American College of Cardiology

EAGLE ET AL., PERIOPERATIVE CARDIOVASCULAR EVALUATION FOR NONCARDIAC SURGERY UPDATE
http://www.acc.org/clinical/guidelines/perio/update/periupdate_index.htm

ACC/AHA Guideline Update for Perioperative Cardiovascular Evaluation for Noncardiac Surgery

A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1996 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery)

This is a Guideline Update of the 1996 Perioperative Guidelines. To highlight the changes, deleted text is indicated by strikeout, and revised text is presented in red. A clean version of the document, with changes fully incorporated, is available for download and print.

II. General Approach to the Patient

Preoperative cardiac evaluation must be carefully tailored to the circumstances that have prompted the consultation and the nature of the surgical illness. Given an acute surgical emergency, preoperative evaluation will might have to be limited to simple and critical tests such as a rapid assessment of cardiovascular vital signs, volume status, hematocrit, electrolytes, renal function, urine analysis, and electrocardiogram (ECG). Only the most essential tests and interventions are appropriate until the acute surgical emergency is resolved. A more thorough evaluation can be conducted after surgery. In some circumstances, surgery is not performed as an emergency procedure, but good care dictates prompt surgery. In patients in whom coronary revascularization is not an option, it is often not necessary to perform a test. Under other, less urgent circumstances, the preoperative cardiac evaluation may lead to a variety of responses. Sometimes this situation may include cancellation of an elective procedure. In this era of managed care and cost containment, the special needs of patients with comorbid disease who undergo surgery must be considered. "Same day" admission, which has become standard for most operations because of cost-containment issues, may lead to an abbreviated preoperative assessment and could result in greater morbidity and higher cost in high-risk patients. Further study of this question is needed.

The cardiac consultant must carefully consider the question that he or she has been asked to answer. A misinterpreted ECG anomaly, atypical chest pain, or a benign arrhythmia in an otherwise healthy patient may require no further workup or special precaution, whereas suspicion of previously unsuspected coronary artery disease (CAD) or heart failure (HF) in a patient scheduled for an elective procedure may justify a more extensive workup ^(4-6).

The consultant must also bear in mind that the perioperative evaluation may be the ideal opportunity to affect long-term treatment of a patient with significant cardiac disease or risk of such disease. The referring physician and patient should be informed of the results of the evaluation and implications for the patient's prognosis. The consultant can also assist in planning for follow-up.

A. History

A careful history is crucial to the discovery of cardiac and/or comorbid diseases that would place the patient in a high surgical risk category. The history should seek to identify serious cardiac conditions such as prior angina, recent or past myocardial infarction (MI), CHF, and symptomatic arrhythmias and also determine whether the patient has a prior history of orthostatic intolerance or if a pacemaker or implantable cardioverter defibrillator (ICD) or a history of orthostatic intolerance has been implanted. Modifiable risk factors for coronary heart disease (CHD) should be recorded along with evidence of associated diseases, such as peripheral vascular disease, cerebrovascular disease, diabetes mellitus, renal impairment, and chronic pulmonary disease. In patients with established cardiac disease, any recent change in symptoms must be ascertained. Accurate recording of current medications and dosages is essential. Use of alcohol and over-the-counter and illicit drugs should be documented.

The history should also seek to determine the patient's functional capacity (Table 1). An assessment of an individual's capacity to perform a spectrum of common daily tasks has been shown to correlate well with maximum oxygen uptake by treadmill testing ⁽⁷⁾. A patient classified as high risk owing to age or known CAD but who is asymptomatic and runs for 30 minutes daily may need no further evaluation. In contrast, a sedentary patient without a history of cardiovascular disease but with clinical factors that suggest increased perioperative risk may benefit from a more extensive preoperative evaluation ^(5,6,8,9). The preoperative consultation may represent the first careful cardiovascular evaluation for the patient in years, and in some instances, ever. For example, inquiry regarding symptoms suggestive of angina or anginal equivalents such as dyspnea or HF may establish or suggest these diagnoses for the first time.

B. Physical Examination

A careful cardiovascular examination should include an assessment of vital signs (including measurement of blood pressure in both arms), carotid pulse contour and bruits, jugular venous pressure and pulsations, auscultation of the lungs, precordial palpation and auscultation, abdominal palpation, and examination of the extremities for edema and vascular integrity. The presence of an implanted pacemaker or ICD can also be confirmed on physical examination. More detailed observations will be dictated by specific circumstances.

