Published: June 2014
One of the most common requests made to physicians is to assess the perioperative cardiac risks of noncardiac surgery. Once the physician estimates the risk of a patient, he or she may be able to apply measures to decrease the risk for the patient and improve the outcome. Frequently in these cases, an opportunity is created for the first time to address cardiac risk factors in the patient undergoing surgery. This opportunity often is limited by time constraints and brief contact with the patient, especially if the surgery is semi-urgent or pre-scheduled at short notice. The major goal is to assess the risk of myocardial infarction (MI), heart failure (HF), or both, which are the most common causes of morbidity and mortality with noncardiac surgery. The mortality rate of patients with perioperative MI is substantial, ranging from 30% to 50%.
Conversely, there are very few cases in which the surgical outcomes and treatments are affected by extensive preoperative cardiac testing. Although preoperative testing is indicated in some cases, it does not always lead to a scientifically tangible improvement in outcome. Indiscriminate and extensive preoperative cardiac testing is an ineffective way of using health care funds and can lead to more unwarranted and risky procedures. In addition to the inappropriate expenditure of resources, unnecessary testing could cause harm to the patient by delaying surgery. For a test to be considered useful it should be accurate, influence outcome, and have a favorable risk-to-benefit ratio. Therefore, it is essential for the physician to identify patients who will benefit most from an in-depth preoperative evaluation. It is important for the physician to explore noncardiac issues (e.g., chronic lung disease, coagulopathy, anemia, renal and cerebrovascular disease, diabetes) that can negatively affect the outcome of the surgery. A preoperative evaluation should be considered an opportunity for a thorough medical evaluation in patients who previously might not have been in contact with the medical system.
There are various factors to be considered when assessing the cardiac risks of anesthesia and surgery. These are generally divided into patient-related and surgery-specific risks, as well as test-specific considerations (Box 1).
|Box 1: Factors to be Considered When Assessing Cardiac Risk|
|Chronic diseases (e.g., coronary artery disease, diabetes mellitus, hypertension)|
|Type of surgery (e.g., vascular, endoscopic, abdominal)|
|Urgency of the operation (e.g., emergent, urgent, elective)|
|Duration of the operation, possibility of blood loss and fluid shifts|
|Sensitivity and specificity of a test|
|Effect on management|
In the late 1970's, Goldman and coworkers1 developed a user-friendly point system that identified risk factors for perioperative fatal and nonfatal cardiac events. This system created four classes of risk, depending on the total points accumulated (Table 1).
|Preoperative third heart sound or jugular venous distention indicating active heart failure||11|
|Myocardial infarction in the past 6 months||10|
|≥5 Premature ventricular complexes/min before surgery||7|
|Rhythm other than sinus||7|
|Age ≥70 years||5|
|Significant aortic stenosis||3|
|Intraperitoneal, intrathoracic, or aortic surgery||3|
|Markers of poor general medical condition (e.g., renal dysfunction, liver disease, lung disease, electrolyte imbalance)||3|
Patients in the lowest risk quartile (0 to 5 points) had less than a 1% risk of postoperative major cardiac complications. In the two middle quartiles with 6 to 25 points, the major cardiac event risk was 9%, and 22% of the patients in the highest risk group (≥26 points) had a major perioperative cardiac event.
One of the limitations of the Goldman criteria1 was the inability to predict the operative risk for patients undergoing vascular surgery because of the low number of such patients included in the original study population. This limitation was addressed by Eagle and colleagues2 in a study of patients undergoing vascular surgery. Multivariate analysis has shown that the following factors predict an adverse event following vascular surgery:
Combining both the clinical data and thallium imaging was more sensitive and specific than either alone in predicting postoperative complications. In this model, the following can be noted:
A modified cardiac index that included a change in the scores allocated to risk factors such as type of operation, age, frequency of premature ventricular contractions, and aortic stenosis (AS) was published by Detsky and associates in 1986.3 However, HF was defined in this study as pulmonary edema determined by chest radiograph or by history of severe respiratory distress and resolution of the symptoms by use of diuretics. In addition, angina was subdivided into four classes according to the Canadian Cardiovascular Society classification.4 The score obtained from the patient’s risk factors, along with the risk associated with the type of surgery, were used to calculate the probability of a cardiac event.
The modified cardiac index was revised by Lee and coworkers,5 who devised a six-point index score for assessing the risk of complications with noncardiac surgery. The Revised Cardiac Risk Index (RCRI) includes the following variables and risks:
Each of the six risk factors was assigned one point. Patients with none, one, or two risk factor(s) were assigned to RCRI classes I, II, and III, respectively, and patients with three or more risk factors were considered class IV. The risk associated with each class was 0.4%, 1%, 7%, and 11% for patients in classes I, II, III, and IV, respectively. We recommend the use of this index because it is simple, has been extensively validated, and provides a good estimate of the preoperative risk.
The American College of Cardiology (ACC) together with the American Heart Association (AHA) have divided markers of perioperative risks into two categories: active cardiac conditions and clinical risk factors (Box 2). Patients presenting with active cardiac conditions generally will need extensive cardiovascular investigation and treatment, as well as postponement or cancellation of their elective surgery (class I recommendation).6
Patients with clinical risk factors need careful assessment, weighed together with functional capacity and procedural risk, to decide on the need for noninvasive cardiac stress testing:
|Box 2: ACC/AHA Perioperative Risk Assessment6|
|Active Cardiac Conditions|
|Unstable coronary syndromes|
|Decompensated heart failure|
|Severe valvular disease|
|Clinical Risk Factors|
|History of ischemic heart disease|
|History of cerebrovascular disease|
|Compensated or prior heart failure|
† The American College of Cardiology National Database Library has defined recent myocardial infarction as >7 days but ≤1 month (30 days); acute myocardial infarction is within 7 days.
