Chapter 9. Evaluation of the Patient with Cardiovascular Disease

1. Perioperative ischemia results from an imbalance of oxygen supply and demand of the myocardium during perioperative stress, rupture of a vulnerable plaque with subsequent thrombosis of a coronary artery, endothelial dysfunction, or some combination of all of these.

2. The history and physical examination, along with limited indicated tests, are often adequate to determine the general perioperative risk of a patient.

3. Preoperative cardiac testing is usually not needed in patients who have good functional status or are undergoing low-risk procedures. Further testing is unlikely to change management, and these patients may proceed directly to surgery.

4. Many interventions, such as control of diabetes, hypertension, stable coronary artery disease, congestive heart failure, and endothelial function, can be well within the scope of an anesthesiologist's skill set. If anesthesiologists wish to consider themselves perioperative clinicians, they must be active in these well-established aspects of medicine.

5. The attending anesthesiologist of a specific patient, not a consultant, is in the best position to balance the risks and determine the best anesthesia techniques for that case. Other clinicians and consultants are often needed to ensure optimization, although increasingly, the anesthesia team can do at least some of this.

Preoperative evaluation continues to change. In the distant past, the attending anesthesiologist would interview the hospitalized patient and family a day or two before surgery. All resources of the medical systems were mobilized to ensure that all aspects of the patient's health were assessed and treated. Surgery was delayed until any cardiac condition was fully addressed.

Practice changed; it then became common for any potentially complex patient to see a consultant to get “clearance” so surgery could proceed with some nebulous guarantee of acceptable risk, or surgery would be cancelled until such clearance could be provided.

Current practice includes surgery and anesthesia techniques that are numerous and changing. Anesthesiologists are able, if needed, to shepherd high-risk patients through high-risk surgery. Consideration is given to striking a balance between risks at either end of a spectrum, weighing the risk of postponing surgery versus proceeding. At one end, postponing surgery for investigations may be higher than the risk of proceeding. At the other end of the spectrum, some surgical procedures are unlikely to cause more stress than the activities of daily living. In essence, the paradigm of preoperative assessment is shifting from predicting risk to actively managing risk.1 The issue is less whether to cancel surgery and more whether indicated cardiac tests and management need to be done preoperatively under the supervision of the perioperative physician or postoperatively in a more elective fashion by primary care clinicians.

As medical therapy of cardiovascular disease improves, routine care is sometimes as effective as interventional therapy in reducing cardiac morbidity and mortality.2 This chapter reviews current understanding of the physiology of perioperative ischemia, followed by examination of how an individual patient's perioperative risk can be assessed. More importantly, it then discusses when and how more detailed risk management should be undertaken.

This chapter focuses on the patient presenting for noncardiac surgery. Many of the studies in this area have been done on patients having vascular or cardiac surgery, but as much as possible, evidence on other surgical populations is presented. Although it will focus on perioperative cardiac complications (PCCs), it is important to remember that the same process is occurring in the rest of the cardiovascular system3 as well as in other organs such as the brain and kidneys.

Gastrointestinal, pulmonary, renal, infectious, wound, and pain control complications are not insignificant in the surgical patient.4 Approximately 50% of patients with cardiac disease develop other complications perioperatively,5 but it is cardiovascular complications that garner the most concern. In patients with cardiovascular disease, known or unknown, perioperative cardiac complications occur at a higher rate. An analysis of the 6 studies assessing the incidence of major cardiac complications in patients with some risk factor for coronary artery disease (CAD) revealed an overall rate of 3.9%.6 Patients with peripheral vascular disease (PVD) are a well-studied group because they are easily identifiable and have a high complication rate. Even vascular patients with no risk factors for CAD have a 5% incidence of PCC.7 The Coronary Artery Revascularization Prophylaxis (CARP) trial found a PCC rate of 15.6% in vascular patients8 and a 27% rate in vascular patients with known CAD.9 Risk of perioperative complications in patients with no evidence of cardiac disease is very low (<1%).10,11 An analysis of 188,110 joint-replacement surgeries showed the excess mortality resulting from complications to be only 0.12%.12

The main reason for performing preoperative assessments is ultimately to get the surgery done. But additionally, in a sometimes fragmented health care system, the preoperative encounter may be one of the few times a careful assessment of a patient's cardiac status will occur. Following a comprehensive assessment, conditions such as atherosclerotic disease, hypertension, congestive heart failure (CHF), chronic obstructive pulmonary disease (COPD), asthma, and diabetes may be identified and treated so the patient's health will be improved for years to come. Some medications, such as antihypertensives, β-blockers (BBs), statins, and hypoglycemics, can be easily initiated with instructions for follow-up with primary care. Anesthesiologists have traditionally been reluctant to prescribe medications whose effects last past the immediate encounter. But primary care clinicians routinely and safely initiate drug therapy with follow-up done weekly or monthly at best. Because the limitations to successful treatment are usually patient access and compliance, these clinicians will enthusiastically welcome any assistance in drug therapy initiation, and willingly assume follow-up. The risk of initiating recommended medications is minimal, manageable, and a necessary part of ethical medical care. To miss these opportunities to forward improvement in the patient's health would be disappointing (Table 9-1).

Dangers of Testing in Low-Risk Populations

In contrast, routine screening investigations of asymptomatic patients should not be undertaken unless a plan for follow-up is in place. Testing a low-prevalence population inevitably produces mild abnormalities that are likely to be either false positives or true positives that have little bearing on the perioperative period. If surgery is delayed even to follow up on just the true positives, the inconvenience and risks of delayed surgery may outweigh the minimal benefits of follow-up. The likelihood of using this data to modify plans is also small.13 The perioperative period is currently too short and isolated for disease screening. Screening is better done by primary care, where follow-up is conducted over time in the context of addressing overall health and where the few true positives will be sifted out from the many false positives. Case finding—the investigation not of asymptomatic patients but of those who have manifested some indication of possible disease—is sometimes appropriate perioperatively.

Each institution must balance how much intervention can be done without exposing the patient and institution to the dangers of providing inadequate follow-up. Generally, what can be done easily and safely should be done; other investigations or interventions that do not have a direct impact on the perioperative period should be deferred to primary care. Ideally, the perioperative team and the primary care practitioners should share information and resources.

Who Should Do the Preoperative Assessment?

As mentioned previously, one method of preoperative assessment is to refer any patient with potential cardiac risk to the cardiology or internal medicine service and await “clearance” for surgery by that specialist. Unfortunately, cardiologists may not be in the best position to judge the risk of surgery, the risk of not proceeding with surgery, the risks of various anesthesia techniques, and the skills of a specific anesthesiologist. Thus it is inappropriate for a cardiologist or any other clinician to “clear” the patient for surgery. It is sometimes very important to get a consultant's advice on how to optimize a patient for surgery and an opinion on whether the risk/benefit of surgery is favorable, but the decision on whether a specific patient can undergo a specific procedure with the given resources can only be done by the anesthesiologist responsible for that patient on the day of surgery. Again, this chapter clarifies what information the anesthesiologist needs to make that final decision and to formulate the anesthesia plan. As a side note, many preoperative clinics use nurse practitioners or physician assistants. If these clinicians are educated and supported by anesthesiologists and consultants knowledgeable in anesthesia and surgical issues, they do a comprehensive and meticulous job of focusing on critical issues. However, because the anesthesiologist is the main consumer of the preoperative assessment, it behooves every anesthesiology department to remain intimately involved in the preoperative assessment process.

Patients with cardiac disease are at increased risk for perioperative morbidity and mortality. The overall rate of perioperative myocardial infarction (MI) in these patients is approximately 5%.6

Several cardiac complications can occur perioperatively (Table 9-2). Because most of these are either the cause or the result of ischemia,13 all of which may ultimately lead to infarction, the emphasis in this chapter is on the causes of ischemia.14

The etiology of perioperative ischemia is clearly multifactorial and defies simple classification, but understanding the theories on the physiology helps us understand the optimal ways to minimize it.15 As the theory changes, so does the treatment. Or as different treatments are proven superior, our theories must change. The physiology of ischemia is another area of perioperative medicine that is entering the third swing of the pendulum, back to one of caution.

Supply-Demand Mismatch

Ischemia traditionally was believed to be due to a supply-demand mechanism at a critical stenosis. Infarction from this source probably occurs only after several hours of ischemia.16 Reduction in the mismatch by reducing heart rate and/or increasing oxygen supply may relieve the ischemia, or a thrombus may form and cause complete coronary occlusion. In that case, thrombolytics, coronary artery bypass grafting (CABG), or percutaneous transluminal coronary angioplasty (PTCA) with stenting is indicated. On this basis, the preoperative treatment of critical stenosis made sense, and early studies such as the Coronary Artery Surgery Study in 1986 showed this was the case.17 Until recently, surgical intervention appeared superior to medical therapy alone.18

Thrombosis from Endothelial Dysfunction

Over the last few decades, there has been increasing appreciation of another mechanism of ischemia: erosion and rupture of unstable endothelial plaques, with resultant abrupt occlusion. These episodes result in ST depression as demonstrated on an electrocardiogram (ECG). They may resolve spontaneously but leave the damaged areas vulnerable to further plaque rupture.19,20 The Task Force on the Redefinition of Myocardial Infarction defines this mechanism as causing type 1 MI and the supply-demand mismatch mechanism as causing type 2 MI.21

In the last decade, several new studies have suggested that plaque rupture of noncritical lesions was a more common cause of ischemia, at least in the case of perioperative ischemia.22 The study by Dawood et al of coronary artery autopsies showed evidence of plaque disruption in 55% of fatal perioperative MIs and in 40% of nonoperative MIs.23 Ellis et al assessed the preoperative angiograms of vascular patients experiencing a postoperative MI and compared them with the findings in a matched group of nonoperative MIs; they did not find a high-grade stenosis (>70%) in any of the patients who had a PCC.24 Cohen and Aretz studied the coronary artery autopsies of 26 cases of fatal postoperative MIs. A thrombus occurred on a stenosis of more than 50% in only 33% of cases25 (Fig. 9-1).

Vasoconstriction can be considered as part of endothelial function, although plaque rupture is probably not involved. This may be a fairly common cause of cardiac ischemia, especially in women.26

These theories are consistent with accumulating evidence, such as that from the CARP trial,27 in demonstrating that medical therapy is sometimes as effective as invasive treatment of critical stenosis. In the CARP study, McFalls et al randomized 510 Veterans Administration patients undergoing abdominal aortic aneurysm repair to either coronary artery revascularization before surgery or no revascularization before surgery. Percutaneous coronary intervention was performed in 59% and bypass surgery in 41%. There was no difference in MI in the first 30 days after surgery or in mortality at 2.7 years.27 A recent post hoc analysis of the high-risk patients in that study came to the same conclusion.8 A major criticism of this study was the exclusion from randomization of patients with greater than 50% stenosis of the left main coronary artery. But the Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial, which between 1999 and 2004 randomized 2287 patients to either percutaneous coronary intervention and medical therapy or medical therapy alone, did include patients with multiple vessel disease, diabetes, and it likewise showed no difference at 4.6 years of follow-up,28 even in advanced age patients.29

Endothelial dysfunction is a result of three broad groups of cardiovascular etiologies: hypercoagulability, inflammation, and sympathomimetic overdrive. These detrimental forces, discussed later, are controlled to some extent by endothelial protective factors (Fig. 9-2).

