Chapter 2. Preoperative and Postoperative Management

Modern advances in patient care have enabled surgeons to treat more challenging and complicated surgical problems. In addition, surgical treatment can be offered to more fragile patients, with successful outcomes. In order to achieve these good results, it is vital to master the scientific fundamentals of perioperative management. The organ system–based approach allows the surgeon to address the patient's pre- and postoperative needs, and ensures that these needs are part of the surgical plan.

The most common neuropsychiatric complications following abdominal surgery are pain and delirium. Moreover, uncontrolled pain and delirium prevent the patient from contributing to vital aspects of his or her care such as walking and coughing, and promote an unsafe environment that may lead to the unwanted dislodgment of drains and other supportive devices, with potentially life-threatening consequences. Pain and delirium frequently coexist, and each can contribute to the development of the other. Despite high reported rates of overall patient satisfaction, pain control is frequently inadequate in the perioperative setting1 with high rates of complications such as drowsiness and unacceptable levels of pain. Therefore, it is mandatory that the surgical plan for every patient include control of postoperative pain and delirium and regular monitoring of the efficacy of pain control.

Pain management, like all surgical planning, begins in the preoperative assessment. In the modern era, a large proportion of surgical patients will require special attention with respect to pain control. Patients with preexisting pain syndromes, such as sciatica or interspinal disc disease, or patients with a history of opioid use may have a high tolerance for opioid analgesics. Every patient's history should include a thorough investigation for chronic pain syndrome, addiction (active or in recovery), and adverse reactions to opioid, nonsteroidal, or epidural analgesia. The pain control strategy may include consultation with a pain control anesthesiology specialist, but it is the responsibility of the operating surgeon to identify complicated patients and construct an effective pain control plan.

Opioid Analgesia

Postoperative pain control using opioid medication has been in use for thousands of years. Hippocrates advocated the use of opium for pain control. The benefits of postoperative pain control are salutary, and include improved mobility and respiratory function, and earlier return to normal activities. The most effective strategy for pain control using opioid analgesia is patient-controlled analgesia (PCA), wherein the patient is instructed in the use of a preprogrammed intravenous pump that delivers measured doses of opioid (usually morphine or meperidine). In randomized trials, PCA has been shown to provide superior pain control and patient satisfaction compared to interval dosing,2 but PCA has not been shown to improve rates of pulmonary and cardiac complications3 or length of hospital stay,4 and there is evidence that PCA may contribute to postoperative ileus.5 In addition, PCA may be unsuitable for patients with a history of substance abuse, high opioid tolerance, or those with atypical reactions to opioids.

Epidural Analgesia

Due to the limitations of PCA, pain control clinicians have turned to epidural analgesia as an effective strategy for the management of postoperative pain. Postoperative epidural analgesia involves the insertion of a catheter into the epidural space of the lumbar or thoracic spine, enabling the delivery of local anesthetics or opioids directly to the nerve roots. The insertion procedure is generally safe, with complication rates of motor block and numbness between 0.5% and 7%,6 and an epidural abscess rate of 0.5 per thousand.7 Potential advantages of epidural analgesia include elimination of systemic opioids, and thus less respiratory depression, and improvement in pulmonary complications and perioperative ileus. There have been several large trials,8–10 a meta-analysis,6 and a systematic review11 comparing PCA with epidural analgesia in the setting of abdominal surgery. These studies indicate that epidural analgesia provides more complete analgesia than PCA throughout the postoperative course. Furthermore, in randomized prospective series of abdominal procedures, epidural analgesia has been associated with decreased rates of pulmonary complications12,13 and postoperative ileus.14,15 Epidural analgesia requires a skilled anesthesia clinician to insert and monitor the catheter and adjust the dosage of neuraxial medication. Some clinicians may prefer correction of coagulopathy before inserting or removing the catheter, although the American Society of Anesthesiologists (ASA) has not issued official guidelines on this issue.