‡May include "stable" angina in patients who are unusually sedentary.
Individuals with known coronary artery disease (CAD) should be classified into a specific risk class according to one of the risk indices cited previously, preferably Lee's RCRI. For patients classified into the low-risk group, we recommend a preoperative ECG and chest radiograph. Postoperative care should include aggressive pain management, and maintenance of euvolemia. Monitoring for ischemia (serial ECGs, cardiac enzyme levels) would only be necessary if the patient has had intraoperative hemodynamic instability. We also recommend an ECG before discharge.
If the patient belongs to the intermediate-risk group, he or she should be managed aggressively with lipid-lowering agents and tight blood pressure (BP) control. Much debate is ongoing concerning the use of noninvasive stress testing in this patient subgroup. In any case, there is not much evidence supporting the use of revascularization before noncardiac surgery.
Retrospective data analyses of patients who have undergone coronary artery bypass grafting (CABG) or percutaneous coronary intervention (PCI) months to years before noncardiac surgery have shown a lower incidence of perioperative complications compared with patients who had medical therapy alone. However, the average mortality rate of CABG in the United States in 2002 was 2.6%, which usually exceeds the risk of noncardiac surgery in these patients. Furthermore, one study has shown that percutaneous angioplasty performed on stable CAD patients due to undergo vascular surgery, with at least one coronary artery having more than 70% stenosis, did not result in any survival benefit over 2.7 years of follow-up.6
Another study has revealed that in-stent thrombosis may complicate noncardiac surgery if PCI is done within 6 weeks of surgery. In addition, some reports have suggested that the benefit of PCI might not be evident until 90 days after the procedure. To preserve the bare-metal stent placed during PCI, the patient has to take aspirin and clopidogrel (Plavix) for at least 1 month, which might delay noncardiac surgery further.
Data suggest that with drug-eluting stents, risks of stent thrombosis off dual antiplatelet therapy are considerable, even 1 year after stent placement. In addition, the percentage of patients who needed revascularization by CABG or PCI was relatively small in most studies. Performing extensive testing to identify these patients is not a cost-effective strategy. Thus, for most intermediate-risk patients, we would recommend medical management without an extensive diagnostic workup (see later discussion of medical therapy).
In high-risk patients (RCRI >2 or with signs and symptoms of CAD), diagnostic catheterization should be carried out, followed by revascularization if indicated, irrespective of the noncardiac surgical plans.
Silent myocardial ischemia occurs commonly in diabetic patients because of diabetic neuropathy, even in patients with well-controlled glycemia. In addition, diabetic patients are more predisposed to infection, poor wound healing, and episodes of hypoglycemia and hyperglycemia, which might negatively affect the outcome of noncardiac surgery. Thus, the diabetic patient needs more aggressive evaluation than the nondiabetic patient. A study of diabetic patients undergoing noncardiac surgery has concluded that diabetic patients are at increased risk for perioperative mortality, mainly due to cardiovascular causes.6 Assessment should include history, physical examination, and noninvasive testing, depending on the patient’s risk factor profile (ECG, noninvasive imaging stress test, and creatinine level). It is recommended that well-controlled glucose levels (fasting 90-126 mg/dL, random <200) be maintained perioperatively by insulin infusion to decrease the risk of wound infection.
The association between age and complications during noncardiac surgery was significant in a prospective cohort analysis done by Polanczyk and colleagues at Brigham & Women's Hospital.7 Advanced age adversely affected the rate of cardiac and noncardiac complications, mortality, and length of stay. Perioperative mortality risk was in the intermediate range (2.6% in patients >80years vs. 0.3% in patients 50-59 years of age; P = .002). However, it is possible that, due to preferential nonsurgical management of older patients, the >80 year-old sample in this study may have been lower-risk than persons of this age in the general population. The fact that this was a single-institution study from a very high quality-of-care hospital may have lowered the event numbers, as well.
It is unclear from the literature whether the criteria of Goldman, Eagle and others1,2,5are sufficient to risk-stratify the very elderly, or whether additional testing and triage could lead to improved outcomes. The RCRI predicts major adverse cardiac events (MACE) more reliably in patients younger than 55 years as compared with patients older than 75 years. Welten and colleagues, in a study on vascular surgery patients older than 18 years (60% of the patients were >66 years and 20% were >75 years) showed that addition of age and the type of surgical procedure to the RCRI improves its predictive value; older patients were at higher risk for MACE, with the highest risk being in the 66 to 75 year age group.8
Likewise, Feringa and colleagues found that advanced age is an independent predictor of in-hospital and long-term mortality in patients older than 65 years undergoing major vascular surgery.9 In addition, the use of aspirin, beta blockers, and statins was associated with 47%, 68%, and 65% relative risk reduction of in-hospital mortality, respectively. The aforementioned drugs and angiotensin-converting enzyme (ACE) inhibitors were associated with reduced incidence of long-term mortality in the same study as well. Despite these benefits, we recommend caution when using beta blockers, diuretics, and other antihypertensive drugs in the elderly, given the reduced clearance of drugs and metabolites in this age group.
The main issue with hypertensive patients is whether they have uncomplicated hypertension or hypertension with end-organ damage (e.g., renal dysfunction, cerebrovascular disease, left ventricular hypertrophy, systolic dysfunction, diastolic dysfunction, or CAD). Patients with hypertension and no evidence of end-organ damage are not at increased risk for major perioperative cardiovascular complications; they may proceed to surgery without further investigations, with the goal of tight BP control.6
Preoperative cardiac testing (e.g., stress echocardiography, scintigraphy) can be considered if chronically-hypertensive patients are planned to undergo high-risk procedures. If the BP is above 180/110 mm Hg, it is recommended to delay surgery until the BP is normalized. BP control can take days to weeks, which is acceptable in the setting of elective surgery. However, if the surgery is urgent, BP can be acutely controlled by infusion of intravenous(IV) antihypertensive medications, such as sodium nitroprusside or labetalol. BP should be lowered slowly to avoid inducing cerebral hypoperfusion.
