Clinical Evaluation of the Patient

A critical first step in patient evaluation is identifying the presence of unstable symptoms. Frequently, the preoperative evaluation is an opportunity to evaluate patients who rarely access the health care system and are not currently receiving treatment for cardiovascular disease. This is particularly true with patients scheduled for vascular surgical procedures, in whom coexisting diabetes or change in lifestyle may mask cardiac symptoms. History, physical examination, electrocardiography, or chest radiography may identify unsuspected hypertension, coronary artery disease, valvular heart disease, myocardial disease, and electrical problems. Each type of condition has its own foundation of knowledge with respect to management in the nonsurgical setting; most of this knowledge applies in the surgical setting.

In most studies, the presence of active congestive heart failure before surgery has been associated with an increased incidence of perioperative cardiac morbidity [1]. Stabilization of ventricular function and treatment of pulmonary congestion is prudent before elective surgery. It is also important to determine the cause of failure of the left side of the heart. Congestive symptoms may be caused by nonischemic cardiomyopathy, coronary heart disease, mitral or aortic valvular insufficiency, or valvular stenosis. Because the type of perioperative monitoring and treatments would be different for different conditions, clarifying the cause of cardiac congestion is important.

In patients with symptomatic coronary disease, the preoperative evaluation may lead to the recognition of a change in the frequency or pattern of anginal symptoms. The presence of unstable angina has been associated with a high perioperative risk for myocardial infarction. The perioperative period is associated with a hypercoagulable state and surges in endogenous catecholamine levels, both of which may exacerbate the underlying process in unstable angina and thus increase the risk for acute infarction.

The preoperative evaluation may also identify patients with unrecognized or occult symptoms. The presence of dyspnea in the elderly may be the only manifestation of cardiovascular disease. Many elderly patients subconsciously restrict activity in such circumstances and do not seek medical attention. Several studies have reported a high incidence of occult coronary artery disease in patients scheduled for vascular surgery [2]. Patients with vascular disease have a lower life expectancy than the general population, largely because of concomitant coronary artery disease. Thus, identification and institution of appropriate long-term treatment may improve outcome [3].

Preoperative cardiovascular testing is most often advocated in patients who cannot exercise. Patients with vascular disease have a particularly high probability for coronary atherosclerosis, and their exercise capacity is often limited by claudication. Similarly, patients having hip surgery are often elderly persons with substantial coexisting disease. Such patients may not stress themselves sufficiently in daily life to provoke symptoms of myocardial ischemia or congestive heart failure. The planned noncardiac surgery may be a stress significantly greater than that encountered in daily life.

Use of Information from the Preoperative Evaluation

Few randomized trials have evaluated the ways in which information from the preoperative evaluation is used and the efficacy of different interventions in reducing morbidity during the perioperative period. Modifications of perioperative care based on the preoperative evaluation can take the form of changes in anesthetic technique, aggressive treatment of hemodynamic perturbations, preoperative medical and coronary interventions, and use of expensive resources such as the intensive care unit, ST-segment telemetry, or invasive monitoring. The accumulated data do not seem to suggest a "best" anesthetic technique; however, information from the preoperative evaluation can help in designing an anesthetic plan.

Although the evidence of their value is insufficient, postoperative intensive care or invasive monitoring is routinely used for high-risk patients with active cardiac disease having extensive surgery. However, the evidence suggests that more intensive monitoring is not needed in the absence of active cardiovascular disease (shown by a negative result on a noninvasive test) [4]. Screening can therefore define the population that would not benefit from intensive monitoring. In an era of rationing of expensive medical resources, savings from the reduced use of postoperative resources may easily offset the costs of screening. Additionally, the identification of very high-risk patients (for example, those with extensive coronary disease or left ventricular dysfunction) would allow for the rational use of these resources for extended periods of time.

Modification of Medications

ß-adrenergic-blocking agents were associated with improved outcome in several studies [5]. In addition to oral and longer-acting intravenous agents, short-acting intravenous agents that require administration in a monitored setting are available. Administration of ß-adrenergic-blocking agents should be considered in patients who appear to be at substantial risk, because of the mounting evidence that these agents reduce the cardiac effects of perioperative stresses, including myocardial ischemia and possibly myocardial infarction.