Analgesia with Nonsteroidal Anti-Inflammatory Drugs

Oral nonsteroidal anti-inflammatory drugs (NSAIDs) have long been used for postoperative analgesia in the outpatient setting, and with the development of parenteral preparations, have come into use in the inpatient population. This class of medication has no respiratory side effects and is not associated with addiction potential, altered mental status, or ileus. In addition, these medications provide effective pain relief in the surgical population. However, use of NSAIDs has not been universally adopted in abdominal surgery due to concerns regarding the platelet dysfunction and erosive gastritis associated with heavy NSAID use. In prospective trials, NSAIDs were found to provide effective pain control without bleeding or gastritis symptoms following laparoscopic cholecystectomy,16 abdominal hysterectomy,17 and inguinal hernia repair.18,19 NSAIDs have also been shown to improve pain control and decrease morphine dosage when used in combination following appendectomy.20

The sensation of pain is very subjective and personal. Accordingly, the surgeon must individualize the pain control plan to fit the needs of each patient. The pain control modalities discussed above can be used in any combination, and the surgeon should not hesitate to use all resources at his or her command to provide adequate relief of postoperative pain.

Postoperative Delirium

Delirium, defined as acute cognitive dysfunction marked by fluctuating disorientation, sensory disturbance, and decreased attention, is an all too common complication of surgical procedures, with reported rates of 11–25%, with the highest rates reported in the elderly population.21,22 The postoperative phase of abdominal surgery exposes patients, some of whom may be quite vulnerable to delirium, to a large number of factors that may precipitate or exacerbate delirium (Table 2-1). These factors can augment each other: postoperative pain can lead to decreased mobility, causing respiratory compromise, atelectasis, and hypoxemia. Escalating doses of narcotics to treat pain can cause respiratory depression and respiratory acidosis. Hypoxemia and delirium can cause agitation, prompting treatment with benzodiazepines, further worsening respiratory function and delirium. This vicious cycle can play out right before the physician's eyes, and if not interrupted, can result in serious complications or death. Preoperative recognition of high-risk patients and meticulous monitoring of every patient's mental status are the most effective ways to prevent postoperative delirium; treatment can be remarkably difficult once the vicious cycle has begun.

Patient factors that are associated with high risk of perioperative delirium include age greater than 70 years, preexisting cognitive impairment or prior episode of delirium, history of alcohol or narcotic abuse, and malnutrition.21,23 Procedural factors associated with high delirium risk include operative time greater than 2 hours, prolonged use of restraints, presence of a urinary catheter, addition of more than three new medications, and reoperation.22

Once the patient's risk for postoperative delirium is identified, perioperative care should be planned carefully to decrease other controllable factors. Epidural analgesia has been associated with less delirium than PCA after abdominal surgery.24 Sedation or “sleepers” should be used judiciously, if at all, with high-risk patients. If the patient requires sedation, neuroleptics such as haloperidol and the atypical neuroleptics such as olanzapine are tolerated much better than benzodiazepines.25 The patient's mental status, including orientation and attention, should be assessed with every visit, and care should be taken to avoid anemia, electrolyte imbalances, dehydration, and other contributing factors.

Once the diagnosis of postoperative delirium is established, it is important to recognize that some of the causes of delirium are potentially life-threatening, and immediate action is necessary. Evaluation begins with a thorough history and physical examination at the bedside by the surgeon. The history should focus on precipitating events such as falls (possible traumatic brain injury), recent procedures, use of opioids and sedatives, changes in existing medications (eg, withholding of thyroid replacement or antidepressants), and consideration of alcohol withdrawal. The vital signs and fluid balance may suggest sepsis, hypovolemia, anemia, or dehydration. The examination should include brief but complete sensory and motor neurological examinations to differentiate delirium from stroke. Pay attention to common sites of infection such as the surgical wound, the lungs, and intravenous catheters. Urinary retention may be present as a result of medication or infection. Deep venous thrombosis may be clinically evident as limb swelling. Postoperative myocardial infarction (MI) may often present as acute cardiogenic shock.

The history and physical examination should then direct the use of laboratory tests. Most useful are the electrolytes, blood glucose, and complete blood cell count. Pulse oximetry and arterial blood gases may disclose hypercapnia or hypoxemia. Chest x-ray may disclose atelectasis, pneumonia, acute pulmonary edema, or pneumothorax. Cultures may be indicated in the setting of fever or leukocytosis, but will not help immediately. Electrocardiogram (ECG) and cardiac troponin may be used to diagnose postoperative MI.