Hypercoagulability

Surgery induces a prothrombotic state that may be variable and hopefully measurable.30 If this state could be measured, it would be valuable for selecting which high-risk patients to treat with anticoagulants. Some patients are probably predisposed to hypercoagulability. The patient group with factor V Leiden is one example where it has not generally been necessary to require perioperative investigation or treatment additional to that which is standard prophylactic practice.31

Inflammation

Atherosclerosis is largely an inflammatory process,32 and surgery initiates further inflammatory response. Current research on the usefulness of markers such as C-reactive protein (CRP) in the prediction of perioperative MI from this source is promising.33 This topic is discussed further later in this chapter.

Sympathetic Stimulation

The obvious sympathetic response to the stress of surgery may be attenuated by neuraxial anesthesia,34,35 clonidine,36 or BBs.37 Reduction in anxiety by pharmacologic and interpersonal intervention is important to reduce all three of the factors of endothelial dysfunction: hypercoagulability, inflammation, and sympathomimetic overdrive.

Protective Factors

A few short episodes of ischemia may induce preconditioning that results in less damage if later ischemia occurs.38 One mouse study suggests that the surgical incision induces protective cardiac preconditioning.39 Nitric oxide induces endothelial stability. The anesthetic nitrous oxide does the opposite. An example of how these factors are interrelated is a study by Zafar et al of the relationship between clinical anxiety, depression, and platelet reactivity measured by optical aggregometry40 (Fig. 9-3).

Multifactorial Stenosis

Countering the evidence that much perioperative ischemia is from noncritical stenosis, a recent ultrasound assessment of culprit lesions20 and several recent analyses of patients presenting to an angio catheterization lab all showed that most infarcts did occur at high-grade stenosis.41-43 Some authors still believe that type 2 supply-demand ischemia is the most common cause of perioperative MI.22 Part of the apparent discrepancy occurs because angiography likely is not the gold standard for detecting sites of critical stenosis. Because atherosclerosis is a systemic disease in which all of the arteries are affected to some degree, but angiography compares the lumen of a segment with the lumen before and after that segment, a severe stenosis in the middle of moderate stenosis will not show up as a critical lesion (Fig. 9-4). There is also evidence of a “Glagov” phenomenon of vascular remodeling where the plaque may increase in size, yet the lumen remains unchanged or even enlarges slightly.44,45 Intravascular ultrasound or cardiac magnetic resonance imaging (MRI) shows promise in more accurately assessing lumens.46,47

The other reality is that high-grade critical stenosis probably forms because of repeated plaque rupture. Progression of lesions probably occurs in a phasic fashion, with the perioperative period a time of potential rapid progression.48 The study by Galal et al with transesophageal echocardiography showed that the high-grade lesions detected preoperatively are not necessarily the sites of perioperative ischemia.49 Whether ischemia occurs because of either acute plaque rupture or preexisting coronary blockage,20 endothelial dysfunction likely is common to both mechanisms. Both identifiable high-grade stenoses and numerous smaller vulnerable plaques are amenable to systemic treatment of this endothelial dysfunction as well as interventional procedures where appropriate.50 BB therapy addresses both mechanisms: from reduced demand and probably more importantly from pleotropic benefits on the endothelium. As medical treatment improves, that treatment is becoming a viable option for treating even high-grade stenosis. This has fairly revolutionary implications in that if medical optimization can be safely and easily applied to any patient at risk, then detailed cardiac investigations or invasive therapies are probably unnecessary, at least in the short term.

Postoperative Ischemia

Advances in anesthesia have meant that the intraoperative period is a controlled and safe period. Most ischemia occurs in the immediate postoperative period during emergence from anesthesia, when catecholamines are surging,51 rather than during anesthesia. Landesberg et al showed that most ischemic events, and most ischemic events that result in infarction, occur immediately postoperatively (Fig. 9-5).52 Several studies have shown that postoperative MIs occur mostly in the first few days after an operation (Fig. 9-6), and are often asymptomatic.3,51,53,54

Prolonged ischemia is more likely to result in infarct than a shorter duration of ischemia. When the endothelium is damaged, acetylcholine causes muscarinic receptor–mediated vasoconstriction instead of endothelium-dependent vasodilatation.55 Thus prolonged ischemia may be the cause, as well as the result, of vasoconstriction and subsequent thrombus formation.22

Mangano et al showed that the risk from surgery may extend up to 2 years beyond the immediate perioperative period.56 Although much emphasis is placed on ischemia, postoperative stress may also result in patients reaching their anaerobic threshold. Once this threshold is met, diastolic dysfunction develops, possibly causing later morbidity, even if no ischemia occurs.57 Cardiac damage is a continuum, not a dichotomy of infarction or no infarction.58 Even asymptomatic ischemia may lead to long-term complications.13 Asymptomatic plaque rupture may heal but may leave the plaque more vulnerable to subsequent rupture.19,20 With these mechanisms in mind, we proceed to an analysis of risk for these events.

Preoperative assessment should occur in two stages: first, through a clinical assessment that defines general risk and/or determines that more detailed assessment will be useful. Only after this stage should more selective tests be ordered. The initial assessment can usually be conducted using questions, hands, and a stethoscope. When previous medical records are added, this basic history and physical examination often provide all the information needed for a given anesthesia plan. If further testing will not result in interventions or a change of anesthesia plan, it should at least be deferred to the postoperative period.

This section reviews the initial stage of general overall assessment of factors that affect perioperative risk. The information resulting from this assessment will determine if the next level of investigations are indicated. Risk indices have been a key part of this assessment, and following is a review of the common ones. A discussion of the latest American College of Cardiology/American Heart Association (ACC/AHA) guidelines is followed here by a more detailed review of individual patient risk factors. Particulars of more intensive investigations that may be indicated are covered in the following section, “Rationale for Selective Testing.”

Comparison of Risk Indices

Beginning with Lee Goldman in 1977,308 a number of collections of risk factors have been proposed to determine which patients are at high risk perioperatively and thus require detailed cardiac investigations. The simple and rather vague American Society of Anesthesiologists (ASA) Physical Classification system may predict risk almost as well as more complex indexes.59 The most commonly used algorithm for determining the need for assessment of risk for cardiovascular disease is the one created by the ACC/AHA.72 It is based largely on what is the best studied and validated index for use in predicting PCC: the Revised Cardiac Risk Index proposed and validated by Lee et al.10,60,61 Risk indices are cost-effective in vascular surgery.62,63 They are derived from populations and most appropriately applied to them.58 Although they do not perform well in defining exact risks in individuals, they can help place a patient in a general risk category.64 A study by Paul et al of four risk factors—history of diabetes, prior angina, previous MI, and history of congestive heart failure (CHF)—in 878 vascular surgery patients showed a good correlation with coronary angiography65 (Fig. 9-7).

Most indices divide patients into one of three risk groups: high, moderate, or low. A patient who has three or more risk factors on the Revised Cardiac Risk Index has a risk for postoperative MI of 11.4%; a patient with no risk factors has a risk of 0.3%.10 Patients scheduled for vascular surgery with no other cardiac risk factors have a PCC rate of 1.6% versus 23.4% in those with 3 or more risk factors.8

High Risk for Perioperative Cardiac Complications

Those patients with a greater than 5% risk of developing PCC are termed high risk. These patients occasionally warrant postponement of surgery to allow assessment of the risk-benefit ratio of proceeding to surgery and/or to allow optimization or special perioperative measures to control risk. In the past, it was believed that these patients required an automatic referral to cardiology. However, as described earlier, if the patient is already known to be optimized or investigation is unlikely to result in any change in perioperative management, then further investigation is unnecessary. Standard medical therapy increasingly is the optimal therapy even for high-risk patients.

Moderate or Indeterminate Risk for Perioperative Cardiac Complications

This group of patients is not clearly either high risk or low risk, and these are the patients who often warrant further risk analysis. Increasingly, however, if these patients are stable, they may proceed to surgery with relatively simple risk reduction. Even if they turn out to be high-risk patients, the management may be no different, and therefore there is no need to change perioperative management.

Low Risk for Perioperative Cardiac Complications

These low-risk patients rarely require preoperative testing.67,68 Tests that might be indicated in routine primary care should not be ordered unless follow-up of results can be arranged. However, the preoperative assessment may be one of the few opportunities in a fragmented health care system to initiate or adjust needed medical care. Each institution must determine the optimal amount of involvement by the perioperative staff in the patient's overall health care. Table 9-3 compares the contents of the most common risk indices.

Rationale for Selective Testing

Selective testing is cost-effective and safe62; indeed, it is statistically and medically inappropriate to test low-risk populations.69 The Bayesian theorem describes how testing a group with a low prevalence of a given condition results in too many false positives for the information to be useful. The purpose of the clinical assessment is to select patients who have a roughly 5% or greater likelihood of developing PCCs70 and in whom further testing will yield a useful change of pretest to posttest probabilities. Applying tests to a lower prevalence group is ineffective. For example, the recent Detection of Ischemia in Asymptomatic Diabetics (DIAD) study showed that even routine investigation of diabetics for cardiac disease using myocardial perfusion testing does not result in any significant reduction in morbidity and mortality if it is applied to asymptomatic diabetics.71 Even angiography is imprecise in a low-risk population.65 The application of this principle to cardiac testing is discussed later in this chapter.

American College of Cardiology/American Heart Association Guidelines

The clinical history and examination will uncover any of several risk factors for PCC, which each need to be assessed in a general fashion. However, more detailed investigation and management is not always needed perioperatively. This decision to investigate further is multifactorial, and the ACC/AHA algorithm, based on the Revised Cardiac Risk Index,10 is a useful aid to determine if such detailed investigation and management are indicated (Fig. 9-8). This aid should be used after a standard clinical examination has identified general issues of concern. The first four steps of the algorithm provide for potential “Reasons to Bypass Detailed Preoperative Assessment.” If none of these reasons to bypass apply—need for emergency surgery, presence of active cardiac conditions, low-risk surgical procedure, and good functional status of the patient—a more detailed assessment of risk factors is indicated. However, this fifth step remains quite flexible in that the end recommendation—“consider testing if it will change management”—applies to most branches of the algorithm, regardless of the number of risk factors. Thus counting the number of risk factors is less important than assessing each risk factor in light of that question, “Would more investigation change management?” Having said this, presence of two unoptimized risk factors is a reasonable simple threshold for ordering more detailed investigation. The first five steps of the algorithm are described here. The fifth step involves assessment of 5 main risk factors:

Step 1. Emergency Surgery

The need to proceed with truly emergent surgery with the information and resources at hand is fairly self-explanatory.