Coronary Revascularization

Finally, the preoperative evaluation may occasionally identify a patient who would benefit from coronary revascularization. Several investigators have proposed that coronary revascularization be done before noncardiac surgery to reduce perioperative risk. Preoperative screening may identify patients who have severe coronary disease. No randomized trials have addressed this issue, and such trials would require very large sample sizes and would have many confounding issues. Several large cohort studies suggest that patients who survive coronary artery bypass grafting have a decreased risk during subsequent noncardiac surgery [6]. Few data support the notion of coronary revascularization solely to improve perioperative outcome, although it is true that revascularization may enhance long-term survival in specific patient subsets. Using the CASS database, Rihal and colleagues [7] found that coronary artery bypass grafting substantially improved survival in patients with both peripheral vascular disease and triple-vessel coronary disease, especially in the group with depressed ventricular function. Therefore, patients with diffuse disease or a substantial left main stenosis amenable to surgery and acceptable risk should have coronary artery bypass grafting before noncardiac surgery. Long-term benefit and reduced threat of perioperative myocardial infarction before noncardiac surgery justify this strategy.

The value of percutaneous transluminal coronary angioplasty is less well established. In several series, a low incidence of cardiovascular complications was seen in patients having prophylactic percutaneous transluminal coronary angioplasty before vascular surgery, but it is difficult to determine the expected complication rate in a comparison group with single- or double-vessel disease [6]. For perioperative myocardial infarctions resulting from plaque rupture and coronary thrombosis in noncritical lesions as seen in the ambulatory setting, single-vessel percutaneous transluminal coronary angioplasty of more critical stenoses would theoretically have minimal benefit. Therefore, current evidence does not support the use of percutaneous transluminal coronary angioplasty beyond established indications for nonsurgical patients.

Risks and Benefits of Coronary Revascularization

Another approach to determining the optimal strategy for medical care in the absence of clinical trials is the construction of a decision analysis. Two decision analyses have been published on cardiac screening of patients without cardiac conditions [8, 9]. Both assumed that patients with clinically significant coronary artery disease would have coronary artery bypass grafting before noncardiac surgery. Both models found that the optimal decision was sensitive to local morbidity and mortality rates within the clinically observed range. These models suggest that preoperative testing for the purpose of coronary revascularization is not the optimal strategy if perioperative morbidity and mortality are low.

The primary cost (both in dollars and morbidity) of preoperative testing and revascularization is the revascularization procedure itself. Therefore, the indications for revascularization and the frequency of its use substantially affect the model. Also, potential long-term benefits of coronary revascularization in this population were not included in the analysis, creating a potential bias against the revascularization arm.

Value of Noninvasive Testing

Several studies have suggested that exercise capacity is an important factor in predicting perioperative outcome. McPhail and colleagues [10] showed that patients having major vascular surgery who could exercise to 85% of their peak maximal heart rate had a lower risk for cardiac-related morbidity than those who could not reach the target heart rate, even in the presence of a positive stress test result.

Although an objective measure of exercise tolerance predicts outcome, history alone is frequently sufficient to delineate those persons with good exercise tolerance. These patients rarely require additional testing, because perioperative management will not be modified on the basis of the results. Therefore, exercise stress testing is rarely indicated before surgery in patients who have excellent functional capacity in daily life.

In patients who cannot exercise, pharmacologic testing has been advocated. Boucher and coworkers [11] first described the use of dipyridamole thallium imaging in a cohort of 48 patients who had had peripheral vascular surgery. They reported eight ischemic cardiac events in 16 patients who had redistribution and no events in patients who had normal scans. The entrance criteria of this study included at least one objective indication of coronary disease (chest pain, abnormal electrocardiogram, or previous myocardial infarction), and patients with indications of extensive disease (unstable angina, high-grade ventricular ectopy, recent myocardial infarction, congestive heart failure, or severe renal or pulmonary disease) were excluded. Therefore, the initial report advocating noninvasive testing was applied to a selected group of patients at moderate risk.

Routine Testing Compared with Selected Testing

After the initial enthusiasm for preoperative dipyridamole thallium imaging, the use of this procedure was expanded to more diverse populations. When these studies are being evaluated, it is important to determine whether the study sample was consecutive or selected on the basis of established criteria. After these initial reports, there was general enthusiasm to apply the test to all patients having a specific, usually vascular, surgical procedure. However, the probability and extent of coronary artery disease vary greatly for consecutive patients having one type of surgery. Even patients having vascular surgery have only a 30% to 40% probability of disease. Thus, it is important to determine whether a study evaluates consecutive testing or selective testing.

Eagle and colleagues [12] evaluated the implications of the prior probability of disease, based on clinical criteria, in determining the value of preoperative testing. Two hundred consecutive patients having vascular surgery were studied, and five clinical risk factors were identified (Q waves on electrocardiogram, angina, diabetes mellitus, age more than 70 years, and ventricular ectopic activity being treated). Patients without clinical risk factors had only a 3% incidence of perioperative morbidity, and noninvasive testing could not further stratify risk. Similarly, patients with three or more clinical risk factors had a morbidity rate of 50%, and noninvasive testing did not further stratify risk. Preoperative dipyridamole thallium imaging was useful in the group at moderate risk (one or two risk factors). This concept, in which the probability of a disease outcome after testing is a function of the probability of disease before testing, is an application of Bayes theorem. It helps explain differences in the predictive values reported in consecutive and selective patient series. Vanzetto and coworkers [13] applied such an approach to a study of consecutive patients having abdominal aortic surgery and reported that only 134 of 517 patients had preoperative testing and that thallium single-photon emission computed tomographic imaging had substantial prognostic value compared with only clinical variables.

L'Italien and colleagues [14] constructed a Bayesian model using clinical risk factors and the results of dipyridamole thallium imaging (Figure 1). The pretest (baseline) probability of an event can be calculated in a manner similar to that used to calculate cardiac risk indices. Each clinical risk factor is associated with a certain numeric weight, and all of the weights can be added to determine overall risk. The results of dipyridamole thallium imaging can then be used to modify the risk and determine the post-test probability of an event. Thus, clinical risk factors can determine whether the test will increase the probability of morbidity above some threshold for action.

View larger version (16K): [in this window] [in a new window]   Figure 1. Clinical algorithm developed using clinical variables and dipyridamole thallium results sequentially to stratify risk in 1081 patients having vascular surgery. Events are perioperative death or myocardial infarction. Clinical variables include advanced age (> 70 years); history of angina, myocardial infarction, diabetes, or congestive heart failure; and previous coronary bypass grafting. Dipyridamole thallium variables were ischemic electrocardiographic changes, fixed defects, and reversible defects. The algorithm shows the Bayesian concept that clinical markers only may substantially alter the probability of disease, eliminating the need for further testing. Testing can therefore be reserved for those patients in whom it would further modify the probability of disease and change management. Adapted with permission from the Journal of the American College of Cardiology [14].

Two recent studies show the reduced value of noninvasive testing when consecutive surgical patients are studied. Mangano and colleagues [15] studied 60 consecutive patients having vascular surgery and reported a positive predictive value of 27% for adverse cardiac events, a negative predictive value of 82%, and no net discriminative ability of the test. Baron and colleagues [16] studied the largest consecutive sample of patients having abdominal aortic surgery (n = 457) and were also unable to show an association between thallium redistribution and perioperative cardiac morbidity. Both studies show the low positive predictive value and clinically significant incidence of morbidity in patients with negative test results if consecutive patients are studied.

Using Other Diagnostic Tests before Noncardiac Surgery

In addition to dipyridamole thallium imaging, dobutamine stress echocardiography, Holter monitoring for silent ischemia, and coronary angiography have been advocated. Considering the risks and costs of coronary angiography, the test should be reserved for patients at very high risk (more than three clinical markers or moderate clinical risk plus highly positive stress test result) in whom coronary revascularization is considered. Dobutamine stress echocardiography has been found to have excellent negative and positive predictive values in several large series. The test is best used in patients at moderate clinical risk, similar to dipyridamole thallium imaging [17]. Although the presence of silent ischemia on preoperative Holter monitoring predicted poor outcome in several series, the test does not appear to stratify risk on the basis of the quantity of ischemia [18, 19]

Further Refinements of Noninvasive Testing

Several investigators have suggested the use of quantitative diagnostic testing to improve the predictive value of a test. Patients with a greater quantity of reversible segments of thallium defects or increased lung uptake were found to have greater perioperative and long-term risk [19]. In a large series of patients having dobutamine stress echocardiography, the heart rate at onset of regional wall motion abnormalities was the best predictor of perioperative risk [17].

American College of Cardiology and American Heart Association Guidelines

Several investigators have proposed algorithms for preoperative evaluation, including the use of noninvasive tests and coronary angiography. Most recently, the American College of Cardiology and the American Heart Association published guidelines on perioperative cardiac evaluation for noncardiac surgery [20]. They proposed an algorithm that incorporates six steps Table 1 and Table 2. First, the clinician must evaluate the urgency of the surgery and the appropriateness of a formal preoperative assessment. Next, he or she must determine whether the patient has previously had revascularization procedure or coronary evaluation. Patients with unstable coronary syndromes should be identified, and appropriate treatment should be instituted. Finally, the decision to have further testing depends on the interaction of the clinical risk factors, surgery-specific risk, and functional capacity.

On the basis of the previous discussion, these guidelines state that the indications for coronary revascularization are essentially identical during the preoperative period and the nonsurgical setting. Therefore, coronary artery bypass grafting should be considered if it is likely to affect symptoms or long-term survival. Percutaneous transluminal coronary angioplasty should be considered for patients with severe ischemia (particularly those with signs of left ventricular dysfunction), poorly controlled angina, or both.

How Little Information Do We Need?

In an era of cost-containment, one goal is to obtain the minimum workup that provides sufficient information for safe surgical and anesthetic procedures. Many patients are evaluated 1 day or less before surgery, which may lead to an abbreviated evaluation. In addition to the potential short- and long-term benefits of the preoperative evaluation outlined above, the evaluation may also give the patient information with which to make an educated decision for or against elective surgery.

Cost Issues

Assessment of the economic effect of preoperative testing depends on the strategy used. On the basis of a survey of cardiovascular anesthesiologists, it has been estimated that more than 1 billion dollars is spent annually on cardiac testing before vascular surgery. The routine use of preoperative testing would be extremely expensive and not cost-effective. However, if a selective testing strategy such as the one outlined here is used, only a fraction of patients having major vascular surgery will have a noninvasive test [13]. Considering the costs associated with a major vascular operation ($20 000 to $40 000), selective testing in those patients in whom care might be substantially modified is likely to be cost-effective and may actually reduce overall costs if the surgical procedures are modified or less intensive monitoring is used.

Most costs associated with preoperative testing strategies are associated with subsequent interventions such as angiography and coronary revascularization. By applying a strategy that limits further testing to persons with evidence of extensive myocardial risk and limiting coronary revascularization to indications that improve long-term outcome, the costs could be viewed as part of the long-term care of the underlying coronary artery disease.

Can We Ever Prevent All Morbidity?

In determining the value of preoperative testing, it is important to acknowledge that perioperative cardiac morbidity is multifactorial in origin. Many physiologic factors have been associated with postoperative ischemia and infarction, including increased myocardial oxygen demand and decreased myocardial oxygen supply from anemia. Importantly, the postoperative period is associated with a hypercoagulable state that may promote the formation of postoperative coronary artery occlusion. Noninvasive testing may identify obstructive coronary lesions distal to which myocardial ischemia may develop, but it cannot predict noncritical stenoses that are susceptible to plaque rupture and development of myocardial infarction.

Conclusions

The preoperative evaluation offers an opportunity to identify unsuspected or occult coronary artery, valvular, arrhythmic, or myocardial disease and to determine the patient's functional status. For many patients, presentation for noncardiac surgery may represent the first chance for cardiac assessment in years, if ever. Noninvasive cardiovascular testing for coronary disease can identify patients with extensive disease and may lead to modifications in perioperative care, but it should only be done in selected patients who are at moderate clinical risk. If not, the tests lose their predictive value because of a high incidence of false-negative and false-positive results. To best use a test, the results should be quantitated, and patients with extensive test abnormalities in particular should be considered for coronary angiography. If coronary revascularization is reserved for those patients in whom it may improve symptoms or enhance long-term outcome, then the overall evaluation should prove cost-effective from the perspective of both perioperative and long-term cardiovascular care. Finally, the foundation of management strategies in perioperative treatment must draw from the medical profession's rich experience in managing patients in the nonsurgical setting.

Dr. Eagle: University of Michigan Medical Center, 3910 Taubman Center, 1500 Medical Center Drive, Ann Arbor, MI 48109.