Hypertensive patients with end-organ damage should be considered for preoperative testing (electrocardiography, noninvasive imaging stress test), especially if they are scheduled for moderate-to high-risk surgery. In patients with hypertension and left ventricular hypertrophy, ischemia may result from supply/demand mismatch in the thickened ventricle rather than from CAD. Kidney dysfunction is a common result of hypertension. An elevated creatinine level is an independent predictor of worse outcome in patients undergoing noncardiac surgery. The serum creatinine level should be determined preoperatively in these patients. Unless hypotension is present, the patient's antihypertensive medications should be continued, even on the day of surgery. Withdrawal of beta blockers and clonidine may be associated with rebound hypertension, with possible adverse operative and postoperative complications.
All patients with prosthetic valves should receive antibiotic prophylaxis before noncardiac surgery. Patients with mitral valve prolapse can undergo surgery without antibiotics specifically intended for valve protection. The decision to repair or replace a diseased valve should be made in the context of indications for valve surgery, independently of whether the patient is to undergo noncardiac surgery.
Aortic Stenosis. Patients with severe AS are at elevated risk for fatal and nonfatal complications during noncardiac surgery, as has been shown in many observational studies. Proceeding with noncardiac surgery in patients with uncorrected severe AS can have a mortality rate of 10%. Therefore, patients with severe symptomatic AS (who have a class I indication for valve replacement anyway) should undergo aortic valve replacement before noncardiac surgery. Balloon aortic valvuloplasty is a palliative option in patients who are not candidates for cardiac surgery. This approach carries considerable risk of complications, provides only temporary symptomatic benefit, and has not been proven to reduce the perioperative risk of noncardiac surgery.
Patients manifesting symptoms/signs of both CAD and AS should undergo appropriate testing (e.g., cardiac catheterization, echocardiography) followed by coronary revascularization and valve replacement before noncardiac surgery. Patients with isolated asymptomatic severe AS and no evidence of CAD can proceed with minor noncardiac surgery; however, care should be taken to avoid BP fluctuations and hemodynamic instability.6
Mitral Stenosis. Patients with severe mitral stenosis should undergo percutaneous valvuloplasty or surgical correction of the stenosis before undergoing noncardiac surgery. For patients with mild to moderate mitral stenosis, care should be taken to avoid tachycardia postoperatively, which can be induced by blood loss or surges in catecholamine levels. Tachycardia causes decreased filling time of the left ventricle, which can lead to decreased cardiac output, pulmonary congestion, and acute decompensated HF (ADHF). If the patient with mitral stenosis is asymptomatic, and has no evidence of pulmonary hypertension or atrial fibrillation, the risk of noncardiac surgery is not substantially higher than for normal patients.
Aortic and Mitral Regurgitation. The presurgical management of patients with regurgitant aortic and mitral valves depends on the severity and chronicity of the regurgitation. Patients with preserved left ventricular ejection fraction (LVEF) and volumes by echocardiography, as well as good functional capacity, can undergo noncardiac surgery without excess risk. For patients with severe regurgitant valvular lesions, few guidelines are available to describe the indications and appropriateness of valve repair or replacement before noncardiac surgery.6
In patients with aortic regurgitation, hemodynamic intraoperative assessment with a pulmonary artery catheter (PAC) is recommended to monitor afterload and to prevent hypotension, which can adversely affect these patients. Patients with severe mitral regurgitation may be treated with ACE inhibitors and diuretics. Any reduction in the ejection fraction should be considered abnormal and signals increased risk for congestive HF.6
Mechanical Prosthetic Valves. Patients with mechanical valves pose the special problem of anticoagulation. Stopping warfarin preoperatively is generally unavoidable in order to reduce bleeding complications, but doing so will concomitantly increase the risk of thromboembolic events. Both aortic and mitral mechanical prosthetic valves have an increased risk of thromboembolism off anticoagulation, compared with native valves. Mechanical mitral valves are at higher risk for embolization off anticoagulation than mechanical aortic valves, because of stasis and lower pressures on the left atrial side.6
Warfarin should be stopped 5 to 7 days before the procedure; if the patient is on aspirin, it should be stopped 1 week before the procedure, unless the surgeon has agreed to operate with the patient on daily aspirin. High-risk patients include those with mechanical mitral valve replacement, Björk-Shiley valves (old-generation valves), history of thromboembolic event in the past year, or at least three of the following four risk factors: atrial fibrillation, embolus at any time, hypercoagulable state, and mechanical prosthesis with LVEF of less than 30%. In high-risk patients, warfarin anticoagulation will be replaced with unfractionated heparin (UFH) or low molecular weight heparin (LMWH) until just before the surgery. UFH is stopped 4 to 6 hours before the procedure, and LMWH is stopped at least 12 hours beforehand.
Resumption of anticoagulation with UFH in the postoperative period is recommended; the heparin drip should be continued as a "bridge" until warfarin anticoagulation reaches its therapeutic target. If the patient is to undergo a minimally invasive procedure, anticoagulation can be withheld long enough to maintain the prothrombin time/international normalized ratio at the low end of the therapeutic range, and then resumed after the procedure.6
The presence of supraventricular or ventricular dysrhythmias preoperatively is considered an independent risk factor for adverse postoperative cardiac events. Patients with these arrhythmias are at an elevated risk of experiencing intraoperative and postoperative arrhythmias. However, they are not at increased risk for perioperative fatal or nonfatal MIs. In general, continuation of their home medications, such as beta blockers, calcium channel blockers, or anti-arrhythmics, is crucial in order to lessen the risk of perioperative arrhythmias and resultant hemodynamic instability. It is impressive how many times our cardiology consult service has been asked to see a patient with postoperative arrhythmia, only to find that his or her outpatient medication had been interrupted. On the other hand, in patients with no evidence of cardiac disease (structural or coronary) and no risk factors for arrhythmias (e.g., electrolyte abnormalities, acid-base disturbances, drug toxicities), perioperative cardiac monitoring or treatment is usually unnecessary.