The following points are worth emphasizing:

• The general appearance provides invaluable evidence regarding the patient's overall status. Cyanosis, pallor, dyspnea during conversation or with minimal activity, Cheyne Stokes respiration, poor nutritional status, obesity, skeletal deformities, tremor, and anxiety are just a few of the clues that can be recognized by the skilled physician.
• In patients with acute HF, pulmonary rales and chest X-ray evidence of pulmonary congestion correlate well with elevated pulmonary venous pressure. However, in patients with chronic HF, these findings may be absent. An elevated jugular venous pressure or a positive hepatojugular reflux are more reliable signs of hypervolemia in these patients ^(10,11). Peripheral edema is not a reliable indicator of chronic HF unless the jugular venous pressure is elevated or the hepatojugular test is positive.
• A careful examination of the carotid and other arterial pulses is essential. The presence of associated vascular disease should heighten suspicion of occult CAD.
• Cardiac auscultation will often provide useful clues to underlying cardiac disease. When present, a third heart sound at the apical area suggests a failing left ventricle, but its absence is not a reliable indicator of good ventricular function ⁽¹¹⁾.
• If a murmur is present, the clinician will need to decide whether or not it represents significant valvular disease. Detection of significant aortic stenosis is of particular importance because this lesion poses a higher risk for noncardiac surgery ⁽¹²⁾. Significant mitral stenosis or regurgitation increases the risk of HF. Aortic regurgitation and mitral regurgitation may be minimal, yet they predispose the patient to infective endocarditis should bacteremia occur after surgery. In these conditions, especially if mitral regurgitation is rheumatic in origin or due to mitral valve prolapse, consideration must be given to endocarditis prophylaxis ⁽¹³⁾.

C. Comorbid Diseases

The consultant must evaluate the cardiovascular system within the framework of the patient's overall health. Associated conditions often heighten the risk of anesthesia and may complicate cardiac management. The most common of these conditions are discussed below:

1. Pulmonary Disease
The presence of either obstructive or restrictive pulmonary disease places the patient at increased risk of developing perioperative respiratory complications. Hypoxemia, hypercapnia, acidosis, and increased work of breathing can all lead to further deterioration of an already compromised cardiopulmonary system. If significant pulmonary disease is suspected by history or physical examination, determination of functional capacity, response to bronchodilators, and/or evaluation for the presence of carbon dioxide retention through arterial blood gas analysis may be justified. If there is evidence of infection, appropriate antibiotics are critical. Steroids and bronchodilators may be indicated, although the risk of producing arrhythmia or myocardial ischemia by beta-agonists must be considered.

2. Diabetes Mellitus
A variety of metabolic diseases may accompany cardiac disease. Diabetes mellitus is the most common. Its presence should heighten suspicion of CAD, particularly because CAD and myocardial ischemia is are more likely in patients with diabetes and more likely to be silent ^(230,231). Older patients with diabetes are more likely to develop HF postoperatively than those without diabetes mellitus even after adjustment for treatment with angiotensin converting enzyme (ACE) inhibitors. Management of blood glucose levels in the perioperative period may be difficult. Fragile diabetic patients need careful treatment with adjusted doses or infusions of short-acting insulin based on frequent blood sugar determinations. More stable diabetic patients may be treated with long-acting insulin or oral hypoglycemics. Historically, it has been acceptable to maintain relatively high glucose levels perioperatively to avoid the attendant risks of hypoglycemic episodes. However, aggressive perioperative glucose control in coronary bypass surgery patients by a continuous, intravenous insulin infusion was superior to intermittent subcutaneous insulin administration in significantly reducing postoperative wound infection ⁽²³²⁾. Similar benefit may occur surrounding noncardiac surgery ⁽²³³⁾. It is acceptable to maintain blood glucose at relatively high levels perioperatively and it is far preferable to tight control with its attendant risks for hypoglycemic episodes.