Step 2. Active Cardiac Conditions

Acute coronary syndrome, decompensated heart failure, significant arrhythmias, and severe valvular disease generally warrant postponement of surgery and urgent attention by emergency or cardiology clinicians. In earlier versions of the ACC/AHA guidelines, these conditions were grouped as Major Clinical Predictors of PCC. But they are more than risk factors; they are manifestations of actual disease. In the latest guidelines they are termed Active Cardiac Conditions,72 and the algorithm states they must be addressed before proceeding with anything other than emergency surgery. That these patients should have preoperative cardiac investigation is one of the rare class I recommendations in the guidelines. The details of management of these conditions are not discussed here.

Step 3. Is the Surgical Procedure Low Risk?

Many low-risk procedures entail risks that do not differ greatly from activities of daily living. If the patient can walk to the surgical suite, he or she has probably demonstrated cardiac function adequate for the stress of surgery. The large study of 18,189 cataract surgery patients by Schein et al showed that forgoing preoperative testing did not result in any change in outcome.73 The critical element of this study, which is sometimes missed, is that all these patients were assessed by their primary care practitioners, and any tests required for general health maintenance were done. This study supports the concept that a clinical examination is the irreplaceable foundation of preoperative assessment, and tests are secondary. A pilot study by Chung et al suggested that in many ambulatory surgery patients no routine tests of any kind are needed.68 The ACC/AHA guidelines categorize surgical risk as shown in Table 9-4. The study by Kumar et at verifies the profound influence of the type of surgical procedure on perioperative risk.74 Noordzij et al performed a recent assessment of postoperative mortality in 3.7 million surgical procedures in the Netherlands that shows an overall death rate of 1.42%. These data demonstrate that there is a wide difference in risk among different procedures and suggest that grouping procedures into 3 risk categories is too simplistic. For example, where both gastric surgery and biliary duct surgery are usually considered medium risk, the overall multivariable adjusted mortality for the former is 6.5% but for the latter is 1.6%.75 The ACC/AHA guidelines no longer make different recommendations for intermediate- versus high-risk surgery.73 Probably a more important differentiator is elective versus urgently needed surgery.

Step 4. Functional Status

Asking about the functional status of the patient is a remarkably simple tool that provides good prediction of risk for PCC, or more specifically, those who are not at risk of PCC. Functional status is a descriptor of how much the previously mentioned risk factors have an impact on the patient. Reilly et al showed that patients who reported they could not climb 2 flights of stairs or walk 4 blocks had twice as many PCCs as those who reported they could.76 Although patient self-description of exercise capacity correlates well with actual functional capacity,77 the accuracy of this test can be improved by actually observing the stair climbing. Girish et al observed maximum stair climbing in 83 surgical candidates.78 The overall rate of postoperative complications, including cardiac events, arrhythmias, reintubation, atelectasis, and pneumonia, was 25%. The rate was 89% in those who could not complete 1 flight of stairs. No patient able to climb seven flights of stairs had a complication.

Climbing 1 flight of stairs suggests having an aerobic capacity of 4 metabolic equivalents77 and probably represents a capacity to endure postoperative stresses without ischemia. The definition of a “flight of stairs” is quite variable: from 8 to 30 steps. By using a threshold of 2 flights of stairs, we signify the ability to climb to the second story of most houses. It is important that the patient be able to climb 2 flights of stairs without developing dyspnea, chest pain, or leg pain; this will minimize the contribution of oxygen-independent energy generation, which results in an overestimation of aerobic capacity in short-duration tests. It is the aerobic capacity not the anaerobic capacity that is critical in handling perioperative stress. Biccard described this concept and other limitations of a stair-climbing test.79

Good functional status usually signifies that whatever risk factors may be present, they are either mild enough or optimized well enough that they will not become a problem during surgery. Investigation and management may be needed, but they are unlikely to change perioperative risks, and therefore such investigation can be deferred to the postoperative period when it can be conducted methodically with careful follow-up.

The recognition that functional status is such a simple yet effective discriminator is in some ways a return to the use of the simple ASA Physical Classification scheme, which suggests that class 3 patients, those whose medical conditions affecting daily activities (as well as those in class 4), are the patients of concern, the others are not. In some ways, the pendulum of practice is swinging away from detailed risk assessment and back to a more general assessment.

Step 5. Clinical Perioperative Risk Factors

Assessment of clinical perioperative risk factors is the crux of preoperative assessment. The 5 risk factors described in the ACC/AHA guidelines have achieved unique status in preoperative assessment because they are clearly associated with PCC.72 The next section discusses those 5 risk factors as well as a few others.

Clinical Risk Factors

History of Ischemic Heart Disease

It has sometimes been assumed that the presence of known CAD automatically required a detailed cardiology consultation to “clear” the patient. Although having CAD is an obvious risk for PCC, it may not be as important a risk factor as previously thought, or at least not one requiring as much invasive intervention preoperatively. Most clinicians have encountered a vibrant middle-aged adult with a remote history of a small heart attack but who now runs marathons and probably has less risk than younger video-gamers.

A cardiologist's “clearance” is of limited usefulness. What is really needed is clarification of the risks involved, an assessment of whether these risks have been minimized as much as possible, as well as an assessment of the stability of the patient. When is CAD considered stable? Certainly either lack of symptoms or optimization and stabilization of symptoms is an important part of this assessment. Left main vessel disease or 3-vessel disease should be carefully assessed even if asymptomatic because progression is more likely to be clinically significant. But as the previous discussion of the physiology of ischemia addressed, even high-grade stenosis may be relatively stable, with preoperative investigation or treatment not adding benefit additional to that provided by routine care such as statin therapy.

Previous Myocardial Infarction

A few decades ago, a patient was considered at high risk for PCC for 6 months after an MI, and for some patients at some increased risk forever. But many MIs were treated with bed rest alone. Today, advances in treatment of acute coronary syndrome have improved the prognosis after MI. The angiography study by Van Belle et al of 56 post-MI patients showed that plaques remain unstable and vulnerable to reocclusion for 4 weeks, even after thrombolysis.80 Cardiac scar tissue heals in 5 to 6 weeks.21 Therefore, the high-risk period is now 6 weeks, with a period of relative risk from 6 to 12 weeks. In this relative-risk period, cardiac function is more important than time elapsed since the event.

Recent Coronary Artery Bypass Graft

In 1997 the Coronary Artery Surgery Study showed that high-risk patients undergoing a variety of noncardiac operations fared better if they had CABG surgery than if they had not. Clinically stable patients who had undergone CABG within the last 5 or 6 years were relatively “protected” from MI complicating noncardiac surgery and thus probably did not warrant routine preoperative stress testing.81 Then, a study by McFalls et al in vascular patients suggested revascularization may not be necessary preoperatively.27 Revascularization should be performed if it would have been indicated irrespective of the planned surgery. Which procedure should be done first requires a case-by-case analysis. This topic is discussed in more detail in the section “Managing Risk: What Interventions?”

Recent Percutaneous Transluminal Coronary Angioplasty

Recommendations for surgery after PTCA have changed dramatically as the type of intervention has changed, and they continue to be refined. In the 1990s, it appeared that PTCA performed a few weeks before an operation resulted in halving the rate of PCC.82 However, other studies quickly showed it was advantageous to wait some time after PTCA before undergoing surgery. Posner et al retrospectively studied 686 matched pairs of patients with CAD undergoing noncardiac operations: one set having undergone PTCA preoperatively and one set who did not undergo revascularization of their CAD. These groups were compared with 2155 normal controls. There was no difference in PCC between the 142 patients who had had PTCA less than 90 days before surgery and the matched patients with CAD who were not revascularized. Patients who had PTCA more than 90 days before the operation had a lower risk of PCC than did CAD patients who were not revascularized, although not as low a risk as that of normal controls.83 On this basis, it is recommended that surgery be delayed until at least 6 weeks after PTCA without stents.

Recent Bare Metal Stent Placement

With the advent of stenting during PTCA (Fig. 9-9), the vulnerable period initially appeared to be unchanged. In 2002 Kaluza et al studied 40 patients who underwent noncardiac surgery less than 6 weeks after PTCA using stents. Ticlopidine and aspirin were continued postoperatively. There were 8 deaths, all in patients who had operations less than 2 weeks after PTCA. Six deaths were from MI; 2 were from bleeding complications.84

A retrospective analysis by Nuttall et al of 899 patients who underwent noncardiac surgery within 1 year of percutaneous intervention (PCI) with bare metal stent placement showed the highest rate of PCC occurred when surgery was performed less than 30 days after PCI and the lowest if done more than 90 days after PCI.85 Thus the ACC/AHA guideline is to delay surgery to at least 90 days after bare metal stent placement.

Recent Drug-Eluting Stent Placement

With the introduction of the drug-eluting stent (DES) in about 2002, evidence quickly emerged to substantiate that although there were clear advantages regarding early stent thrombosis, there was an increase in late stent thrombosis, especially if antiplatelet therapy was discontinued.86 The late stent thrombosis was due to delayed endothelialization,87 and although late stent thrombosis was rare at about 1% to 2%, it was fatal half the time. Thus in 2007 the ACC/AHA issued a scientific advisory about the importance of perioperative continuation of antiplatelet therapy for at least 6 months, and preferably 12 months after placement of a DES, including during the entire perioperative period88 (Fig. 9-10).

Further studies show that antiplatelet therapy is more important than time between stent placement and surgery in reducing stent thrombosis.89 Although dual antiplatelet continuation for 12 months is mandatory, clopidogrel treatment beyond 12 months may not be necessary.90,91 A new generation of DES using everolimus shows promise of reduced stent thrombosis.92

Diabetes

The second clinical perioperative risk factor is diabetes. Patients with asymptomatic type 2 diabetics have the same risk of MI as patients with previous MI.93 Diabetes is considered a risk factor for PCC by most sources94,75,95 (Fig. 9-11), although some studies have not shown evidence in this regard.10,96,97 Patients with diabetes who do have perioperative MIs have more complications than patients without diabetes.9 Although diabetes is a risk factor for cardiac disease, screening asymptomatic persons with diabetes for ischemia has not proven beneficial.66,71 The American Diabetes Association recommends screening asymptomatic persons for diabetes only in high-risk patients.98 Hyperglycemia is a risk factor for PCC; whether strict control of diabetes reduces risk is discussed in the section “Medications.”

Congestive Heart Failure/Cardiomegaly

The third clinical perioperative risk factor, CHF, is both a cause and a result of ischemia. Discovery of unknown or decompensated CHF is one of the main purposes of conducting preoperative assessment. Decompensated CHF is an unstable cardiac condition that always warrants preoperative optimization, but even stable CHF warrants respect. Hammill et al, in a retrospective multivariate analysis of 159,327 Medicare patients undergoing major noncardiac surgery, showed that patients with the diagnosis of CHF have 51% more perioperative morbidity than those with CAD but no CHF99 (Fig. 9-12).