Conduction disturbances should be dealt with preoperatively. If the patient has delayed conduction such as left bundle branch block, right bundle branch block, or first-degree atrioventricular block, it is unlikely to progress to complete heart block perioperatively. Patients with delayed conduction and heart block that, are asymptomatic and have no history of syncope do not require implantation of a temporary or permanent pacemaker. Patients with advanced heart block (second-degree Mobitz 2, third-degree) will need a temporary or permanent pacemaker prior to surgery.
When assessing a patient with a pacemaker or implantable cardioverter defibrillator (ICD), it is crucial to identify the device's type, mode, and indication for implantation. Other device-related and patient-related information should be collected preoperatively (Box 3). A device check, or interrogation, is also recommended preoperatively.
|Box 3: Issues to be Addressed Preoperatively in Patients with Pacemakers or Implantable Cardioverter Defibrillators|
|Identification of the type of device|
|Determination of pacing mode and/or defibrillatory algorithm|
|Knowledge of primary indication for implantation|
|Details of when device was implanted|
|When and where device was last interrogated|
|Anatomic position of current active pulse generator|
|Reset mode information|
|Confirmation of satisfactory thresholds|
The major issue of concern in patients with permanent pacemakers or ICDs is the potential for electromagnetic interference. The most common causes of interference in the hospital are listed in Box 4. The most common source of such interference during noncardiac surgery is electrocautery (electrocutting more than electrocoagulation). The intensity of electromagnetic interference from cauterization is related to the distance and direction of the current to the pacemaker generator and leads. If the cautery is to be used in close proximity to the generator, care should be taken to avoid loss of ventricular pacing, causing asystole. In such cases, temporary transcutaneous or transvenous pacing should be initiated preoperatively. If possible, the surgeon should use bipolar cautery, which, unlike unipolar cautery, disperses energy over a small surface area. She or he should use the lowest possible amplitude and apply the current in bursts rather than continuously. If the patient has an ICD, the device may interpret electrocautery as ventricular fibrillation, leading to an unnecessary shock. To avoid such problems, defibrillatory function should be turned off immediately prior to surgery, and turned on just afterwards. Pacing function should remain on. An external defibrillator should be immediately available and switched on, with defibrillation patches in place on the patient. If the patient takes antiarrhythmic or rate-controlling medication, it should be resumed postoperatively.
|Box 4: Most Common Causes of Interference with Pacemaker/Implantable Cardioverter Defibrillator Function in the Hospital|
|Magnetic resonance imaging|
|Transcutaneous electrical nerve stimulation|
|Drugs that interfere with pacemaker thresholds|
Because of the large amount of energy delivered, external cardioversion or defibrillation are other possible sources of electromagnetic interference in patients undergoing noncardiac surgery. Distinct problems with the operation of pacemakers and ICDs have been reported (Box 5). However, because of the isolation of the circuitry in titanium pacemakers, the introduction of noise protection algorithms, and the use of bipolar leads, the incidence of these complications is decreasing over time.
Some device functions need to be inactivated (rate-responsive pacing, ICD functions) before procedures; other devices need to be reprogrammed before surgery (e.g., in pediatric patients, patients with hypertrophic cardiomyopathy [HCM] or HF). In patients with HF, preoperative echocardiography, along with pacemaker interval programming, is advisable. Some patients have very slow or absent native cardiac rhythm, and as such are partially or completely pacemaker-dependent. The devices in such patients should be switched to an asynchronous pacing mode, which for single-lead ventricular pacemakers is VOO (ventricular pacing, no sensing, no response to sensing), and for dual-lead pacemakers is DOO (atrial and ventricular pacing, no sensing, no response to sensing). Once surgery is completed, the device should be reprogrammed back to its original mode.
In patients undergoing extracorporeal shock wave lithotripsy (ESWL), shock waves have the potential to trigger or inhibit myocardial depolarization, and more rarely, cause damage to the device itself. To avoid these issues, it is recommended that the device be temporarily reprogrammed to VOO or VVI mode during ESWL. The lithotripter will generally be set to trigger on the R-wave. In this way, shock waves will be synchronized with QRS complexes and unlikely to affect cardiac activity. In order to avoid damage to the device, the lithotripter's focal point should be maintained at least 6 inches (15 cm) away from the device.
|Box 5: Possible Problems with Pacemaker/Defibrillators During Surgery|
|Resetting to a backup, reset, or noise reversion pacing mode|
|Temporary or permanent inhibition of pacemaker output and/or defibrillatory function|
|Increasing pacing rate|
|Firing of defibrillator|
|Myocardial injury at the lead tip, causing failure to sense, capture, or both|
|Damage to the device's circuitry, resulting in pacing failure|
Patients with ADHF are at increased risk for perioperative complications. Goldman and colleagues1,2,5 have assigned the highest score in the cardiac risk index to signs of heart ADHF: jugular venous distention and the presence of S3. Shortness of breath, weight gain, lower extremity edema, crackles (rales) on lung exam, diffuse infiltrates on chest x-ray, and an elevated serum B-type natriuretic peptide (BNP) level are other subjective and objective evidence for the diagnosis of ADHF. However, the challenge remains, not only in the preoperative management of patients with known chronic HF, but also in the identification of patients with undiagnosed HF. In addition to diagnosing ADHF and assessing its severity, the ACC and AHA guidelines stress the need to identify the cause of the HF6 even though no study has proved a survival difference from HF of different causes.
