From Surgery, (Ed:Norton) CH.17
Preoperative Cardiovascular Assessment
More than 3 million patients with coronary artery disease (CAD) undergo surgery each year in the United States.
Among them, 50,000 patients sustain a perioperative myocardial infarction (MI). The incidence may be increasing
because of an aging population. Overall mortality for perioperative MI remains nearly 40%. Aortic and peripheral
vascular surgery, orthopedic surgery, and major intrathoracic and intraperitoneal procedures are more frequently
associated with perioperative cardiac mortality than are other types of surgery. Absent a history of heart disease, men
are at increased risk above 35 years of age, whereas women are at increased risk after age 40. Cardiac mortality risk
increases markedly in patients over age 70. Cigarette smoking also confers increased risk.
Crucial to the task of risk-benefit analysis is the prospective identification of the patient at risk for a perioperative
cardiac complication. Unfortunately, although the presence of CAD is not difficult to demonstrate by screening
techniques, there is little evidence that prophylactic coronary revascularization, whether by open surgery or
angioplasty, can reduce risk before noncardiac surgery. Routine noninvasive testing is expensive, and clinical criteria
may be nearly as good for the identification of patients at high risk. Until recently, it has been unclear whether medical
management in preparation for surgery accomplishes much unless the patient has decompensated disease (e.g.,
congestive heart failure, recent MI),34 but new evidence indicates that perioperative β-blockade can reduce
cardiovascular mortality even when started immediately preoperatively. 35,36
In addition to the presence of CAD, the perioperative history and physical examination must ascertain the presence of
valvular heart disease (particularly asymptomatic aortic stenosis [AS]), congestive heart failure (CHF), or arrhythmias.
Congestive heart failure is strongly predictive of perioperative pulmonary edema and other complications. A
prospective study of 254 predominantly hypertensive diabetic patients who underwent elective general surgery
operations revealed a 17% incidence of perioperative CHF among patients with cardiac disease (previous MI,
valvular disease, or CHF).37 Patients with both diabetes and heart disease were at especially high risk. In contrast,
CHF developed in fewer than 1% of patients without prior cardiac disease.
Severe AS (defined as a pressure gradient >50mmHg) must be detected preoperatively because the risk of
perioperative mortality has been estimated at 13%. The increased mortality results from a limited capacity to increase
cardiac output in response to stress, vasodilation, or hypovolemia. Patients with AS tolerate poorly the development
of hypovolemia, tachycardia, or new-onset atrial fibrillation. Moreover, left ventricular hypertrophy decreases
ventricular compliance and leads to decreased diastolic filling. Elective aortic valve replacement before noncardiac
surgery may be indicated in severe AS, even in the absence of symptoms. Patients with less-critical AS require
invasive hemodynamic monitoring in the perioperative period and caution with the use of afterload-reducing agents.
Aortic and mitral insufficiency subject the left ventricle to high-volume loads that may impair contractility, but the risk
is comparatively small compared to that conferred by AS. Occult ventricular dysfunction may be present in the
asymptomatic patient, and therefore close monitoring is required, but patients can be expected to tolerate surgery
well if they are not in CHF. Patients with mitral stenosis or hypertrophic cardiomyopathy are at intermediate risk of
perioperative pulmonary edema, especially with tachycardia and decreased left atrial emptying. Perioperative fluid
shifts of little consequence to the healthy patient may wreak havoc in the setting of mitral stenosis. Hypovolemia and
a resultant low-flow state may occur despite relatively high pulmonary vascular pressures, but overzealous volume or
blood administration may cause pulmonary edema rapidly.
Atypical or unstable chest pain requires careful evaluation. Stable chest pain does not increase perioperative risk, but
unstable disease (e.g., new-onset or crescendo angina, a recent MI, or recent or current CHF) certainly warrants both
evaluation and stabilization. The preoperative evaluation of a patient with angina should determine whether the
patient's disease and symptoms are truly stable. If so, surgery may proceed with the maintenance of an effective
antianginal regimen during and after operation. Similarly, asymptomatic or only minimally symptomatic patients who
have previously undergone coronary bypass grafting tolerate surgery well.
TABLE 17.1. Risk Stratification Parameters and Criteria for Cardiac Events Following Noncardiac Surgery.
View Large
A recent MI is the single most important risk factor for perioperative infarction (Table 17.1). The risk is greatest within
the early aftermath following an infarction, probably the first 30 days. Estimates of the risk of anesthesia following an
MI range as high as a 27% reinfarction rate within 3 months, 11% between 3 and 6 months, and 5% after a 6-month
interval. Patients who suffer nontransmural (non-Q-wave) infarctions appear to be at identical risk. However, cardiac
risk management strategies may be succeeding. With intraoperative hemodynamic monitoring, the risk may be
reduced to as low as 6% within 3 months of the first MI and only 2% incidence within 3 to 6 months. Elective surgery
should be postponed for 6 months following an acute MI. When major emergency surgery is necessary, it should be
performed with intraoperative hemodynamic monitoring. When operation is urgent, as for a potentially resectable
malignant tumor, it can be undertaken from 4 to 6 weeks after infarction if the patient has had an uncomplicated
recent course and the results of noninvasive stress testing are favorable.
The Cardiac Risk Index System (CRIS) is an accepted system that was developed from a cohort of patients aged 40
years or more who underwent noncardiac surgery.38 Risk classes (I-IV) are assigned on the basis of accumulated
points (Table 17.2). According to CRIS, any elective operation is contraindicated if the patient falls within class IV.
One benefit of CRIS is that more than one-half of the total points are potentially controllable (e.g., treating CHF
reduces the score by 11 points, delaying surgery for a recent MI decreases it by 10 points), thereby reducing risk.
Further study of CAD serves primarily to quantify risk in patients with identified risk factors. Whether patients with no
cardiac risk factors should undergo additional preoperative testing is still debated. Algorithms from the American
College of Cardiology/American Heart Association (ACC/AHA) Task Force on Practice Guidelines can be used to
guide the evaluation (Table 17.3; Figs. 17.1-17.3).39
TABLE 17.2. Cardiac Risk Index System (CRIS).
View Large
The routine resting electrocardiogram (ECG) remains the primary screening modality for virtually all patients over age
40 who are to undergo general anesthesia. It is undeniably cost-effective but may be normal in many patients with
CAD. However, evidence of a prior MI (Q-wave 0.04s or wider and at least one-third the height of the R-wave) is
nearly indisputable evidence of CAD. A wide array of other tests have been employed for the preoperative
assessment of cardiac risk, including ambulatory ECG, exercise ECG, stress echocardiography, radionuclide
imaging, and coronary angiography. Noninvasive tests are sufficiently sensitive to identify most patients at increased
risk. Exercise ECG (exercise stress testing) is the historical standard to unmask myocardial ischemia. The sensitivity
for detection of CAD ranges up to 81%, whereas specificity varies up to 96%, depending on the testing protocol.
Testing has important prognostic value when ST segment depression of 1.5mm or greater occurs early during testing,
is sustained into the recovery period, is associated with a submaximal increase in heart rate or blood pressure, or is
accompanied by angina or an arrhythmia. However, false-negative studies are problematic. Moreover, the test has
limited value as a screening procedure for healthy, asymptomatic individuals.
Radionuclide cardiac imaging is popular for preoperative evaluation of cardiac disease, most commonly with thallium
perfusion scanning, which can be performed at rest, during exercise, or during a pharmacological exercise equivalent
(e.g., dipyridamole) for patients who cannot exercise (e.g., those with peripheral vascular disease, lower-extremity
orthopedic problems). Myocardial perfusion imaging using intravenous
CAD, the reversibility of the lesions, and the stress response of
201Th
201Th
analyzes the extent and localization of
in the coronary circulation. The isotope is taken
up by myocytes in a manner analogous to potassium. Rapid uptake allows visualization of ischemic or unperfused
myocardium. Normal coronary blood flow is relatively homogeneous, such that perfusion deficits cannot be detected
in the resting state unless severe (90% or greater) coronary artery stenosis is present. Heterogeneity can therefore
be enhanced by superimposed myocardial stress, which reflects ischemia. Because myocardial clearance of
rapid, redistribution during reperfusion of ischemic myocardium can also be observed. The accuracy of
201Th
is
201Th
perfusion scans is limited by lower sensitivity with lesser degrees of coronary stenosis. Single-vessel disease
involving the circumflex or right coronary circulations may not be detected, and disease in the left anterior descending
artery may go unrecognized if redistribution occurs in other segmental circulations. Although the negative predictive
value is high (90%), the presence of redistribution during reperfusion is identified so often, particularly in vascular
surgical patients, that its positive predictive value is low (30%). It is also possible to estimate the left ventricular
ejection fraction (LVEF), which portends increased risk when below 35% however it is measured (e.g.,
echocardiography).
TABLE 17.3. Evaluation Steps Corresponding to ACC/AHA Guideline Algorithms for Perioperative
Cardiovascular Evaluation of Noncardiac Surgery.ª
View Large
FIGURE 17.1 American College of Cardiology/American Heart Association
guideline algorithm for evaluation of cardiac risk before noncardiac surgery.
Patients with major clinical predictors of risk may have to have surgery postponed
or cancelled or undergo an invasive evaluation. See Table 17.5 for additional
information. (Reprinted from Eagle et al.,39 with permission.)
View Large
FIGURE 17.2 American College of Cardiology/American Heart Association
guideline algorithm for evaluation of cardiac risk before noncardiac surgery.
Patients with intermediate clinical predictors of risk or who are about to undergo
high-risk surgery may have to have noninvasive testing before surgery. See
Tables 17.3 and 17.5 for additional information. Four metabolic equivalents
(METs) are equivalent to climbing one flight of stairs with a bag of groceries.
(Reprinted from Eagle et al.39 with permission.)
View Large
Stress echocardiography (usually with infusion of dobutamine) may be even more accurate than
201
Th scanning
according to a meta-analysis of the recent literature.40 Dobutamine echocardiography is less expensive than a
201Th
perfusion scan and has the advantage of additional imaging possibilities. Valvular function can be assessed, wall
motion and wall thickening can be quantified, and an estimate of LVEF can be made from measurements of endsystolic and end-diastolic areas. Dobutamine echocardiography should probably be considered the provocative test
of choice for moderate- to high-risk patients. Echocardiographic estimates of ventricular function correlate well with
angiographic and radionuclide data. Such information can be of great value as reduced LVEF (<35%) correlates
strongly with perioperative myocardial events. Some patients may be evaluated more safely at rest than under
pharmacological stress. An equivocal or positive result from noninvasive testing is an indication for cardiac
catheterization (Table 17.4; see Figs. 17.1-17.3).
FIGURE 17.3 American College of Cardiology/American Heart Association
guideline algorithm for evaluation of cardiac risk before noncardiac surgery.
Patients with minor or no clinical predictors who are about to undergo high-risk
surgery may have to have noninvasive testing before surgery. See Tables 17.3
and 17.5 for additional information. Four metabolic equivalents (METs) are
equivalent to climbing one flight of stairs with a bag of groceries. (Reprinted from
Eagle et al.,39 with permission.)
View Large
View Large
Adjustment of Cardiovascular Medications
To minimize risk, the patient must be in optimal medical condition. Such optimization is ultimately the responsibility of
the operating surgeon but may often be undertaken by the referring physician or a consultant. Congestive heart
failure, poorly controlled hypertension (diastolic blood pressure >110mmHg), and diabetes mellitus must be stabilized
before an elective procedure is undertaken. In general, cardiovascular medications should be continued through the
perioperative period. Continuation of antihypertensive therapy throughout the perioperative period does not contribute
to hemodynamic instability, although the data are conflicting regarding whether continuation of antihypertensive
therapy actually decreases morbidity. Discontinuation of antihypertensive therapy does pose potential hazards.
Rebound hypertension may be precipitated when centrally acting α 2-adrenergic agonists (e.g., clonidine) are withheld
abruptly. Congestive heart failure may recur if angiotensin-converting enzyme inhibitors or angiotensin receptor
blockers are withheld. Diuretic therapy may cause hypovolemia or hypokalemia, but neither problem poses major
difficulties if recognized and treated. There is widespread agreement that β-adrenergic blockade should not be
discontinued abruptly. Abrupt discontinuation may be associated with a hyperadrenergic withdrawal syndrome
characterized by unstable angina, tachyarrhythmias, MI, or sudden death.
Tachycardia is common in the perioperative period, whether from pain, discomfort from positioning or an indwelling
catheter, hypovolemia, or many other reasons. Less evident but equally important is ECG evidence of myocardial
ischemia, which may be inapparent unless it is specifically sought by ST-segment trend analysis.41 The diagnosis of
perioperative MI can be elusive because most are silent clinically, many are nontransmural (non-Q-wave) and
therefore have minimal accompanying ECG changes, and the incision of muscle may elaborate creatine
phosphokinase and confound enzymatic diagnosis. Current ACC/AHA recommendations are to screen for MI in
patients without evidence of CAD only if signs of cardiovascular dysfunction develop. For patients with CAD
undergoing high-risk operations, an ECG at baseline, immediately postoperatively, and daily for the first 2
postoperative days should be obtained. Measurements of cardiac enzymes are best reserved for patients at high risk
or those who demonstrate ECG or hemodynamic evidence of myocardial dysfunction. 39 Increasingly, measurement of
serum troponin concentration is supplanting creatine phosphokinase determination.
Several studies suggested that both short- and long-term survival can be improved by β-adrenergic blockade for
patients who undergo noncardiac surgery.35,36 In one study,35 200 patients with CAD and at least two risk factors
were randomized to receive placebo or atenolol at 5 to 10mg i.v, 30min before major noncardiac surgery, which
lasted an average of 6h. Atenolol at 5 to 10mg every 12h i.v. or 50 to 100mg p.o. daily was continued until discharge
or for a maximum of 7 days. There was no difference in in-hospital MI or death rate, but overall mortality and deaths
from cardiovascular disease were reduced significantly at 6 months and 2 years. At 2 years, reduction of the relative
risk for death was 48%, and there was a 15% absolute increase in event-free survival at 2 years, from 68% to 83%. In
another study,36 a randomized multicenter trial of bisoprolol versus standard perioperative care was conducted in
high-risk patients (clinical indicators and the results of dobutamine echocardiography) not already taking β-blockers
and about to undergo major vascular surgery. The medication was started 1 week before surgery and continued for
30 days thereafter. Those who received β-blockers had statistically lower rates of perioperative (30-day) MI and
death.
On the other hand, continuation of calcium channel antagonists in the perioperative period is controversial. Rebound
phenomena associated with abrupt drug discontinuance are less common than with β-blockers, but patients receiving
combined therapy with β-adrenergic and calcium channel blockers are at increased risk of conduction abnormalities
and depressed ventricular function. Digoxin therapy for chronic CHF, particularly with complicating supraventricular
tachyarrhythmias, should be continued.
Preoperative Preparation in the Intensive Care Unit
Preoperative admission to the ICU for final preparation for surgery has fallen out of favor. Preoperative "optimization"
is no longer utilized routinely because of the chronic shortage of ICU beds and nurses and a lack of supporting data.
Only patients who need active therapy, such as aggressive fluid resuscitation, should be considered candidates for
preoperative ICU admission. Patients likely to benefit from preoperative evaluation in the ICU include those with
unstable angina, severe valvular heart disease or decompensated CHF, shock, or perhaps severe renal disease.
Preoperative Pulmonary Evaluation
Patients with a history of lung disease or those for whom a pulmonary resection is planned may benefit from
preoperative assessment and optimization of pulmonary function. Late postoperative pulmonary complications are
leading causes of morbidity and mortality, second only to cardiac complications as causes of death after surgery.
Prolonged postoperative decreases in functional residual capacity (FRC) and forced vital capacity (FVC) are
associated with atelectasis, decreased pulmonary compliance, increased work of breathing, and tachypnea at low
tidal volumes. Poor cough effort and impaired airway reflexes increase susceptibility to retained secretions, bacterial
invasion, and pneumonia. Older age, upper abdominal and thoracic incisions, neurosurgical procedures, emergency
operations, prolonged operative time, increased severity of underlying pulmonary disease (chronic obstructive
pulmonary disease [COPD] or chronic bronchitis), alcohol abuse, cigarette smoking, poor preoperative nutrition, and
preoperative blood transfusion are independent risk factors for major pulmonary morbidity 42 (Table 17.5).
Pulmonary morbidity may be anticipated after thoracotomy. Resection of a lung tumor requires removal of functional,
albeit abnormal, tissue from patients who have limited pulmonary reserve. Operability is assessed by evaluation of
baseline pulmonary function, including the contribution to overall pulmonary function of the tissue proposed for
resection (by split-function studies). In contrast, preoperative assessment before nonthoracic surgery should focus on
identification of chronic airway obstruction, possible preoperative intervention to minimize risk, and the choice of
surgical incision in the case of celiotomy and laparoscopic surgery as an alternative. However, few data suggest that
outcome is improved by optimization of pulmonary function before elective procedures. There is no doubt, however,
that prophylaxis of venous thromboembolic complications is important for patients at risk.
View Large
Preoperative chest radiography, of no value routinely, may be of value for dyspneic patients with underlying lung
disease or to serve as a basis for comparison. Radiographic indicators of possible airflow obstruction include
depression of the right hemidiaphragm at or below the seventh rib anteriorly on a conventional posteroanterior view, a
cardiac silhouette with a transverse dimension less than 11.5cm, and a retrosternal air space greater than 4.4cm on a
lateral view. Substantive airflow obstruction may be associated with a normal X-ray.
Most laboratory studies, other than the serum albumin concentration, are of little benefit for prediction of pulmonary
morbidity (Table 17.5). An elevated serum bicarbonate concentration suggests chronic respiratory acidosis, whereas
polycythemia may suggest chronic hypoxemia. A room air arterial oxygen tension (PaO 2) less than 60mmHg
correlates with pulmonary hypertension, whereas a PaCO2, greater than 45mmHg is associated with increased
perioperative morbidity. Spirometry before and after bronchodilators is simple and safe to obtain. Analysis of forced
expiratory volume in 1s (FEV1) and FVC usually provides sufficient information for clinical decision making. Dyspnea
is assumed to occur when FEV1 is less than 2l, whereas an FEV1 less than 50% of the predicted value correlates with
exertional dyspnea. In COPD, the FVC decreases less than the FEV 1, resulting in an FEV1/FVC ratio less than 0.8.
Spirometry correlates with the development of postoperative atelectasis and pneumonia, particularly if FEV 1 is less
than 1.2l or less than 70% of predicted, if FVC is less than 1.7l or less than 70% of predicted, or if FEV 1/FVC is less
than 0.65. If spirometric parameters improve by 15% or more after bronchodilator therapy, then such therapy should
be continued. If pulmonary resection is planned, then split-lung function can be determined. An FEV1 of
approximately 800ml from the contralateral lung is required to proceed with pneumonectomy. For abdominal surgery,
there is no indication for evaluation beyond spirometry and arterial blood gas analysis.
View Large
Preoperative pulmonary toilet is of unproven benefit and probably accomplishes more by patient education than by
actually improving gas exchange. However, lung expansion exercises (e.g., spirometry) are of proven benefit in the
postoperative period (Table 17.6).43 Chronic bronchodilator therapy should be continued perioperatively. A short
course of oral antibiotics can treat acute bronchitis before surgery if the sputum is purulent or tenacious, but
expectorants and mucolytic agents are of no value and may actually promote bronchospasm. Cessation of cigarette
smoking has been advocated for those who smoke more than 10 cigarettes per day, but the benefit is uncertain
(Table 17.6). Short-term abstinence (48h) decreases the carboxyhemoglobin concentration to that of a nonsmoker,
abolishes the effects of nicotine on the cardiovascular system, and improves mucosal ciliary function. Sputum volume
decreases after 1 to 2 weeks of abstinence, and spirometry improves after about 6 weeks.
Prophylaxis of Venous Thromboembolism
The morbidity and mortality of venous thromboembolism make consideration of prophylaxis mandatory for every
major operation. Several risk factors have been identified (Table 17.7), including increasing age, obesity, previous
thromboembolic disease, varicose veins, cigarette smoking, major surgery (especially pelvic, urological, orthopedic,
and cancer surgery), and several hematologic disorders; risk is increased further by the presence of several risk
factors (Table 17.8).44 The lowest-risk patients are those who are undergoing only minor surgery and who have no
risk factors; prophylaxis is not usually needed. However, risk is increased somewhat for any patient over the age of
40 who undergoes general anesthesia for more than 30 min. Many prophylactic regimens have proven efficacy for
patients at moderate to high risk, and the morbidity is acceptable 45, therefore, standard regimens are employed
increasingly for virtually all patients (Tables 17.8 and 17.9).
TABLE 17.7. Risk Factorsª for the Development of Venous Thromboembolism in the Perioperative Period.
View Large
TABLE 17.8. Risk Stratification Scheme and Incidence of Venous Thromboembolic Events.
View Large
View Large
TABLE 17.10. Effectiveness of Various Modalities for Risk Reduction in Prophylaxis of Perioperative Deep
Venous Thrombosis.
View Large
For general surgery patients, prophylaxis options of proven benefit in prospective trials include low-dose
unfractionated heparin (LDUH), low molecular weight heparin (LMWH), intermittent pneumatic compression, and oral
warfarin (Table 17.10). Subcutaneous heparinoids must be administered 2h before induction to be maximally
effective, making them somewhat inconvenient for use in ambulatory or same-day admission settings. Moreover, it is
recommended that LMWH should be administered for 1 week postoperatively in orthopedic surgery patients, but it is
unknown whether this applies to general surgery patients as well (Table 17.9). Metaanalysis of more than 30
randomized controlled trials comparing LMWH to LDUH in general surgery patients demonstrated comparable
efficacy for the prevention of thromboembolic phenomena but at a consequence of a slightly higher incidence of
minor wound bleeding.45 Low molecular weight heparin is inadvisable for a patient with recent neurosurgery,
gastrointestinal bleeding, or renal insufficiency and has been reported to cause spinal or epidural hematomas in
patients with epidural catheters. It is recommended that an epidural catheter should be removed at least 12h before
instituting LMWH for any indication. Concomitant LMWH and epidural catheterization are contraindicated.
Patients at high risk require aggressive prophylaxis; multimodality therapy (e.g., anticoagulation plus intermittent
pneumatic compression) is common (see Table 17.9), although supporting data are scant. The increased cost and
risks of combined-modality prophylaxis are offset by the potential devastation of a pulmonary embolism. Although
expensive and lacking any randomized prospective trials, prophylactic placement of vena cava filters is popular. In
longitudinal outcome studies, filters are more than 96% effective, and the incidence of complications is low. High-risk
patients undergoing high-risk surgery may be appropriate candidates, such as those who have a history of deep
venous thrombosis or pulmonary embolism or certain patients with multiple trauma. Newer, potentially retrievable
filters may expand the indications by reducing long-term complications further, but experience remains limited.
Evaluation of the Risk of Bleeding
Evaluation of the patient's risk of bleeding requires a careful history and physical assessment to be cost-effective
because the routine screening tests of hemostasis have a low yield. Important historical data include whether the
patient or a relative has had a prior episode of bleeding or a thromboembolic event and whether the patient reports
prior transfusions, prior surgery, heavy menstrual bleeding, easy bruising, frequent nosebleeds, or bleeding gums
after brushing the teeth. Coexistent liver or kidney disease, poor diet, excessive alcohol use, ingestion of aspirin,
other nonsteroidal antiinflammatory drugs, lipid-lowering drugs (possible vitamin K deficiency), and anticoagulant
therapy (usually warfarin) must be ascertained. Answers to these questions should uncover most potential problems
with hemostasis. If the history is completely negative and the patient has had a previous hemostatic challenge from
surgery or trauma, then an important hemostatic defect is extremely unlikely. A mild coagulopathy in the previously
unchallenged patient is not excluded; however, the consequences once such a mild defect is unmasked can be
managed readily.
