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.

V. Supplemental Preoperative Evaluation

A. Shortcut to the Decision to Test

The preoperative guidelines (ACC/AHA) are fairly straightforward about recommendations for patients about to undergo emergency surgery, the presence of prior cardiac revascularization, and the occurrence of major cardiac predictors. However, the majority of patients have either intermediate or minor clinical predictors of increased perioperative cardiovascular risk. Table 5A presents a shortcut approach to a large number of patients in whom the decision to recommend testing before surgery can be difficult. Basically, if two of the three listed factors are true, the guidelines suggest the use of noninvasive cardiac testing as part of the preoperative evaluation. In any patient with an intermediate clinical predictor, the presence of either a low functional capacity or high surgical risk should lead the consulting physician to consider noninvasive testing. In the absence of intermediate clinical predictors, noninvasive testing should be considered when both the surgical risk is high and the functional capacity is low. The guidelines define minor clinical predictors as advanced age, abnormal ECG, rhythm other than sinus, history of stroke, or uncontrolled systemic hypertension. These factors do not by themselves suggest the need for further testing, but when combined with low functional capacity and high-risk surgery, they should lead to consideration of preoperative testing. In making the decision to obtain noninvasive testing, there will occasionally be some practical circumstances when testing will be obtained after surgery, particularly if the results will not affect perioperative care. This test information may also be useful in predicting long-term risk of cardiac events (also see Section X). More specifically, identification of high-risk patients whose long-term outcome would be improved with medical therapy or coronary revascularization procedures is a major goal of preoperative noninvasive testing. Numerous studies using different preoperative noninvasive techniques before noncardiac surgery have demonstrated the ability to detect patients at increased risk of late cardiac events ^{(254,261,265,267-270)} (see Figure 2).

B. Resting Left Ventricular Function

1. Summary of Evidence
Resting ventricular function has been evaluated preoperatively before noncardiac surgery by radionuclide angiography, echocardiography, and contrast ventriculography ^(23,96-105). Of seven eight studies that demonstrate a positive relation between decreased preoperative ejection fraction and postoperative mortality or morbidity, four five were prospective ^{(96,97,100,103,271)} and three retrospective ^(98,99,103). The greatest risk of complications was observed in patients with an LVEF at rest of less than 35%. In the perioperative phase, poor left ventricular systolic or diastolic function is mainly predictive of postoperative HF, and in critically ill patients, death. It is noteworthy, however, that resting left ventricular function was not found to be a consistent predictor of perioperative ischemic events.

Recommendations for Preoperative Noninvasive Evaluation of Left Ventricular Function

Class I
Patients with current or poorly controlled HF. (If previous evaluation has documented severe left ventricular dysfunction, repeat preoperative testing may not be necessary.)

Class IIa
Patients with prior HF and patients with dyspnea of unknown origin.

Class III
As a routine test of left ventricular function in patients without prior HF.

C. Assessment of Risk for CAD and Functional Capacity

1. The 12-Lead ECG
In patients with established or documented coronary disease, the 12-lead rest ECG contains important prognostic information that relates to long-term morbidity and mortality ^(272-275). The magnitude and extent of Q waves provide a crude estimate of LVEF, and are a predictor of long-term mortality ^(276,277). Horizontal or downsloping ST-segment depression greater than 0.5 mm, left ventricular hypertrophy with a "strain" pattern, and left bundle-branch block in patients with established coronary disease are all associated with decreased life expectancy ^(272-280). The resting 12-lead ECG does not identify increased perioperative risk in patients undergoing low-risk surgery ⁽²⁸¹⁾, but certain ECG abnormalities (above) are clinical predictors of increased perioperative and long-term cardiovascular risk in clinically intermediate- and high-risk patients. In particular, the presence of left ventricular hypertrophy or ST-segment depression on preoperative 12-lead ECG predicts adverse perioperative cardiac events ⁽²⁸²⁾.

Recommendations for Preoperative 12-Lead Rest ECG

Class I
Recent episode of chest pain or ischemic equivalent in clinically intermediate- or high-risk patients scheduled for an intermediate- or high-risk operative procedure.

Class IIa
Asymptomatic persons with diabetes mellitus.

