CHAPTER 70 PREOPERATIVE EVALUATION OF THE HIGH-RISK SURGICAL PATIENT

CHAPTER 70 PREOPERATIVE EVALUATION OF THE HIGH-RISK SURGICAL PATIENT BHIKEN I. NAIK DEANE MURFIN LISA THANNIKARY OVERVIEW The perioperative mortality rate for elective surgical procedures is low, ranging from 0.001% to 1.9% (1). This incidence increases, however, depending on the type of surgery, whether it is emergent in nature, the severity of concurrent disease, and the patient s age. Browner et al. (2), in a prospective cohort study of 474 men between 38 and 89 years of age at a Veterans Medical Center, reported a mortality rate of 5% during major noncardiac surgery. Multivariate analysis demonstrates that hypertension, limited functional capacity, and renal dysfunction are independently associated with increased perioperative mortality. Other studies have shown that both age and the American Society of Anesthesiologist (ASA) physical status classification systems are good predictors of perioperative complications (Table 70.1). Specifically, age greater than 70 years and ASA physical status greater than or equal to II are strong predictors of postoperative pulmonary complications (3). The ASA physical status classification remains the most widely used perioperative patient classification. Its simplicity is both its strength and weakness. Its strength is based on its ability to be applied to all age groups, medical conditions, and degrees of health. The weakness of the ASA classification system is its inability to distinguish among disorders of different systems and to cumulate risk based on multiple disorders. In an attempt to provide a multidimensional model of perioperative risk, Holt and Silverman (4) have devised an integrative model using various risk factors. In its simplest form, it provides a successive listing of the ASA physical status, surgical risk/invasiveness, physical factors affecting mask ventilation, intubation predictors, and a list of optional risk indicators. The acronym ASPIRIN is applied to this model. Although not validated in large studies, the ASPIRIN model provides an integrated framework for the assessment of the perioperative patient. The approach to the high-risk patient begins with preoperative identification, stratification, and modification of risk factors. This is achieved initially by the preoperative history and physical examination, which may be cursory in the event of a life-threatening emergency or more thorough if an elective procedure is planned. The data obtained from the history and physical examination allow for the application of Bayesian decision making that is, using preoperative testing based on clinical risk categorization. As a result, rational use of preoperative testing, particularly in this era of cost containment, can be achieved. Almanaseer et al. (5) demonstrated a 6.7% and 9.4% absolute reduction in stress thallium/echocardiogram and dobutamine echocardiogram testing, respectively, following the implementation of the American College of Cardiology/American Heart Association (ACC/AHA) guidelines for preoperative cardiac risk assessment. They also demonstrated a 19% increase in the use of beta-blockers following implementation of the ACC/AHA guidelines. Froehlich et al. (6) analyzed the impact of implementing the ACC/AHA guideline on resources utilization for aortic surgery. Initiation of the preoperative guideline reduced mean preoperative evaluation cost from $1,087 to $171, with no change in the incidence of myocardial infarction or death. Evidence-based preoperative evaluation allows for appropriate and cost-effective resource use, without increasing the risk of perioperative complications (6). CARDIOVASCULAR SYSTEM Noncardiac Surgery Of the approximately 44 million patients undergoing noncardiac surgery in the United States yearly, 30% either have, or are at risk for, coronary artery disease (CAD). The presence of CAD increases the incidence of perioperative myocardial ischemia, with a 2.8-fold increase in adverse postoperative cardiac events (7). Therefore, in an attempt to identify high-risk cardiac patients presenting for noncardiac surgery, both Goldman et al. (8) and Lee et al. (9) devised cardiac risk indices. Based on the points accrued during risk stratification, patients have either no testing performed or are referred for noninvasive testing or angiography. However, the predictive value of the cardiac risk index is poor in patients undergoing major vascular surgery. Vascular surgery patients represent a unique cohort, as the incidence of CAD in this population group is disproportionately higher than in the general population. Hertzer et al. (10), in a landmark study, evaluated 1,000 patients with coronary angiography prior to vascular surgery. The primary vascular diagnoses were abdominal aortic aneurysm, cerebrovascular disease, and lower extremity ischemia. Severe correctable CAD was demonstrated in 25% of the cohort whereas 6% of the study group demonstrated severe inoperable CAD; only 8% of the patients had no evidence of CAD. Furthermore, over the last decade, the management of the patient presenting with an ST-segment elevation myocardial infarction (STEMI) has evolved. Early aggressive reperfusion therapy and post-mi risk stratification are the current cornerstones of therapy. In an attempt to provide current evidence-based recommendations to manage the cardiac patient presenting for noncardiac surgery, the ACC/AHA Task Force on Practice Guidelines convened a panel of experts and published guidelines 1059

1060 Section VIII: The Surgical Patient TABLE 70.1 AMERICAN SOCIETY OF ANESTHESIOLOGIST PHYSICAL STATUS CLASSIFICATION Physical status I II III IV V VI E Definition Healthy patient Mild systemic disease; no functional limitation Severe systemic disease; definite functional limitation Severe systemic disease that is constant threat to life Moribund patient; unlikely to survive 24 hours with or without surgery Brain-dead patient; organ donor Emergency procedure on the perioperative cardiovascular evaluation for noncardiac surgery (11). The guideline was subsequently revised in March 2002; the update is available at the following Web site: www. acc.org/qualityandscience/clinical/statements.htm The important aspects of the guideline are to identify highrisk patients, appropriately stratify them according to their risk category, and perform preoperative testing in a rational and cost-effective manner. The guideline emphasizes that no test should be performed unless it is likely to influence patient treatment. The ACC/AHA guideline is an eight-step algorithm that incorporates clinical predictors based on the patient s history and physical examination, surgery-specific risk, and the functional capacity or exercise tolerance (Fig. 70.1). In the event of an emergency procedure, patients should be taken to surgery with risk stratification performed after the surgical procedure is completed. No preoperative testing is warranted under these circumstances. If an elective or urgent procedure is planned, the decision making proceeds down steps 2 and 3 of the algorithm. If the patient has had a coronary revascularization procedure, either coronary artery bypass grafting (CABG) or percutaneous coronary intervention (PCI) performed within the last 5 years with no recurrent symptoms or signs, then no further workup is necessary. A cardiac evaluation performed