3. Renal Impairment
Azotemia is commonly associated with cardiac disease and is associated with an increased risk of cardiovascular events. often complicates its management. Maintenance of adequate intravascular volume for renal perfusion during diuresis of a patient with HF is often challenging. Excessive diuresis in combination with initiation of ACE inhibitors or angiotensin receptor blockers may result in an increase in blood urea nitrogen and plasma serum creatinine concentrations. In patients with known vascular disease, a small increase in blood urea nitrogen and creatinine may suggest the presence of renal artery stenosis. However, small increases in blood urea nitrogen and serum creatinine concentrations are not an indication to discontinue these drugs, because they have been shown to improve survival in patients with HF due to systolic dysfunction. Preoperative evaluation of the patient on dialysis or after renal transplantation should essentially be the same as that for those patients not afflicted with these conditions. Many are elderly and have heart problems similar to the general population. However, a significant number of are diabetic, and similar patients are quite predisposed to CHD. They should have adequate dialysis preoperatively to prevent pulmonary edema and the consequence of impaired oxygenation or tendency to bleed due to significant azotemia. With the transplant patient, the major issue is management of immunosuppression in the perioperative period. Pre-existing renal disease (preoperative serum creatinine levels between 1.4 and 2 mg per dl or above) has been identified as a risk factor for postoperative renal dysfunction and increased long-term morbidity and mortality compared with patients without renal disease ⁽²³⁴⁾. In coronary artery bypass patients who are more than 70 years old, preoperative creatinine levels greater than 2.6 mg per dl place the patient at much greater risk for chronic dialysis postoperatively than those with creatinine levels below 2.6 mg per dl ⁽²³⁵⁾. Intuitively, one might extrapolate these findings to those older patients with comparable creatinine levels who undergo major noncardiac surgical procedures. One large study has shown that a preoperative creatinine level greater than 2 mg per dl is a significant, independent risk factor for cardiac complications after major noncardiac surgery ⁽²³⁶⁾.

4. Hematologic Disorders
Anemia imposes a stress on the cardiovascular system that may exacerbate myocardial ischemia and aggravate HF ⁽¹⁴⁾. Preoperative transfusion, when used appropriately in patients with advanced CAD and/or HF, may reduce perioperative cardiac morbidity. However, with current concern about possible transmission of human immunodeficiency virus and hepatitis through the use of blood products, a conservative approach with respect to transfusion is warranted. Hematocrits less than 28% are associated with an increased incidence of perioperative ischemia and postoperative complications in patients undergoing prostate and vascular surgery ^(237-239).

Polycythemia, thrombocytosis, and other conditions that increase blood viscosity may increase the risk of thromboembolism and/or hemorrhage. Appropriate steps to reduce these risks should be considered and tailored to the individual patient's particular circumstances.

D. Ancillary Studies

The consultant should review all pertinent available laboratory data. In this era of cost containment, the laboratory data available may be minimal. Therefore, the consultant may require additional tests such as blood chemistries and a chest X-ray on the basis of history and physical examination. Blood levels of cardiac drugs, including digoxin, should be obtained only when there are specific indications, such as changing renal function, recent change in dose, or symptoms suggesting toxicity.

In patients referred for cardiac consultation, anThe ECG is frequently obtained as almost always indicated as part of a preoperative evaluation in all patients over a specific age or undergoing a specific set of procedures. In fact, an abnormal ECG report is often the reason that consultation is requested. If not, the ECG is almost always indicated as part of a cardiac consultation. Metabolic and electrolyte disturbances, medications, intracranial disease, pulmonary disease, etc., can alter the ECG. Conduction disturbances, such as bundle-branch block or first-degree atrioventricular block, may lead to concern but usually do not justify further workup. The same is often true of asymptomatic ventricular arrhythmias, even in the presence of structural heart disease ^(240,241). in the absence of significant underlying structural heart disease. On the other hand, subtle ECG clues can point the way to a clinically silent condition of major import.

The basic clinical evaluation obtained by history, physical examination, and review of the ECG usually provides the consultant with sufficient data to estimate cardiac risk. In an attempt to codify those clinical and laboratory factors that influence outcome, numerous investigators have developed risk indices over the past 25 years based on multivariate analyses ^(12,15-24). Although some authors have suggested a scoring system that assigns more weight to some factors than others and sums these to arrive at a composite risk ^(12,22,24), most recent articles have suggested simpler criteria ^(15-21)(236). For example, Lee et al derived and validated a "simple index" for the prediction of cardiac risk for stable patients undergoing nonurgent major noncardiac surgery ⁽²³⁶⁾. Six independent risk correlates were identified: ischemic heart disease (defined as history of MI, history of positive treadmill test, use of nitroglycerin, current complaints of chest pain thought to be secondary to coronary ischemia, or ECG with abnormal Q waves); congestive HF (defined as history of HF, pulmonary edema, paroxysmal nocturnal, dyspnea, peripheral edema, bilateral rales, S3, or X-ray with pulmonary vascular redistribution); cerebral vascular disease (history of transient ischemic attack or stroke); high-risk surgery (abdominal aortic aneurysm, other vascular, thoracic, abdominal, or orthopedic surgery); preoperative insulin treatment for diabetes mellitus; and preoperative creatinine greater than 2 mg per dl. Increasing numbers of risk factors correlated with increased risk, yet the risk was substantially lower than described in many of the original indices. These improvements in outcome most likely reflect selection bias with respect to who presents for elective surgery and advances in surgical technique and anesthesia and in the management of CAD both perioperatively and in general.