Class IIb
1. Patients with prior coronary revascularization.
2. Asymptomatic male more than 45 years old or female more than 55 years old with two or more atherosclerotic risk factors.
3. Prior hospital admission for cardiac causes.

Class III
As a routine test in asymptomatic subjects undergoing low-risk operative procedures.

2. Exercise Stress Testing for Myocardial Ischemia and Functional Capacity
The aim of supplemental preoperative testing is to provide an objective measure of functional capacity, to identify the presence of important preoperative myocardial ischemia or cardiac arrhythmias, and to estimate perioperative cardiac risk and long-term prognosis. Poor functional capacity in patients with chronic CAD or those convalescing after an acute cardiac event is associated with an increased risk of subsequent cardiac morbidity and mortality ⁽³⁷⁾. Decreased functional capacity may be caused by several factors, including inadequate cardiac reserve, advanced age, transient myocardial dysfunction from myocardial ischemia, deconditioning, and poor pulmonary reserve.

In evaluating the role of exercise testing to assess patients undergoing noncardiac procedures, it is useful to summarize what is known about ECG exercise testing in general. The sensitivity gradient for detecting obstructive coronary disease is dependent on severity of stenosis and extent of disease as well as criteria used for a positive test. As many as 50% of patients with single-vessel coronary disease and adequate levels of exercise can have a normal exercise ECG ⁽³⁸⁾. The mean sensitivity and specificity of exercise testing for obstructive coronary disease are 68% and 77%, respectively ⁽³⁹⁾. The sensitivity and specificity for multivessel disease are 81% and 66%, and for 3-vessel or left main coronary disease, 86% and 53%, respectively ⁽⁴⁰⁾.

Weiner et al ⁽³²⁾ studied 4,083 medically treated patients in CASS and identified a high-risk patient subset (12% of the population) with an annual mortality rate greater than or equal to 5% per year when the exercise workload was less than Bruce stage I and the exercise ECG showed ST-segment depression greater than or equal to 1 mm. A low-risk subset (34% of the population) who were able to complete or do more than Bruce stage III with a normal exercise ECG had an annual mortality rate of less than 1% per year over 4 years of follow-up ⁽³²⁾. Similar results have been reported by others ^(41,42).

Summary of Evidence
The use of preoperative exercise ECG testing to estimate coronary disease presence and risk of perioperative events in patients undergoing major noncardiac surgery was reviewed through a Medline search of the English literature on exercise and peripheral vascular disease from 1975 to September 1994 (106-116). Table 6 lists publications in which exercise test results and perioperative events were reported. In most series, very-high-risk patients (recent MI, unstable angina, HF, and serious ventricular arrhythmias) were excluded. McPhail et al ⁽¹¹³⁾ reported on preoperative exercise treadmill testing and supplemental arm ergometry in 100 patients undergoing surgery for peripheral vascular disease or abdominal aortic aneurysm. Of the 100 patients, 30 were able to reach 85% of age-predicted heart rate maximum, and only two had cardiac complications (6%). In contrast, 70% of the population were unable to reach 85% of age-predicted heart rate or had an abnormal exercise ECG. In this group the cardiac complication rate (MI, death, HF, or ventricular arrhythmia) was 24% (17 patients).

The data in Table 6 indicate a peak exercise heart rate greater than 75% of age-predicted maximum can be expected in approximately half of patients who undergo treadmill exercise, with supplemental arm ergometry when necessary for patients limited by claudication ⁽¹⁰⁷⁾. The frequency of an abnormal exercise ECG response is dependent on prior clinical history ^(107,110). Among patients without a cardiac history and with a normal resting ECG, approximately 20% to 50% will have an abnormal exercise ECG. The frequency is greater (35% to 50%) in patients with a prior history of MI or an abnormal rest ECG. The risk of perioperative cardiac events and long-term risk is significantly increased in patients with an abnormal exercise ECG at low workloads ^{(107,108,113)}.