within the last 2 years with no deterioration in cardiac status also negates the need for further workup. Steps 5 to 8 of the algorithm integrate the clinical predictor, surgery-specific risk, and the functional capacity to determine whether the patient warrants further cardiac workup. The presence of major clinical predictors demands intense further workup and may result in delay or cancellation of elective surgery. Intermediate clinical predictors increase the risk of perioperative cardiovascular complications, whereas minor clinical predictors have not been proven to independently increase cardiac risk (Table 70.2). The nature and duration of the surgical procedure is a strong predictor of cardiovascular morbidity and mortality. Aortic, major vascular, and prolonged procedures associated with significant fluid shifts have a greater than 5% cardiac risk. Intermediate-risk procedures, which include intrathoracic, major orthopedic, intraperitoneal, head and neck, and prostate surgery, have a cardiac risk that is less than 5%. Endoscopic procedures, and cataract, breast, and superficial procedures are associated with minimal risk (less than 1%), and further workup is necessary only if the patient has major clinical predictors (Table 70.3). It is important to factor institutional and surgical expertise when evaluating the surgery-specific risk. Pronovost et al. (12) analyzed outcomes from abdominal aortic surgery in nonfederal acute care hospitals in Maryland. Mortality varied among hospitals from 0% to 66%, based on several factors including hospital and surgeon volume. Postoperatively, the absence of daily rounds by an ICU physician increased the risk of cardiac arrest, acute renal failure, sepsis, and reintubation (12). Finally, the functional capacity of the patient must be evaluated, as it is a strong predictor of perioperative outcome (13). Functional capacity is expressed in metabolic equivalents (METs), where one MET is 3.5 ml/kg/min of oxygen consumption in a 70 kg, 40-year-old man at rest. Increasing levels of activity correlate with increasing METs, with strenuous sports requiring greater than 10 METs, whereas activities of daily living require between 1 and 3 METs (Table 70.4). According to the ACC/AHA guidelines, the inability to perform at least 4 METs is associated with increased perioperative cardiac risk. Based on the aforementioned triad of clinical predictors, functional capacity, and surgery-specific risk, a decision is made whether the patient can proceed to surgery or whether additional investigations to delineate the ischemic burden are required (Table 70.5) (14). Delineation of the ischemic burden can be broadly achieved by two methods. The first method involves coronary vasodilatation and induction of a steal phenomenon by pharmacologic agents, followed by a nuclear imaging technique to determine the degree of myocardial ischemia. The second method involves increasing myocardial oxygen demand and evaluating electrocardiographic or echocardiographic data for evidence of ischemia. Myocardial oxygen demand can be increased either by exercise stress testing or pharmacologically with dobutamine or atropine. In light of their limited functional capacity, vascular patients can rarely complete exercise stress testing. Therefore, dipyridamole-thallium nuclear imaging or dobutamine stress echocardiography remains the mainstay of noninvasive testing for this cohort of patients. The negative-predictive value of both tests is high. However, the positive-predictive value of dobutamine stress echocardiography is higher (14). To increase the predictive value of nuclear imaging, several criteria have been proposed that help to differentiate the low-risk scan from the high-risk scan. These include the size of the defect, increased lung uptake, and the presence of left ventricular cavity dilation (15). Once the degree of myocardial ischemia is quantified, patients can either undergo perioperative medical optimization or revascularization by either percutaneous coronary intervention or surgery. It is important to note that to obtain benefit from a preoperative coronary intervention, the risk of noncardiac surgery must supersede the combined risk of both coronary catheterization and subsequent revascularization procedure. Eagle et al. (16) evaluated the Coronary Artery Surgery Study (CASS) database for patients requiring noncardiac surgery. CASS registry enrollees had coronary artery disease and were randomized to either optimal medical therapy or CABG. In patients undergoing high-risk surgery, prior CABG was associated with fewer postoperative deaths (1.7% versus 3.3%, p = 0.03) and MIs (0.8% versus 2.7%, p = 0.002) compared to medical management. In patients undergoing vascular surgery, the

Chapter 70: Preoperative Evaluation of the High-Risk Surgical Patient 1061 STEP 1 Need for noncardiac surgery Urgent or elective surgery Emergency surgery Operating room No Postoperative risk stratification and risk factor management STEP 2 Coronary revascularization within 5 yr? No Yes Yes Recurrent symptoms or signs? STEP 3 Recent coronary evaluation Yes Recent coronary angiogram or stress test? Favorable result and no change in symptoms Operating room No Clinical predictors Unfavorable result or change in symptoms STEP 5 STEP 4 Major clinical predictors** Intermediate clinical predictors Major or no clinical predictors Major Clinical Predictors** Consider delay or cancel noncardiac surgery Consider coronary angiography Go to step 6 Go to step 7 Unstable coronary syndromes Decompensated CHF Medical management and risk factor modification Subsequent care dictated by findings and treatment results Significant arrhythmias Severe valvular disease STEP 6 Clinical predictors Intermediate clinical predictors Intermediate Clinical Predictors Functional capacity Poor (<4 METs) Moderate or excellent (>4 METs) Mild angina pectoris Prior MI Compensated or prior CHF Surgical risk High surgical risk procedure Intermediate surgical risk procedure Low surgical risk procedure Diabetes mellitus Renal insufficiency STEP 8 Noninvasive testing Noninvasive testing Low risk Operating room Postoperative risk stratification and risk factor reduction High risk Invasive testing Consider coronary angiography Subsequent care dictated by findings and treatment results STEP 7 STEP 8 Clinical predictors Functional capacity Surgical risk Noninvasive testing High surgical risk procedure Noninvasive testing Poor (<4 METs) Minor or no clinical predictors Low risk Intermediate or low surgical risk procedure Moderate or excellent (>4 METs) Operating room Postoperative risk stratification and risk factor reduction Minor Clinical Predictors Advanced age Abnormal ECG Rhythm other than sinus Low functional capacity History of stroke Uncontrolled systemic hypertension High risk Invasive testing Consider coronary angiography Subsequent care dictated by findings and treatment results FIGURE 70.1. Stepwise approach to preoperative cardiac assessment. (Reproduced with permission from Eagle KA, Berger PB, Calkins H, et al. ACC/AHA Guideline Update for Perioperative Cardiovascular Evaluation for Noncardiac Surgery: a report of the American Heart Association/American College of Cardiology Task Force on Practice Guidelines (Committee to Update the 1996 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery). Circulation. 2002;105:1257 1267.