Cardiomegaly is a risk factor for PCC.100 Routine preoperative assessment of left ventricular function is a IIA recommendation (reasonable, may benefit) in the ACC/AHA guidelines.72 CHF can usually be detected with a careful history and physical examination, although subclinical levels of cardiomyopathy may develop postoperatively, which lead to late morbidity. In high-risk groups there may be benefit to preoperative echocardiography even in asymptomatic patients.50 Testing for B-type natriuretic peptide (BNP) may assist in monitoring mild changes.49,101

Renal Insufficiency

Renal insufficiency was added to the list of intermediate clinical predictors of PCC by the 2002 update of the ACC/AHA guidelines,102 and it is the fourth of the clinical perioperative risk factors. An analysis by Lee et al showed increased risk with creatinine more than 2 mg/dL in the derivation cohort but not the validation cohort.10 A recent validation of the Lee Risk Index by Hoeks et al verified that renal insufficiency along with cerebrovascular disease and insulin-dependent diabetes are the predominant contributors to predicting late mortality.103 Often, the only optimization that can be done for renal insufficiency is to ensure that a reversible cause such as a drug reaction has not been missed. It is wise to document preoperatively the etiology and level of insufficiency or failure so that the cause of perioperative changes can be appropriately attributed.

Cerebrovascular Disease

The fifth clinical perioperative risk factor is a history of cerebrovascular disease. History of stroke or transient ischemic attack (TIA) was an independent correlate of PCC in both the derivation and the validation cohort of the Revised Cardiac Risk Index,10 as well as in studies conducted by Boersma et al and Kertai et al.104,105 Optimization includes appropriate anticoagulation.

Carotid stenosis, in addition to being a marker for cardiovascular disease, may predispose a patient to perioperative stroke. Therefore, whether or not carotid surgery should precede other operations has been a dilemma. The risk of serious stroke in patients undergoing a nonvascular operation is less than 1%. It is no higher in patients with bruits than in those without bruits.106 In 2004 the Asymptomatic Carotid Surgery Trial showed benefit from carotid endarterectomy performed in patients younger than 75 years who had asymptomatic stenosis greater than about 70%, but that was probably achievable only in certain centers.107,108 There is no evidence to support prophylactic carotid endarterectomy before general surgery.109 Medical therapies have improved in recent years to provide benefits similar to surgery.110-112 Thus in asymptomatic carotid stenosis, medical therapy for peripheral vascular disease should be optimized, but no further screening in this low-risk group is needed. Even a carotid bruit is not an indication for screening in asymptomatic patients. The practice guidelines of the American Society of Neuroimaging as well as the Society of Vascular and Interventional Neurology recommend screening for carotid stenosis only in patients undergoing CABG and in symptomatic vascular patients.113

Valvular Lesions

Asymptomatic valvular conditions are generally of no specific concern perioperatively. Aortic stenosis, however, is of grave concern to the anesthesiologist, and the discovery of this condition is a primary objective of preoperative assessment. The history may give evidence of some recent change in exercise capacity if this lesion is becoming significant. However, where aortic stenosis is a possibility, the threshold for investigation of a heart murmur should be very low because the rate of PCC can be very high.114 New murmurs and probably any murmurs in an older patient warrant echocardiographic assessment. Physiologic murmurs are common but do not preclude development of aortic stenosis later in life. If aortic stenosis is discovered, the risk can be minimized with advance planning and careful technique.115,116 Detection of any new diastolic murmur requires further investigation.

Infective endocarditis (IE) prophylaxis was simplified in updated guidelines.117,118 The risk of IE is so low, the benefit of prophylaxis is so limited, and the concern of developing antibiotic resistance is significant enough that the recommendations have been scaled back considerably. There are no longer any class I recommendations (should be performed) for IE prophylaxis. Basically, the class IIa recommendations (reasonable) are that prophylaxis is recommended for high-risk patients (those with prosthetic valves, repairs with prosthetic material, history of IE, some congenital heart conditions, some heart transplant) undergoing invasive dental procedures. Only severe mitral regurgitation warrants prophylaxis, although this is somewhat controversial.119 Prophylaxis for these high-risk patients who are undergoing invasive respiratory or infected-skin procedures is a class IIb recommendation (may be considered) with level of evidence C (consensus only). A similar recommendation is made for high-risk patients with established genitourinary (GU) or gastrointestinal (GI) infection undergoing GI or GU tract procedures, but this is mostly to prevent wound infection or sepsis; it is unlikely to prevent IE. The British guidelines include a slightly stronger recommendation for prophylaxis in GU and GI procedures.120 The European guidelines are slightly more restrictive than those of the ACC/AHA in that no heart transplant patients are considered high risk, and prophylaxis in any procedure except dental is considered class III (not recommended).121 Most needs for prophylaxis of subacute bacterial endocarditis are satisfied as part of usual surgical prophylaxis.

Peripheral Vascular Disease

Peripheral vascular disease is a condition that shows up in 2 locations on most assessments of perioperative risk. Vascular surgery has been repeatedly demonstrated to be high risk because of the comorbidities and risk the patient brings to the procedure,122,123 as well as those of the procedure itself. But some vascular surgical procedures such as endovascular carotidectomy are proving to be of low to moderate risk. Therefore vascular patients should be individualized according to the specific procedure and almost inevitable comorbidities. As with CAD, if optimal medical treatment is in place, further investigation is probably not needed. However, these are often very complex patients who are not optimized and may need assistance organizing overall care. Aspirin should be continued, although the overall benefit of this is under some debate.124

Other Coronary Artery Disease Risk Factors

Several other conditions or markers are associated with CAD, but the association with PCC is still being clarified. With further studies, some of these conditions may become promoted to the previous category of clinical perioperative risk.

Smoking

Smoking is a clear risk factor for cardiac disease and poor wound healing. Smoking cessation is one of the most effective ways to reduce cardiovascular risk. The unsolved question is how long before surgery cessation is necessary for there to be some benefit. There is concern that the stress of withdrawal may add to perioperative catecholamine surge. The issue of preoperative cessation is discussed in the section on smoking cessation.

Hyperlipidemia

Hyperlipidemia is a major risk factor for CAD, and improving the lipid profile leads to a reduction in risk. Initially it was assumed that because lipid profiles change so slowly, management of lipids was not a practical perioperative issue. However, treatment with statins is proving to have some fairly rapid benefits, although this is probably unrelated to improvement of hyperlipidemia. Perioperative statins are discussed in detail in the next section.

Arrhythmias

Arrhythmias are a cause and an effect of ischemia. Significant arrhythmias such as ventricular tachycardia are considered unstable conditions needing preoperative attention, but arrhythmias in and of themselves do not generally precipitate problems unless the heart is already compromised. Therefore, arrhythmias are of more concern if there is other evidence of cardiac disease. In otherwise healthy persons, conditions that may occur which are of perioperative concern are Wolff-Parkinson-White disease and prolonged QT syndrome. These conditions have a bearing on which drugs are selected to treat potential arrhythmias.

Prolonged QT Interval

Patients with a long QT interval—a corrected QT interval (QTc) greater than 470 milliseconds for males or 480 milliseconds for females—are at increased risk for perioperative cardiac arrhythmias and morbidity and mortality. Congenital long QT syndrome affects 1 in 5000 people and occurs because of cardiac ion channel abnormalities.125 Acquired long QT syndrome is more common and usually occurs secondary to medication side effects or electrolyte disturbances. A prolonged QT interval increases the likelihood of torsades de pointes ventricular tachycardia and ventricular fibrillation. In patients with newly identified long QT interval, elective surgery should be delayed for patient optimization or workup. The strategies for management of patients with long QT syndrome include suppressing sympathetic nervous system activity through establishing β-blockade, avoidance of sympathomimetics, and ensuring adequate pain control and anxiolysis. Drugs that may lengthen the QT interval (typically antiarrhythmics, antibiotics, or antiemetics) should be avoided (Table 9-5). Some inhaled anesthetics have been associated with prolongation of QT interval, resulting in recommendations for anesthesia maintenance with isoflurane or propofol infusion. Hypothermia should be avoided. The QT interval should be monitored and a defibrillator should be available during the perioperative period. Table 9-5 and other information on prolonged QT is available at www.torsades.net.126

Hypertension

Although hypertension has clearly been shown to be a risk factor for cardiovascular disease,127,128 the evidence that hypertension actually results in increased PCC is limited. A meta-analysis of 30 observational studies demonstrated an odds ratio for the association between hypertension and PCC of 1.35 (range: 1.17-1.56129; Fig. 9-13). This difference is not clinically significant in the short term. However, even if hypertension in and of itself is not a significant perioperative risk factor, the cardiac conditions that often accompany hypertension are. These include left ventricular hypertrophy, renal failure, and stroke. Uncontrolled hypertension is an unoptimized risk factor that should be addressed preoperatively.

The understanding of the nature of hypertension and its relation to cardiovascular disease is changing. Peripheral arterial pressure, which is what we usually measure, does not reflect large-artery stiffness, pulse wave reflection, and central artery pressure; these are a major part of the hypertension pathology, especially in the elderly.130,131 The initial pulse wave from systole travels down the vasculature and is reflected back at bifurcation points. If these reflected waves return to the central artery before diastole, there is a summation wave that increases the stress on the heart. Measurements of central aortic pressures, although impractical at this time, are probably better predictors of cardiac morbidity and mortality than peripheral blood pressure. Some medications such as BBs may improve peripheral pressure but actually increase central pressures.132,133 Part of the atherosclerotic process is a stiffening of the vasculature that results in augmentation of a reflection of the systolic pulse. Angiotensin-converting enzyme inhibitors (ACEIs) may reduce this augmentation.134

Even white-coat hypertension is associated with increased cardiovascular risk and endothelial dysfunction,135 but evidence as to whether this affects overall cardiac risk, let alone perioperative risk, is mixed.128,136-139 Hypertension likely has a different pathophysiology in the young, obese, and elderly. Management of preoperative high blood pressure is discussed in the section “Management of Hypertension.”

Anemia

In addition to disrupting the supply-demand balance, anemia is a marker for other comorbidities. Anemia is associated with increased postoperative morbidity,140 and measures other than transfusion may be needed to improve outcomes.141 There is a trend toward greater tolerance of perioperative anemia except in patients with CAD.142 If significant anemia is identified far enough in advance of the operation, transfusion may still be avoided through the use of preoperative erythropoietin and iron, intraoperative isovolemic hemodilution, and special intensive care unit (ICU) maneuvers.143

Age

Age older than 70 years was an independent risk factor for PCC in some multivariate analyses9,59,60,144,145 but not in others.10,74,105 Age in and of itself is probably not a significant risk factor in the absence of associated comorbidities, but increased age may magnify the impact of other risk factors.146 For example, if an MI does occur, mortality is higher in older patients.147 Nevertheless, the COURAGE trial did not show any benefit in adding percutaneous cardiac intervention to optimal medical therapy, even in patients older than 65 years.148

Biomarkers

B-type natriuretic peptide (BNP) is a biomarker of inflammation and mild myocardial damage that is abnormal in advanced heart failure.149,150 BNP has a high negative predictive value and might be useful to select out some high-risk patients who do not need echocardiograms to rule out ventricular dysfunction.151 Recent studies show that preoperative levels may predict myocardial injury152-154 and hospital length of stay.155 As we become more aware of the effect of mild cardiac failure, BNP may be a useful marker of perioperative deterioration.101

High-sensitivity C-reactive protein (CRP) was shown by the Women's Health Initiative, which analyzed more than 14000 women, to be predictive of adverse cardiovascular events.156 CRP is particularly predictive of postoperative morbidity in cardiac surgical patients.157,158 The US Preventive Services Task Force states that CRP is useful in CAD risk stratification.159 Measurement of CRP was instrumental in the somewhat controversial JUPITER (Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin) trial on statin use in healthy patients with normal cholesterol.160-162

CRP is probably more than a biomarker in that it is actually part of the inflammation cascade. If so, then decreasing CRP may directly improve endothelial function.163 Even so, at-risk patients should already have been identified by clinical examination and optimal treatment with statins instituted.159

Albumin

Low albumin is a marker of overall debilitation and acute-phase reactants, and thus it serves as a very good prognostic test for PCC. Unfortunately, this test result does not offer much guidance in perioperative management, aside from suggesting a longer delay in surgery to allow improved nutrition.