Table 1 lists clinical predictors of increased perioperative risk of MI, HF, and death established by multivariate analyses ^(12,15-24). In clinical practice, more weight is attachedshould be given to active conditions than to dormant ones, while the degree of deviation from the norm is used as an implicit modifier. Although the scoring systems may assist some practitioners in defining specific risk categories, there was general consensus among committee members that clinical factors could be placed into the following three categories:

• Major predictors, when present, mandate intensive management, which may result in delay or cancellation of surgery unless it is emergent.
• Intermediate predictors are well-validated markers of enhanced risk of perioperative cardiac complications and justify careful assessment of the patient's current status.
• Minor predictors are recognized markers for cardiovascular disease that have not been proven to independently increase perioperative risk.

A history of MI or pathological abnormal Q waves by ECG is listed as an intermediate predictor, whereas an recent acute MI (The American College of Cardiology National Cardiovascular Database Registry Library (defined as at least one documented MI less than or equal to 7 days before the examination) or recent MI (greater than 7 days but less than or equal to 1 month before the examination) with evidence of important ischemic risk by clinical symptoms or noninvasive studys recent MI as greater than 7 days but less than or equal to 1 month [30 days is a major predictor. This definition reflects the consensus of the ACC Cardiovascular Database Committee. In this way, the separation of MI into the traditional 3- and 6-month intervals has been avoided ^(12,25). Current management of MI provides for risk stratification during convalescence ⁽²⁶⁾. If a recent stress test does not indicate residual myocardium at risk, the likelihood of reinfarction after noncardiac surgery is low. Although there are no adequate clinical trials on which to base firm recommendations, it appears reasonable to wait 4 to 6 weeks after MI to perform elective surgery.

Table 2 presents a validated method for assessing functional capacity from a carefully obtained history. This method represents an important aspect of evaluating overall cardiac risk and planning appropriate preoperative testing.

Table 3 stratifies the risk of various types of noncardiac surgical procedures. This risk stratification is based on several reported studies ^{(12,15,21,22,25,28-30)}. It is clear that major emergent operations in the elderly (i.e., those violating a visceral cavity and those likely to be accompanied by major bleeding or fluid shifts) place patients at highest risk. Vascular procedures are higher risk and, primarily because of the likelihood of associated coronary disease, justify careful preoperative screening for myocardial ischemia in many instances. This aspect of decision making is covered more extensively in Section IV.

E. Stepwise Approach to Perioperative Cardiac Assessment

Figure 1 presents in algorithmic form a framework for determining which patients are candidates for cardiac testing. For clarity, categories have been established as black and white, but it is recognized that individual patient problems occur in shades of gray. The clinician must consider several interacting variables and give them appropriate weight. Furthermore, there are no adequate controlled or randomized clinical trials to help define the process. Thus, collected observational data and expert opinion form the basis of the proposed algorithm. However, since publication of the Perioperative Cardiovascular Evaluation Guidelines in 1996 ⁽²⁴²⁾, several studies have suggested that this stepwise approach to the assessment of CAD is both efficacious and cost-effective ^(243-246).

Step 1 (Figure 1). The consultant should determine the urgency of noncardiac surgery. In many instances, patient or surgery-specific factors dictate an obvious strategy (i.e., immediate surgery) that may not allow for further cardiac assessment or treatment. In such cases, the consultant may function best by providing recommendations for perioperative medical management and surveillance. Selected postoperative risk stratification is often appropriate in patients with elevated risk for long-term coronary events who have never had such an assessment before. This is usually initiated after the patient has recovered from blood loss, deconditioning, and other postoperative complications that might confound interpretation of noninvasive test results.