Echocardiographic Assessment. The utility of echocardiography as a means of screening for HF in patients undergoing noncardiac surgery has been investigated.6 It was found that after adjusting for all confounding variables, parameters measured by echocardiography (e.g., LVEF, wall motion score) were not independent predictors of adverse cardiovascular outcomes. In addition, the LVEF had low sensitivity, low positive predictive values, and a likelihood ratio close to 1 for the end points examined. Thus, echocardiography does not add much to the risk-assessment tools used by clinicians prior to a noncardiac procedure.6 Dobutamine stress echocardiography (DSE) adds to the clinical predictors of risk used for a patient's evaluation preoperatively, since it can reveal inducible ischemia. An abnormal DSE finding has high sensitivity for predicting postoperative cardiac complications in patients undergoing nonvascular surgery.8 However, use of this diagnostic test should conform to consensus guidelines, and it should not be ordered indiscriminately.
Drug Therapy. Some authors have advocated the use of beta blockers in patients with HF undergoing noncardiac surgery, despite the limited number of HF patients involved in studies investigating the effect of beta blockade on perioperative complications.10 In addition, the need for noninvasive stress testing for HF patients having a score of 3 points or more on the RCRI index is recommended. In terms of drug management, we recommend continuing the same medications in asymptomatic HF patients. If patients have symptoms of ADHF, optimization of therapy should be attempted prior to surgery, since symptomatic patients have twice the perioperative complication rates of asymptomatic patients.8 Chronic HF patients using beta blockers should be kept on these medications before surgery, and continued on them afterwards. However, we recommend that beta blockers not be started immediately before surgery in HF patients. Beta blocker therapy can take months to uptitrate appropriately before its benefit is realized.
As for digoxin, preoperative initiation is not recommended, and whether or not to continue digoxin in patients taking it chronically should be determined on a case-by-case basis. Patients with New York Heart Association (NYHA) class III or IV HF have been shown to benefit from chronic spironolactone treatment; however, its use perioperatively remains unclear because the evidence is nonexistent. In patients undergoing noncardiac surgery, the use of ACE inhibitors the morning of surgery has been found to result in more frequent episodes of hypotension, whereas stopping ACE inhibitors before surgery led to bouts of hypertension afterwards. For most individuals, we recommend continuing the ACE inhibitors preoperatively. However, in patients with low-normal baseline BPs, it is reasonable to skip the dose of ACE inhibitors the morning of surgery to prevent hypotension during the operation.
PAC. The utility of PACs in patients undergoing high-risk surgery has been investigated by Sandham and colleagues.11 No benefit was found in patients who underwent surgery with the use of a PAC as opposed to patients without one. Of patients who had surgery with a PAC, 7.8% died versus 7.7% of those with no PAC; 13.4% of patients in the standard care group and 12.4% in the catheter group had a NYHA classification of III or IV. In light of the available evidence, we advise against the routine perioperative use of PACs in noncardiac surgery patients.
Studies assessing the risk of patients with pulmonary hypertension undergoing surgery are lacking. The risk of adverse events increases with higher NYHA classification. The optimal and safe type of anesthesia suitable for this population has not been determined, but expert opinion seems to favor epidural anesthesia whenever possible. Anesthesia should be administered by an experienced cardiovascular anesthesiologist in a center familiar with treating these high-risk patients. Patients receiving oral or inhaled treatment for pulmonary hypertension should be shifted to IV treatment if the expected withholding period of the drug is more than 12 to 24 hours. Early ambulation is preferable after surgery, and deep vein thrombosis prophylaxis with UFH or LMWH is advised in cases involving prolonged immobilization.
Patients with Eisenmenger’s syndrome should follow up routinely at a tertiary care center. Perioperative mortality associated with noncardiac surgery in patients with Eisenmenger’s syndrome is approximately 20%.6 The perioperative management of these high-risk patients involves avoiding fasting, volume depletion, and hypotension. If hypotension does occur, the patient should receive an alpha adrenergic agonist (e.g., methoxamine, phenylephrine) or IV fluids if the patient is hypovolemic. Endocarditis antibiotic prophylaxis should be routinely followed. All IV lines should be equipped with air filters to avoid paradoxical air embolism. The hematocrit level should be kept above normal, perhaps requiring transfusion of red blood cells, because a normal hematocrit value may not provide adequate end-organ oxygenation for patients with Eisenmenger's syndrome. An intra-arterial cannula should be used to monitor BP and oxygenation. The anesthetic technique should be chosen in order to minimize the likelihood of hypotension. The patient should be monitored closely in an intensive care unit (ICU) setting after surgery.
Patients with congenital heart disease were found to have an elevated risk of postoperative complications when undergoing noncardiac surgery compared with matched individuals without congenital heart disease;6 however, this overall risk is still not particularly high (5.8%). As part of the preoperative assessment of patients with congenital heart disease, care should be taken to ensure that the cardiac defect is limited to the heart, or is part of a well-characterized systemic syndrome. Syndromes involving the heart can involve the airway, gastrointestinal tract, or the central nervous system. The patient’s course in the hospital should be managed carefully to help avoid prolonged intubation, subglottic stenosis, worsened vascular access, or thrombosed vessels. Patients with congenital heart disease are predisposed to erythrocytosis because of the chronic cyanotic state that characterizes some conditions. As a result, volume depletion can lead to hyperviscosity and red blood cell sludging, with the possible result of cerebrovascular complications. Proper preoperative hydration, lowering the transfusion threshold, minimizing the fasting preoperative period, allowing water sips up to 2 hours before the operation, and scheduling the patient as the first case are some measures that can be taken to minimize complications in congenital heart disease patients.