Many hematologic tests can be omitted safely from the preparation protocol if there is no clinical suspicion of a
coagulopathy, even before major surgery. Notable exceptions are those operations known to affect coagulation,
including cardiopulmonary bypass procedures, prostatectomy, and possibly peripheral vascular surgery. Laboratory
testing is over-utilized and almost never yields a finding of importance in patients with a negative history and physical
examination.46 Many screening coagulation tests can generate false-positive results that do not translate into
increased surgical bleeding. Minor problems, such as oozing from the subcutaneous portion of the incision, can be
managed readily by a number of techniques. In the absence of clinical suggestion of a bleeding disorder, the chance
that a patient will have a major clotting disorder during surgery has been estimated to be less than 0.01%. Even when
indicated, the usual screening tests (prothrombin time [PT], activated partial thromboplastin time [aPTT], and platelet
count) identify abnormalities of importance in only 0.2% of patients. False-positive results are especially common with
the aPTT; one study found the test to be abnormal 14% of the time but consequential in only 16% of positives (2.2%
overall). Similarly, prolongation of the template bleeding time (e.g., after aspirin ingestion) does not correlate with
increased operative blood loss. If a clinically important coagulopathy is identified, therapeutic strategies for
management of various coagulation disorders in preparation for surgery are listed in Table 17.11.
TABLE 17.11. Preoperative Management of Selected Coagulation Disorders.
View Large
Management of the Therapeutically Anticoagulated
Patient
It is often necessary to operate on an anticoagulated patient. In such circumstances, it is desirable to reverse the
patient's anticoagulation temporarily so that hemostasis can be optimized. Procoagulant therapy may sometimes
obviate the need for surgery by stopping the bleeding (e.g., gastrointestinal hemorrhage). Previously, perioperative
anticoagulant management was needed for patients with a metal prosthetic heart valve, but now chronic atrial
fibrillation is the most common indication for long-term warfarin. The approach can be individualized, based on the
urgency and magnitude of the surgery to be performed and the strength of the indication for anticoagulation.
Most patients who take warfarin and who are to undergo ambulatory or same-day admission elective surgery can be
managed simply by discontinuing the warfarin several days before surgery. Most such patients are on a stable dose
of warfarin and are sophisticated regarding their medication and diet because of the need for frequent monitoring of
their anticoagulation. The timing of the medication adjustment depends on the degree of anticoagulation determined
by preoperative testing, which in turn depends on the indication for the anticoagulation. For example, a patient with a
valve prosthesis can be maintained chronically at an international normalized ratio (INR) of 2.5 to 3.0. If there is
concern that the patient should not be without anticoagulation, then the patient can be heparinized systemically with
an infusion of unfractionated heparin or placed on LMWH. The heparin infusion is discontinued approximately 4h
preoperatively (the half-life of heparin is about 90 min), and surgery proceeds with good hemostasis. However, data
are insufficient for a definitive recommendation regarding LMWH.
Clopidogrel, a potent selective inhibitor of adenosine diphosphate-mediated platelet aggregation, is prescribed
increasingly for prophylaxis of thrombosis of drug-coated stents placed for management of coronary artery occlusive
disease. The effect of clopidogrel on platelets is immediate and irreversible, so optimally the drug should be withheld
for 5 to 7 days prior to elective surgery. However, there is increased risk of stent occlusion without clopidogrel for at
least 6 months after stent placement, so benefit and risk must be considered, including referral of elective surgery for
the 6-month period.
In most circumstances, there is less urgency for reanticoagulation than is generally appreciated. Protection of a
cardiac valve prosthesis is the most urgent indication, but a metallic valve can be left without anticoagulation for at
least 72h and perhaps as long as 1 week (especially in the aortic position), although such a long interval is seldom
necessary. High-risk patients or those unable to take warfarin by mouth can be heparinized safely as early as 12h
after almost any operation with secure hemostasis, except neurosurgical procedures and some operations for major
trauma. Patients who take clopidogrel appear to be at risk for postoperative bleeding for up to 2 weeks even if
clopidogrel is withheld for several days after surgery, so the drug should be reintroduced with particular caution.
Steroid Prophylaxis
It is traditional that patients who are on a maintenance glucocorticoid regimen, or who have received corticosteroids
within the past 6 months, should receive supplemental "stress dose" steroid prophylaxis owing to concern that a
hypophysis-pituitary-adrenal axis suppressed by exogenous steroids may not respond to surgical stress. Large doses
(100mg hydrocortisone i.v, every 8h or equivalent) were given for undefined periods without any monitoring, despite
the fact that normal adrenal glands, stimulated maximally, increase their output from about 35 to 150mg cortisol/day
on average, and that exogenous high-dose steroids have deleterious effects on wound healing, host defenses,
carbohydrate metabolism, and other systems. Moreover, there has been no accounting for variability in the stress
response (i.e., that a hernia repair causes less stress than an esophagogastrectomy or that laparoscopic procedures
appear to be less stressful than their open counterparts). A CBG (there are virtually no class I data) has suggested
that such doses are far too high and are given for too long. 47 A minor surgical stress (e.g., inguinal herniorraphy) may
not need steroid supplementation at all, whereas a major stress (e.g., esophagogastrectomy) may need only 150 to
200mg hydrocortisone/day for 1 to 2 days.
Of course, there must be a high index of suspicion for adrenal insufficiency in such patients, which can be
precipitated by postoperative events, such as infection. The diagnosis is best made by a stimulation test using
cosyntropin. A baseline cortisol concentration is drawn, and 0.01 or 0.25mg cosyntropin is administered
intravenously; the merits of the two doses are under debate. The serum cortisol concentration is repeated 30 to
60min after the challenge. Glucocorticoids can be given immediately thereafter as indicated, pending the results. The
diagnosis is confirmed if neither of the values exceeds 15ng/ml or the stimulated cortisol concentration does not
increase by at least 9ng/ml. Patients respond hemodynamically within 12 to 24h of starting glucocorticoid (50 to 75mg
hydrocortisone every 8h or equivalent), but it may take several days to correct the electrolyte abnormalities or for
fever to dissipate.
Resuscitation: The Interface Between Preoperative and
Postoperative Care
Fluid resuscitation is an ubiquitous issue in the perioperative period, one that is multifaceted and context sensitive.
Among the many facets are the choice of fluid, the monitoring modalities, and the choice of endpoint. Whether to use
isotonic crystalloid or colloid solutions for resuscitation has been a matter of debate for more than three decades.
However, one meta-analysis of fluid resuscitation for shock indicated that albumin administration is associated with
excess mortality.48 Considering that human albumin solutions are nearly 100-fold more expensive than equivalent
volumes of crystalloid, there no longer seems to be any justification for the administration of colloids for fluid
resuscitation unless the patient is markedly hypoalbuminemic (<2g/dl).
Among the many considerations in addition to the volume status are whether shock is present and
vasopressor/inotropic therapy is needed in addition to fluid,49 whether the patient is bleeding, whether relevant
preexisting disease is present (e.g., CHF, chronic renal insufficiency, diabetes mellitus), or if there has been an acute
decompensation of vital organ function (e.g., acute MI, acute respiratory distress syndrome [ARDS], or acute tubular
necrosis [ATN]). The patient may require resuscitation to prepare for surgery (e.g., small bowel obstruction,
penetrating trauma to the torso), or resuscitation may be the primary therapeutic modality (e.g., acute alcoholic
pancreatitis, anatomically stable pelvic fracture with no associated injuries). Postoperative patients often require fluid
resuscitation, especially if intraoperative fluid requirements have been underestimated, evaporative losses are high
because a body cavity (especially both chest and abdomen) is open for a prolonged period, the patient is
hypothermic, or there has been osmotic diuresis from hyperglycemia or the administration of mannitol.
The goal of fluid resuscitation is to support the circulation to provide sufficient oxygen and metabolic fuel for cellular
function, thereby facilitating the resolution of the inflammatory response 50 and preventing or ameliorating the
development of the multiple-organ dysfunction syndrome.51