Step 2 (Figure 1). Has the patient undergone coronary revascularization in the past 5 years? If the patient has had complete surgical revascularization in the past 5 years or percutaneous coronary angioplasty intervention (PCI) from 6 months to 5 years previously, and if his or her clinical status has remained stable without recurrent signs or symptoms of ischemia in the interim, the likelihood of perioperative cardiac death or MI is extremely low ⁽³¹⁾. Further cardiac testing in this circumstance is generally not necessary.

Step 3 (Figure 1). Has the patient undergone a coronary evaluation in the past 2 years? If an individual has undergone extensive coronary evaluation with either noninvasive or invasive techniques within 2 years, and if the findings indicate that coronary risk has been adequately assessed with favorable findings, repeat testing is usually unnecessary. An exception to this rule is the patient who has experienced a definite change or new symptoms of coronary ischemia since the prior coronary evaluation.

Step 4 (Figure 1). Does the patient have one of the unstable coronary syndromes or major clinical predictors of risk (Table 1)? In patients being considered for elective noncardiac surgery, the presence of unstable coronary disease, decompensated HF, hemodynamically significant arrhythmias, or severe valvular heart disease usually leads to cancellation or delay of surgery until the cardiac problem has been clarified and appropriately treated. Examples of unstable coronary syndromes include recent previous MI with evidence of important ischemic risk by clinical symptoms or noninvasive study, unstable or severe angina, and new or poorly controlled ischemia-mediated HF. Many patients in these circumstances are referred for coronary angiography to assess further therapeutic options.

Step 5 (Figure 1). Does the patient have intermediate clinical predictors of risk (Table 1)? The presence or absence of angina pectoris, prior MI by history or ECG, compensated or prior HF, preoperative creatinine greater than 2 mg per dl or diabetes mellitus helps to further stratify clinical risk for perioperative coronary events. For patients with or without these intermediate clinical risk predictors, consideration of functional capacity (as determined by history of daily activities) and level of surgery-specific risk (Table 3) allows a rational approach to identifying which patients may most benefit from further noninvasive testing.

Functional status has been shown to be reliable for perioperative and long-term prediction of cardiac events ^{(32,33,34)(243,247,248)}. If the patient has not had a recent exercise test, functional status can usually be estimated from the ability to perform the activities of daily living ⁽²⁴⁷⁾. Functional capacity can be expressed in metabolic equivalent (MET) levels; the oxygen consumption (VO₂) of a 70-kg, 40-year-old man in a resting state is 3.5 mL per kg per minute or 1 MET. For this purpose, functional capacity has been classified as excellent (greater than 107 METs), good (7 to 10 METs), moderate (4 to 7 METs), poor (less than 4 METs), or unknown. Multiples of the baseline MET value provide a uniform terminology across different exercise protocols to express aerobic demands for specific activities. Maximum and submaximum levels of work differ per unit of time according to the exercise protocol used. Thus, six minutes of a Naughton protocol is not equivalent to six minutes on a standard Bruce protocol in terms of work performed and energy expended. Multiples of the baseline MET value can be used to express aerobic demands for specific activities, and the use of the MET level is a more meaningful and useful expression of exercise capacity than attempting to express functional capacity in terms of protocol time and stages reached during an exercise test. The predicted MET level for a certain activity is influenced by the degree of conditioning and genetic predisposition. Perioperative cardiac and long-term risks are increased in patients unable to meet a 4-MET demand during most normal daily activities ⁽²⁴⁷⁾. In one series of 600 consecutive patients undergoing major noncardiac procedures, perioperative myocardial ischemia and cardiovascular events were more common in patients reporting poor exercise tolerance (inability to walk 4 blocks or climb 2 flights of stairs) even after adjustment for baseline characteristics known to be associated with increased risk ⁽²⁴⁷⁾. The likelihood of a serious complication occurring was inversely related to the number of blocks that could be walked (p=0.006) or flights of stairs that could be climbed (p=0.01). Examples of leisure activities associated with less than 4 METs are baking, slow ballroom dancing, golfing with a cart, playing a musical instrument, and walking at a speed of approximately 2 to 3 mph. Activities that require more than 4 METs include moderate cycling, climbing hills, ice skating, roller blading, skiing, singles tennis, and jogging. The Duke Activity Status Index (Table 2) contains questions that can be used to estimate the patient's functional capacity ^(7,33). Use of the Duke Activity Status Index or other activity scales ⁽³⁴⁾ and knowledge of the MET levels required for physical activities, as listed above, provide the clinician with a relatively easy set of questions to estimate whether a patient's functional capacity will be less than or greater than 4 METs (Table 2). At activity levels less than 4 METs, specific questions to establish risk gradients are less reliable. Furthermore, a clinical questionnaire only estimates functional capacity and does not provide as objective a measurement as exercise treadmill testing or arm ergometry. Other activity scales have been advocated, including the Specific Activity Scale ⁽²⁴⁹⁾.