There is a limited amount of data discussing the perioperative complications of HCM patients undergoing noncardiac surgery. A study investigating the outcomes of patients with HCM undergoing noncardiac surgery, compared with matched controls, showed a significantly higher rate of in-hospital death or MI for patients with HCM (6.7% versus 2.5%).12 In a subset of HCM patients-those with hypertrophic obstructive cardiomyopathy-care should be taken to avoid hypovolemia, decreased vascular resistance, increased venous capacitance, and the use of catecholamines, because they will all tend to increase left ventricular outflow tract obstruction. Worsened obstruction leads to increased myocardial oxygen demand, decreased cardiac output, diminished coronary blood flow, and systemic hypotension which, if unaddressed, can rapidly result in hemodynamic collapse. During the perioperative period, it is critical to continue hypertrophic obstructive cardiomyopathy patients' medications such as beta blockers and calcium channel blockers, which help to maintain a lower left ventricular outflow tract gradient.
Since the 1980s, the prevalence of obesity in the United States and the rest of the world has increased dramatically. In 2005, 31% of all Americans older than 20 years had a body mass index greater than 30. The use of bariatric surgery as a therapeutic option for weight reduction has increased 10 times from the 1990s to 2004 (140,000 bariatric surgeries done in 2004). Given the risks inherent in bariatric surgery, some investigators classify this surgery as a moderate- to high-risk surgery.13 In a retrospective case-control study of noncardiac surgery patients with morbid obesity, elevated body mass index itself was not found to portend a higher risk of cardiovascular complications or mortality.14 However, given the increased number of risk factors that are often associated with morbid obesity, and the risks inherent in the surgery itself, careful preoperative assessment of the morbidly obese patient should be performed to reduce mortality and morbidity.13
We recommend a detailed history (screening for angina, paroxysmal nocturnal dyspnea, orthopnea, and palpitations), including estimation of the functional capacity of the patient. Given the challenging nature of auscultation in such patients, the physical examination should be focused on gathering evidence of cor pulmonale, left ventricular dysfunction, peripheral arterial disease, and venous insufficiency. In addition, the morbidly obese patient should be evaluated routinely for obstructive sleep apnea (OSA) by historical features such as apneic episodes and daytime sleepiness, and by physical examination findings of increased neck circumference and waist-to-hip ratio. Some investigators recommend routine evaluation with a sleep study (polysomnography) before bariatric surgery. OSA increases complications (arrhythmias, MI, ICU admissions), cost of postoperative care, and length of hospital stay. Some investigators recommend the use of continuous positive airway pressure for these patients in the perioperative period. This would be expected to decrease the rate of postoperative complications.15
In addition, the patient should be managed by a multidisciplinary team including an anesthesiologist, a nutritionist, a surgeon and a cardiologist or pulmonologist. In patients with OSA, cautious use of analgesia should be exercised postoperatively because some agents (such as narcotics) induce respiratory depression. The second most common complication postoperatively in such patients is pulmonary embolism; therefore, we recommend prophylactic anticoagulation in this population, taking into account the weight, renal function and bleeding risk of each individual patient.16
In addition to assessing the risk imposed by the patient's various medical conditions, one must also take into consideration the timing of the surgery (elective vs. urgent vs. emergent), as well as the risk-level of the procedure itself. In a study intended to develop a predictive model for operative risk, patients undergoing various types of surgeries (except cardiac surgery and cesarean section) were enrolled. Multivariate analysis has revealed that the complexity of the surgical procedure (low, moderate, or high risk) according to the modified Johns Hopkins surgical criteria (Box 6) and the mode of surgery (elective, urgent, or emergent) were two of the four predictors of in-hospital death. This is secondary to the fact that the more urgent the surgery, the less time is available to adequately assess and reduce the patient’s risk. The other two factors were age and American Society of Anesthesiologists’ grade (Table 2).
|Box 6: Modified Johns Hopkins Surgical Criteria|
|Minimal to Mild Risk Independent of Anesthesia|
|Minimal to Moderately Invasive Procedure|
|Potential Blood Loss Less than 500 mL|
|Moderately to Significantly Invasive Procedures|
|Potential Blood Loss of 500-1,500 mL|
|Moderate Risk to Patient Independent of Anesthesia|
|Highly Invasive Procedure|
|Potential Blood Loss More than 1,500 mL|
|Major to Critical Risk to Patient Independent of Anesthesia
||Usual Postoperative ICU Stay with Invasive Monitoring|
ICU, intensive care unit; PACU, post-anesthesia care unit.
|II||Mild systemic disease; no functional limitation|
|III||Severe systemic disease; definite functional limitation|
|IV||Severe systemic disease that is constant threat to life|
|V||Moribund patient; unlikely to survive 24 hours with or without operation|
In patients undergoing elective surgery, two risk factors have been found to significantly influence cardiovascular mortality within 30 days of the operation: prior MI and renal failure. In a case-control study on patients who underwent urgent or emergent surgical procedures, a history of congestive HF was the only significant predictor of 30-day mortality on multivariate analysis.6
The ACC and AHA have jointly classified surgeries into various categories of risk (Box 7). High-risk procedures have an adverse cardiac event risk higher than 5% and include emergent major procedures, major vascular surgeries (except carotid endarterectomy, which is intermediate risk), and prolonged procedures with fluid shifts and possible blood loss. Low-risk procedures have a risk lower than 1% and include all endoscopic procedures, superficial procedures, and cataract and breast surgeries. The remaining procedures are classified as intermediate risk, greater than 1% but less than 5%. In addition to this stratification, the operative experience of the surgeon and volume of the medical center influence the cardiovascular outcomes, especially in vascular surgeries.
|Box 7: Cardiac Risk* Stratification for Noncardiac Surgical Procedures|
|High Risk (reported cardiac risk often >5%)|
|Emergent major operations, particularly in older patients|
|Aortic and other major vascular surgeries|
|Peripheral vascular surgery|
|Anticipated prolonged surgical procedures associated with large fluid shifts, blood loss, or both|
|Intermediate Risk (reported cardiac risk generally >1% but <5%)|
|Head and neck surgery|
|Intraperitoneal and intrathoracic surgery|
|Low Risk (reported cardiac risk generally <1%)†|
* Combined incidence of cardiac death and nonfatal myocardial infarction.
