experienced one episode of sustained ventricular tachycardia at a mean of 4.1 days after surgery, with an in-hospital mortality rate of 25%. Previous myocardial infarction, severe congestive heart failure, low ejection fraction, noncollateralized occluded vessels on angiogram, and bypass grafting to an infarcted zone were the most predictive of risk for ventricular tachycardia following coronary surgery.¹⁴

Levine's study of 218 patients with automatic implantable cardioverter-defibrillator (AICD) placed for tachyarrhythmias found that nearly 50% of patients had their first AICD fire (discharge) for syncope, presyncope, or sustained ventricular tachycardia/ventricular fibrillation. Ejection fraction of less than 25% was noted to have an earlier first fire and shortened survival after AICD discharge. Beta blockers and coronary artery bypass grafting were associated with longer interval to first fire, but only coronary artery bypass grafting bestowed prolonged survival.¹⁵

Pacing Complications

Malpositioned leads in various positions (into the left ventricle through a patent foramen ovale or displaced lead into coronary artery sinus, etc.) have been reported and can be fairly easily diagnosed by x-ray or EKG tracings. Some patients may also present with cerebral embolism from inadvertently placed left ventricular leads.¹⁶ Echocardiography is useful in the diagnosis of such malpositioned leads. Infected transvenous permanent pacemakers can be diagnosed by transthoracic or transesophageal echocardiography. However, transesophageal echocardiography detected lead and valve vegetations in 7 out of 10 patients, compared to only 2 out of 10 by transthoracic echocardiography.¹⁷

Management in the large majority of these situations involves replacement of the leads. Pacing lead extractions become necessary for dysfunction or incompatibility in 12% of patients, lead endocarditis in 32%, Accufix lead in 14%, pulse generator pocket infection in 22%, or lead erosions in 19%. Jarwe, in his study of 128 lead extractions, noted that 95% of the leads could be completely extracted, 2% partially, and 4% could not be removed. Four complications of two tricuspid leaflet tears, lead migration, and femoral hemorrhage were seen.¹⁸ Femoral lead extraction has now become safe and efficient.

Infections of the pacing systems are infrequent, with Staphylococcus being the predominant organism in generator pocket infections. Endocarditis induced by pacemaker leads can arise de novo or as an extension of generator pocket infection. Primary treatment option includes removal of the entire pacing system under antibiotic cover.¹⁹ Explantation, sterilization, and reimplantation have also been successfully employed. New leads are placed, old leads removed, and the sterilized generator reimplanted, producing excellent results in 17 patients over 14 years.²⁰ Runaway pacemakers have been reported, which usually indicates battery end-of-life and requires generator replacement.²¹

	PERICARDIAL COMPLICATIONS

Top Previous Next References

 
Postpericardiotomy syndrome, pericardial effusions, tamponade, and constrictive pericarditis are some of the well-known late pericardial complications after cardiac surgery.

Postpericardotomy Syndrome

Postpericardiotomy syndrome and post–myocardial infarction pericardial syndrome (Dressler's space syndrome) are characterized by fever with a pleural/pericardial pain component associated with eosinophilia and atypical lymphocytosis. The usual treatment algorithm is nonsteroidal anti-inflammatory agents followed by steroids for persistent effusions. Colchicine has been used in some resistant cases of postpericardiotomy effusions with success. Postperfusion pericardial syndrome is caused by cytomegalovirus inflammation of the pericardium and is characterized by sore throat without pleural/pericardial or eosinophilic accompaniment. Elevated cytomegalovirus titers (Ig M) are usually present.²²

Pericardial Effusions

Postpericardiotomy effusions have an incidence of 58% to 64% following coronary grafting and valvular surgery. In one study, loculated effusions (57%) were more frequent than diffuse (42%), with tamponade in 1.5%; however, aortic root surgery predisposed to a 6-fold increase in effusions to 31.6%.²³ This certainly warrants surveillance echocardiography every 6 months for 2 years postoperatively, to exclude tamponade physiology, especially for those on anticoagulant therapy. Postpericardiotomy effusions are frequently seen 1 to 2 weeks postoperatively in patients receiving aspirin or warfarin and contribute to 35% of late effusions. Symptoms of postpericardiotomy syndrome, pericardial rub, atrial arrhythmias, cardiac enlargement, and pleural effusions are the most commonly noted. A higher incidence of pericardial and pleural effusions is associated with the use of internal mammary as a conduit.^23–25 Use of anticoagulation contributed to 86% of early and 65% of late pericardial effusions. Diagnosis and treatment by echocardiography is safe and effective. Echo-guided pericardiocentesis is successful in draining 89% to 97% of all effusions and nearly 96% of loculated effusions with a major complication rate of 2% (chamber laceration and pneumothorax). With use of a flexible catheter for extended drainage, incidence of recurrent effusion can be decreased by 50%.^26,27 The subxiphoid approach has a higher recurrence rate while the transthoracic approach is more invasive. These more invasive methods may be reserved for situations in which pericardiocentesis is not feasible.

Refractory supraventricular tachyarrhythmias and chylopericardium (due to operative trauma during mammary artery harvesting) are two unusual presentations of postpericardiotomy effusion.^28,29 A higher incidence of pericardial effusion causing tamponade (8.9%) is seen in heart transplant patients. A weight mismatch (recipient weight greater than donor weight), lack of previous sternotomy, rejection, and cyclosporine use have been implicated. Forty-four percent were treated successfully by pericardiocentesis and 28% required pericardiectomy in one series.³⁰

Constrictive Pericarditis

Constrictive pericarditis as a late complication of cardiac surgery has an incidence of 0.2% to 2.0%. The interval between surgery and presentation is an average of 82 days (14 to 186). Risk factor analysis reveals normal left ventricular function, warfarin administration, and inadequate drainage of early postoperative pericardial effusions as the significant causes.^31,32 Diagnosis is established by echocardiography, tomography, and catheterization showing pericardial thickening, constrictive physiology, and characteristic early diastolic dip and late plateau pattern (square root sign). Leaving the pericardium open at initial surgery does not influence the incidence of constrictive pericarditis.³³