Surgery-Specific Risk (Table 3, Figure 1). The surgery-specific cardiac risk of noncardiac surgery is related to two important factors. First, the type of surgery itself may identify a patient with a greater likelihood of underlying heart disease. Perhaps the best example is vascular surgery, in which underlying CAD is present in a substantial portion of patients. The second aspect is the degree of hemodynamic cardiac stress associated with surgery-specific techniques. Certain operations may be associated with profound alterations in heart rate, blood pressure, vascular volume, pain, bleeding, clotting tendencies, oxygenation, neurohumoral activation, and other perturbations. The intensity of these coronary and myocardial stressors helps determine the likelihood of perioperative cardiac events. This is particularly evident in emergency surgery, where the risk of cardiac complications is substantially elevated.

Examples of noncardiac surgeries and their surgery-specific risks are given below. Higher surgery-specific cardiac risk (e.g., combined perioperative MI and/or death rate equal to or greater than 5%) is present in patients undergoing aortic surgery, peripheral vascular surgery, and anticipated prolonged surgical procedures associated with large fluid shifts and/or blood loss involving the abdomen, and thorax, head, and neck. Intermediate-surgical-risk procedures (combined MI and/or death risk 1% to 5%) include urologic, orthopedic, and uncomplicated abdominal, head, neck, and thoracic surgery. Urologic and orthopedic surgery would be at the lower end of this risk group. Low-risk procedures include cataract resection, dermatologic operations, endoscopic procedures, and breast surgery (Table 3). Patients undergoing low-risk procedures do not require further evaluation. Some require endocarditis prophylaxis.

Step 6 (Figure 1). Patients without major but with intermediate predictors of clinical risk (Table 1) and with moderate or excellent functional capacity can generally undergo intermediate-risk surgery with little likelihood of perioperative death or MI. On the other hand, patients with poor functional capacity or those with a combination of only moderate functional capacity and higher-risk surgery are often considered for further noninvasive testing. This is especially true for patients possessing two or more of the above intermediate markers.

Step 7 (Figure 1). Noncardiac surgery is generally safe for patients with minor or no clinical predictors of clinical risk (Figure 1) and with moderate or excellent functional capacity (equal to or greater than 4 METs), regardless of surgical type. Patients with poor functional capacity facing higher-risk operations (vascular surgery, anticipated long and complicated thoracic surgery, abdominal surgery, and head and neck surgery) may be considered for further testing on an individual basis.

To reiterate, it is important to emphasize that the concept of "medical clearance" for surgery is short-sighted. The real issue is to perform an evaluation of the patient's current medical status, make recommendations concerning the diagnosis and medical management (e.g., use of beta blockers) of the patient with significant cardiac riskthe risk of a cardiac problem over the entire perioperative and postoperative period, and provide a clinical risk profile that the patient, anesthesiologist, and surgeon can use to make management decisions. At times it is appropriate for the consultant to recommend preventive measures that will decrease the patient's cardiovascular risk for years to come. The overall goal of cardiac assessment should be a consideration of both the impending surgery and the long-term cardiac risk, independent of the decision to go to surgery ⁽³⁵⁾. It is almost never appropriate to recommend coronary bypass surgery or other invasive interventions such as coronary angioplasty in an effort to reduce the risk of noncardiac surgery when they would not otherwise be indicated.

Step 8 ^{(Figure 1)}. The results of noninvasive testing can then be used to determine further perioperative management. Such management may include intensified medical therapy or cardiac catheterization, which may lead to coronary revascularization or potentially to cancellation or delay of the elective noncardiac operation. Alternatively, results of the noninvasive test may lead to a recommendation to proceed directly with surgery (Figure 1). In some patients, the risk of coronary angioplasty or corrective cardiac surgery may approach or even exceed the risk of the proposed noncardiac surgery. In rare some instances, this approach may be appropriate, however, if it also significantly improves the patient's long-term prognosis (Table 4). Table 4 has now been changed to "Recommendations for Coronary Angiography in Perioperative Evaluation Before (or After) Noncardiac Surgery" (Section V).