Resuscitative measures may be required if life-threatening causes of delirium are suspected. Airway control, supplemental oxygen, and fluid volume expansion should be considered in patients with unstable vital signs. The patient should not be sent out of the monitored environment for further tests, such as head computed tomography (CT), until the vital signs are stable and the agitation is controlled. Treatment of postoperative delirium depends on treatment of the underlying causes. Once the underlying cause has been treated, delirium may persist, especially in elderly or critically ill patients, who regain orientation and sleep cycles slowly. In these patients, it is important to provide orienting communication and mental stimulation during the day, and to promote sleep during the night. The simplest ways are the most effective: contact with family members and friends, use of hearing aids, engagement in activities of daily living, and regular mealtimes. Sleep can be promoted by keeping the room dark and quiet throughout the evening, and preventing unnecessary interruptions. If nighttime sedation is required, atypical neuroleptics or low-dose serotonin reuptake inhibitors such as trazodone are better tolerated than benzodiazepines. If agitation persists, escalating doses of neuroleptics (or benzodiazepines in the setting of alcohol withdrawal) can be used to control behavior, but hidden causes of delirium must be considered.

Risk Assessment

It has been estimated that 1 million patients have a perioperative MI each year, and the contribution to medical costs is $20 billion annually.26 Thoracic, upper abdominal, neurological, and major orthopedic procedures are associated with increased cardiac risk. Diabetes, prior MI, unstable angina, and decompensated congestive heart failure (CHF) are most predictive of perioperative cardiac morbidity and mortality, and patients with these conditions undergoing major surgery warrant further evaluation27 (Table 2-2). Patient factors conferring intermediate risk include mild angina and chronic renal insufficiency with baseline creatinine ≥2 mg/dL.28 It is worth noting that women were underrepresented in the studies on which the American College of Cardiology and the American Heart Association (ACC/AHA) guidelines are based.29 A retrospective study in gynecological patients found that hypertension and previous MI were major predictors of postoperative cardiac events, as opposed to the ACC/AHA guidelines, which indicate that they are minor and intermediate criteria, respectively.30 Therefore, vascular surgical patients are at highest risk because of the prevalence of underlying coronary disease in this population.27,31 Other high-risk procedural factors include emergency surgery, long operative time, and high fluid replacement volume; these are associated with a more than 5% risk of perioperative cardiac morbidity and mortality. Intraperitoneal procedures, carotid endarterectomy, thoracic surgery, head and neck procedures, and orthopedic procedures carry an intermediate risk, and are associated with a 1–5% risk of a perioperative cardiac event.28

Perioperative evaluation to identify patients at risk for cardiac complications is essential in minimizing morbidity and mortality. Workup should start with history, physical examination, and ECG to determine the existence of cardiac pathology. Screening with chest radiographs and ECG is required for men over 40 and women over 55. According to the ACC/AHA guidelines, indications for preoperative cardiac testing should mirror those in the nonoperative setting.32 The preoperative evaluation should include the surgeon, anesthesiologist, primary care physician, and possibly a cardiologist. Cardiology consultations are recommended for patients with major clinical predictors, those with intermediate clinical predictors and poor functional status undergoing intermediate-risk procedures, or those undergoing high-risk procedures with poor functional status or intermediate clinical predictors (Table 2-3). Overall functional ability is the best measure of cardiac health. Patients who can exercise without limitations can generally tolerate the stress of major surgery.33 Limited exercise capacity may indicate poor cardiopulmonary reserve and the inability to withstand the stress of surgery. Poor functional status is the inability to perform activities such as driving, cooking, or walking less than 5 km/h.