1062 Section VIII: The Surgical Patient TABLE 70.2 CLINICAL PREDICTORS OF INCREASED PERIOPERATIVE CARDIOVASCULAR RISK (MYOCARDIAL INFARCTION, CONGESTIVE HEART FAILURE, DEATH) MAJOR Unstable coronary syndromes Acute or recent MI a with evidence of important ischemic risk by clinical symptoms or noninvasive study Unstable or severe b angina (Canadian class III or IV) c Decompensated heart failure Significant arrhythmias High-grade atrioventricular block Symptomatic ventricular arrhythmias in the presence of underlying heart disease Supraventricular arrhythmias with uncontrolled ventricular rate Severe valvular disease INTERMEDIATE Mild angina pectoris (Canadian class I or II) Previous MI by history or pathologic Q waves Compensated or prior heart failure Diabetes mellitus (particularly insulin-dependent) Renal insufficiency MINOR Advanced age Abnormal ECG (left ventricular hypertrophy, left bundle-branch block, ST-T abnormalities) Rhythm other than sinus (e.g., atrial fibrillation) Low functional capacity (e.g., inability to climb one flight of stairs with a bag of groceries) History of stroke Uncontrolled systemic hypertension MI, myocardial infarction; ECG, electrocardiogram. a The American College of Cardiology National Database Library defines recent myocardial infarction as greater than 7 days but less than or equal to 1 month (30 days). b May include stable angina in patients who are unusually sedentary. c Campeau L. Grading of angina pectoris. Circulation. 1976;54:522 523. From Eagle KA, Berger PB, Calkins H, et al. ACC/AHA Guideline Update for Perioperative Cardiovascular Evaluation for Noncardiac Surgery: a report of the American Heart Association/American College of Cardiology Task Force on Practice Guidelines (Committee to Update the 1996 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery). Circulation. 2002;105:1257 1267, with permission mortality benefit was similar to the high-risk cohort; however, there was a 7.9% reduction in the perioperative MI rate. Therefore, among high-risk patients with multivessel CAD and evidence of significant myocardial ischemic burden, preoperative CABG confers a survival benefit and decreases the incidence of perioperative MI. With regard to the coronary intervention, there does not appear to be any difference in mortality or MIs in patients with multivessel disease randomized to either CABG or percutaneous coronary angioplasty (PTCA) presenting for noncardiac surgery (17). Currently, angioplasty is followed by placement of either a bare-metal or a drug-eluting stent; stents reduce both the acute risk of major complications and long-term restenosis rate. Following placement of a stent, patients require antiplatelet therapy to prevent in-stent thrombosis. Antiplatelet therapy is maintained for 1 to 12 months, depending on whether a baremetal or a drug-eluting stent is placed. The presence of antiplatelet therapy adds a new dimension of complexity to the patient presenting for noncardiac surgery following PCI. The risk benefit ratio of preventing thrombosis of the stent versus the risk of catastrophic perioperative bleeding must be carefully weighed. Kaluza et al. (18) reported 7 myocardial infarctions, 11 major bleeding episodes, and 8 deaths in 40 consecutive patients presenting for noncardiac surgery following placement of a stent. All deaths and MIs, as well as 8 of the 11 bleeding episodes, occurred within 2 weeks of coronary stent placement. Wilson et al. (19) reported a 4% incidence of death, MI, or stent thrombosis among 207 patients at the Mayo Clinic. Furthermore, they documented no adverse events in the 39 patients undergoing surgery 7 weeks after stent placement. It appears from these two important studies that the greatest risk of adverse cardiovascular events and bleeding complications occur within 2 weeks of stent placement. Elective surgery should be delayed for greater than 6 weeks to allow for endothelialization of the stent and discontinuation of antiplatelet therapy. In the event of urgent surgery and severe CAD, angioplasty alone with no stent placement can be performed. This obviates the need for prolonged antiplatelet therapy and the risk of perioperative bleeding. Perioperative β-blockade Therapy Of the pharmacologic agents that have been used during the perioperative period, β-blockade therapy remains the most studied. β-blockers have several salutary effects that decrease the

Chapter 70: Preoperative Evaluation of the High-Risk Surgical Patient 1063 TABLE 70.3 CARDIAC RISK a STRATIFICATION FOR NONCARDIAC SURGICAL PROCEDURES High (Reported cardiac risk often greater than 5%) Emergent major operations, particularly in the elderly Aortic and other major vascular surgery Peripheral vascular surgery Anticipated prolonged surgical procedure associated with large fluid shifts and/or blood loss Intermediate (Reported cardiac risk generally less than 5%) Carotid endarterectomy Head and neck surgery Intraperitoneal and intrathoracic surgery Orthopedic surgery Prostate surgery Low b (Reported cardiac risk generally less than 1%) Endoscopic procedures Superficial procedures Cataract surgery Breast surgery TABLE 70.4 ESTIMATED ENERGY REQUIREMENTS FOR VARIOUS ACTIVITIES 1 MET Can you take care of yourself? Eat, dress, or use the toilet? Walk indoors around the house? Walk a block or two on level ground at 2 to 3 mph (3.2 to 4.8 km/h)? 4 METs Do light work around the house like dusting and washing dishes? Climb a flight of stairs or walk up a hill? Walk on level ground at 4 mph (6.4 km/h)? Run a short distance? Do heavy work around the house like scrubbing floors or lifting or moving heavy furniture? Participate in moderate recreational activities like golf, bowling, dancing, doubles tennis, or throwing a baseball or football? >10 METs Participate in strenuous sports like swimming, singles tennis, football, basketball, or skiing? a Combined incidence of cardiac death and nonfatal myocardial infarction. b Do not generally require further preoperative cardiac testing. From Eagle KA, Berger PB, Calkins H, et al. ACC/AHA Guideline Update for Perioperative Cardiovascular Evaluation for Noncardiac Surgery: a report of the American Heart Association/American College of Cardiology Task Force on Practice Guidelines (Committee to Update the 1996 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery). Circulation. 2002;105:1257 1267, with permission MET, metabolic equivalent. Adapted from the Duke Activity Status Index and AHA Exercise Standards. From Eagle KA, Berger PB, Calkins H, et al. ACC/AHA Guideline Update for Perioperative Cardiovascular Evaluation for Noncardiac Surgery: a report of the American Heart Association/American College of Cardiology Task Force on Practice Guidelines (Committee to Update the 1996 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery). Circulation. 2002;105:1257 1267, with permission. risk for cardiovascular morbidity and mortality in a select cohort of patients. β-blockers help to correct the imbalance between myocardial oxygen demand and supply. They have additional plaque-stabilizing, antiarrhythmic, anti-inflammatory, and altered gene expression effects (20). Their beneficial effects on perioperative mortality has been assessed in two much-discussed studies (21,22). Although both studies have design flaws, they demonstrate both short-term and potentially long-term beneficial effects of perioperative β-blockade. However, two recent trials have failed to demonstrate a beneficial effect with β-blockade therapy. The Diabetic Postoperative Mortality and Morbidity (DIPOM) trial, involving 921 diabetic patients undergoing noncardiac surgery, did not demonstrate a significantly decreased risk of death and cardiac complications with metoprolol use (23). In the Metoprolol after Vascular Surgery (MaVS) trial, vascular patients scheduled for abdominal aortic aneurysm reconstruction or infrainguinal or extra-anatomic revascularization were randomized to either metoprolol or placebo 2 hours prior to surgery (24). The study drug was continued until hospital discharge or a maximum of 5 days. This trial demonstrated no difference in cardiac mortality, nonfatal MI, or new congestive heart failure between the two groups. Although the aforementioned trials demonstrated no benefit in certain cohorts of patients, is there potential harm in initiating perioperative β-blockade? Lindenauer et al. (25) con- ducted a retrospective cohort study of patients 18 years of age or older undergoing noncardiac surgery at 329 hospitals. Propensity score matching was used to adjust for differences between patients who received perioperative. β-blockade and those who did not receive such therapy. In-hospital mortality was compared using multivariable logistic modelling. The Revised Cardiac Risk Index (RCRI) score was used to assess the association between β-blocker therapy and the risk of inhospital death. In patients with a RCRI of 0 or 1, perioperative β-blockade was associated with no benefit and possible harm (RCRI 0: odds ratio [OR] 1.43; 95% confidence interval [CI] 1.29 1.58). The beneficial effects of β-blockade were seen only in patients with a RCRI of 2 or more (OR 0.9; 95% CI 0.75 1.08). In this study, β-blockade therapy appears to have a beneficial effect only in a select high-risk patient group. Currently, the best evaluation of the level of evidence for perioperative β-blockade is the ACC/AHA 2006 Guideline Update on Perioperative Cardiovascular Evaluation for Noncardiac Surgery: Focused Update on Perioperative Beta-Blocker Therapy (26). Based on the patient s cardiac risk and the nature of the surgery, three classes of recommendations are made from the current evidence. Insufficient data is available regarding the use of β-blockade therapy in low cardiac risk patients undergoing intermediate or high-risk surgery. In addition, the role of β-blockade therapy in low-risk surgery has not been defined.

1064 Section VIII: The Surgical Patient TABLE 70.5 SUMMARY OF AMERICAN COLLEGE OF CARDIOLOGY AND AMERICAN HEART ASSOCIATION GUIDELINES FOR CARDIAC EVALUATION BEFORE NONEMERGENT, NONCARDIAC SURGERY Clinical predictors Intermediate clinical predictors Minor clinical predictors Major clinical Poor functional Good functional Poor functional Good functional Risk of surgery predictors capacity capacity capacity capacity High Emergent major operations, particularly in the elderly Aortic and other major vascular surgery Peripheral vascular surgery Anticipated prolonged surgical procedures associated with large fluid shift and/or anticipated blood loss Intermediate Carotid endarterectomy Head and neck surgery Intraperitoneal and intrathoracic surgery Orthopedic surgery Prostate surgery Low Doscopic procedure Superficial procedures Cataract surgery Breast surgery Postpone or delay surgery Postpone or delay surgery Postpone or delay surgery Testing indicated Testing indicated Possible testing Testing indicated Testing not indicated Testing not indicated Testing indicated Testing not indicated Testing not indicated Testing not indicated Testing not indicated Testing not indicated From Akhtar S, Silverman DG. Assessment and management of patients with ischemic heart disease. Crit Care Med. 2004;32(Suppl):S126, with permission. In an attempt to clarify the role of β-blockade therapy in these groups of patients, two randomized controlled trials are in their recruitment phase. The PeriOperative ISchemic Evaluation trial (POISE) is designed to evaluate the efficacy of 30 days of controlled-release metoprolol to prevent major perioperative cardiovascular events in patients undergoing all types of noncardiac surgery (27); a total recruitment of 10,000 patients is planned. The DECREASE IV trial is designed to evaluate the efficacy of combination therapy with fluvastatin and bisoprolol in 6,000 patients scheduled to undergo noncardiac, nonvascular surgery (28). It is hoped that these trials with large study samples will clarify the role of β-blockade therapy in the low- and intermediate-risk patient. Cardiac Surgery Perioperative and long-term risk evaluation in cardiac surgery is complicated by several factors. These include procedural factors, patient factors, and data collection. Cardiac surgery, with its many confounding variables, requires large patient numbers for studies to be statistically relevant. Appropriate and meaningful data collection is a relatively recent phenomenon (29,30). However, this collection effort has been hampered by the reluctance to publish data on high-risk subgroups and the inclusion of data from low-output centers that are not part of a larger data collection network. Risk factors for cardiac surgery are identified by examining multiple databases and large case series. In one large database, 19 independent variables have been identified (31). However, there are no standardized definitions for risk thresholds; many of the assessments of statistical risk are based on odds ratios. In addition, multiple risk factors frequently coexist, making risk profiling for the individual patient difficult. Patients are presenting cumulatively with more risk factors; however, the impact of the individual risk factor appears to be decreasing. Data accrued over the last two decades suggest a steady improvement in cardiac surgical outcomes (32). This improvement is attributed to improving surgical technique, perioperative care, and patient selection. There is still, however, large variation in surgical technique across the spectrum of cardiac surgical procedures, which may explain the significant interunit variation in outcomes (32 34). Preoperative Evaluation Cardiac risk profiling begins with a thorough clinical examination and a review of the completed special investigations. Additional investigations will be guided by the presence and severity of other organ dysfunction. Cardiac risk evaluation can be performed by risk assessment tools. These tools are based on large databases, such as

Chapter 70: Preoperative Evaluation of the High-Risk Surgical Patient 1065 the EuroSCORE and the Society for Thoracic Surgeons (STS) database (29,31). The value of these databases is that standardized definitions are used to classify patients. The risk assessment tools have two important objectives: 1. Identifying independent risk factors for morbidity and mortality in valvular, coronary, and thoracic aortic surgery. 2. Risk prediction modelling through multivariate logistic regression analysis with a view to assessing individual patient risk, comparing and auditing individual units, and appropriate resource allocation. The STS database working group has recently published their analysis of independent risk factors in valvular surgery (31). Table 70.6 represents the information submitted by North American centers for 409,904 cardiac valvular procedures for the decade starting in 1994 and ending 2003. The risk factors are stratified according to procedural risk and patient-related factors. Statistical analysis techniques have been used to generate scoring systems. The Parsonnet score, developed in the late 1980s, predicts risk for CABG and valvular surgery based on an additive score of weighted risk factors (35). This score tends to overestimate mortality in modern clinical practice (32). In 1999, Nashef et al. (29) published an additive-weighted scoring system called EuroSCORE based on and validated using the EuroSCORE database. This system has an improved correlation with modern practice but still overestimates the mortality in higher-risk patients. The Parsonnet and EuroSCORE have modified versions applicable at the bedside. Recently, a score based on Bayesian modeling has been advocated by the United Kingdom s Society of Cardiothoracic Surgeons (32). This system uses only nine weighted variables and has the best correlation and receiver operator curve of all three systems tested. Scoring systems have a role in predicting both perioperative and long-term mortality and intensive care unit (ICU) resource use (30,32,36 38). Furthermore, they provide a framework to direct clinical examination and special investigations, thereby facilitating the process of identifying and modifying preoperative risk. Patients with a prohibitive perioperative risk can be better identified by these scoring systems. EuroSCORE is a compilation of risk factors as weighted by the Parsonnet, EuroSCORE, and Society of Cardiothoracic Surgeons databases. The percentages quoted in Table 70.7 are for individual risk factors in each section. Risk factors may be additive and/or synergistic. Preoperative Risk Modification Preoperative risk modification involves optimization of comorbidity and limiting cardiopulmonary bypass-related myocardial injury. Comorbidity. Cardiac patients have multiple comorbidities; the most common are renal dysfunction, diabetes mellitus, and congestive cardiac failure. Renal dysfunction and failure are significant risk factors in both valvular and CABG surgery (39). The severity of renal dysfunction preoperatively correlates well with mortality postoperatively. Kuitunen et al. (40) TABLE 70.6 PERIOPERATIVE RISK FACTORS FOR VALVULAR CARDIAC SURGERY Surgical factors Patient factors High risk Odds ratio Odds ratio Aortic root replacement 2.78 Isolated tricuspid replacement 2.26 4 Comorbidities 2 Emergent operation 2.11 Multiple valve replacement 2.06 Moderate risk Concurrent operation 1.58 Reoperation 1.61 Age 70 years 1.88 Valve replacement vs. repair 1.52 3 Comorbidities 1.68 Endocarditis 1.59 Low risk Coronary artery disease 1.58 Isolated mitral replacement 1.47 Year <1999 1.34 2 Comorbidities 1.41 Isolated pulmonic replacement 1.29 CHF 1.39 Isolated aortic replacement 1 Female gender 1.37 Ejection fraction <0.35 1.34 Average per comorbidity 1.19 Procedural risk is compared to the lowest-risk procedure Isolated Aortic valve replacement The odds ratio for perioperative mortality during valvular surgery is compared to the lowest risk valvular procedure (aortic valve replacement; odds ratio 1, mortality 5.6 %). CHF, congestive heart failure. From Rankin JS. Determinants of operative mortality in valvular heart surgery. J Thorac Cardiovasc Surg. 2006;131:547, with permission.

1066 Section VIII: The Surgical Patient TABLE 70.7 PERIOPERATIVE RISK FACTORS FOR CABG ACCORDING TO SCORING SYSTEM Parsonnet Score EuroSCORE UKSCTS complex Bayes HIGH RISK (17% 40% MORTALITY) Cardiogenic shock Postinfarct septal rupture Emergency surgery Acute renal failure Thoracic aortic surgery LV EF <30% Acute structural defect, e.g., VSD LV EF <30% Dialysis >80 y old ARF Creatinine >200 μmol/l (2.3 mg/dl) Preop IABP or inotropes Preop ventilation Active endocarditis Reoperation Age 75 y SIGNIFICANTLY ELEVATED RISK (9% 17% MORTALITY) 75 79 y old CABG plus major symptoms Age >75 y Third reoperation Emergency One or more previous operations Dialysis dependency Systolic PAP >60 mm Hg BSA <1.70 m 2 Rare circumstances Recent MI <90 days Recently failed intervention <24 h IVI nitrates on arrival in theatre PA pressure 60 mm Hg Serum creatinine >200 μmol/l (2.3mg/dL) AV pressure gradient 120 mm Hg Neurology affecting ADL Peripheral vascular disease Age 70 74 y ELEVATED RISK (3% 9% MORTALITY) MV surgery PAP <60 mm Hg LV EF 30% 50% Age 71 75 y AV surgery gradient <120 mm Hg Chronic pulmonary disease Left main stem disease Recently failed intervention >24 h Female Diabetes Aneurysmectomy Age 65 69 y Urgent surgery Second reoperation LV EF 30% 49% Age 70 74 y Hypertension LV EF <30% Hypertension BMI >35 IABP preoperative Female UKSCTS, United Kingdom Society of Cardiothoracic Surgeons; VSD, ventricular septal defect; LV EF, left ventricular ejection fraction; IABP, intra-aortic balloon counter pulsation; CABG, coronary artery bypass grafting; PAP, pulmonary artery pressure; BSA, body surface area; MI, myocardial infarction; PA, pulmonary artery; IVI, intravenous infusion; AV, atrioventricular; ADL, activities of daily living; MV, mitral valve; BMI, body mass index. reported that patients with an increase in plasma creatinine of one and half times from baseline with short periods of oliguria had a 90-day mortality of 8%. However, in the anuric patient with a threefold increase in creatinine, mortality increased to 32%. Patients with cardiogenic shock or an emergent indication for surgery often have acute renal failure; these patients have a high risk of death perioperatively. All scoring systems categorize their mortality above 40%, irrespective of the reason for the cardiogenic shock or the surgery required (30 32). These poor outcomes make it difficult to ascertain the beneficial effects that preoperative dialysis will confer. Patients presenting for coronary revascularization with dialysis-dependent end-stage kidney disease have been extensively studied (41,42). These patients appear to have a major survival benefit when coronary revascularization is performed under cardiopulmonary bypass. This survival advantage is lost when either off-pump CABG or percutaneous coronary intervention is performed (42). Survival rates at 8 years in dialysisdependent patients were 45.9% for CABG, 32.7% for PCI, and 29.7% for no surgical intervention. Patients with non dialysisdependent renal insufficiency have a significant incidence of postbypass renal failure (43). The STS CABG database suggests that they have higher perioperative mortality than patients with dialysis-dependent kidney failure (44). Prophylactic dialysis in nondialysis-dependent renal insufficiency may decrease the incidence of postbypass acute renal failure (45). In addition, these patients are fluid restricted and are often on diuretic therapy. Marathias et al. (46) reported that preoperative rehydration, with 1 ml/kg/hour of 0.45% saline, nearly halved the incidence of acute renal failure and decreased the need for postoperative dialysis. Diabetics undergoing cardiac surgery are at increased risk of prolonged ventilation, postoperative sepsis, renal failure, and cognitive dysfunction (47). The perioperative management of diabetes and hyperglycemia in cardiac surgery is controversial. Insulin has been used in two strategies: tight glycemic control and as part of glucose-insulin-potassium regimens (GIK). The

Chapter 70: Preoperative Evaluation of the High-Risk Surgical Patient 1067 studies evaluating these strategies have concentrated on the intraoperative and postoperative periods. The implementation of tight glycemic control in the postoperative setting improves mortality significantly: 8.0% versus 4.6% for tight glycemic control (48). This study, by Van den Berghe et al., was performed on predominantly postoperative cardiac patients and revealed significant reductions in length of stay, renal failure, and nosocomial sepsis. Intraoperative insulin infusion has also been shown to reduce postoperative complications in diabetic CABG patients (49,50). The use of a GIK infusion in the setting of ongoing myocardial ischemia and infarction results in significant improvements in myocardial preservation and contractile function (51). The technique has been applied to both coronary and valvular surgery with mixed results. The trials showing modest benefit initiated GIK preoperatively, used high doses of insulin, and continued the infusion through cardiopulmonary bypass and reperfusion (51). Meta-analysis of GIK therapy indicates that trials using tight glycemic control gave the best results. This observation needs validation by other randomized trials. Clinical experience indicates that ongoing myocardial ischemia and poor diabetic control frequently occur preoperatively. Evidence from the intraoperative and postoperative periods suggests that preoperative initiation of GIK, combined with tight glycemic control, may significantly decrease postoperative complications. Congestive cardiac failure represents a complex neurohumoral syndrome that develops in response to altered cardiac function. The stages of this condition have been classified by the ACC/AHA (Table 70.8) (52). In the perioperative period, decompensated stage C or D heart failure represents an independent risk factor for cardiac complications (31). The syndrome covers a spectrum of patients: from those with cardiogenic shock and an ejection fraction less than 30% to those with stable but inotrope-dependent cardiac function. These patients are at high risk for postoperative complications (29 32,35). Decompensated cardiac failure in valvular or coronary heart disease may represent a progression of the primary disease process. Under these circumstances, surgery offers the only chance to improve the biomechanical cardiac dysfunction and attenuate the maladaptive myocardial response. Mortality in medically managed decompensated cardiac failure can be improved by using new classes of drugs such as β-type natriuretic peptide and the calcium channel sensitizers (53 56). These agents have an inotrope-sparing effect, shorten hospital stay, and improve TABLE 70.9 INDICATIONS FOR INTRA-AORTIC BALLOON COUNTERPULSATION Ongoing unstable angina refractory to medical therapy Acute myocardial ischemia/infarction associated with percutaneous transluminal angioplasty (PTCA) Perioperative low cardiac output syndrome Cardiogenic shock after myocardial infarction Congestive heart failure Bridge to transplant Ischemic ventricular septal defect Acute mitral valve insufficiency Poorly controlled perioperative ventricular arrhythmias medium-term survival. The role of these agents in the perioperative setting has not been assessed, although small studies show promising results. The role of nonsurgical therapy is in the long-term prevention of progression of structural heart disease, prevention of remodeling, and modification of underlying risk factors (57). Myocardial Preservation Interventions. Several nonpharmacologic and pharmacologic interventions can be initiated preoperatively to improve intraoperative and postoperative outcomes. Intra-aortic balloon counter pulsation (IABP) is a nonpharmacologic intervention that can be commenced preoperatively in the appropriate group of patients (Table 70.9). Pharmacologic therapies include preoperative statins, β-type natriuretic peptide, calcium channel sensitizers, and antioxidant therapy. Intra-aortic balloon counterpulsation (IABP). The optimal use of the IABP in cardiac surgery is controversial. Preoperative IABP may offer improved myocardial perfusion and stability during induction and maintenance of anesthesia, prior to cardiopulmonary bypass. Good evidence exists for IABP in CABG patients with ischemia or with an ejection fraction less than 25% undergoing nonelective operation or reoperation, or who have New York Heart Association (NYHA) class III to IV symptoms. The evidence is less clear for patients without ongoing ischemia but who are undergoing reoperation, have isolated left main disease or a low ejection fraction, or who are undergoing procedures other than an isolated CABG (58). The efficacy of IABP for valvular surgery is poor and is associated with a twofold increase in mortality regardless of timing of use. TABLE 70.8 ACC/AHA CLASSIFICATION OF HEART FAILURE Stage A Stage B Stage C Stage D Patients at high risk of developing heart failure (HF) because of the presence of conditions that are strongly associated with the development of HF. Such patients have no identified structural or functional abnormalities of the pericardium, myocardium, or cardiac valves and have never shown signs or symptoms of HF Patients who have developed structural heart disease that is strongly associated with the development of HF but who have never shown signs or symptoms of HF Patients who have current or prior symptoms of HF associated with underlying structural heart disease Patients with advanced structural heart disease and marked symptoms of HF at rest despite maximal medical therapy and who require specialized interventions

1068 Section VIII: The Surgical Patient This probably reflects the fact that the ventricular dysfunction is either nonreversible or only partially reversible (59). IABP improves cardiac output by approximately 20% if set to maximal efficiency. However, the insertion and use of an IABP is not a benign procedure. Limb ischemia occurs in 8% to 42% of patients, and 30% of those will require a surgical intervention. Limb ischemia is more common when a Seldinger technique is used for insertion as opposed to surgical exposure of the femoral vessels and placement under direct vision (59). In conclusion, the best 5-year results for IABP are in patients undergoing isolated CABG (51%), whereas those undergoing CABG with aortic valve replacement have the lowest actuarial survival (34%). Further research is required to better define the role of this intervention in cardiac surgery (59). HMG CoA reductase inhibitors. HMG CoA reductase inhibitors, commonly known as statins, have several beneficial effects on arteriosclerosis and vascular graft disease. Their mechanism of action is via a lipid-dependent and a lipid-independent pathway. Inhibition of atherogenesis, thrombosis, and inflammation and maintenance of endothelial integrity are all attributed to this class of drug (60,61). The efficacy of statins in the reduction of graft stenosis and progression of atheroma in native vessels post-cabg is well documented (62). However, in many of these trials, statin therapy was initiated after surgery. The early beneficial effects of statins are on endothelium recovery and in inflammation in coronary vessels (63). In a large prospective longitudinal study, Collard et al. (64) evaluated the effect of preoperative statin therapy on cardiac mortality following CABG. Preoperative statin therapy was associated with a 1.1% absolute reduction in mortality (OR 0.25; CI 0.07 0.87). Interestingly, cessation of statin therapy after surgery was associated with an increased in-hospital and late cardiac mortality. This suggests that preoperative statin therapy must be considered prior to CABG and may become a standard of care in the future. Brain natriuretic peptide. Nesiritide is a recombinant form of brain natriuretic peptide that decreases pulmonary artery pressures and myocardial oxygen consumption, and increases coronary blood flow and urine output. Nesiritide is used in two clinical settings: inotrope-resistant cardiac failure and postcardiac surgery patients with high pulmonary pressures and low cardiac output syndrome. Salzberg et al. (53) published a case series of 14 patients with severe mitral regurgitation and pulmonary pressures above 60 mm Hg undergoing cardiac surgery. Their predicted mortality based on EuroSCORE was 26%. These patients received a nesiritide infusion preoperatively, with the goal of reducing pulmonary pressures by 25%. The infusion was discontinued intraoperatively and restarted on return to the ICU. There was no reported mortality among the patients receiving nesiritide (53). These results need to be confirmed in a properly powered study, but the evidence for preoperative use in heart failure patients with pulmonary hypertension is promising. Calcium sensitizers. Levosimendan is a calcium-sensitizing inodilator that improves myocardial contractility without increasing oxygen demand. It also decreases pulmonary vascular resistance in patients with heart failure. The LIDO trial showed it to be more effective than dobutamine in the management of severe congestive heart failure insofar as hemodynamic and mortality benefit (54). These findings have been verified by other large double-blind randomized trials (55). Experience with levosimendan in cardiac surgery is currently limited to small studies. These studies looked at the physiologic effects of levosimendan in patients with good left ventricular function, poor ventricular function, or acute myocardial ischemia with hemodynamic compromise. In all three groups, low-dose levosimendan infusions improved cardiac output and decreased systemic vascular resistance, myocardial oxygen demand, and inotropic requirements coming off bypass (54). This drug offers enormous promise in the preoperative period in patients with severe congestive cardiac failure and poor cardiac output. Antioxidants. Reactive oxygen species (ROS), both within myocardial cells and those derived from the systemic circulation, are thought to overwhelm local endogenous antioxidant systems during bypass. They initiate cellular damage, necrosis, and apoptosis during cardiopulmonary bypass and reperfusion. There have been many attempts to provide external sources of antioxidants or to improve endogenous antioxidant systems, and these approaches are supported by a large body of animal studies (65). Allopurinol, which inhibits xanthine oxidase, a significant source of ROS outside the myocardium, has been studied in ten human CABG trials. Eight of these trials showed improved hemodynamic markers and less cardiac enzyme release (65). However, despite these encouraging data, allopurinol has not received widespread support. Superoxide dismutase, desferrioxamine, mannitol, vitamins C and E, and N-acetylcysteine are additional antioxidants with encouraging results in small human trials. They demonstrate decreased surrogate markers of tissue damage, although no outcome improvements have been shown. More research is required to define the role of antioxidant therapy in cardiac surgery. There has been an improvement in the ability to identify and categorize the high-risk cardiac patient presenting for cardiac surgery. As more data are accrued, risk profiling is becoming more accurate. This will allow for cost-effective implementation of promising preoperative interventions in the appropriate patient. PULMONARY SYSTEM The risk of postoperative pulmonary complications varies widely and according to the definitions applied. The risk evaluation process also differs between cardiothoracic and noncardiothoracic surgery. In elective noncardiothoracic surgery, postoperative pulmonary complications vary between 1.7% and 2.6%, whereas in valvular heart surgery, the Society of Thoracic Surgeons database reports a pulmonary complication rate of 8.9% (66 68). The definition of pulmonary complications in these studies included respiratory failure, atelectasis, pneumonia, and pulmonary edema. Pulmonary thromboembolic disease has been specifically excluded in these studies. The contribution of postoperative pulmonary complications to morbidity, mortality, and length of stay is not dissimilar to the cardiac complication profile (66,67,69 71). However, in the subgroup of patients older than 70 years of age, pulmonary complication is a better predictor of long-term mortality than cardiac risk factors (72). Predicting the likelihood of

Chapter 70: Preoperative Evaluation of the High-Risk Surgical Patient 1069 postoperative pulmonary complications requires preoperative pulmonary risk stratification. Smetana et al. (3) in a systematic review identified and categorized preoperative risk factors that predicted postoperative pulmonary complications following noncardiothoracic surgery. Preoperative Evaluation The initial evaluation for risk factors requires a thorough history and clinical examination. Following the history and clinical examination, patients can be classified into two groups: those with known pulmonary disease and those with suspected pulmonary disease. Both groups require an assessment of their functional classification and the degree of pulmonary reversibility. Following the history and physical examination, laboratory and special investigations are guided by the database established from the clinical evaluation. Laboratory investigations with a good predictive value for postoperative pulmonary complications include blood urea nitrogen greater than 7.5 mmol/l (21 mg/dl) and creatinine level greater than 133 μmol/l (1.5mg/dL) (73 75). However, the most powerful predictor of pulmonary outcome is a serum albumin level. Albumin less than 3.0 g/dl correlates with an increased 30-day perioperative morbidity and mortality (76). The utility and cost effectiveness of routine preoperative chest radiography has been extensively debated. An abnormal chest radiograph does predict postoperative complications; however, only 4.9% of radiographs in patients younger than 50 years of age will be abnormal. Among routine preoperative chest radiographs ordered, only 0.1% to 3% will alter management (77,78). A focused history and physical examination should identify the patient who is likely to have an abnormal preoperative chest radiograph; this is supported by a recent practice guideline issued by the American College of Physicians suggesting that (3): 1. Only patients with known cardiopulmonary disease should have a routine preoperative chest radiograph. 2. Patients older than 50 years undergoing procedures with high pulmonary risk should have a preoperative chest radiograph. These procedures include aortic surgery (thoracic or abdominal), neurosurgery, abdominal surgery, and prolonged surgery. Spirometry has been evaluated as a predictive tool for pulmonary disorders in noncardiothoracic surgery. There are, unfortunately, no studies to guide spirometry evaluation in the perioperative period for restrictive pulmonary disorders. In obstructive pulmonary disorders, there are conflicting data on the utility of spirometry; however, it may identify patients at higher risk for postoperative pulmonary complications (3). No threshold or prohibitive value has been defined for spirometry indices in obstructive pulmonary disease, probably related to the evidence that long-term prognosis for chronic obstructive pulmonary disease (COPD) is better predicted by the BODE severity scoring system (79). This system uses a holistic approach assessing the body mass index, degree of airway obstruction, symptom scoring, and exercise testing to assess severity and predict outcome. In lung resection surgery, however, spirometry forms the cornerstone of the evaluation process in both Europe and North America (80,81). Here, the perioperative risk of morbidity and mortality is directly related to a three-legged physiologic testing algorithm. Spirometry data are an integral part of the respiratory mechanics evaluation process. Other parameters evaluated are the cardiopulmonary reserve and the lung parenchymal function (Fig. 70.2). A simplified algorithm integrating these parameters assists in the preoperative workup for lung resection surgery (Fig. 70.3). A forced expiratory volume (FEV 1 ) greater than 80% of predicted or greater than 2 L allows pneumonectomy without further investigation. An FEV 1 greater than 1.5 liters allows lobectomy without further investigation. If any of these criteria are not met, then a predicted postoperative FEV 1 (ppofev 1 ) and carbon monoxide diffusion capacity (ppodlco) need to be calculated. This allows for further risk stratification: 1. Patients with ppofev 1 and ppodlco greater than 40% can be resected. 2. If the ppofev 1 or ppodlco is less than 40% and >30%, then a VO 2max needs to be assessed. If the VO 2max is greater than 15 ml/kg/min, then resection can continue. 3. If the VO 2max is less than 15 ml/kg/min, then the risk is prohibitive unless V:Q scanning indicates that the lung pathology is predominantly involving the area to be resected. 4. A VO 2max less than 10 ml/kg/min or ppofev 1 less than 30%, and ppodlco less than 30% are all prohibitive risks (79,80). When the information gathered from history, physical examination, and special investigations is examined, a risk profile can be constructed from the guidelines published by the American College of Physicians (Table 70.10) (3,82). Important points that need to be highlighted from Table 70.10 are as follows: Lung Resection Evaluation Respiratory Mechanics FEV 1 (ppo >40%) MVV, RV/TLC, FVC Cardiopulmonary Reserve VO 2 max >15 ml/kg/min Stair climb >2 flights, 6-min walk test, Exercise SpO 2 drops <4% Lung Parenchymal Function DLCO (ppo >40%) Blood gas: PaO 2 >60 PaCO 2 <45 FIGURE 70.2. The best validated test is shown in the first box. Alternative tests are shown below. DLCO, total diffusing capacity for carbon monoxide; FEV1, forced expiratory volume at 1 second; FVC, forced vital capacity; MVV, maximal voluntary ventilation; PaO 2, arterial partial pressure of oxygen; PaCO 2, arterial partial pressure of carbon dioxide; ppo, predicted postoperative value based on the number of lung segments remaining after resection; RV/TLC, residual volume divided by total lung capacity; SpO 2, pulse oximetric oxygen saturation; VO 2, oxygen uptake/consumption.