Depression

Many studies show a that depression and major affective disorder are predictors of CAD, imposing about a 2-fold increase in risk.164-166 The relationship is likely multifactorial with mechanisms such as compliance to treatment and antiplatelet activity of antidepressants completing part of the picture.167,168 Successful treatment of depression is correlated with reduced risk of postoperative death.169 Again, this serves to remind us of the importance of cooperation with primary care or consultant specialists in treating the whole patient, not just the organs and chemistry.

Genetic Polymorphisms

Genetic testing is increasingly commonplace, and there is great interest in how this might be applied to perioperative risk stratification.170 Factor V Leiden is an example of a common genetic polymorphism. However, with the possible exception of cardiac surgery, routine antithrombotic measures are sufficient to counteract the effects of this condition, and routine preoperative testing for this is not recommended.31

Some studies show an association of specific genetic polymorphisms with atherosclerotic complications after CABG,171,172 whereas other studies have not shown strong associations.173,174 The Multicenter Study of Perioperative Ischemia Research Group showed that certain polymorphisms for platelet glycoprotein IIIa as well as the degree of platelet activation are related to levels of troponin after CABG, suggesting that this platelet polymorphism contributes to perioperative myocardial injury.175 Inflammatory response is also altered by genetic variability of key inflammatory genes.176-178 Genetic variability in β-adrenergic responses may assist in determining who should receive perioperative BBs and who should not.179-181

Although not routinely used at this time, genetic testing is a significant frontier in perioperative medicine.

As routine medical therapy becomes almost as effective as more invasive interventions, whether a patient has moderate or severe disease is often less critical to short-term management.

When Not to Start Detailed Preoperative Investigation

If extensive medical investigation would not change perioperative management, it is a poor use of time and resources. It is important to also remember that the patient is an active partner in risk management. For example, if a patient refuses to undergo invasive cardiac intervention, and medical optimization is likely all that will be done, then testing may be a moot point.

This is the main question to be answered in preoperative assessments: “Is the patient as optimized as possible?” Conducting a good functional status assessment often provides evidence that the patient's constellation of medical conditions is adequately optimized for surgery. Thus good functional status is one of the “shortcuts” that precludes the need for preoperative investigation (step 4 of the ACC/AHA guidelines).

A common scenario is a patient with stable cardiac disease for whom it would be “nice” to have more information. A stress echocardiogram is arranged, and the result is indeterminate. The cardiologist, when faced with the possibility that the patient has serious cardiac disease, wants to be sure nothing is missed. This motivation may stem as much from the concern for litigation as from a logical medical indication. Thus an angiogram is scheduled, and it shows a stenosis amenable to stenting. A stent is placed and clopidogrel started, which means that surgery should be delayed for up to 12 months.88 All this is done for cardiac disease that may have been less hazardous than the indication for an operation. Delay in surgery is inconvenient at least, probably unnecessary, and possibly dangerous. The CARP study27 and, more recently, the DECREASE (Dutch Echocardiographic Cardiac Risk Evaluation Applying Stress Echocardiography) V study182 show there is no benefit in preoperative versus postoperative cardiac intervention in vascular surgery (Fig. 9-14).

More recently the COURAGE trial, which randomized 2287 patients with CAD to PCI versus medical therapy, raises further doubt about the benefit of putting stable cardiac patients on the “cardiac train” of progressively increasing investigation and intervention.183 (Fig. 9-15). The results were the same whether the end point was all-cause mortality or only cardiac events.28

These findings are consistent with the DECREASE II trial that showed that although the Revised Cardiac Risk Index correlated strongly with postoperative cardiac complications, stress testing did not improve outcome. Median follow-up was 2 years. It must be noted that even the untested subjects received maximal medical therapy including titrated BB therapy.184 If this result is true for vascular surgery, it is likely even more true for lower-risk operations and lower-risk patients184 (Fig. 9-16).

Thus if a patient is generally stable (good functional status), it is often best to keep a patient off the cardiac train of a stress test, angiogram, angioplasty, and an antiplatelet agent because, once on board, he or she is unavailable for elective surgery until the end of the journey. Before purchasing the “ticket” of a stress test, the clinician and patient must be willing to go the full journey before scheduling the procedure.

When to Undertake Detailed Preoperative Cardiac Investigation

However, there certainly are times when detailed preoperative testing is appropriate. Just as a good surgeon can handle most conditions that might be encountered but still prefers radiologic investigations so as to be as forewarned as possible, so the anesthesiologist may benefit from knowledge of the balance of risks and the range of physiologic reserves.

A simplification of the previously discussed ACC/AHA guidelines is that when there are 2 or more unoptimized clinical risk factors,105,185 when surgery is assessed as more than low risk, and when good functional status is not present, consideration of further testing is appropriate for optimization. The tests are needed not primarily because the patient falls above a threshold for testing but because the patient's place on the risk continuum and adequacy of optimization are unknown. This testing can sometimes be done by the anesthesia team without cardiac consultation.

The main questions to be answered are these: Can the patient's health be improved? Can the risk of PCC be reduced? Are special perioperative measures needed? Can the procedure be done at an ambulatory surgery center?

What Tests?

Second-level tests, for purposes of discussion, are divided into resting cardiac tests, cardiac stress tests, and postoperative tests. Invasive tests, such as angiography, or alternatives such as intravascular ultrasonography,186 coronary artery calcium by electron-beam tomography,159 cardiac computed tomography,187,188 or cardiac MRI47,189 are not discussed. These specialized tests are rarely ordered or interpreted without a cardiology consultation.

Resting Cardiac Tests

Electrocardiogram

Routine preoperative electrocardiograms have limited value.68,190 The specificity of abnormalities is low, resulting in so many false positives requiring further testing and delay of surgery that the ACC/AHA guideline considers routine ECGs a class III recommendation (not recommended because probable harm is greater than benefit).72 Some of the abnormalities, such as major Q waves, ST-segment alterations, T-wave changes, Mobitz type II or higher blockade, left bundle-branch block, and atrial fibrillation, are associated with increased risk of PCC. But even these abnormalities do not add predictive value to that provided by the clinical examination.66 A careful clinical history and physical examination is the best screening test. If that assessment shows that the patient was asymptomatic, the optimal preoperative therapy is likely to be medical therapy, which would have been instituted even in the absence of an abnormal ECG. Infarction cannot be diagnosed on the basis of ECG changes unless there is also an elevation in troponin.21 Troponin levels are rarely drawn in absence of symptoms. A study by Liu et al study suggests that even ECG changes in patients older than age 70 are not predictive of PCCs.190 Thus most jurisdictions are selective, with overall ECG use rate as low as 13%.191

In contrast, selective ECGs in higher-risk patients do have value. The ACC/AHA guidelines recommend preoperative ECGs in patients with at least one clinical risk factor undergoing vascular surgical procedures, or patients with known coronary heart disease, peripheral arterial disease, or cerebrovascular disease undergoing at least intermediate-risk surgery.72 The recent study by Correll et al adds some evidence-based guidance. Their results showed that age older than 65 years, history of heart failure, high cholesterol, angina, MI, or severe valvular disease are associated with clinically significant abnormalities on ECG.192 Notice that this list does not include diabetes, which has often been considered an indication for preoperative ECGs, to detect silent MIs. The DIAD study showed that even screening with myocardial perfusion testing did not confer a reduction in cardiac events.71 The much less precise ECG presumably has even less utility in patients with asymptomatic diabetes.

If the ECG uncovers disease, it is usually in a setting where a careful history has not been taken. If, for whatever reason, a hospital is not able to provide consistent clinical examinations, screening ECG may have some value.

One argument for obtaining preoperative ECGs is to document any baseline abnormality and thereby assist in determining if an abnormality discovered postoperatively represents a perioperative change.193 No published evidence supports this argument, although it is logical. For such baseline purpose, an ECG less than 12 months old is likely adequate if there have been no clinical changes.

Two-Dimensional Echocardiography

Transthoracic (2D) echocardiography is used to detect major wall-motion anomalies, measure ejection fraction, and assess valvular lesions. Certainly identification of severe valvular disease is important. Left ventricular dysfunction is a predictor of future cardiovascular events and increased overall mortality.194,195 In the case of high-risk surgery and poor exercise tolerance in addition to suspicion of heart failure, knowledge of cardiac capacity will be useful to guide intraoperative management. If there is concern about ischemia, it is preferable to select one of the tests described next.

Cardiac Stress Tests

Cardiac stress tests involve two components: a stressor and a test. These can be mixed in any way, but the most common combinations are treadmill ECG, dobutamine echocardiography, and adenosine radionuclide perfusion.

Treadmill Exercise Testing

The simplest of the stress tests, treadmill ECG testing is a good first-line test in those patients able to tolerate the exercise.82,112 It is not useful in patients unable to exercise, with baseline ECG changes (left bundle-branch block, pacemaker, Wolf-Parkinson-White disease, ST elevation at rest), or with a history of revascularization. Treadmill exercise testing produces 30% false positives in women.196 In practice, if a patient is able to do treadmill testing, he or she probably has already demonstrated adequate functional capacity to preclude the need for preoperative testing. If further testing is needed, a more sensitive test, such as stress echocardiography, is generally preferred.

Older et al showed that cardiorespiratory exercise testing for anaerobic threshold is useful for identifying high-risk patients who are unlikely to develop diastolic heart failure postoperatively and therefore are unlikely to need postoperative ICU admission.57 Unfortunately, few institutions can logistically arrange these tests preoperatively.