†Does not generally require further preoperative cardiac testing.
A considerable number of studies have attempted to identify medications that could minimize the perioperative risk of patients undergoing noncardiac surgery. These studies involved beta blockers, lipid-lowering agents, clonidine, and other drugs (verapamil, diltiazem).
Several older studies showed a decreased incidence of death and MI during and after noncardiac surgery in patients who were given perioperative beta blockers. This benefit was most accentuated in patients who were intermediate- or high-risk. No benefit was found in low-risk patients. Some studies even reported harm from the use of beta blockers in low-risk populations. Other historical studies that started therapy hours before surgery did not find benefit from beta blockers.10,17 The benefit was mainly found in nonblinded studies such as the DECREASE family of trials, which titrated the dose of beta blockers over many days to a target heart rate close to 65 beats/min.18,19
In the more recent double-blinded, placebo-controlled, randomized POISE trial,20 patients received 100 mg metoprolol succinate or placebo 2 to 4 hours prior to their planned surgery, then continued to receive 100 to 200 mg long-acting metoprolol or placebo daily for 30 days after surgery. The results of POISE showed that, compared with those taking placebo, patients receiving metoprolol had fewer perioperative MIs. However, those taking metoprolol were also more likely to have a stroke, or to die. These findings led to a more conservative set of guidelines for perioperative beta blocker usage.21 Individuals who were already on beta blockers represented the only class I indication, mainly to avoid the harmful effects of beta blocker withdrawal (tachycardia and hypertension). Beta blocker titration to heart rate and BP were "probably recommended" for patients undergoing vascular surgery who were at high cardiac risk due to CAD or ischemia on stress testing (class IIa), and they were "reasonable" for patients undergoing vascular surgery or intermediate-risk surgery with more than one clinical risk factor (class IIa).
In 2011, the validity of data from the DECREASE trials was called into question. As such, evidence supporting initiation of perioperative beta blockers has become even weaker. Unless future data can demonstrate a clear benefit for starting beta blockers perioperatively, our recommendation would be to give these drugs only to patients who have already been taking them at home, since a recent meta-analysis of nine secure trials found that the 30-day relative risk of all-cause mortality was 27% higher in patients treated with beta blockers, compared with those taking placebo.22
The use of lipid-lowering agents has been advocated by some investigators as a possible means of reducing perioperative cardiac complications, presumably via atherosclerotic plaque stabilization. One retrospective study of patients undergoing vascular surgery found that statin use was associated with a reduced incidence of the composite end point (death, nonfatal MI, and ischemia).9 However, statin therapy was not associated with a statistically significant difference in nonfatal MI or death. This study was limited by a retrospective design and by nonspecified dose and duration of statins. Another observational study reviewed 780,000 records of adult patients who had undergone noncardiac surgery, comparing those taking statins in the hospital to those who were not.23 With adjustment for baseline factors, patients on statin drugs had an in-hospital mortality rate of 2.18% versus 3.15% for those not on statins. This amounted to a statistically significant adjusted odds ratio of 0.62 for hospital mortality, favoring statin treatment.
At present, we do not have randomized controlled trial data to support preoperative initiation of statin drugs. By current ACC/AHA guidelines, the only class I indication for perioperative statin therapy is for those patients who were already taking statins at home. For patients undergoing vascular surgery, with or without clinical risk factors, perioperative statin use is felt to be "reasonable" (class IIa recommendation).21
The evidence for benefit of alpha-2 agonists in the perioperative setting has been shown in two meta-analyses and one randomized trial, in which it was found that clonidine given before noncardiac surgery reduced the incidence of perioperative ischemia and mortality. However, this benefit was found in only subgroup of patients (those undergoing vascular surgery) and not in others; thus, we cannot extrapolate the findings in this study to other surgical patients. Until evidence for benefit of starting this medication preoperatively has been established, we do not recommend using clonidine unless the patient had been taking it chronically. There exists a class IIb recommendation in the guidelines to consider use of alpha-2 agonists for perioperative control of hypertension in patients with known CAD or at least one clinical risk factor.21
Some patients may lead a sedentary lifestyle secondary to orthopaedic conditions, rheumatologic conditions, or morbid obesity; this lifestyle prevents clinicians from properly assessing the occurrence of symptoms related to cardiac supply-and-demand mismatch. In addition to the use of the RCRI to risk-stratify these patients, some laboratory tests can provide evidence for ventricular overload and increased left ventricular end-diastolic pressure.