Other rare causes such as tuberculosis, viral, chlamydial, and fungal infections, and autoimmune reactions have also been reported. Treatment by pericardiectomy is the most durable. Use of 0.1% dilute sodium, hyaluronic acid (NaHA), and 0.1% carboxymethylcellulose (CMC) or tissue protective bioabsorbable polymer membrane solutions has been shown experimentally to decrease pericardial and pleural adhesions to 20% while control groups had an 80% incidence.^34,35

	RESPIRATORY COMPLICATIONS

Top Previous Next References

 
Chest Wall Complications

Sternal or chest wall pain and its inadequate control have a debilitating effect on a patient's recovery and long-term functional status. Mueller et al studied prospectively the location, distribution, and intensity of pain after cardiac surgery in 200 patients. The pain intensity was at its maximum on postoperative days 1 and 2 and decreased to a minimum by day 7; however, the location of pain shifted from midsternal early postoperatively to shoulder pain by day 7. Age had an impact on pain, with patients less than 60 years having a higher intensity.³⁶ Long-term pain (beyond four weeks) tended to be more osteoarthritic and musculoskeletal with a small percentage developing paresthesias at the costochondral junctions of the left anterior chest, at the site of internal mammary artery harvest.

In a series of 288 patients, a 9% incidence of sternal fractures with a higher association with mammary artery harvest was noted. Of the patients with sternal fracture, 10% had major respiratory compromise from pain, and epidural anesthesia was the most effective in managing these patients.³⁷ In another study, 13% patients had rib fractures, with 60% involving the first rib (following mammary harvest) and the remaining comprised of costotransverse disarticulations. Rib fractures negatively influence postoperative rehabilitation and overall postsurgical satisfaction with quality of life.³⁸

Gust et al performed a randomized trial comparing patient-controlled anesthesia (PCA) with nurse-controlled anesthesia and the effect on atelectasis, rehabilitation, mobility, and overall satisfaction. The study concluded that PCA significantly decreases postoperative pulmonary atelectasis and respiratory complications, and increases return to functional status.³⁹ Please see the later section on quality of life for further return to work, functional status, and overall satisfaction details.

Sternal Wound Complications

Serious sternal wound infection and dehiscence occurred in 1.86% of a study group of 2579 consecutive cardiac surgical procedures reported by Ottino et al. In univariate analysis for risk factors, age, sex, type of surgical procedure, antibiotic prophylaxis, duration of ventilation, reoperation duration of CBV, blood transfusions, reexploration, rewiring, and ICU length of stay were significant. Multiple regression analysis showed that hospital environment, interval between admission and surgery, reoperation, reexploration, rewiring, and blood transfusions were important in predicting sternal instability, infection, and dehiscence.⁴⁰ Stahle et al studied 13,285 cardiac procedures and noted a 1.5% incidence of sternal infection, mediastinitis, and dehiscence. The CABG group had a 1.7% and the valve group a 0.7% incidence. Risk factors in their study were female sex, diabetes, bilateral IMA use, postoperative dialysis, reexploration, and transfusions. Patients with sternal wound complications showed no significant increase in early mortality but had worse long-term survival even after risk adjustments.⁴¹

Coagulase-negative staphylococcus is the most common organism involved. In the Cleveland Clinic study of 22,180 cardiac procedures, 23% of wound infections were from coagulase-negative staphylococci. Coagulase-negative staphylococcal sternal wound infections were superficial in 56% of the cases, 27% were deep, and 70% had mediastinitis, with 14% having simultaneous blood stream infection. Ninety-two percent of these infections were methicillin resistant. Management included reexploration (39%), flap operation (12%), and sternectomy (5%). Mean interval from operation to onset of sternal infection was 24 days with a mean parenteral antibiotic requirement of 22 days, resulting in an additional direct cost per patient of $20,000 for the entire admission.⁴² Streptococcus, Klebsiella, Enterobacter, Enterococcus, Pseudomonas, and rarely Candida albicans are the other organisms cultured. Obesity and use of ACE inhibitors (the side effect of dry cough with ACE I inhibitors caused sternal instability and dehiscence) have also been implicated.⁴³ Use of ACE type II receptor inhibitors decreased the incidence.

Symptoms, physical examination, and culture of wound drainage are sufficient to confirm the diagnosis. Computed tomography has a 67% sensitivity and 83% specificity and is not the most accurate diagnostic modality.⁴⁴ Management involves several approaches of reinforced sternal wiring, prosthesis insertion for strength of closure, various flap operations, etc. Reinforced sternal closure produces a statistically significant lower rate of dehiscence.⁴⁵ Pericostal sutures decreases incidence of dehiscence to 0.38%.⁴⁶

Use of cannulated screws with wire passed through the screw cannula seems to reduce stresses in the sternum.⁴⁷ The Ley prosthesis (a 0.5-mm-thick titanium alloy plate for stabilizing the sternum), when compared to controls, significantly reduced the incidence of reexploration and rewiring, thus reducing cardiac surgery ward length of stay and decreasing overall cost.⁴⁸ Open chest and delayed sternal closure are required in situations of cardiac edema, intractable bleeding, and arrhythmias.

The incidence of superficial wound infections (1.6%), mediastinitis (1%), and dehiscence (2.4%) following the use of these modalities was similar to controls, indicating that they may be safely used as adjuncts to drainage, debridement, and refixation of the sternum.⁴⁹

Spontaneous right ventricular rupture following sternal dehiscence is a catastrophic complication that occurs between sternal debridement and a flap procedure, with an incidence of 0.07% of all cardiac procedures. Mortality from this complication can be decreased by freeing the right ventricle from the underside of the sternum at the initial sternal debridement followed by flap closure at first instance when primary closure is not feasible.⁵⁰

Muscle flap reconstruction for major sternal infection comprises pectoralis (unilateral or bilateral), rectus abdominis, and/or greater omental flaps. Pectoralis flap advancement provides bulk but causes unsightly deformity and limits functional status. Rectus abdominis flaps are superior functionally but are limited by use of mammary arteries for coronary grafting, since the IMA forms a major part of the rectus blood supply. A delay period after mammary harvest has been suggested to improve collateral blood supply. Omental flaps have been used when rectal flaps are not feasible and have provided bulk and cosmesis. Reinfection rates after muscle flap reconstructions are very low. Freedom from reinfection and reintervention after muscle flap is 72% to 74%, compared to 12% after rewiring alone.^51–54

Airway and Pulmonary Complications

Durand et al studied pulmonary function tests as predictors of pulmonary outcomes after cardiac surgery, and showed that decreased preoperative vital capacity (less than 2.5 liters) and FeV1 less than 1.5 liters were the most reliable predictors of unfavorable long-term pulmonary outcomes.⁵⁵ Bevelaqua et al studied complications after cardiac surgery in patients with severe preoperative pulmonary impairment. Over 65% of these patients had obstructive disease and the remaining restrictive disease. Compared to controls, this entire group had a significantly higher incidence of atelectasis, effusion, bronchospasm, pneumonia, and pneumothorax. It was also found that in these patients, although pulmonary morbidity was higher, overall hospital mortality was comparable to controls. Restrictive disease patients fared better than obstructive. Valve replacement carried a higher morbidity compared to isolated coronary surgery. Perioperative pulmonary function tests can be used to accurately predict postoperative risk for pulmonary complications.⁵⁶

Several studies on ARDS after cardiac surgery place its incidence at 0.4% to 2.5% with a mortality of 15% to 34%. Redo surgery, poor preoperative pulmonary function, transfusions, emergent surgery (shock), smoking, diabetes, poor left ventricular function, and renal failure were predictors of ARDS following cardiac surgery.^57–59

Morbidity from prolonged ventilatory support (greater than 14 days) was studied by Lo Cicero et al in 581 open heart surgery patients, showing an incidence of 9.9%. Overall mortality was 43% in those who required prolonged ventilation. Of the survivors, 55% required only an endotracheal tube and were extubated within 14 days, while 45% needed tracheostomy. Of the tracheostomy patients, 26% were extubated, 37% died, and 37% remained on mechanical ventilation. Complication rate for patients with endotracheal tubes was 65% compared to 37% for those with tracheostomy.⁶⁰ Common tracheostomy complications included sternal infection, stomal infection, minor hemorrhage, tracheal stenosis, tube displacement, delayed wound healing, and soft tissue hemorrhage.⁶¹ Early tracheostomy (14 days) conferred better pulmonary toilet and outcomes.

Vocal cord dysfunction is an overlooked complication after respiratory support for cardiac surgery. Its incidence is about 1.9% of all cardiac surgery. It is usually self-limiting and masquerades as respiratory insufficiency. A small percentage of patients with more permanent vocal cord dysfunction require tracheostomy.⁶²

	GENITOURINARY COMPLICATIONS

Top Previous Next References

 
Renal Dysfunction

Several studies have defined by multivariate analysis specific pre-, intra-, and postoperative risk factors that increase chances of acute and long-term renal dysfunction after cardiac surgery. These studies found that 10% to 20% of patients developed transient oliguria and elevation in serum creatinine, with 75% of these spontaneously returning to normal renal function. However, 1% to 5% progressed to renal failure requiring dialysis. Mortality in this group ranged from 38% to 52%.

Preoperative risk factors include age greater than 65 years, diabetes, poor left ventricular function (ejection fraction less than 35%), class III and class IV heart failure, emergent surgery, and preoperative renal dysfunction. Intraoperative causes include prolonged bypass time (greater than 140 minutes), transfusions, reexplorations for bleeding, hypothermic circulatory arrest, use of full-dose aprotinin, low perfusion pressure, low flow rates during bypass, and high on-bypass hematocrits (greater than 35%).

Postoperative renal dysfunction, defined as a decrease by 25% in creatinine clearance from baseline or a value of less than 40 mL/min, results from postoperative hypovolemia, anemia, hypotension, low output syndrome, redo cardiac surgery, excessive use of vasoconstrictor therapy, myoglobinuria, jaundice, sepsis, and pericardial tamponade.^63,64 Off-pump cardiac surgery has in at least one study been shown to attenuate transient renal injury compared to traditional on-pump grafting.⁶⁵

Use of full-dose aprotinin has anecdotally been reported to cause postoperative renal dysfunction from an exaggerated adverse reaction. However, at least two well-known large studies, from Lemmer et al and Mora Mangano et al, evaluated the effect of full-dose aprotinin in coronary grafting on-pump and in other cardiac surgery requiring hypothermic circulatory arrest, and neither study showed any significant renal toxicity or dysfunction, even though a transient postoperative rise in serum creatinine was noted.^66,67

A concerted effort on the part of the surgeon, anesthesiologist, and perfusionist in addressing each of the listed risk factors is essential to significantly decrease the incidence of postoperative renal dysfunction. If a high likelihood of renal dysfunction is suspected despite the above precautions, intraoperative conventional hemofiltration or modified ultrafiltration can be undertaken. A few studies randomizing conventional hemofiltration (continuous veno-venous hemofiltration or continuous arterio-venous hemofiltration) to modified ultrafiltration have shown that a significant rise in postoperative hematocrit, mean blood pressure, and cardiac index occurred with modified ultrafiltration.^68–70 This improved hemodynamics and resulted in better clinical outcomes and mortality rates in that group of patients. Modified ultrafiltration is known to attenuate systemic inflammatory response by removing serum interleukin-8 plasma.⁷¹

Cardiac surgery requiring cardiopulmonary bypass has been performed in patients with prior kidney transplantation. The majority of these patients retained full function of the renal graft when measures were taken to ensure that hypotension, fluid overloading, and prolonged bypass were avoided. Some patients may need their immunosuppression to be reinstituted with a bolus dose because CPB dilutes their plasma levels. Less than 2% of these patients experience graft rejection. Perioperative incidence of renal failure requiring dialysis is less than 12%, although a transient rise in serum creatinine may be noted.⁷²

Urethral Complications

Urethral complications are seen mainly in those older than 60 years, with an incidence of 4% per year. Factors implicated in development of urethral stricture following cardiac surgery are a combination of urethral hypersensitivity and bulbar and penile urethral ischemia during cardiopulmonary bypass. Pressure necrosis at the urethral orifice of the urinary bladder also occurs from traction on the balloon of the urinary catheter exerted by its weight (collection bag) or from manner of anchoring.

A randomized trial of urethral catheterization to suprapubic catheterization showed a 4-fold increase of stricture with urethral catheterization. Avoiding overinflation of the catheter balloon and traction on the catheter may decrease incidence of this complication.⁷³

	GASTROINTESTINAL COMPLICATIONS

Top Previous Next References

 
Due to the absence of early specific clinical signs, gastrointestinal complications after cardiac surgery are often diagnosed late with resultant high mortality. Incidence of all intra-abdominal side effects after cardiopulmonary bypass is about 1% to 2%, but mortality ranges from 12.5% to greater than 90% depending upon the specific complication^74–76 and the surgical intervention needed.

Common in-hospital late complications (beyond 7 to 10 days postoperatively) include bowel ischemia from embolization or low flow, upper or lower gastrointestinal bleeding from gastritis, peptic ulceration and diverticular disease, and diarrhea from pseudomembranous colitis, pancreatitis, cholecystitis, and septic rupture of spleen. In a study of 4923 patients in which 1.3% had GI complications, the most frequent were GI bleeding (40%), pancreatitis (34%), acute cholecystitis (11%), perforated duodenal ulcer (8%), ischemic bowel (5%), and diverticulitis 2%.⁷⁷ In another study of 4473 patients in which 0.78% had GI complications, mortality rate by complication was GI bleeding 45%, intestinal infarction 67%, and pancreatitis 100%.⁷⁸

A third study of 6281 patients, evaluating mortality from GI complications requiring intervention based on type of cardiac surgery, showed that CABG had an incidence of 0.4% to 1% mortality from GI complications, valvular surgery 0.8%, cardiac transplantation 6%, acute aortic dissection 9%, and reoperative surgery 6% to 10%. Major associated risk factors were sepsis (odds ratio 38.7), renal failure (odds ratio 3.5), and prolonged ventilatory support (odds ratio 2.7).⁷⁹ A retrospective analysis of 4463 patients with a 1.9% incidence of GI complications noted a mortality rate of 70% for medical management of the GI complication versus 50% for surgical management.⁸⁰

Risk factors and predictors of intestinal ischemia and GI complications identified by multivariate analysis in two large series include age older than 70 years, duration of bypass, reoperation, need for transfusion of more than 2 units of blood, heart failure class IV, triple-vessel coronary disease associated with significant peripheral vascular disease, use of intra-aortic balloon counterpulsation, aortic athero emboli, reexploration for bleeding, postoperative low cardiac output, and significant inotropic support. In all patients with GI complications, a combination of at least four of the above predictors was present.⁸¹

Studies on gut barrier and mucosal function during anesthesia and cardiopulmonary bypass evaluated time sequence between intraoperative and postoperative endotoxemia, intramucosal pH changes, mediator and intestinal permeability, and acute phase proteins and their relationship to postoperative GI complications such as cholecystitis, hepatic, pseudomembranous colitis, and postoperative infections. Intraluminal pH of the stomach dropped significantly (6.98), plasma levels of endotoxin (greater than 0.2 endotoxin units/mL) increased (14-fold), interleukin-6 (IL-6) increased on the second postoperative day (2-fold), and C-reactive protein increased (4-fold) in patients who developed GI and infective complications. Ischemia in splanchnic circulation either intraoperatively (CBV) or postoperatively (low cardiac output) seemed to be the most likely stimuli.^82,83

Intestinal Complications

Mesenteric ischemia is a rare but catastrophic complication (0.1% of all cardiac surgery) with mortality rates of as high as 91%. Nonocclusive ischemia (45% to 60%) and embolic events are the common causes. Physical examination findings of ileus or acute abdomen with elevated lactate and amylase levels should usually arouse suspicion.

Radiologic and angiographic studies are used to confirm diagnosis. Interventions include intra-arterial papavarine injection, embolectomy (9%), bowel resection (36%), and exploratory laparotomy by itself in 55%. Average in-hospital mortality is 80% to 90% for bowel resection.^84–86 Acute diverticulitis is a rare complication with an incidence of 2% to 5% of all GI complications and conservative management is advocated, usually with good outcomes.⁸⁷

Hepatic Complications

Synthetic and secretory dysfunction of liver, icterus, and hepatic encephalopathy are rare complications of cardiac surgery utilizing cardiopulmonary bypass. Risk factors predicting hepatic dysfunction and adverse outcome include preoperative hepatic dysfunction evidenced by elevated SGOT, SGPT, prothrombin time, conjugated and unconjugated bilirubin, thrombocytopenia, right heart failure with severe hepatic congestion, portal hypertension, cardiac cirrhosis, hypotension, hypoxemia, and amount of transfusions. Age, sex, underlying cardiac lesions, and the presence or absence of hepatitis B are not predictive of postoperative jaundice. Classes B and C of Child's classification of liver dysfunction are the strongest predictors of mortality following cardiac surgery. Elective surgery with Child's class A including CABG and valve replacements survived the surgery, but succumbed late to GI complication such as bleeding or hepatic encephalopathy, etc. Histology shows central necrosis of hepatocytes with sinusoidal dilation. Good pre-, intra-, and postoperative management of systemic perfusion (with high flow rates), optimization of heart and liver function, and pharmacologic/mechanical circulatory support helps to produce better outcomes in patients with liver dysfunction.

Onset of fulminant hepatic dysfunction postoperatively invariably leads to multiorgan failure and is uniformly fatal.^88–91 Conventional tests of hepatic function do not offer precise evaluation of functional liver reserve. The two most common pharmacologic means of assessing hepatic reserve in patients undergoing cardiac surgery are the antipyrine plasma clearance test and the indocyanine green clearance test. Drop in antipyrine clearance (0.4 N mL/min/kg) closely correlated (coefficient r = 0.699) with ability to predict postoperative hyperbilirubinemia, while indocyanine green (coefficient r = 0.477) was slightly less sensitive and changes depended on cardiac index. These tests can be used to identify patients at risk for postoperative hyperbilirubinemia and hepatic dysfunction.^92,93

Pancreatic Complications

A study of 5621 cardiac surgery patients showed an incidence of 0.44% of pancreatic complications.⁹⁴ Hyperamylasemia is common after cardiac procedures (up to 32% of patients). About 20% of patients with hyperamylasemia (less than 1000 IU/L) had no serum lipase elevations and the elevated amylase was predominantly the salivary isoenzyme. Of the patients with elevated amylase, 10% had subclinical pancreatitis with mild elevations in serum lipase but had a self-limiting GI course. About 3% of patients with elevated amylase and lipase had overt clinical and diagnostic (tomography) signs of severe pancreatitis.⁹⁵ Patients with hyperamylasemia had a higher mortality of 9% compared to controls. Acute pancreatitis was by far the common complication (60%) with pancreatic necrosis next. Mortality increased up to 44% with these complications. The risk factors include pre-/postoperative hypotension, excessive use of inotropic support, renal failure, prolonged ventilatory support, intra-aortic balloon usage, and administration of calcium in doses of 800 mg or more/cm² of body surface area in the perioperative period.⁹⁶ Hyperamylasemia was more commonly seen with nonpulsatile perfusion (in 70% of patients) versus 32% of patients with pulsatile perfusion.⁹⁷ Modifying the above risk factors in the perioperative period can result in better outcomes. A 30-fold increase in incidence of pancreatitis is noted in heart/lung transplant recipients due to the combined effects of CPB and immunosuppression.⁹⁸

Splenic Complications

Splenic injury and rupture following cardiac surgery are very rare complications, sometimes noted in patients with myeloid dyscrasias and enlarged spleen or in those with erosion from fulminant pancreatitis or operative trauma exacerbated by full heparinization.⁹⁹

Biliary Complications

Acalculous and calculous cholecystitis have both been described following cardiac surgery and cardiothoracic organ transplantation. In a study of 645 cardiothoracic organ transplantation patients, 5.7% had symptomatic cholecystitis with all of them containing gallstones. Close to 45% of these patients required emergent cholecystectomy, with one mortality. All patients were female with a higher body mass index and a significant proportion had common bile duct stones also.¹⁰⁰ In posttransplant patients, early diagnosis and cholecystectomy with screening for choledocholithiasis produce the best results.

In nontransplant cardiac surgery, acalculous cholecystitis (biliary-based dyskinesia) is as frequent as calculous cholecystitis. Sepsis appears to be a major associated risk with acalculous cholecystitis. Conservative medical management with interval cholecystectomy appears a reasonable approach in nondiabetics. Acalculous cholecystitis also is a frequently associated GI complication in HIV-positive and AIDS patients undergoing cardiac surgery. Aggressive management in this immunosuppressed group produces the best outcomes.

	VASCULAR COMPLICATIONS

Top Previous Next References

 
Monitoring and Access Complications

Catheter access for monitoring and therapeutic reasons is well established. The placement and maintenance of these devices are also associated with a set of late complications. Arterial and venous monitoring and indwelling catheter complications have been described in several studies pertaining to surgical and medical intensive care patients. A brief review of long-term complications is presented here.

ARTERIAL CATHETERS

Frezza et al studied 2119 patients in medical and surgical critical care units with femoral and radial artery monitoring catheters in situ. Site of catheter insertion, intervals to the catheter change, number of changes, bleeding, infection, and ischemia were among the complications studied. Arterial catheters were present in 48% of MICU patients and 33% of SICU patients; 45% of MICU versus 11.5% of SICU patients had femoral catheters; and 78% of SICU and 52% of MICU patients had radial catheters. Catheters were changed in 9.5% of MICU versus 13% of SICU patients. The most common complication was vascular insufficiency (3.4% in MICU and 4.6% in SICU) followed by bleeding (1.8% in MICU and 2.6% in SICU). The catheter infection rates were 0.4% and 0.7%, respectively, with radial and femoral arterial sites having similar infection rates of 43% and 50%. The study concluded that catheter infection rates, timing, number of catheters, and site changes made no significant difference in complication rates.^101–103

PULMONARY ARTERY CATHETERS

Catheter colonizations and bacteremia with pulmonary artery catheters were studied by Singh et al in 51 critically ill patients having 52 arterial and 37 pulmonary catheterizations. Daily cultures of blood and catheter sites showed a colonization rate of 10% with a 45% bacteremia risk. Presence of concurrent infection and use of antibiotics did not change catheter colonization. Femoral catheters had more colonization than pulmonary or radial.^104,105

Thrombosis rates of arterial and pulmonary catheters were compared with heparinized and nonheparinized solutions by Zevola et al. Thrombosis rates of pulmonary catheters were no different in the two subgroups; however, arterial catheters failed less often when maintained with heparinized solutions.¹⁰⁶

CENTRAL VENOUS CATHETERS

Central venous line infections in the SICU were studied by Charalambous et al in 232 consecutive catheter insertions. Their analysis showed that 49% had no microbial growth, 17% were colonized, and 34% were infected. Univariate analysis showed catheter site, placement in operating room versus SICU, and catheter used for monitoring versus fluid and nutrition were all significant factors. Internal jugular insertion site was the single most important predictor of infection in the multivariate analysis. Of the catheter infections, 68% were monomicrobial and 32% were polymicrobial. A concurrent bacteremia was present in 45% of the patients with infected catheters, and death from catheter-related sepsis occurred in 7%. Gram-positive bacteria were found in 86% while others had gram-negative bacteria or Candida albicans.¹⁰⁷

Conduit Harvesting

INTERNAL MAMMARY ARTERY

El-Ansary et al compared the musculoskeletal and neurologic complications of internal mammary artery and saphenous vein harvests and noted an incidence of 78.5% and 45%, respectively, in patients at 3 to 6 weeks following cardiac surgery. The musculoskeletal effects relate to use of retraction and the neurologic side effects pertaining to anterior chest wall relate to internal mammary artery takedown with electrocautery causing injury to intercostal nerve branches.¹⁰⁸

A higher incidence of pleural changes including thickening occurred in internal mammary artery harvest compared to saphenous vein graft, although incidence and size of left-sided pleural effusions were similar in both groups.¹⁰⁹ Incidental malignancy in internal mammary artery lymph nodes is found very rarely. Undiagnosed lymphomas or carcinoma of the breast or lung are the usual causes. Abnormally enlarged internal thoracic artery lymph nodes should be histopathologically examined.¹¹⁰

Chylothorax following median sternotomy and internal mammary artery harvest has an incidence of less than 0.25%. It occurs from injury to lymphatic drainage from the thymus following median sternotomy or from injury to the thoracic duct during internal mammary artery harvest. Conservative management is sufficient in the majority; however, minimally invasive procedures such as video-assisted thoracic surgery (VATS) have provided good results.^111–113

Phrenic nerve dysfunction is not uncommon after coronary surgery, and a study of risk factors revealed that the use of pericardial ice slush for myocardial protection was the most important predictor followed by mammary artery harvest.¹¹⁴ An elevated hemidiaphragm with atelectasis is diagnostic. Phrenic nerve dysfunction is usually transient and resolves in less than 3 to 6 weeks. However, in phrenic transsection, diaphragmatic plication may in rare instances become necessary to overcome respiratory failure.

Steal phenomenon from side branches after use of internal mammary artery as conduit has been described, and various forms of therapy have been attempted. Its occurrence is rare and the pathogenesis is controversial. Several studies have used modalities such as Doppler echo and thallium scintiscan to determine the predisposing factors and the extent of steal from the side branches. Selective muscular vasodilation is known to be the most important factor that evokes the steal phenomenon when the first intercostal and/or certain large-caliber pericardial side branches are left in situ during mammary artery harvest. Approximately 25% of patients with a patent internal mammary artery with ischemia in its distribution are noted to have significant side branch collaterals. Marked decrease in distal mammary graft flow with large side branches on arteriography is diagnostic. Percutaneous intervention with transcatheter coil embolization has been successful in management of the unligated side branches. VATS ligation is another safe and minimally invasive option.^115–119

GASTROEPIPLOIC ARTERY

This artery has been used as an in situ and as a free graft. Short- and mid-term patency rates have been approximately 95% and 92%, respectively. In 104 patients, complications of the following types occurred, requiring catheter-based interventions: 3 occlusions, 5 stenoses, 3 competitive flow, and 8 instances of exercise-induced ischemia.¹²⁰ Intraoperative Doppler flow assessment of flow before artery harvest of 25 mL per minute or less suggests unsuitability for grafting.¹²¹

RADIAL ARTERY

Sensation abnormalities or thumb weakness reflect median nerve and radial nerve injury during radial artery harvest. In a study of 615 patients, Denton et al reported thumb weakness in 5.5% and sensation abnormality in 18.1%. Over a period of 8 to 9 months postoperatively, only 12.1% of patients reported symptoms without improvement. Statistical association was noted with diabetes, peripheral vascular disease, smoking, and elevated creatinine level.¹²² Both clinical and laboratory studies have shown that a calcium antagonist in combination with nitroglycerin is more potent than either calcium antagonist or nitroglycerin alone in prevention of radial artery vasospasm in the long term.¹²³

Aortic Complications

Postoperative aortic complications of dissection or enlargement have been reported in patients undergoing aortic valve replacement or ascending aortic surgery. Patients with an aortic diameter of 4.0 cm or greater with hypertension were most at risk. The incidence was approximately 0.25% in 8 patients out of 2205 in a study by Milano et al. They concluded with recommendations for aortic root replacement along with other planned aortic surgery in patients with a preoperative aortic diameter greater than 5.0 cm. Congenital bicuspid aortic valve, young age at aortic valve replacement, aortic regurgitation, and fragility or thinning of the aortic valve were strong multivariate analysis predictors for aortic dissection and enlargement.^124–126

Aneurysms and pseudoaneurysms of the saphenous vein graft–aortic anastomosis site are rare, but may present as paracardiac, hilar, or mediastinal masses on chest radiographs. Fistulous connections with chambers of the heart and cardiac compression are also reported. Mediastinitis appears to be one of the most commonly associated predisposing factors. Rupture of pseudoaneurysms at aortic cannulation sites secondary to mediastinitis has also been documented.^127–129

	HEMATOLOGIC COMPLICATIONS

Top Previous Next References

 
Hemolysis

Chronic hemolysis from a well-functioning bioprosthesis or mechanical valve is rare. Paravalvular leak is the most common cause for clinically significant hemolysis. A patient-prosthesis mismatch (for example, a 19-mm mechanical valve in a body surface area greater than 1.8 cm²) or an underlying hemolytic blood dyscrasia such as spherocytosis or thalassemia is another rare cause for hemolysis. Reconstructive patches (Dacron or Hemashield), when subjected to turbulent blood flow in and out of cardiac chambers, pledgets in the blood stream, or other stresses, are other flow-related causes for hemolysis.¹³⁰ Ventricular assist devices, pulsatile pumps, and centrifugal or axial flow pumps produce hemolysis enough to elevate bilirubin levels significantly for up to three weeks after implantation, after which there is a gradual downtrend. Intra-aortic balloon usage has also been reported to produce significant hemolysis based on the duration of usage.

Diagnosis is usually made by serial hemoglobin, serum LDH, haptoglobin, and serum and urine bilirubin levels. Patients with hemolysis can be treated conservatively until there is a drop in hemoglobin level to 5 or 6 g/dL (hematocrit of 20%), if the increase in reticulocyte count is over 10%, reflecting adequate bone marrow compensation. Onset of significant bilirubinemia (greater than 4 g/dL) and a packed cell transfusion requirement of 1 to 2 units per week are usually indications for surgical reintervention. A red cell fragility test can be used to detect underlying hemolytic anemia disorder.

Heparin-Induced Thrombocytopenia and Thrombosis

Heparin-induced thrombocytopenia (HIT) and thrombosis (HITT) are the result of autoantibodies directed against heparin-platelet factor 4 complexes that cause platelet activation and aggregation. Based on the platelet count, HIT type I is a milder form, with counts greater than 100,000/dL, and type 2 is the more clinically significant form, with counts less than 100,000/dL. Incidence of type 2 HIT is approximately 5% in patients undergoing cardiac surgery. HITT has been reported to cause stroke, myocardial infarction, graft closure, aortoiliac thrombosis with leg ischemia, mesenteric and deep vein thrombosis, and prosthetic valve thrombosis.^131–136

Management options may be based on urgency of surgery. A 6- to 8-week observation period decreases the antibody titer significantly, allowing surgery to be performed at full-dose heparin. Serial platelet counts and platelet aggregation tests are mandatory during this period. If surgery cannot be postponed, low molecular weight heparin (LMWH) can be used; however, a cross-reaction risk is then present. Danaparoid (Organon), a mixture of dermatans and heparins that has anticoagulant activity but minimal cross-reaction, is a third option. However, anticoagulation with danaparoid is excessive and difficult to control and reverse. The final option is plasmapheresis. Autoantibodies to heparin can be reduced or eliminated by a series of plasmapheresis treatments.

Neoplastic Hematologic Disorders

Patients with malignant hematologic disorders suffer from an increased risk in undergoing cardiac surgery. Non-Hodgkin's and Hodgkin's lymphoma, multiple myeloma, polycythemia, Waldenström's macroglobulinemia, myelodysplasia, chronic lymphocytic leukemia, and aplastic anemia are some disorders encountered in cardiac surgical practice. Elective cardiac surgery in these patients is complicated by a bleeding diathesis that requires several units of packed cell platelets and FFP concentrate. This predisposes to postoperative multiorgan failure, respiratory insufficiency, and long-term infections. Therefore, risk for morbidity must be carefully weighed against indications for surgery in this group.¹³⁷

Immunologic Alterations

Cardiopulmonary bypass is known to produce a transient immunosuppressive phase immediately postoperatively due to its effect on circulating lymphocytes and monocytes. Total lymphocyte count on bypass drops dramatically. T-cell numbers are significantly decreased, particularly CD4 cells (to as low as 251 cells/dL), while the ratio of CD3 and CD8 cells is retained. Only B cells and natural killer (NK) cells are increased. Expression of class II major histocompatibility antigens from monocytes is also decreased.^138,139 This results in a transient immunosuppression that peaks at postoperative day 1 and gradually returns to normal by day 3.

The outcome of human immunodeficiency virus (HIV) infection after CPB is still unclear. Conflicting reports, mostly anecdotal, claim that a small percentage of HIV patients progress to AIDS after CPB, but there is no conclusive evidence so far. In fact, in one study of a 74-month follow up, no such progression was noted.^140,141 Vijay et al evaluated outcomes in HIV patients undergoing cardiac surgery based on preoperative CD4 count. Elective surgery for noninfective indications in patients with a CD4 count greater than 200 had immediate postoperative outcomes similar to those of the non-HIV population. However, emergent surgery for infective etiology with a CD4 count of less than 200 was uniformly fatal.¹⁴²

	NEUROLOGIC COMPLICATIONS

Top Previous Next References

 
Neurologic complications after cardiac surgery fall into the categories of neurologic deficits, cognitive dysfunction, or encephalopathy. Neurologic deficits include hemiparesis, hemiplegia, and visual defects. The prevalence of stroke is far less than that of neurocognitive dysfunction. In young patients, the risk of stroke after cardiac surgery is 0.5% but gradually increases to 5% at 65 years and to 8% in those over 75 years of age for coronary surgery, and averages about 8% for valvular surgery. Reoperations and combined CABG plus valve replacement can have an incidence of up to 18%.^143–145

Neurologic Deficits

Aortic atheroma and prior history of stroke are the two most important predictors of stroke following cardiac surgery. Patients who have limited or no atheroma have a less than 2% incidence while those with a grade 4 or grade 5 atheroma have a 40% stroke rate.¹⁴⁶ Patients with carotid stenosis of less than 50% have a 1% or less incidence, while those with 90% stenosis or greater have a 6.5% stroke rate.¹⁴⁷ Other important risk factors for stroke include age, diabetes, peripheral vascular disease, renal failure, aneurysmal disease of the abdominal aorta, left main coronary stenosis, emergent operation, heart failure class, duration of bypass, number of aortic anastomoses, LV venting, intra-aortic balloon pump (IABP), reexplorations, and intraoperative detection of aortic atheroma. Use of a nonmembrane oxygenator and an arterial filter of greater than 40-micron porosity have also been implicated as risk factors.^148,149

Abraham et al reported the incidence of stroke in off-pump and on-pump groups as 1.2% and 3.6 %, respectively. But the off-pump group had a significantly higher percentage of redos and calcified aortas (26.4% redos in the off-pump group versus 8.7% in on-pump, and 79% versus 2.9% for calcified aortas, respectively), reiterating that the 3-fold reduction of stroke in off-pump may be significant.¹⁵⁰ Other studies have shown no significant difference in perioperative stroke rate between off-pump and on-pump coronary surgery. Use of low-dose heparin combined with lack of the protective effect of CPB (platelet destruction) on coagulation and blood pressure changes during off-pump surgery contribute to cerebral no-reflow phenomenon accounting for some delayed postoperative TIA and stroke incidence. This may account for the lack of significant stroke rate differences between off-pump and on-pump in the long term.

Several therapeutic measures can be adopted during surgery to reduce the risk of nonembolic stroke and neurocognitive dysfunction. EEG and transcranial Doppler (TCD) have been used to monitor adequacy of CPB physiology and cerebral perfusion; however, cumbersomeness of EEG and lack of specificity with TCD have decreased their utility in the operating room.

INVOS-cerebral oximetry (Somanetics Corporation, Troy, MI) is a device that monitors cerebral venous oxygen saturation and has been utilized at several centers as an adjunct to detect cerebral hypoperfusion. This allows certain intraoperative measures to be taken, such as increasing flow rates, increasing mean arterial pressure and hematocrit, increasing PaCO₂, and deepening the level of anesthesia to augment cerebral oxygenation, which decrease cerebral oxygen demand and thus eliminate nonembolic causes of stroke and cognitive dysfunction. Use of cerebral oximetry in off-pump surgery to decide on timing of conversion to on-pump surgery is recommended based on a decrease in cerebral saturation (less than 50% saturation correlates with postoperative neurocognitive dysfunction) during off-pump cardiac manipulation for graft placement.^151,152

Certain mechanical devices such as various cannula tips, aortic nets combined with aortic cannulas (EMBOL-X, Mountain View, CA), have shown that a significant embolic load can be removed during the critical portions of surgery that are notorious for showering emboli.¹⁵³ A holistic approach to addressing and optimizing all risk factors provides the best chance for a favorable neurologic outcome.

Neurocognitive Function

Evaluation of neurocognitive function consists of eight tests consolidated into five domains: attention, cognitive speed, memory, executive function, and fine motor function. Disturbances in these areas are more pronounced after deep hypothermia and circulatory arrest than after other types of cardiac operations with significantly shorter bypass times. Cognitive defects are subtle and transient in most patients; however, significant defects persist long term and primarily affect functional recovery and overall patient satisfaction, although it is a rare patient who does not return to work because of cognitive deficits. Nevertheless, a 50% or greater negative change in memory, fine motor function, and attention at 1 week after surgery, as determined by a battery of commonly used tests, was a strong predictor of poor performance at 6 weeks and these disabilities persisted up to 6 months to 5 years in some patients.¹⁵⁴

Newman found that cognitive deficits are evident in approximately 53% of patients at discharge, decreasing to 36% at 6 weeks, and 24% at 6 months. Linguistic function is the best preserved of all cognitive functions. Such neurocognitive decline has been shown to be dramatically reduced in off-pump surgery, compared to a 90% incidence immediately postoperatively in on-pump surgery.^155,156

Encephalopathy

Encephalopathy (delirium) is seen in 30% to 35% of patients, with confusion being the most common presentation in the absence of focal neurologic deficits. Most episodes peak within 24 hours after surgery and only 10% of patients have symptoms by the 10th day. Encephalopathy is associated with preoperative substance abuse syndromes, metabolic conditions, and dementia, but not with age, alcoholism, narcotics, or sedatives.^157,158 Please refer to the next section for further details on return to work, productive life, and overall satisfaction following neuropsychological disturbances.

	QUALITY OF LIFE AFTER CARDIAC SURGERY

Top Previous Next References

 
Functional benefits are a group of three subsets of quality of life instruments used to assess patients before and after cardiac surgery. They are physical functioning (fewer incapacitated days per month and fewer activity restrictions), sexual functioning (through increased energy, decreased pain and worry), and role functioning (ability to return to work, participate in social activities, resume pursuit of hobby, and so forth).¹⁵⁹

Coping and emotional response in cardiac surgery patients can be assessed by a coping checklist and a profile of mood state. There is a significant postoperative decrease in dependent coping strategies such as blaming oneself, wishful thinking, and seeking social support. Preoperative emotional state is the single most important predictor of postoperative coping and emotion.¹⁶⁰

In a study of gender-matched and procedure-matched patients of two groups (one younger than 50 years and the other older than 75 years), incidence and consequences of mental disturbances in elderly patients following cardiac surgery were compared to younger patients. Mental confusion was more prevalent in the older group, 22.6% versus 11.8% in the younger group. Late mortality was significantly worse and the quality of life diminished in the elderly who were confused in the perioperative period.¹⁶¹

Another study of 241 patients correlated perioperative neurocognitive function with the quality of life 5 years after surgery. Lower 5-year overall quality of life correlated well with lower general health and less productive working status and neurocognitive decline postoperatively.¹⁶² Patients who suffered from posttraumatic stress disorder in the postoperative phase consistently rated their life satisfaction lower than the controls in the long term.¹⁶³

Several studies have assessed general quality of life and level of satisfaction in the elderly using the Barthel mobility index and the Duke activity index, and found that 99.1% were satisfied with their operation, and 96.5% were in heart failure class I and II with dramatic improvement in functional status following cardiac surgery, thus justifying surgery in this age group.^164,165 A study of survival and functional status after cardiac surgery and long-term intensive care stay showed that the geriatric depression scale assessment put 91% as normal with only 8% in a severely depressed state. Although in-hospital mortality was high at 34%, all survivors had above average functional state and quality of life, justifying the surgery.¹⁶⁶

With multiple regression analysis, it has become clear that the strongest predictors of return to work are preoperative employment, educational level, higher family income, early surgery, and less postoperative angina and fatigue.

The overall reemployment rate after surgery was 78% in patients who had the above favorable predictors. These results suggest that the determinants of return to work are largely present before surgery, and that a patient's attitude and expectations along with early surgery play an important role.^167,168

	SUMMARY

Top References

 
Over time, the spectrum of surgical procedures available to patients with cardiac and great vessel disease continues to diversify and to mature. In parallel, our ability to match a widening range of surgical procedures with the individual needs of the patient has evolved as well. The relationships among preoperative patient-related risk factors and procedure-related perioperative care form the basis of the ongoing process of surgical care. Through a precise understanding of the long- and short-term benefits of a given surgical procedure, the risks for early and late complications can be evaluated to ensure that responsible clinical decision making occurs.

The close relationships among our nonsurgical colleagues, patients, and families as well as with the other members of the surgical care team facilitate the prevention of many of the above-described late complications of cardiac surgery. In addition, this teamwork also supports the early detection and successful treatment of many late surgical sequelae.