Dobutamine Stress Echocardiography

This test has the advantage of requiring relatively little time to complete. It is a functional test in that it shows how cardiac function is affected by underlying pathology. Image quality is variable and usually limited in obese patients. Meta-analyses show that the receiver operating curve of dobutamine stress echocardiography is better than that of other tests.197 An analysis by Beattie et al, which included 68 studies of 10,049 patients, evaluated stress echocardiography to be superior to radionucleotide imaging because the negative predictive of stress echocardiography was superior.198 Because accuracy and availability vary between locations, each institution may have a different preferred test.82

A common dilemma is whether a dobutamine stress echocardiogram should be postponed in the case of a patient taking BBs. If the test is being done to assess whether CAD is present and whether β-blockade might be indicated, then testing without BBs has theoretical advantage. However, because continuation of BB has no significant effect on sensitivity or specificity,199 and there is risk with stopping β-blockade, in most cases continuation is a better course. Stopping BBs for the investigation might only prove that the BBs should not have been stopped.

Galal et al compared preoperative dobutamine echocardiography, intraoperative transesophageal echocardiography (TEE), and postoperative cardiac events. Although new wall-motion abnormalities noted on TEE were more predictive of postoperative events, neither of the echocardiographs reliably predicted the anatomic site of the postoperative event.49 Thus echocardiograms may be mostly markers of more systemic atherosclerosis, with limited ability to predict the specific site of thrombosis.

Radionucleotide Imaging

This test may be more sensitive than others in this category because it detects flow anomalies that do not yet cause wall-motion abnormalities; however, the test is also less specific. It does have the advantage of allowing use of vasodilation, which is a gentler stressor of the heart or aneurysms that may be present. Compared with echocardiography, its accuracy is less affected by obesity but is altered by large breasts.

Cardiac Magnetic Resonance Imaging

This test can combine flow studies with wall-motion studies in remarkable resolution. It is useful for complex high-risk cases but is time consuming and does not yet have a clear role in preoperative screening.47

Tests of Endothelial Dysfunction

As described earlier, endothelial dysfunction is one of the unifying mechanisms of PCC. Several new biomarkers and tests for this are showing prognostic value for cardiovascular disease in general and may prove to be useful in the perioperative period. One of these biomarkers is brachial artery flow-mediated dilation. In this “stress test” of the vasculature, nitroglycerin dilation is the “stressor,” and ultrasonic observation of the brachial artery is the test.200-202

Postoperative Tests

Because most ischemia occurs postoperatively and is silent, it makes sense to monitor for it, at least in high-risk patients.203 Postoperative ECG changes suggesting ischemia may be predictive of postoperative morbidity and mortality even in low-risk patients.204,205

Postoperative troponin elevation indicates increased risk of cardiac events.9,129,206,207 This test has become the gold standard for identifying postoperative ischemia. Le Manach et al studied the postoperative troponin levels of 1136 abdominal aortic aneurysm surgical patients and found that a slightly elevated level was associated with delayed postoperative MI.16 Discovery of elevated postoperative troponin may represent an opportunity to reduce asymptomatic ischemia.

After general risk level has been determined and more specific risks identified, that information must be used to reduce the perioperative risk where practical. Although consultation and cooperation with cardiology specialists are needed for many aspects of optimization, the anesthesiologist assigned to a case knows best what information is needed or not needed to balance risks, obtain informed consent, and formulate an anesthesia plan.

This section focuses mostly on preoperative and a bit on postoperative interventions. Interoperative management is addressed in other chapters.

Many of the medical interventions for reducing perioperative risk are part of the secondary prevention of cardiac disease. Although the perioperative team must limit itself largely to the management required in the perioperative period, this is an ideal time to institute or reinforce long-term cardiac care, especially the simple proven interventions such as control of hypertension, diabetes, and endothelial dysfunction.

Preoperative Invasive Interventions: When Should Coronary Artery Bypass Grafting or Percutaneous Transluminal Coronary Angioplasty Be Done Preoperatively?

Glance, in 1999, in a decision tree model, found that proceeding directly to vascular surgery resulted in the poorest 5-year survival (77.4%) compared with 3 screening strategies: (1) nuclear stress test in everyone (86.1% 5-year survival); (2) coronary angiography in everyone (87.9% 5-year survival); and (3) angiography in high-risk patients, nuclear stress in intermediate-risk patients, and no testing in low-risk patients (86.0% 5-year survival). Although the 5-year survival rate was similar in all of the screening strategies, the cost ($/year of life saved) of selective screening was half that of either nonselective strategy. There was no difference in 30-day outcome between any of the 4 strategies.62

The CARP trial, published in 2004, showed that preoperative medical therapy is as effective as invasive treatment of vascular surgery patients with moderate disease. Similarly, the COURAGE trial showed that percutaneous coronary intervention added no additional benefit to medical therapy at 4.6 years of follow-up, even in patients with multiple vessel disease or diabetes.28 The recent retrospective study of 23,991 surgical patients undergoing preoperative stress testing in Ontario, Canada, who were propensity matched to a cohort without testing, showed only slight benefit from testing (improved 1-year survival hazard ratio of 0.92; 95% confidence interval [CI], 0.86 to 0.99, p = 0.03) and reduced mean hospital stay (difference: −0.24 days, 95% CI, −0.07 to −0.43; p < 0.001). The authors conclude that much of the benefit of preoperative stress testing is from the medical management that results.69 It might be argued that this medical management could and should have been instituted regardless of stress testing.

The DIAD study has shown that even in patients with diabetes, ischemia detected with stress testing often resolves over time with medical therapy alone.208 Biccard and Rodseth's meta-analysis separating PCTA from CABG suggests there may be benefit to preoperative CABG in some situations.209

Coronary interventions should generally be performed as indicated by guidelines used in nonsurgical settings. Whether or not an operation should be delayed depends more on the urgency of the surgery than on the presence of CAD. Because of the evidence previously cited that preoperative surgical intervention is not necessarily needed, and also because of improving medical therapy options, the finding of mild reversible ischemia on stress testing does not necessarily require proceeding to angiography. It is not necessarily predictive of PCC.210 It may be appropriate to stop the cardiac train at this point if medical therapy is optimal. In essence, the destination has been reached, and it is best to get off the train.

Management of Hypertension

Hypertension as a risk factor for cardiovascular disease was discussed previously in “Other Coronary Artery Disease Risk Factors.” Management is usually elective, and an operation does not need to be delayed unless blood pressure is quite high. There are no evidence-based guidelines on the threshold of hypertension that warrant delay of surgery. The oft-quoted threshold for postponing surgery of 180/110 mm Hg is the result of a recommendation by Goldman and Caldera in their 1979 study that included 5 such patients, none of whom had any complications.211 Roghi et al did show an increased risk of cardiac complications in vascular surgery patients whose diastolic blood pressure was higher than 110 mm Hg despite taking two or more antihypertensive medications,63 but evidence to support postponing surgery because of severe hypertension is surprisingly absent.212

The important aspect of management of severe hypertension (>180/110 mm Hg) is the search for end-organ damage (cardiac ischemia, renal dysfunction, cerebral impairment). If there is any evidence of end-organ damage, the patient has an active cardiac condition that needs emergency care. This should be managed according to ICU protocols.213 In the absence of these, the condition is termed a hypertensive urgency, and the treatment goal is to gradually improve blood pressure control. Indeed, rapid reduction in blood pressure has added risks. If these patients are sent to the emergency department, they will be assessed with a clinical examination, ECG, and simple blood tests to rule out end-organ damage. If there is no such damage, oral antihypertensive therapy will be adjusted and the patient will be sent home. The anesthesiologist can often do this as well as the emergency department and at significantly less cost and inconvenience to the patient. There is probably no need to delay an operation if there is no end-organ damage and improvement in long-term management is underway.129

Blood pressure control must be continued through and beyond the perioperative period.214 Hypertension is the classic condition that may present preoperatively and whose treatment will result in long-term health benefits. Treatment of hypertension results in improvement in endothelial function within hours134 and reduction in cardiovascular events in 6 months.215 The reasons for treating hypertension may not stem so much from perioperative benefits as from the clear lifelong benefits. Patient compliance is often suboptimal, and perioperative attention to this condition by the anesthesiologist will assist greatly in long-term management. Compared with the potent drugs used daily for anesthesia, antihypertensives have great margins of safety. Anesthesiologists need to learn to write prescriptions where needed for long-term hypertension management. Primary care physicians can assume responsibility for titration of medication initiated perioperatively.

Smoking Cessation

Smoking is one of the strongest modifiable risk factors for cardiovascular disease. Whether the stress of smoking cessation outweighs the short-term benefits of preoperative cessation is unclear. An early study by Warner et al showed that it took more than 6 months to see a benefit from preoperative smoking cessation. This study is often quoted as showing that there was actually an increase in pulmonary complications during the first 2 months of cessation efforts.216 But the CIs for pulmonary complications were much too wide to support this apparent trend. Although sputum production does not decrease for at least 2 months, it is likely that benefits from smoking cessation increase with time.217 Møller et al studied 120 joint-replacement patients who were randomized to either a 6-week preoperative cessation program that included weekly meetings and free nicotine replacement medication or to standard care.217A The overall complication rate was 18% in the smoking intervention group and 52% in controls (p = 0.0003). The wound-related complication rate was 5% in the intervention group and 31% in the controls (p = 0.001). There were no cardiovascular complications in the intervention group and 10% in the controls (p = 0.08).

Results of decision analytical modeling showed that it would be cost-effective to introduce a smoking cessation program even to patients undergoing surgery for lung cancer.218 It is likely that smoking cessation programs would pay for themselves in the savings from reduced postoperative complications, to say nothing of the long-term benefits of smoking cessation.

β-Blockers

Recommendations on the use of BBs continue to swing from enthusiasm to caution and back again. Over a decade ago, a few small studies in high-risk patients showed dramatic benefit from the use of perioperative BBs (Fig. 9-17). But there were only 83 serious adverse events in all the studies.219

These are efficacious medications, but like virtually all drugs, they have significant side effects. Recent studies have shown more adverse effects and probably less benefit than was previously appreciated. Effective use requires appreciation of new understanding in the mechanism and indications of BBs: which agents to use in which patients.

Understanding of the mechanisms of action of BBs continues to evolve. Reductions of heart rate and blood pressure are only two of the actions of BBs, and these may not always be beneficial.133 The pleiotropic effects such as vasodilatation, reduction in inflammation, and improvement of endothelial function are likely the more useful mechanisms, and these must be maximized while minimizing adverse mechanisms.

Titration of heart rate is important, if not for benefit, at least for prevention of adverse effects.220,221 Titration to a low heart rate is necessary for benefit after MI,222,223 whether this is necessary for benefit perioperatively, as shown by Raby et al37 and Poldermans et al,184 or for treatment of hypertension is coming under some debate.224 It is possible that in some patients, lower heart rate may increase central blood pressure.133

BBs probably reduce postoperative asymptomatic ischemia and resultant long-term morbidity. Although this may not affect immediate PCC,225,226 it reduces long-term PCC for up to 2 years.14 BBs also reduce the likelihood of fibrillation with ischemia. Some short-acting BBs augment anesthetic effects.227,228

Some contraindications to use of BBs have not turned out to be appropriate. Asthma and COPD are not contraindications to initiation of cardioselective β-blockade unless the patient is having an acute attack.229,230 In fact, the cardioselective BBs may actually improve pulmonary function.231 Patients with diabetes benefit from BBs, although other cardiac medications may be preferable.99

BBs are no longer recommended for all perioperative patients. Studies such as the POISE (Perioperative Ischemic Evaluation) trial have reminded us that they must be used selectively.220 They have little or no benefit in low-risk populations.184,232 In the latest ACC/AHA update on perioperative use, the only class I perioperative recommendation is for continuation in patients who are already receiving BBs for treatment of conditions that have ACC/AHA class I guideline indications for BBs.233 BB withdrawal has risks and must not be done in the perioperative period. Thus we need to review those indications, which have been or are in the areas of hypertension, CAD, CHF, and diabetes.