In contrast to the above studies of patients with vascular disease, in a general population of patients of whom only 20% to 35% had peripheral vascular disease and were undergoing noncardiac surgery, Carliner et al ⁽¹¹⁴⁾ reported exercise-induced ST-segment depression greater than or equal to 1 mm in 16% of 200 patients older than 40 years (mean age, 59 years) being considered for elective surgery. Only two patients (1%) had a markedly abnormal (ST-segment depression of 2 mm or more) exercise test. Of the 32 patients with an abnormal exercise test, five (16%) died or had a nonfatal MI. Of 168 patients with a negative test, 157 (93%) did not die or have an MI. In this series, however, the results of preoperative exercise testing were not statistically significant independent predictors of cardiac risk.

Table 5 provides a prognostic gradient of ischemic responses during an ECG-monitored exercise test as developed for a general population of patients with CAD ⁽¹¹⁸⁾. The onset of a myocardial ischemic response at low exercise workloads is associated with a significantly increased risk of perioperative and long-term cardiac events. In contrast, the onset of a myocardial ischemic response at high exercise workloads is associated with significantly less risk. The prognostic gradient is also influenced by the age of the patient, the extent of the coronary disease, and the degree of left ventricular dysfunction, hemodynamic response to exercise, and presence or absence of chronotropic incompetence. ACC/AHA guidelines concerning the indications for and interpretation of exercise stress testing are available ⁽⁴³⁾.

3. Nonexercise Stress Testing
The two main techniques used in preoperative evaluation of patients undergoing noncardiac surgery who cannot exercise are to increase myocardial oxygen demand (pacing, intravenous dobutamine) and to induce hyperemic responses by pharmacological vasodilators such as intravenous dipyridamole or adenosine. The most common examples presently in use are dobutamine stress echocardiography and intravenous dipyridamole/adenosine myocardial perfusion imaging using both thallium-201 and technetium-99m. Adensoine can also be used as an alternative to dipyridamole in these types of studies.

4. Myocardial Perfusion Imaging Methods
Summary of Evidence
A computerized search (Medline) of the English literature from 1975 to 1999 identified 27 publications describing the use of stress myocardial perfusion testing in the preoperative evaluation of patients before both vascular and nonvascular surgery. These pPublications that report the results of stress myocardial perfusion testing before both vascular and nonvascular surgery are summarized in Table 7. and were chosen from a total of 397 407 articles identified using key words (key words: dobutamine or dipyridamole or exercise test or stress and thallium radioscopes or radionuclide imaging with preoperative care or risk factors or all risk or predictive value of tests or prognosis or prospective studies or noncardiac and surgery or preoperative and evaluation). Included were mostly prospectively recruited patient studies, a majority of which involved patients undergoing vascular surgery. Cardiac events in the perioperative period were defined, for the purpose of this table, as MI or death from cardiac causes, and information about events and scan results had to be available. The percentage of patients with evidence of ischemic risk as judged by thallium redistribution ranged from 23% to 69%. The positive predictive value of thallium redistribution ranged from 4% to 20% in reports that included more than 100 patients. In more recent publications, the positive predictive value of thallium imaging has been significantly decreased. This is probably related to the fact that in recent years, scintigraphic information obtained is actively used to select patients for therapeutic interventions such as coronary revascularization, as well as to adjust perioperative medical treatment and monitoring and to select different surgical procedures. The negative predictive value of a normal scan remains uniformly high at approximately 99% for MI and/or cardiac death. Although the risk of a perioperative cardiac event in patients with fixed defects is higher than in patients with a normal scan, it is still significantly lower than the risk in patients with thallium redistribution.

In a meta-analysis of dipyridamole thallium imaging for risk stratification before vascular surgery, Shaw et al ⁽²⁸³⁾ reported that a total of 10 studies involving 1994 patients referred for testing before elective vascular surgery demonstrated significant prognostic utility for this scintigraphic technique. In addition, they noted that the positive predictive value of perfusion imaging was correlated with the pretest cardiac risk of the patients. Overall, a reversible myocardial perfusion defect predicted perioperative events, and a fixed thallium defect predicted long-term cardiac events. Of note, the addition of semiquantitative analysis of perfusion imaging improved the clinical risk stratification based on a relationship of increasing event rates in patients with larger defects.

The need for caution in routine screening with dipyridamole thallium stress test of all patients before vascular surgery has been raised by Baron et al ⁽¹³³⁾. In this review of 457 patients undergoing elective abdominal aortic surgery, the presence of definite CAD and age greater than 65 years were better predictors of cardiac complications than perfusion imaging.

This issue of routine testing has been evaluated by 2 studies that prospectively evaluated preoperative cardiac risk assessment with a methodology that generally follows the guidelines outlined in this review. In a report by Vanzetto et al ⁽²⁸⁴⁾, 517 consecutive patients were evaluated before abdominal aortic surgery. If no major or fewer than 2 intermediate clinical cardiac risk factors were present, patients (n=317) went directly to elective surgery. The authors noted a 5.6% incidence of cardiac events (death/MI) in those patients with 1 risk factor and a rate of 2.4% in those with no cardiac risk factors. All high-risk patients (n=134, 2 or more cardiac risk factors) underwent dipyridamole-thallium SPECT imaging, and those with a normal scan (38%) had a cardiac event rate of 2% in contrast to a rate of 23% in 43 patients (36%) demonstrating reversible thallium defects. Bartels et al ⁽²⁴³⁾ also reported that patients (n=203) referred for elective vascular surgery who had no clinical intermediate or major clinical risk factors had a 2% incidence of cardiac events. Those patients with either intermediate risk factors and a functional capacity of less than 5 METs or high clinical risk (10 of 23 patients) underwent stress-thallium imaging. The remaining patients had intensified medical therapy before elective surgery. The cardiac event rates were 9% in the intermediate-risk group and 5% in the high-risk group, but the overall cardiac mortality rate was only 1% in the patients who underwent the ACC/AHA guideline workup. Another recent report ⁽²⁸⁵⁾ also used the clinical risk factor parameters to divide vascular surgery patients into low-, intermediate-, and high-cardiac-risk groups. Those authors did not include functional capacity measurements but noted a 0% death or MI rate in the perioperative period among the low-risk patients (n=60). These additional reports support the use of the perioperative risk assessment guidelines, especially in the confirmation that cardiac patients with low clinical risk can typically undergo elective surgery with a low event rate.

In several publications by Hendel et al ⁽¹²⁸⁾, Lette et al ⁽¹²⁹⁾, and Brown et al ⁽¹³¹⁾, the scoring or quantitation of scan abnormalities had a significant impact on improving risk assessment and positive predictive value. The data suggest that as the size of the defect increases to a moderate (20% to 25% of left ventricular mass) degree, the cardiac risk significantly increases. The use of techniques to quantitate the extent of abnormality and the current routine use of quantitative gated SPECT perfusion imaging to evaluate LVEF will probably improve the positive predictive nature of myocardial perfusion imaging. This would also impact the potential role of interventions such as cardiac catheterization and revascularization. Although there are few published reports using adenosine myocardial perfusion imaging in the preoperative risk assessment of patients before noncardiac surgery, its usefulness appears to be equivalent to that of dipyridamole. ACC/AHA guidelines concerning indications for and interpretation of stress testing with myocardial perfusion imaging are available ⁽¹⁴¹⁾.

5. Dobutamine Stress Echocardiography
Summary of Evidence
Several reports have documented the accuracy of dobutamine stress echocardiography to identify patients with significant angiographic coronary disease ^(141-146). The use of dobutamine stress echocardiography in preoperative risk assessment was evaluated in six 12 studies, all published since 1991 and identified by a computerized search of the English language literature (Table 8) ^{(105,147-151,263,266,286-289)}. The populations included predominantly, but not exclusively, patients undergoing peripheral vascular surgical procedures. Only two studies blinded the physicians and surgeons who treated the patients to the dobutamine stress echocardiographic results ^(105,149). In the remaining studies, the results were used to influence preoperative management, particularly the decision whether or not to proceed with coronary angiography or coronary revascularization before elective surgery. Each study used similar, but not identical, protocols. The definition of a positive and negative test result differed considerably, based on subjective analysis of regional wall motion; i.e., worsening of pre-existing wall-motion abnormalities was considered by some investigators as a positive and by others as a negative finding. The end points used to define clinical outcome varied and included both "soft" (i.e., arrhythmia, HF, and ischemia) and "hard" (i.e., MI or cardiac death) events.

The data indicate that dobutamine stress echocardiography can be performed safely and with acceptable patient tolerance. The range of positive test results was 239% to 50%. The predictive value of a positive test ranged from 17% to 43% for all events and 707% to 2325% for hard events (MI or death). The negative predictive value ranged from 93% to 100%. In the series by Poldermans et al ⁽¹⁰⁵⁾, the presence of a new wall-motion abnormality was a powerful determinant of an increased risk for perioperative events after multivariable adjustment for different clinical and echocardiographic variables. Several studies suggest that the extent of the wall-motion abnormality and/or wall-motion change at low ischemic thresholds infusion rates of dobutamine is especially important. These findings have been shown to be predictors of long-term ^{(151,286,290)} and short-term ⁽²⁶⁸⁾ outcome. Although hypotension during dobutamine testing is generally not well correlated with the degree of underlying CAD, in one recent study, hypotension was an independent predictor of perioperative complications ⁽²⁶⁸⁾.The published experience of dobutamine stress echocardiography to assess perioperative risk before vascular and nonvascular surgery is relatively small compared with the published literature on exercise testing or intravenous dipyridamole myocardial perfusion imaging. The summary of evidence supports the use of dobutamine echocardiography for assessing preoperative risk in properly selected patients, especially those undergoing peripheral arterial revascularization.Updated ACC/AHA guidelines concerning the indications for and interpretation of echocardiography (including stress echocardiography) will be published shortly.

6. Stress Testing in the Presence of Left Bundle-Branch Block
(Moved original text from after "Recommendations for Exercise or Pharmacologic Stress Testing") The sensitivity and specificity of exercise thallium scans in the presence of left bundle-branch block are reported to be 78% and 33%, respectively, and overall diagnostic accuracy varies from 36% to 60% ^(152,153). In contrast, the use of vasodilators in such patients has a sensitivity of 98%, a specificity of 84%, and a diagnostic accuracy of 88% to 92% ^(154-156). Pharmacological stress testing with adenosine or dipyridamole is preferable to dobutamine or exercise imaging in patients with pre-existing left bundle-branch block. The tachycardia induced during exercise and conceivably also during dobutamine infusion may result in reversible septal defects even in the absence of left anterior descending artery disease in some patients. This response is unusual with either dipyridamole or adenosine stress testing. Exercise should not be combined with dipyridamole in such patients, and synthetic catecholamines will also yield false-positive results ⁽¹⁵⁷⁾. Therefore, the preoperative evaluation of CAD in patients with left bundle-branch block should be performed by means of vasodilator stress and myocardial perfusion studies.

Recommendations for Exercise or Pharmacological Stress Testing
Class I
1. Diagnosis of adult patients with intermediate pretest probability of CAD.
2. Prognostic assessment of patients undergoing initial evaluation for suspected or proven CAD; evaluation of subjects with significant change in clinical status.
3. Demonstration of proof of myocardial ischemia before coronary revascularization.
4. Evaluation of adequacy of medical therapy; prognostic assessment after an acute coronary syndrome (if recent evaluation unavailable).

Class IIa
Evaluation of exercise capacity when subjective assessment is unreliable.

Class IIb
1. Diagnosis of CAD patients with high or low pretest probability; those with resting ST depression less than 1 mm, those undergoing digitalis therapy, and those with ECG criteria for left ventricular hypertrophy.
2. Risk assessment/prognosis of patients with pre-excitation syndrome, electronically paced ventricular rhythm, rest ST depression >1mm, LBBB.
3.2. Detection of restenosis in high-risk asymptomatic subjects within the initial months after PCI.

Class III
1. For exercise stress testing, diagnosis of patients with resting ECG abnormalities that preclude adequate assessment, e.g., pre-excitation syndrome, electronically paced ventricular rhythm, rest ST depression greater than 1 mm, or left bundle-branch block.
2. Severe comorbidity likely to limit life expectancy or candidacy for revascularization.
3. Routine screening of asymptomatic men or women without evidence of CAD.
4. Investigation of isolated ectopic beats in young patients.

Stress Testing in the Presence of Left Bundle Branch Block (Moved original text to after "Dobutamine Stress Echocardiography, Summary of Evidence") The sensitivity and specificity of exercise thallium scans in the presence of left bundle branch block is reported to be 78% and 33%, respectively, and overall diagnostic accuracy varies from 36% to 60% (152, 153). In contrast, the use of vasodilators in such patients has a sensitivity of 98%, a specificity of 84%, and a diagnostic accuracy of 88% to 92% (154-156). Pharmacological stress testing with adenosine or dipyridamole is preferable to dobutamine or exercise imaging in patients with preexisting left bundle branch block. The tachycardia induced during exercise and conceivably also during dobutamine infusion may result in reversible septal defects even in the absence of left anterior descending artery disease in some patients. This response is unusual with either dipyridamole or adenosine stress testing. It should also be noted that exercise should not be combined with dipyridamole in such patients and that synthetic catecholamines will also yield false-positive results (157). Therefore, the preoperative evaluation of CAD in patients with left bundle branch block should be performed using vasodilator stress and myocardial perfusion studies.

7. Ambulatory ECG Monitoring
Summary of Evidence
The use of preoperative ambulatory ECG monitoring to estimate coronary disease presence and risk of perioperative events in patients undergoing major noncardiac surgery was reviewed through a Medline search of the English literature on preoperative and myocardial ischemia and surgery from 1976 to September 1994. The predictive value of preoperative ST changes on 24- to 48-hour ambulatory ECG monitoring for cardiac death or MI in patients undergoing vascular and nonvascular surgery has been reported by several investigators. The frequency of abnormal ST-segment changes observed in 869 patients reported in seven series was 25% (range, 9% to 39%) ^(19,158-162). The positive and negative values for perioperative MI and cardiac death are shown in Table 9. In two recent studies, it had a predictive value similar to dipyridamole thallium imaging ^(160,163).

Although the test has been shown to be predictive of cardiac morbidity, there are several limitations. Differences in the study protocols (24 vs. 48 hours, ambulatory vs. in-hospital) may account for the variability in the predictive value of the test. Preoperative ambulatory ECG monitoring for ST-segment changes cannot be performed in a significant percentage of patients because of baseline ECG changes. The test, as currently used, only provides a binary outcome and therefore cannot further stratify the high-risk group in order to identify the subset for whom coronary angiography should be considered ⁽¹⁶³⁾.

D. Recommendations: When and Which Test

In most ambulatory patients, the test of choice is exercise ECG testing, which can both provide an estimate of functional capacity and detect myocardial ischemia through changes in the ECG and hemodynamic response. Treadmill exercise stress testing in patients with abdominal aortic aneurysms greater than 4 cm in diameter is relatively safe. In a series of more than 250 patients studied in this circumstance, a single patient developed subacute aneurysm rupture 12 hours after testing and was successfully repaired is associated with a 0.4% mortality rate in patients with abdominal aortic aneurysms greater than 4 cm ⁽²⁹¹⁾. In patients with important abnormalities on their resting ECG (e.g., left bundle-branch block, left ventricular hypertrophy with "strain" pattern, or digitalis effect), other techniques such as exercise echocardiography or exercise myocardial perfusion imaging should be considered. The sensitivity and specificity of exercise thallium scans in the presence of left bundle-branch block are reported to be 78% and 33%, respectively, and overall diagnostic accuracy varies from 36% to 60% ^(152,153). In contrast, the use of vasodilators in such patients has a sensitivity of 98%, a specificity of 84%, and a diagnostic accuracy of 88% to 92% ^(154-156). Exercise should not be combined with dipyridamole in such patients, and synthetic catecholamines can also yield false-positive results ⁽¹⁵⁷⁾.

In patients unable to perform adequate exercise, a nonexercise stress test should be used. In this regard, dipyridamole myocardial perfusion imaging testing and dobutamine echocardiography are the most common tests. Intravenous dipyridamole should be avoided in patients with significant bronchospasm, critical carotid disease, or a condition that prevents their being withdrawn from theophylline preparations. Dobutamine should not be used as a stressor in patients with serious arrhythmias or severe hypertension or hypotension. For patients in whom echocardiographic image quality is likely to be poor, a myocardial perfusion study is more appropriate. Soft tissue attenuation can also be a problem with myocardial perfusion imaging. If there is an additional question about valvular dysfunction, the echocardiographic stress test is favored. In many instances, either stress perfusion or stress echocardiography is appropriate. In a meta-analysis of dobutamine stress echocardiography, ambulatory electrocardiography, radionuclide ventriculography, and dipyridamole thallium scanning in predicting adverse cardiac outcome after vascular surgery, all tests had a similar predictive value, with overlapping confidence intervals ⁽¹⁶⁴⁾. The expertise of the local laboratory in identifying advanced coronary disease is probably more important than the particular type of test. Figure 3 illustrates an algorithm to help the clinician choose the most appropriate stress test in those various situations.

Currently the use of ambulatory electrocardiography as a preoperative test should be restricted to identifying patients for whom additional surveillance or medical intervention might be beneficial. The current evidence does not support the use of ambulatory electrocardiography as the only diagnostic test to refer patients for coronary angiography.

For certain patients at high risk, it may be appropriate to proceed with coronary angiography rather than perform a noninvasive test. For example, preoperative consultation may identify patients with unstable angina or evidence for residual ischemia after recent MI for whom coronary angiography is indicated. In general, indications for preoperative coronary angiography are similar to those identified for the nonoperative setting. Table 4 The following recommendations provides a summary of indications for preoperative coronary angiography in patients being evaluated before noncardiac surgery. These are adapted from the ACC/AHA guidelines for coronary angiography published in 1987 (36) 1999 ⁽²⁹²⁾.

Recommendations for Coronary Angiography in Perioperative Evaluation Before (or After) Noncardiac Surgery

Class I: Patients With Suspected or Known CAD
1. Evidence for high risk of adverse outcome based on noninvasive test results.
2. Angina unresponsive to adequate medical therapy.
3. Unstable angina, particularly when facing intermediate-risk* or high-risk* noncardiac surgery.
4. Equivocal noninvasive test results in patients at high-clinical risk† undergoing high-risk* surgery.

Class IIa
1. Multiple markers of intermediate clinical risk † and planned vascular surgery (noninvasive testing should be considered first).
2. Moderate to large ischemia on noninvasive testing but without high-risk features and lower LVEF.
3. Nondiagnostic noninvasive test results in patients of intermediate clinical risk† undergoing high-risk* noncardiac surgery.
4. Urgent noncardiac surgery while convalescing from acute MI.

Class IIb
1. Perioperative MI.
2. Medically stabilized class III or IV angina and planned low-risk or minor* surgery.

Class III
1. Low-risk* noncardiac surgery with known CAD and no high-risk results on noninvasive testing.
2. Asymptomatic after coronary revascularization with excellent exercise capacity (greater than or equal to 7 METs).
3. Mild stable angina with good left ventricular function and no high-risk noninvasive test results.
4. Noncandidate for coronary revascularization owing to concomitant medical illness, severe left ventricular dysfunction (e.g., LVEF less than 0.20), or refusal to consider revascularization.
5. Candidate for liver, lung, or renal transplant more than 40 years old as part of evaluation for transplantation, unless noninvasive testing reveals high risk for adverse outcome.

*Cardiac risk according to type of noncardiac surgery. High risk: emergent major operations, aortic and major vascular surgery, peripheral vascular surgery, or anticipated prolonged surgical procedure associated with large fluid shifts and blood loss; intermediate risk: carotid endarterectomy, major head and neck surgery, intraperitoneal and intrathoracic surgery, orthopedic surgery, or prostate surgery; and low risk: endoscopic procedures, superficial procedures, cataract surgery, or breast surgery.

†Cardiac risk according to clinical predictors of perioperative death, MI, or HF. High clinical risk: unstable angina, acute or recent MI with evidence of important residual ischemic risk, decompensated HF, high degree of atrioventricular block, symptomatic ventricular arrhythmias with known structural heart disease, severe symptomatic valvular heart disease, or patient with multiple intermediate-risk markers such as prior MI, HF, and diabetes; intermediate clinical risk: Canadian Cardiovascular Society class I or II angina, prior MI by history or ECG, compensated or prior HF, diabetes mellitus, or renal insufficiency.