The recommendations in this guideline concentrate on summarizing the evidence supporting the clinical evaluation (history, physical examination, and electrocardiography) and noninvasive evaluation of the patient [1.1, 1.2]. Studies on clinical and noninvasive testing were considered to be of "strong," "fair," or "weak" evidence [2.6]. If a recommendation has no weighting for strength of evidence, no studies were found that related to that clinical decision point. Perioperative risk estimates refer to the reported incidence of perioperative myocardial infarction and cardiac death [2.4]. The literature on noninvasive risk stratification has focused on patients undergoing vascular surgery. These patients have a higher risk for perioperative cardiac events and may not be representative of most patients (>90%) undergoing major noncardiac surgery [1.3]. We make separate recommendations for patients having vascular and nonvascular surgery, according to the quality of evidence available for each. For management issues, evidence is still lacking. In these situations, the clinician must use available evidence from the nonoperative setting combined with clinical judgment and patient preferences [5].

Clinical Assessment

When assessing a patient's risk for major cardiac events during or after a noncardiac operation, the clinician uses clinical evaluation to determine the risk for fatal and nonfatal cardiac events and may refine the risk assessment for intermediate-risk patients through noninvasive testing.

In patients who have previously had angiography results indicating mild coronary artery disease or successful coronary revascularization and have no new clinical symptoms of concern, the risk for perioperative cardiac events is probably similar to that in patients without coronary artery disease (weak evidence) [3.1]. If new or worrisome coronary symptoms develop, the patient's risk status needs to be reevaluated.

Although many risk indices exist, clinical data collection should concentrate on variables that enable the clinician to classify the patient as low-risk, intermediate-risk, or high-risk. The modified Cardiac Risk Index (outlined in Appendix Figure 1) should be performed in all patients (strong evidence) [3.3-3.6]. Class II or III on the modified Cardiac Risk Index predicts a high risk for perioperative cardiac events (10% to 15%; strong evidence for patients having vascular surgery and those having nonvascular surgery) (Appendix Figure 2). Low Cardiac Risk Index scores (class I) still do not reliably identify patients who have low risk for perioperative cardiac events [3.7]. Information on "low-risk" variables should be collected for these patients.

View larger version (28K): [in this window] [in a new window]   Appendix (Figure 1). Suggested algorithm for the risk assessment and management of patients at low or intermediate risk for perioperative cardiac events, usually myocardial infarction and death. Boxed phrases indicate recommended actions. The italicized words beside the boxes indicate the level of evidence supporting the recommendation. If no italicized word is present, no evidence exists for or against use. DTI = dipyridamole thallium imaging; DSE = dobutamine stress echocardiography.

View larger version (37K): [in this window] [in a new window]   Appendix (Figure 2). Suggested algorithm for the management of patients at high risk for perioperative cardiac events. Boxed phrases indicate recommended actions. The italicized words below the boxes indicate the level of evidence supporting the recommendation. If no italicized word is present, no evidence exists for or against use. AHA = American Heart Association; CHF = congestive heart failure.

Low risk [<3%] is predicted by the presence of 0 or 1 cardiac risk factors, as defined by the "low-risk" variables in Appendix Figure 1) (strong evidence for patients having vascular surgery) [3.8]. Low-risk patients may proceed directly to surgery without further noninvasive testing because no testing method has been shown to refine risk assessment in this group.

Between patients at very high risk and those at very low risk, a substantial proportion of patients (30% to 60% of patients having vascular surgery) are at intermediate risk for perioperative cardiac events (3% to 10%) (strong evidence) [3.4-3.8].

We make no recommendations for or against the clinical assessment of functional status; this assessment has not been shown to add to clinical risk index evaluation in the operative setting [3.9].

Noninvasive Testing

Noninvasive tests available for further risk stratification include those that assess left ventricular function (radionuclide angiography), cardiac ischemia (exercise or pharmacologic stress testing and ambulatory electrocardiographic monitoring), or both (dobutamine stress echocardiography) [4.1].

Noninvasive assessment of resting left ventricular ejection fraction is unlikely to add risk discrimination to the clinical examination. We recommend against the use of either radionuclide angiography (fair evidence) or transthoracic echocardiography (strong evidence) to predict perioperative risk [4.2].

Exercise stress testing cannot be done in a significant proportion of patients who are undergoing vascular surgery (30% to 70%) or in patients with diseases that impair walking ability (such as stroke and hip and knee disease). Among patients who can adequately complete the testing, studies have, in general, shown a poor predictive value for patients having vascular surgery and patients having nonvascular surgery. We recommend against using this testing to predict perioperative risk (fair evidence) [4.3].

Dipyridamole thallium imaging and dobutamine stress echocardiography, especially if the results are negative, provide useful information on stratification for patients having vascular surgery who are determined to be at intermediate risk according to clinical evaluation (Appendix Figure 1) [4.4.1, 4.4.2]. We therefore recommend using either dipyridamole thallium imaging or dobutamine stress echocardiography for further risk stratification in intermediate-risk patients having vascular surgery (strong evidence). In patients having nonvascular surgery, most studies have shown that neither dipyridamole thallium imaging nor dobutamine stress echocardiography reliably predict the occurrence of adverse perioperative cardiac events [4.4.3]. Both dipyridamole thallium imaging and dobutamine stress echocardiography carry the risk for minor morbidity, incur the cost of testing, and may result in potentially harmful downstream interventions. Given the lack of demonstrated benefit and the potential harm, we recommend against the use of dipyridamole thallium imaging and dobutamine stress echocardiography for risk stratification in low-risk patients who are undergoing vascular surgery (fair evidence for dipyridamole thallium imaging, strong evidence for dobutamine stress echocardiography) and low- and intermediate-risk patients who are undergoing nonvascular surgery (weak evidence for both tests).

Ambulatory electrocardiographic monitoring does not have good predictive ability among patients undergoing nonvascular surgery and is not recommended for this group (strong evidence) [4.5]. Because ambulatory electrocardiographic monitoring showed good risk discrimination among patients undergoing vascular surgery in one study of strong quality but not in a second, equally strong study, further research is needed before ambulatory electrocardiographic monitoring can be recommended for patients undergoing vascular surgery [4.5].

No studies of coronary angiography findings as predictors of perioperative risk have been reported. Because angiography carries procedural risks for stroke and death, we recommend that it be reserved for patients who are being independently considered for revascularization and that it not be used separately as a test for risk stratification [4.6].

Management Strategies

Once the patient's risk status is clarified, four key questions remain: Are there modifiable operative risk factors [5.2, 5.4]? If the patient is at high risk, can the elective operation be modified or canceled [5.2]? Does the patient have an indication for coronary revascularization [5.3]? If so, what is the optimal sequence of angiography, coronary revascularization, and noncardiac surgery (Appendix Figure 2)?

Patients with unstable coronary syndromes should, if the urgency and gravity of their surgical condition permit, proceed directly to cardiac catheterization to determine whether clinically important coronary artery disease is present and operable [5.2]. For high-risk patients who are not candidates for coronary revascularization, the consultant may wish to 1) delay the upcoming operation, if the patient's over-all status can be improved, or 2) cancel the operation altogether, if the risk posed by the patient's surgical condition is lower than the patient's cardiac risk [5.2].

The use of coronary revascularization as a preventive step to reduce cardiac risk before noncardiac surgery has not been studied prospectively [5.3.1]. A large retrospective analysis suggests that any decrease in risk produced by having coronary artery bypass graft surgery before having noncardiac surgery was offset by the risk posed by the angiography and coronary artery bypass grafting [5.3]. Analyses from two formal decision models also suggest that no net risk reduction occurs with coronary artery bypass graft surgery done before noncardiac surgery [5.3.1]. No prospective evaluations of prophylactic preoperative angioplasty and coronary stent placement are not available [5.3.3]. The clinician deciding whether to refer the intermediate- or high-risk patient for revascularization must weigh the added risks posed by angiography and coronary artery bypass grafting (or percutaneous transluminal coronary angioplasty or stent procedure) against the potential reduction in short-term (perioperative) and long-term risk [5.3]. The lack of demonstrable benefit and the potential increase in risk lead us to recommend that if the goal is to optimize short-term prognosis, the patient should proceed directly to surgery without further cardiac investigation. (Table 1) (Table 2)

View this table: [in this window] [in a new window]   Table 1. *Modified Cardiac Risk Index (1)

View this table: [in this window] [in a new window]   Table 2. Low-Risk Variables

Some patients have an independent need for coronary bypass. In these patients, the planned noncardiac surgery may trigger investigation of the possibility of coronary revascularization with a view to reducing long-term morbidity and mortality [5.3.2]. Determining whether the patient is a candidate for coronary revascularization should therefore be done on the same clinical grounds that govern the need for coronary revascularization in the nonoperative setting. The optimal timing (preoperative, intraoperative, or postoperative) of coronary revascularization to improve long-term outcomes is unknown. In these circumstances, we recommend an individualized approach that weighs patient preferences, the urgency of the upcoming operation, and the urgency of the patient's coronary status.

Authors' addendum: In a randomized trial (strong evidence) published after the American College of Physicians reviewed and approved these guidelines, perioperative ß-blocker therapy was shown to reduce long-term (6-month) mortality with few side effects. Thus, we recommend the perioperative use of atenolol in patients with coronary artery disease or risk factors for coronary artery disease (as per the criteria of Mangano and colleagues, described in Appendix Figure 1), unless the patient has significant contraindications, such as asthma [5.4].

Because of the lack of controlled evidence, no recommendations can be made either for or against the use of other medical therapies, such as digitalis, nitrates. or calcium-channel antagonists [5.4]. Therapy with these drugs should be continued or started on the basis of the same criteria that would apply in the nonoperative setting. Controlled trials suggest that pulmonary artery catheter use may be beneficial in selected patients (fair evidence) [5.3].

Comparison with Guidelines from Other Medical Societies

Recent reviews of preoperative cardiovascular assessment have provided recommendations for stratification and management [1.1]. These reviews and this paper concur in several important respects. All agree that for stratification strategies, patients should routinely be assessed clinically for known operative risk factors (the clinical variables or indices recommended vary) [3] and noninvasive testing should not be used routinely [4]. For intermediate-risk patients having vascular surgery, all reviews agree that thallium imaging and possibly dobutamine stress echocardiography add to perioperative risk discrimination [4]. All reviews concur that coronary artery bypass is not likely to improve short-term outcomes and that indications for coronary artery bypass grafting are identical to those in the nonoperative setting [5].

Unlike other reviews, we recommend against further noninvasive testing for nonvascular patients who are deemed to be at intermediate risk by clinical evaluation [5]. The differences in recommendations developed by different medical societies in this clinical situation (perioperative assessment) stem largely from differences in the values used to extrapolate evidence from one population to another [2.1]. The American College of Physicians clinical efficacy assessment process considers blinding (independent interpretation of the test results and the outcome) and method of patient selection to greatly influence the quality and generalizability of study results. Further well-designed studies could clarify these areas and alter recommendations [7]. The most pressing areas of research include the determination of which noninvasive tests can improve risk stratification among intermediate-risk patients undergoing nonvascular surgery and the study of the circumstances in which coronary revascularization before noncardiac surgery is warranted.

Dr. Detsky: Suite 427, Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario M56 1X5, Canada.