Hypertension

A few years ago, BBs were given as one of the first-line therapies for all hypertensive patients. When this indication was combined with the apparent perioperative benefits, it was an easy decision to start them preoperatively. However BBs have fallen to as low as fourth-line therapy in hypertension.133,234 As described in the previous section on hypertension, the treatment of large-artery stiffness, pulse wave reflection, and central artery pressure is more important than just treating peripheral artery pressure. And although BBs reduce peripheral blood pressure, they may change the timing of the reflected pulse wave, resulting in an increase, not a decrease in central pressures.132 Thus BBs may be largely used for treating blood pressure numbers that are not good indicators of cardiac risk and are not effectively treating important pathology. Hypertension is no longer a reason to start BBs preoperatively, with the following exception.

The latest ACC/AHA update recommends, at a IIa (reasonable) level, the use of perioperative BBs in all patients with CAD or two or more cardiac risk factors undergoing vascular or intermediate-risk surgery.221

Coronary Artery Disease

β-Blockers continue to be one of the first-line options for treatment of hypertension in patients with CAD.235 The 2007 ACC/AHA guidelines on treatment of chronic stable angina state that for hypertensive patients with well-established CAD (but no history of MI or acute coronary syndrome), it is useful to add blood pressure medication as tolerated, treating initially with BBs and/or ACEIs. This is a class I recommendation, albeit with level C (very limited) evidence. For those patients who have had an MI or acute coronary syndrome, irrespective of hypertension, it is a class I recommendation to start and continue BB therapy indefinitely. This is supported by level A evidence.235

There is evidence that risks may outweigh benefit if BBs are given to an unstable patient.236 This may be partly because newer cardiac medications and interventions that were not available during previous studies have reduced the additional benefit possible from BBs, whereas the adverse effects persist. Nevertheless, the benefits of BBs in patients with CAD237 are still unchallenged. Thus if such a patient presents to surgery without β-blockade, it should be started, albeit at a low dose for subsequent titration.

Congestive Heart Failure

Use of BB in CHF is another area of changes between enthusiasm and caution. At one time it was considered malpractice to use BBs in someone with CHF, but eventually it was realized they should be first-line agents for this condition.238 Recently, some dangers have come to light regarding beginning use of BBs in initially unstable heart failure patients and patients with diabetes, and this has resulted in more conservative approaches. But BBs continue to be strongly recommended for most patients with CHF.239 The ACC/AHA 2009 Guidelines for the Diagnosis and Management of Heart Failure in Adults makes a class I recommendation with level A evidence that bisoprolol, carvedilol, or sustained-release metoprolol succinate be used for all stable patients with current or prior symptoms of heart failure and reduced left ventricular ejection fraction.240 Contraindications are allergy, major side effects from BBs, or recent use of inotropes for stabilization.

Diabetes

Early concerns that BBs should not be used in patients with diabetes because they blocked counterregulatory hormones and masked symptoms changed to general recommendation for use because of the apparent cardioprotective effect in this high-risk group. Recently, however, observations that at least some BBs actually promote the development of diabetes241-243 as well as increase doubt about the true extent of the cardioprotective effect has led to more cautious use in these patients. Diabetes on its own is not a reason to start BBs preoperatively.

BBs are still useful agents, and we must use them appropriately. All the recent studies showing limited benefits from BB use older medications. Third-generation agents such as carvedilol244 and nebivolol,245 which have actions of vasodilatation, reduction in platelet activation, and endothelial function enhancement, as well as fewer side effects, may improve the safety profile.133 The DECREASE IV study showed significant reduction in PCCs in intermediate-risk patients at 30 days postoperatively with the use of titrated bisoprolol, a β1 selective agent246 (Fig. 9-18).

Not all individuals respond to BBs equally. Evidence is emerging on which genotypes respond best to β-blockade,247 although this is not yet a practical option for selecting patients. Most studies showing benefit from perioperative BBs start the medication many days before surgery. At the other end of the continuum, prolonged use after the perioperative period may result in diminishing benefit but mounting risk. Chronic BB use does not seem to provide the same benefit as that of acute perioperative use.248 There are no studies comparing the different timing of perioperative BB use. More research is clearly needed. The rule of thumb is the sooner the better, so as to allow titration to maximum effect without side effects. But whereas early institution is ideal, encountering this issue shortly before surgery does not justify doing nothing; it does warrant starting with a low dose and arranging follow-up. Above all, BBs must not be discontinued perioperatively.249,250

In summary, perhaps perioperative BBs have not turned out to be quite as wonderful as initially believed, but they are clearly beneficial. Like all medications, there is a risk-benefit ratio that requires some selectivity. BBs that are being used for appropriate indications should not be stopped perioperatively. Patients who present for intermediate- or high-risk surgery, and who have class I indications to be on BBs (such as a history of CAD or CHF), but are not on them, should be started. Initial dose should be low, with plans for follow-up and titration postoperatively if necessary. If the anesthesiologist is uncomfortable with this type of protocol, the hospital system may need to employ a nurse practitioner, physician assistant, or primary care physician who is knowledgeable in these medical practices. In complex or unstable cases, a cardiology consult may be appropriate. Class II recommendations can be addressed more electively after surgery.

Statins

Inhibitors of 3-hydroxy-3-methylglutaryl coenzyme A reductase (statins) have demonstrated efficacy in decreasing cardiovascular morbidity and overall mortality with chronic therapy and in the perioperative setting. Statins inhibit the rate-limiting step in cholesterol synthesis, decreasing serum cholesterol levels, improving high-density lipoprotein (HDL) to low-density lipoprotein (LDL) cholesterol ratios, and stabilizing vulnerable plaques. Large randomized placebo-controlled studies have shown that statin therapy is effective in the primary and secondary prevention of coronary heart disease and MI and decreases overall mortality in at-risk populations.251 The Heart Protection Study (2002) also demonstrated a decrease in ischemic stroke from statin use even in patients with normal LDL.252

The benefits of statin therapy cannot wholly be explained by, and in fact often appear to be independent of, the lipid-lowering effects of statin therapy. The Prove-It Timi22 (2005) data demonstrate that statin therapy has anti-inflammatory effects that are predictive of cardiac outcomes.156 It has been shown that statins can improve endothelial function in both coronary and peripheral arteries after just 24 hours of therapy.253,254 In a small study, 171 patients presenting with non–ST-segment elevation acute coronary syndrome subject were randomized to 2 doses of atorvastatin or placebo within the 12 hours before percutaneous coronary intervention, and then followed by atorvastatin and clopidogrel in both groups. There were significantly improved outcomes in the pre-procedure atorvastatin group255 (Fig. 9-19).

The pleiotropic effects of statin therapy are likely due to downregulation of membrane-bound g-proteins, such as Rho/Rho kinase (ROCK), and suppression of their subsequent pathways, decreasing inflammatory intermediates and increasing nitric oxide availability. Isoprenoids are cholesterol intermediates inhibited by statin therapy that are necessary to translocate and anchor the g-proteins to the cell membranes. It has been demonstrated that statin therapy inhibits ROCK activity in humans but that this effect is smaller when lipids are lowered to a similar level with a combination of a statin and ezetimibe, a cholesterol-absorbing agent.256,257

Although the data regarding perioperative statin use are less robust, they do indicate a protective effect. The strongest perioperative data reflect patients undergoing major vascular surgery. Durazzo et al (2004) conducted a small clinical trial in which 100 patients were randomized to receive atorvastatin 20 mg or placebo for 31 days (average) prior to undergoing vascular surgery. Cardiovascular events were 3-fold higher in the placebo group as compared with the statin group at 6 months258 (Fig. 9-20). The DECREASE III trial (2009) randomized 497 patients to fluvastatin extended-release (XL) 80 mg or placebo for 37 days (average) prior to major vascular surgery. Statin therapy resulted in a 43% relative reduction in MI (number needed to treat [NNT] =12), and a 53% relative reduction in death (NNT = 19).259 Meta-analysis performed in 2006 regarding use of statin therapy in major vascular surgery examined 7 studies, including one randomized controlled trial and one prospective cohort study, and demonstrated postoperative mortality was lower with statin therapy (1.7% vs 6.1%). MI was also reduced with statin therapy (2.9% vs 6.2%).260

In this same study, a meta-analysis of statin therapy in patients undergoing cardiac surgery examined 7 studies including one randomized controlled trial. Statin therapy was identified as being associated with decreased postoperative mortality (1.9% vs 3.1%). Postoperative MI, however, was increased with statin therapy (4.6% vs 3.6%).260 The likely explanation for this increase in postoperative MI is postoperative interruption of statin therapy resulting in a rebound overexpression of ROCK and other g-protein–mediated pathways. Le Manach et al analyzed 669 patients in a prospective cohort in which statin therapy was continued postoperatively (median delay of 1 day until statin continuation) or discontinued (median delay of 4 days until statin continuation). Continuation of statin therapy was associated with a decrease in MI (odds ratio [OR] = 0.38), whereas discontinuation of statin therapy was associated with an increase in MI (OR = 2.1) compared with controls.261

Lindenauer et al retrospectively studied 780,591 patients who underwent a variety of noncardiac surgical procedures, of which fewer than 10% were vascular cases. Of the studied patients, 77,082 received lipid-lowering agents, which included statins in 70,145 (91%) of these patients. Using propensity-matched groups and conditional logistic regression, they showed that treated patients had a risk of mortality with an adjusted OR of 0.62. The lowest quintile of risk did not benefit from lipid-lowering therapy.15

The ACC/AHA 2007 Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery makes the following recommendations for perioperative statin therapy.72

1. For patients currently taking statins and scheduled for noncardiac surgery, statins should be continued. (Class I Recommendation, Level of Evidence: B)

2. For patients undergoing vascular surgery with or without clinical risk factors, statin use is reasonable. (Class IIa Recommendation, Level of Evidence: B)

3. For patients with at least one clinical risk factor who are undergoing intermediate-risk procedures, statins may be considered. (Class IIb Recommendation, Level of Evidence C)

Use of extended-release statins during the perioperative period, as was done in the DECREASE III trial is another strategy to minimize disruption of therapy.