One such blood test is BNP. In one study, serum BNP measurements were made in 1,590 patients undergoing noncardiac surgery. In a striking correlation, none (0/954) of the patients who had a BNP level less than 100 pg/mL had a postoperative cardiac complication, compared with 5% (16/349) of the patients with BNP between 100 and 200 pg/mL, 13% (28/223) of those with BNP level between 200 and 300 pg/mL, and 81% (52/64) of those with BNP greater than 300 pg/mL. The authors found that a cutoff value of 189 pg/mL most effectively stratified the patients into low- and high-risk groups. In fact, a high-risk BNP level predicted the occurrence of postoperative cardiac events more reliably than the widely-utilized Goldman index.24
In addition to its utility in predicting the occurrence of short-term postoperative complications, serum N-terminal proBNP (NT proBNP) was found by other investigators to predict the long-term occurrence of cardiac complications after major vascular surgery (abdominal aortic aneurysm repair or lower-extremity bypass surgery).25 At 6 months of followup, patients with NT proBNP values greater than 319 pg/dL had a 4-fold higher risk of mortality and an 11-fold higher risk of MACE than those with NT proBNP less than 319 pg/dL.
Feringa and colleagues found that elevated levels of NT proBNP are associated with troponin T release and myocardial ischemia in patients undergoing major vascular surgery.26 In this study, the optimal value of NT proBNP to predict the risk of myocardial ischemia and troponin T release was 270 ng/L. The association found between NT proBNP and cardiac complications and mortality was independent of comorbidities, medication use, and cigarette smoking.
The major limitation of the aforementioned studies is that they primarily included patients undergoing major vascular surgery; hence the applicability of these results to other patient populations and surgeries is unclear. In patients undergoing major vascular surgery, we recommend stratifying patients according to the RCRI and measuring NT proBNP. Patients with an elevated NT proBNP level would be expected to benefit from a more rigorous diagnostic workup, compared to those with lower levels. Also, such high-risk patients should be managed aggressively with lifestyle modifications and medications (such as diuretics and afterload-reducing agents) both preoperatively and during long-term follow up, in order to reduce the likelihood of adverse cardiac events.(Table 3)
|Study||Test||Cutoff Value||OR/HR of MI||OR/HR of Death/Nonfatal MI||OR/HR of all Cardiac Complications|
|Dernellis et al21||BNP||≥189 pg/mL||NA||NA||28.78|
|Feringa et al23||NT proBNP||≥270 ng/L||1.49*||1.59*||NA|
|Feringa et al22||NT proBNP||≥319 ng/L||NA||4†||10.9‡|
|Feringa et al24||HbA1c||≥7%||2.8||3.6||5.6|
|Feringa et al24||IGT||5.6-7.0 mmol/L§||2.2||2||1.9|
|Feringa et al24||DM||≥7 mmol/L§||2.6||2.7||3.1|
BNP, B-type natriuretic peptide; DM, diabetes mellitus; HbA1c, hemoglobin A1c; HR, hazard ratio; IGT, impaired glucose tolerance; MI, myocardial infarction; NA, not available; NT proBNP, N-terminal pro–B-type natriuretic peptide; OR, odds ratio.
*Odds ratio for each 1 ng/L rise in the natural logarithm of baseline NT proBNP.
†Hazard ratio for all-cause mortality.
‡Hazard ratio for major adverse cardiac events.
§Fasting glucose values.
Diabetes mellitus and impaired glucose tolerance (IGT) are associated with an increased risk of cardiovascular events. The vast majority of diabetic patients we encounter in the perioperative setting have type 2 diabetes. The presence of abnormal glucose homeostasis before noncardiac surgery has been shown by many investigators to be a marker of poor outcomes during the postoperative period.
In one study involving patients undergoing major vascular surgery, patients with IGT and diabetes mellitus experienced higher rates of MI, troponin T release, 30-day cardiac complications, and cardiac mortality compared to patients with normoglycemia.27 Patients with a HbA1c higher than 7% had a worse outcome compared with patients with HbA1c lower than 7%.
A similar study was done in patients who were preop for noncardiovascular surgery to assess for a correlation between baseline glucose elevation and risk of perioperative complications.28 IGT was associated with a three-fold increased mortality in patients undergoing nonvascular surgery compared with normoglycemic controls. Patients who had glucose levels in the diabetic range had four-fold increased cardiovascular mortality as compared with normoglycemic patients.28 However, given the fact that this study was retrospective in design, the results would need to be confirmed by a prospective trial.
In another study, an oral glucose tolerance test (OGTT) was performed on prospective vascular surgery patients to diagnose new cases of diabetes mellitus and IGT, as well as to investigate the relationship between OGTT results and perioperative complications.29 IGT and diabetes mellitus were detected in 25.7% and 10.6% of the population studied, respectively. The patients who had a positive OGTT had higher rates of myocardial ischemia, MI, and mortality compared with normoglycemic subjects.29
In patients with no history of diabetes, but with cardiac risk factors (e.g., hypertension, dyslipidemia), we recommend checking a fasting blood glucose level prior to noncardiac surgery. If the level is abnormal, then it should be confirmed by another fasting blood sugar. If the patient is found to have diabetes, he or she should be reclassified according to the RCRI and managed accordingly. In patients with no cardiac risk factors, we recommend taking a random finger stick glucose measurement. Patients who show abnormal results (>200 mg/dL) should be subjected to fasting blood glucose measurement. Patients with IGT or diabetes mellitus should undergo extensive lifestyle modifications, and insulin treatment if needed, depending on the blood glucose levels. Preferably, therapy should be initiated in the hospital. However, oral hypoglycemic drugs should be withheld the morning of the surgery to prevent intraoperative hypoglycemia.
Preoperative insulin doses are generally cut in half. During fasting, full doses would tend to cause hypoglycemia, but total withdrawal of insulin can result in hyperosmotic nonketotic state in type 2 diabetics, or diabetic ketoacidosis in type 1 diabetics. In addition, blood sugar should be monitored closely during the course of the operation; hyper- and hypoglycemic episodes should be treated appropriately (see Table 3), and there are numerous available algorithms which can help maintain intraoperative normoglycemia.
Cardiac events are among the most worrisome complications of noncardiac surgeries. Fortunately, a few steps can help minimize the risks: