Table of Contents

Medical preoperative evaluation consists of evaluating the severity of medical problems in surgical patients, determining how best to manage these problems during the forthcoming surgical procedure, and recommending what needs to be done to stabilize the patient in the perioperative period. Preoperative risk assessment is a critical part of this process and one that is increasingly being performed in the outpatient setting. This chapter will briefly discuss general principles of preoperative evaluation and assessment of cardiac and pulmonary risk. Readers interested in a more in-depth investigation of these topics are encouraged to pursue the references listed at the end of the chapter.

Preoperative risk assessment starts by identifying the type of surgery to be performed and the “type” of patient who is to have it. It is these two factors which determine risk of complications - even a patient with severe coronary artery disease is at relatively low risk from cataract surgery, and even a patient without coronary artery disease is at relatively high risk from a pneumonectomy. The chance of complications from a specific type of surgery can be estimated from large prospective observational studies, and can be expressed as the pre-test probability of complications. The chance of a patient with specific risk factors having complications at surgery can also be estimated from clinical trials, and can be expressed as a likelihood ratio. As pretest odds x likelihood ratio = posttest odds, these data can be used to estimate surgical risk.

If the medical consultant is confident in the assessment of the risk from surgery and the risk conferred by the patient’s health, there is no need for further testing. (The absence of cardiac stress testing is not equivalent to the absence of data!) It is patients who are judged to be at “intermediate” risk, or those whose risk is unknown who should be referred for further evaluation. 

The first part of this equation, the risk to patients from specific types of surgery, has been extensively studied. As a rule, operative death is uncommon, occurring in about 0.3 percent of all operations. “Average risk” implies a perioperative mortality of one percent or less, “significant risk” implies one to ten percent and “high risk” ten to twenty percent.[i] While generalizations about risk are frequently made from clinical series, it is clearly important to examine the types of patients selected for study. For example, in series of consecutive unselected patients, mortality from a particular type of operation is considerably lower than in series of patients with known coronary artery disease, or in those undergoing surgery for conditions associated with coronary artery disease - such as peripheral vascular disease.In general, procedures associated with higher mortality and complication rates include major vascular surgery (femoral-popliteal bypass, aortic aneurysm repair), cardiac surgery, intraperitoneal surgery, intrathoracic surgery and craniotomy. Emergency surgery of any type carries a higher risk of mortality and postoperative complications (Table 1).

The second part of the risk equation is determined by the patient’s health. The American Society of Anesthesiologists (ASA) classification gives a global impression of the clinical state of the patient which correlates well with surgical outcome: I=normal,II =mild-moderate systemic disease, III =severe systemic disease, not life-threatening,IV =severe systemic disease that is life-threatening, V = expected to die in 24 hours. In addition to overall health, specific characteristics have been identified which may predispose patients to poor surgical outcomes. 

Cardiac risk is the most-feared and most-studied complication of surgery. Three well-known indexes of cardiac risk are routinely used when assessing the preoperative patient - those of Goldman et al;[ii] Detsky et al;[iii] and Larsen et al;[iv] which are shown in Table 2. The Goldman risk index, published in 1977, was based on a series of 1001 unselected patients over the age of 40 undergoing major noncardiac surgery. In multivariate analysis, nine preoperative factors were found to be associated with life-threatening cardiac complications and perioperative cardiac death: MI within 6 months, S3 gallop or jugular venous distension, age over 70, rhythm other than sinus on preoperative EKG, more than 5 PVCs/minute, important aortic stenosis, poor general medical status, emergency surgery and intraperitoneal, intrathoracic or aortic surgery. A point score system weighted the factors and allowed physicians to divide patients into four risk groups. Group 1, having 0 to 5 points, would have an estimated 0.7 percent of complications and 0.2 percent of death. Group 4, with 26 or more points, had estimated risks of 22 percent and 56 percent. The index has been validated in large prospective series of general surgery patients.[v] Detksy and Larsen conducted prospective studies in the 1980s which confirmed and refined the original Goldman index, adding categories of angina and prior history of congestive heart failure. Patient risk factors are reviewed in Table 3.

Goldman (continued) Detsky (continued) Larson (continued)
 

surgery type	intraperitoneal, intrathoracic or aortic	3	emergency surgery	10	emergency surgery	3
	emergency surgery	4			aortic surgery	5
					other intraperitoneal/ intrapleural surgery	3

CORONARY ARTERY DISEASE:

Concerns about perioperative ischemia and infarct have inspired studies of additional preoperative testing, such as echocardiography, exercise stress testing and dipyridamole thallium scintigraphy. “Low risk” patients are those with no known coronary artery disease, good cardiac functional status and low scores on the cardiac risk indices. In these patients, further testing adds little to preoperative assessment.[vi] Cardiac stress testing is most useful when the patient’s history is unclear, functional status is poor or there is a history of new or unstable chest pain. 

While some authors argue that all patients with known coronary artery disease should be evaluated with echocardiography and cardiac stress testing prior to surgery, there are no outcomes data to support this approach. Patients with stable coronary artery disease - including those who have had previous coronary artery bypass surgery - are not at increased risk, and further testing in the absence of symptoms is usually unnecessary in active patients.[vii] Studies have demonstrated that exercise tolerance is as good as exercise stress testing in predicting perioperative complications in patients with stable coronary artery disease.[viii] A prospective cohort study analyzed the use of preoperative transthoracic echocardiography in 339 men with known or suspected coronary heart disease undergoing major noncardiac surgery and found that the preoperative echocardiograms did not predict postoperative ischemia, infarction or death. Echocardiography was no better than clinical risk factors in predicting perioperative congestive heart failure and arrhythmia.[ix]

There may be subsets of patients with coronary artery disease who will benefit from further preoperative testing. Patients undergoing vascular surgery are of particular concern because of the risks of these procedures and because the presence of peripheral vascular disease is often associated with clinically significant coronary artery disease.[x] This population has been extensively studied and multiple series have shown that dipyridamole thallium scintigraphy has a high negative predictive value - i.e. that life-threatening cardiac complications are extremely rare in the presence of a normal thallium scan[xi] - although these trials are limited by the selected nature of their subjects. In a trial of 200 selected patients referred for thallium scintigraphy prior to vascular surgery, Eagle et al. found that for patients at intermediate risk, the combination of clinical and thallium information predicted perioperative complications better than either single approach.[xii] In contrast, Baron et al. studied 457 unselected consecutive patients undergoing abdominal aortic repair and found that thallium redistribution was not significantly associated with adverse perioperative outcome.[xiii] Perioperative beta-blockade for high-risk patients undergoing vascular surgery is now considered standard of care, and is discussed below.

Other patients who may benefit from further preoperative testing include those with multiple “low risk” variables, such as those identified by Eagle et al.⁷ and Vanzetto et al.[xiv] These include age over 70, history of angina, diabetes, Q waves on electrocardiogram, a history of ventricular ectopy (Eagle) as well as a history of myocardial infarction, ST-segment ischemic abnormalities during resting EKG, hypertension with severe LVH and a history of congestive heart failure (Vanzetti). Appendix A illustrates the suggested use of these low risk variables.

While there is no consensus about the extent of preoperative testing that is appropriate when assessing cardiac risk, certain general principles are evident. If a test is unlikely to effect posttest probability of poor outcome, it is an unnecessary component of preoperative risk assessment. Thus, for minimally risky procedures (hernia repair, breast biopsy, cataract surgery) or minimally risky patients (patients with good functional status and few or no risk factors from the Goldman, Detsky or Larsen indices) further investigation is unlikely to change patient management. An algorithm for perioperative evaluation for major noncardiac surgery is outlined in Figure 1. The algorithm is presented graphically in Appendix A.

If preoperative evaluation suggests that a patient is at high risk for cardiac complications from surgery, options include canceling the procedure or deferring it while medical or surgical therapy of the patient’s coronary artery disease is implemented. There are no randomized trials of preoperative therapy or prophylactic revascularization. 

Perioperative management of patients at higher risk has traditionally involved intensive care unit monitoring, pulmonary artery catheterization and intravenous nitroglycerin, none of which has been clearly demonstrated to improve surgical outcomes.[xv]^,[xvi] The American College of Cardiology/American Heart Association recommendations for these interventions are reviewed in reference 7. 

In contrast, perioperative beta-blockade has been very successful. A prospective, randomized, double-blind, placebo-controlled trial of perioperative beta blockers was published by Magnano et al. in 1996.[xvii] Two hundred patients with known or suspected coronary artery disease undergoing noncardiac surgery were randomized to atenolol or placebo for the duration of their hospital stay. Overall mortality from deaths due to cardiac causes was significantly lower in the treatment group at six months, twelve months and twenty-four months, as were combined cardiovascular outcomes. Event-free survival was significantly higher in the atenolol group. Polderman et al. subsequently randomized 112 high-risk patients undergoing vascular surgery to bisoprolol vs. standard care; the risk of death from cardiac causes or nonfatal myocardial infarction was dramatically reduced by beta-blockade (from 34 percent in the control group to 3.4 percent in the bisoprolol group).[xviii] We recommend that all high-risk patients not already on beta-blockers and without strong contraindication (such as bronchospasm) receive these agents perioperatively, beginning approximately two weeks prior to surgery and continuing for at least two weeks post-op. 

CONGESTIVE HEART FAILURE:

As demonstrated in the cardiac risk indices, clinically significant congestive heart failure is an important risk factor for perioperative mortality. Decreased cardiac functional status or evidence of pulmonary congestion, JVD or an S3 gallop on exam are all associated with increased risk of perioperative complications. Echocardiography, however, does not seem to add to the information gathered from history and physical exam when assessing preoperative risk.

Preoperative treatment of congestive heart failure should be relatively gradual, as dehydration is associated with intraoperative hypotension - itself a risk factor for perioperative complications. For patients with decompensated congestive heart failure, it is prudent to delay elective surgery for at least a week while gentle diuresis is attempted. 

HYPERTENSION:

Hypertension is a common finding among adults undergoing surgery and mild to moderately elevated blood pressure does not threaten surgical outcome.[xix] Patients with a diastolic blood pressure above 110 mg Hg, however, are at increased risk of serious perioperative complications. In such patients, elective surgery should be postponed for several weeks while hypertension is controlled and rapid drops in blood pressure should be avoided. Antihypertensive medications should be continued through the morning of surgery and consideration given to parenteral alternatives while the patient is NPO. 

VALVULAR DISEASE:

The surgical outcomes of patients with valvular heart disease are directly related to their cardiac functional status. Class I and II patients tolerate surgery well; perioperative morbidity and mortality is much higher for patients with limited functional status. Regurgitant lesions are better-tolerated than stenotic ones. The presence of hemodynamically significant aortic stenosis dramatically increases perioperative risk, and patients suspected of having AS should have preoperative echocardiography. The presence of severe or critical AS makes many types of surgery prohibitively dangerous; valve replacement should proceed major surgery in such cases. Similarly, symptomatic mitral stenosis is associated with a higher incidence of perioperative complications - including sudden death.

Endocarditis prophlyaxis is indicated for patients with prosthetic valves and valvular disease undergoing procedures likely to cause bacteremia; oral, gastrointestinal, respiratory and genitourinary surgery, as well as those involving incision and drainage of an infected site.

Pulmonary function is altered in patients undergoing surgery. Decreased functional residual capacity, vital capacity and cough contribute to aspiration, atelectasis and pneumonia, frequent causes of operative morbidity. Surgical procedures that carry higher risk of pulmonary complications include thoracic and upper abdominal surgery, procedures which require prolonged anesthesia (greater than two hours) and - obviously - lung resection. The presence of obstructive lung disease, a smoking history with productive cough, and hypercapnia are all patient characteristics associated with higher pulmonary risk. Epstein et al. have developed a cardiopulmonary risk index, that modifies the Goldman index to include pulmonary risk factors;[xx] it has been validated only among patients undergoing pulmonary resection.

In marked contrast to the literature on preoperative cardiovascular evaluation, there are relatively few prospective studies of preoperative pulmonary evaluation. As definitions of such complications vary, estimates of their frequency range from 9 to 76 percent.[xxi] Preoperative assessment of patients undergoing pulmonary resection is generally extensive, but studies of this population are not generalizable to other types of surgery. In general, authors agree that in patients with no history or symptoms of clinically significant lung disease and a normal lung exam, no further studies are required. Functional status correlates with pulmonary function,[xxii] and in active patients a history and physical exam are usually sufficient to estimate operative risk. Routine preoperative pulmonary function testing is not recommended,[xxiii] and a study of routine preoperative chest X-rays in adult patients admitted for vascular surgery found that they were of no help in improving patient outcomes.[xxiv]

For patients with known asthma or chronic obstructive pulmonary disease (COPD), the goal is to maximize respiratory function; to adjust medical regimens to bring patients to their “personal best.” While an FEV1 of less than 500 cc or an FVC of less than one liter are generally considered prohibitively dangerous findings, severity of disease or of PFT abnormalities do not strictly correlate to risk of postoperative complications.[xxv] Patients with uncharacterized lung disease may benefit from preoperative PFTs and a specific diagnosis, particularly before thoracic or upper abdominal surgery, although type and duration of surgery and the patient’s functional status remain the most important predictors of operative outcome.[xxvi]

Perioperative interventions have been shown to reduce the incidence of pulmonary complications. Patients who stop smoking two months prior to surgery have significantly fewer pulmonary complications than those who continue to smoke or stop less than eight weeks before admission.[xxvii] Incentive spirometry and chest physiotherapy have been shown to reduce pulmonary morbidity. Adequate analegesia and early mobilization are strongly recommended. Table 4 is adapted from Smetana’s recent review of preoperative pulmonary evaluation[xxviii] and summarizes risk reduction strategies. 

TABLE 4: Pulmonary risk reduction 
 

Preoperative: Øencourage smoking cessation for at least 8 weeks Øtreat airflow obstruction in patients with COPD or asthma Øadminister antibiotics and delay surgery if respiratory infection is present Øbegin patient education regarding post-op lung-expansion manuevers

Intraoperative: Ølimit duration of surgery to less than three hours Øuse spinal or epidural analgesia* Øavoid use of pancuronium Øuse laparascopic procedures when possible

Postoperative: Øuse deep-breathing exercises or incentive spirometry Øuse continuous positive airway pressure Øuse epidural analgesia* Øuse intercostal nerve blocks*

* often recommended but variable efficacy in literature

Although severe anemia and thrombocytopenia are associated with perioperative complications, the chance of these abnormalities being discovered in a healthy patient with no history of disease is extremely small. Mild anemia does not predict poor operative outcome and while it is traditional to recommend that patients be transfused for hematocrit less than 30 percent, this may be unnecessary for patients with chronic anemia. Hematocrit less than 24 percent was associated with increased morbidity in a 1988 study.[xxix] Similarly, while severe thrombocytopenia (less than 50,000) is associated with increased bleeding complications screening asymptomatic patients for platelet abnormalities is unlikely to be productive and routine preoperative coagulation profiles are not recommended. A history of bleeding diathesis, cirrhosis, hematologic malignancy or easy bruisablility should prompt assessment of platelet count and prothrombin time. 

Patients on antiplatelet medications (such as aspirin) or who are chronically anticoagulated fall into two categories. Those needing “tight control” - i.e. those with mechanical heart valves - can be placed on heparin preoperatively. Those in whom “loose control” is acceptable - patients on aspirin for CAD or warfarin for CVA prophylaxis - can discontinue anticoagulation a week prior to surgery and resume the medications on postoperative day one. NSAIDS should also be discontinued five to seven days before surgery.

Prophylaxis of deep venous thrombosis is particularly important after surgery. Patients undergoing pelvic or lower extremity surgery are at highest risk, particularly those having hip or knee replacement. For healthy patients under the age of 40 undergoing general surgery, early ambulation is sufficient. For older patients, elastic stockings and low-dose heparin (5,000 units SQ bid) are recommended. Orthopedic patients with hip fractures or undergoing hip replacement are prophylaxed with warfarin or low-molecular-weight heparin. 

It is important to consider every medication a patient is taking, and its implications for the perioperative period. Diabetics will require adjustment of insulin or oral hypoglycemics; type one diabetics should be followed by the inpatient medical consult team. Patients on chronic steroids will require stress-dose steroids. Patients on antihypertensive medications may require parenteral equivalents while NPO. Anti-ischemic regimens can be changed to transdermal or parenteral equivalents. A careful review with the patient should also include alcohol use and evaluation of the potential for alcohol withdrawal while hospitalized. 

Good communication is an essential feature of preoperative evaluation. Findings and recommendations should always be discussed with the referring surgeon, ideally in person. Notes should be brief, focused and specific. The goal of preoperative risk assessment is to determine if a patient is at average or increased risk for a specific procedure, or to recommend diagnostic testing if this determination cannot yet be made. As no patient is clear of risk, the phrase “medical clearance” is misleading and should not be used by a medical consultant. The patient should understand that medical consultation has been requested by their surgeon, that the two services are working together as a team to optimize their care and that the final decision on whether or not to operate will be made by the surgeon.