Angiotensin-Converting Enzyme Inhibitors/Angiotensin Receptor Blockers

The cardiovascular benefits of ACEIs have been demonstrated in studies such as the Anglo-Scandinavian Cardiac Outcomes Trial–Blood Pressure Lowering Arm (ASCOT-BPLA).242 The effects could not be attributed to blood pressure reduction alone224 and may be due to improvement of arterial function.134

Angiotensin receptor blockers (ARBs) have similar benefits.242,262 Although it would seem that these medications should be continued perioperatively, there have been reports for some time of intractable intraoperative hypertension attributed to ACEIs and ARBs. Some studies have not shown any relation between preoperative ACEIs and intraoperative hypotension,263 but most do.264,265 A retrospective study by Comfere et al of 267 hypertensive patients described increased intraoperative hypotension if ACEIs and ARBs are not stopped, but that it is usually easily managed.265 That study did not show an increase in the incidence of severe hypotension, although other studies have.266 Shahzamani et al observed a dose-related relation between hemodynamic stability and preoperative ACEIs given to CABG patients.267

Angiotensin is important for reversal of hypotension in volume-depleted states, especially during general or epidural anesthesia.264 Therefore it is advisable to stop ACEIs and ARBs 12 to 24 hours before surgery.

Calcium Channel Blockers

Although there is some evidence for perioperative benefit from use of calcium channel blockers,268,269 they are generally second-line agents that should be continued; routine perioperative addition is not indicated.270 The dihydropyridine calcium channel blockers tend to cause a reflex tachycardia and should not be used.271

Diuretics

Whether or not to discontinue diuretics should be determined on a case-by-case basis. The benefit of withholding them is a reduction in the dehydration induced by conservative preoperative fluid restrictions. The benefit of continuing them is in the control of heart failure,272 and they should be continued in patients whose optimization is tenuous.

Use of diuretics is another aspect of medicine that has undergone 3 swings of the pendulum. Initially one of the few medications available for treating hypertension and heart failure, they were replaced by what appeared to be better options. But recent reexamination of studies shows the thiazides to be at least as good as other options.273 Chlorthalidone is the thiazide most studied, although it is underused in the United States.

Clonidine

Clonidine is an old drug with several beneficial perioperative effects. In addition to possessing interesting analgesic properties, it appears to reduce preoperative ischemia36,274,275 (Fig. 9-21). Clonidine may be a good alternative for those patients for whom BBs are not indicated. The European guidelines classify the perioperative use of clonidine as a IIb recommendation for vascular surgery patients.270

Nitroglycerin

A small prospective study by Dodds et al did not show any benefit to use of preoperative nitroglycerin.276 A small case-control study by Sear et al suggested slightly increased risk.226 There are no randomized studies of postoperative benefit of preoperative nitroglycerin.

Antiplatelet and Anticoagulant Agents

Balancing antiplatelet cardioprotection in the procoagulant milieu of perioperative stress with the increased risk of surgical bleeding must be done on a case-by-case basis. Aspirin should be stopped 5 days before an operation, and clopidogrel should be stopped 5 to 10 days before.277 In vascular surgery cases, aspirin is generally continued throughout the perioperative period.277,278 In some cases, continuation of both aspirin and clopidogrel provides the best balance of risks.279 Anekstein et al showed that femur fracture repair on patients taking aspirin resulted in the transfusion of an average of 0.5 more units of blood.280 In many cases this is a small increase in risk compared with the benefit.

Patients with CAD should ideally never stop aspirin.277,281 There is a rebound phenomenon that results in about a 3-fold increase in the risk of PCC.282 A cohort analysis of patients presenting with acute coronary syndrome showed that about 20% of those who had ischemic events within 30 days had recently stopped aspirin.283 An 81-mg dose is as protective as higher doses.284

Oscarsson et al, in a small prospective randomized trial in 220 higher-risk patients, showed that preoperative aspirin use resulted in a 7.2 % absolute risk reduction in PCCs and no difference in bleeding complications. The study was not sufficiently powered to allow firm conclusions, but any further trials requiring a planned stoppage of aspirin in high-risk patients would likely be considered unethical.285 Stopping aspirin in patients with CAD must only be done if the risk is justifiable.

Patients with atrial fibrillation can usually have aspirin and warfarin stopped unless the CHADS2 score is high, about 4 or more (Table 9-6). Stroke or TIA within the past year generally warrants either continuation or bridging therapy.

Patients on anticoagulants because of valvular heart disease and prosthetic valves likewise require a balancing of the risk of bleeding from continued anticoagulation versus the increased risk of thromboembolism resulting from a cessation of therapy. Patients who are at high risk for thromboembolism should stop warfarin 5 days before surgery and institute bridging therapy with unfractionated or low molecular weight heparin. Patients who are at moderate risk should likewise stop warfarin 5 days preoperatively and institute bridging therapy, although the evidence for this is less robust. Patients at low risk, for example those with bileaflet mechanical valve and no other risk factors, do not need bridging therapy.277 The ACC/AHA guidelines suggest that for these patients warfarin can be stopped 48 to 72 hours before surgery, as long as the international normalization ratio (INR) is reduced to below 1.5.286 Because the evidence defining the perioperative management of anticoagulation in this patient population is so sparse, the specifics of cessation or bridging with heparin of a patient with any of the risk factors in Table 9-7 must be individualized.287 A rough guideline is that a patient with two risk factors should have therapeutic doses of intravenous heparin started when the INR has been allowed to fall below 2.0 (typically 48 hours before surgery). Heparin should be adjusted to maintain the activated partial thromboplastin time at 55 to 70 seconds, stopped 4 to 6 hours before the procedure, restarted as early after the operation as possible, and continued until the INR is therapeutic on warfarin.286 Although there have been some concerns that low-molecular-weight heparin for bridging in patients with mechanical valves does not provide enough protection,288 its use is becoming common with no reported problems.289 The European guidelines recommend twice-daily dosing at therapeutic rather than prophylactic levels.290 Ongoing studies may soon provide guidance in this area.

Antiplatelet Agents and Recent Coronary Stents

Generally, dual antiplatelet therapy must be continued for at least 6 weeks after bare metal stent placement291 and 12 months after placement of DESs. Heparin bridging is of unknown utility; it does not provide the level of antiplatelet activity that must be maintained. In fact, heparin might actually decrease platelet activity.292 Some initial evidence has shown that bridging with an intravenous glycoprotein (GP)IIb/IIIa receptor blocker such as tirofiban may be beneficial293,294 (Fig. 9-22).

Diabetes Medications

In the past, metformin was stopped up to 1 week preoperatively because of fear of metabolic acidosis and renal insult. A recent Cochrane analysis revealed no cases of fatal or nonfatal lactic acidosis in 70,490 patient-years of metformin use.295 Metformin is proving to be valuable in improving endothelial function.296,297 Stopping use of this drug preoperatively is probably counterproductive.

Ideally, no patient should proceed to an elective operation without stable glucose control. Although there is strong evidence that perioperative hyperglycemia increases risk of surgical wound infections in cardiac surgery298,299 and fairly good evidence that hyperglycemia increases overall risk of CAD,300 a recent Cochrane review concludes there is no evidence that tight perioperative glucose control is advantageous,301 and it may actually be detrimental.302

A study by Duncan et al of intraoperative glucose levels in 4302 cardiac patients shows a U-shaped curve of morbidity from glucose control; that is, morbidity decreases as glucose levels decrease until they reach near normal, at which point the morbidity becomes as high as with hyperglycemia303 (Fig. 9-23).

Antibiotics

Clinicians should refer to established guidelines for the use of preoperative antibiotics to prevent surgical infections304 and to prevent endocarditis.117 Effective prophylaxis against surgical infections requires that the antibiotic be administered within the hour before incision. Antibiotics for endocarditis prophylaxis are often overused and should be given only when indicated.

Nutritional Supplements

Many patients take nutritional supplements, herbal medications, and/or well-advertised placebos. Because these are not regulated, it is difficult to know what is really in them. Although most supplements are probably innocuous, some may have detrimental perioperative effects, and so patients are usually asked to stop all of them. However some supplements have beneficial effects or may at least have produced a new homeostasis of coagulation or endothelial function that should not be upset. For example, vitamin D,305 long-chain n-3 polyunsaturated fatty acids,306 and isoflavones163 have positive effects on endothelial function, and benefits of continuing them likely outweigh adverse effects. Recent updated guidelines from the American Society of Regional Anesthesia and Pain Medicine on regional anesthesia in patients receiving antithrombotic or thrombolytic therapy state that “The use of herbal medications does not create a level of risk that will interfere with the performance of neuraxial block. We recommend against mandatory discontinuation of these medications or avoidance of regional anesthetic techniques in patients in whom these medications have been administered (Grade 1C).”307 Time may prove that some supplements should be actively started in most patients.

The patient's perioperative plan can often be created on the basis of good interviewing and examination skills. Specialized tests should only be applied as needed to refine management decisions. Perioperative management now entails more than just identifying high-risk patients and referring them back to other specialists. It is possible for the anesthesiologist to actually reduce perioperative risk through relatively simple medical interventions. The extent and timing of interventions can be tailored to each patient according to specific perioperative risks. Anesthesiologists are well positioned to reduce perioperative risk efficiently and institute simple forms of secondary disease management.

• Boden WE, O'Rourke RA, Teo KK, et al. Impact of optimal medical therapy with or without percutaneous coronary intervention on long-term cardiovascular end points in patients with stable coronary artery disease (from the COURAGE trial). Am J Cardiol. 2009;104(1):1-4.
• Chung F, Yuan H, Yin L, Vairavanathan S, Wong DT. Elimination of preoperative testing in ambulatory surgery. Anesth Analg. 2009;108(2):467-475.
• Correll DJ, Hepner DL, Chang C, Tsen L, Hevelone ND, Bader AM. Preoperative electrocardiograms: patient factors predictive of abnormalities. Anesthesiology. 2009;110(6):1217-1222.
• Douketis JD, Berger PB, Dunn AS, et al. The perioperative management of antithrombotic therapy: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest. 2008;133(6 Suppl):299S-339S.
• Fleisher LA, Beckman JA, Brown KA, et al. 2009 ACCF/AHA focused update on perioperative beta blockade incorporated into the ACC/AHA 2007 guidelines on perioperative cardiovascular evaluation and care for noncardiac surgery: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation. 2009;120(21):e169-276.
• Oscarsson A, Gupta A, Fredrikson M, et al. To continue or discontinue aspirin in the perioperative period: a randomized, controlled clinical trial. Br J Anaesth. 2010;104(3):305-312.
• Schouten O, Boersma E, Hoeks SE, et al. Fluvastatin and perioperative events in patients undergoing vascular surgery. N Engl J Med. 2009;361:980-989.
• Schouten O, van Kuijk J-P, Flu W-J, et al. Long-term outcome of prophylactic coronary revascularization in cardiac high-risk patients undergoing major vascular surgery (from the randomized DECREASE-V Pilot Study). Am J Cardiol. 2009;103(7):897-901.
• Wijeysundera DN, Beattie WS, Austin PC, Hux JE, Laupacis A. Non-invasive cardiac stress testing before elective major non-cardiac surgery: population based cohort study. BMJ. 2010;340:b5526.