Intraoperative risk factors include operative site, inappropriate use of vasopressors, and unintended hypotension. Intra-abdominal pressure exceeding 20 mm Hg during laparoscopy can decrease venous return from the lower extremities and thus contribute to decreased cardiac output,34 and Trendelenburg positioning can result in increased pressure on the diaphragm from the abdominal viscera, subsequently reducing vital capacity. Intraoperative hypertension has not been isolated as a risk factor for cardiac morbidity, but it is often associated with wide fluctuations in pressure, and has been more closely associated with cardiac morbidity than intraoperative hypotension. Preoperative anxiety can contribute to hypertension even in normotensive patients. Those patients with a history of hypertension, even medically controlled hypertension, are more likely to be hypertensive preoperatively. Those with poorly controlled hypertension are at greater risk of developing intraoperative ischemia, arrhythmias, and blood pressure derangements, particularly at induction and intubation. Twenty-five percent of patients will exhibit hypertension during laryngoscopy. Patients with chronic hypertension may not necessarily benefit from lower blood pressure during the preoperative period because they may depend on higher pressures for cerebral perfusion. Those receiving antihypertensive medications should continue them up until the time of surgery. Patients taking beta-blockers are at risk of withdrawal and rebound ischemia. Key findings on physical examination include retinal vascular changes and an S4 gallop consistent with left ventricular (LV) hypertrophy. Chest radiography may show an enlarged heart, also suggesting LV hypertrophy.

Noninvasive cardiac testing is used to define risk in patients known to be at high or intermediate risk, and detect those with CHF or dyspnea. It is most useful in intermediate-risk patients. No special laboratory tests are necessary unless there is evidence of active ischemia. A baseline ECG is necessary to identify any new ECG findings, to rule out active ischemia, and as a baseline for comparison during the postoperative period. The baseline ECG will be normal in 25–50% of patients with coronary disease, but no history of MI. A 12-lead ECG should be obtained in patients with chest pain, diabetes, prior revascularization, prior hospitalization for cardiac causes, all men age 45 or older, and all women aged 55 with two or more risk factors. High- or intermediate-risk patients should also have a screening ECG. A lower-than-normal ejection fraction demonstrated on echocardiography is associated with the greatest perioperative cardiac risk, and should be obtained in all patients with symptoms suggesting heart failure or valvular disease. Tricuspid regurgitation indicates pulmonary hypertension and is often associated with sleep apnea. The chest x-ray is used to screen for cardiomegaly and pulmonary congestion, which may signify ventricular impairment.

Exercise testing demonstrates a propensity for ischemia and arrhythmias under conditions that increase myocardial oxygen consumption. Numerous studies have shown that performance during exercise testing is predictive of perioperative mortality in noncardiac surgery. ST-segment changes during exercise including horizontal depression greater than 2 mm, changes with low workload, and persistent changes after 5 minutes of exercise are seen in severe multivessel disease. Other findings include dysrhythmias at a low heart rate, an inability to raise the heart rate to 70% of predicted, and sustained decrease in systolic pressure during exercise.

Unfortunately, many patients are unable to achieve adequate workload in standard exercise testing because of osteoarthritis, low back pain, and pulmonary disease. In this case, pharmacological testing is indicated with a dobutamine echocardiogram. Dobutamine is a beta-agonist that increases myocardial oxygen demand and reveals impaired oxygen delivery in those with coronary disease. Echocardiography concurrently visualizes wall motion abnormalities due to ischemia. Transesophageal echocardiography may be preferable to transthoracic echocardiography in obese patients because of their body habitus, and has been shown to have high negative predictive value in this group.35 Nuclear perfusion imaging with vasodilators such as adenosine or dipyridamole can identify coronary artery disease and demand ischemia. Heterogeneous perfusion after vasodilator administration demonstrates an inadequate response to stress. Wall motion abnormalities indicate ischemia and an ejection fraction lower than 50% increases the risk of perioperative mortality. Angiography should only be performed if the patient may be a candidate for revascularization.

Coronary Disease

Most perioperative MIs are caused by plaque rupture in lesions that do not produce ischemia during preoperative testing.36 This presents an obvious challenge for detecting patients at risk. Stress testing has a low positive predictive value in patients with no cardiac risk factors, and has been associated with an unacceptably high rate of false-positives.37

Preoperative optimization may include medical management, percutaneous coronary interventions (PCI), or coronary artery bypass grafting (CABG).38 The ACC/AHA guidelines (Fig. 2-1) recommend coronary revascularization prior to noncardiac surgery in the following situations:

1. The combined risk of the two procedures does not exceed the risk of the surgical procedure alone.

2. Revascularization should reduce the risk of the noncardiac surgery more than the risk of the revascularization procedure itself.

3. Revascularization should not unduly delay the noncardiac surgery, especially if it is urgent.

Patients warranting emergent CABG will be at greatest risk for that procedure. A recent study from the Veterans Administration Hospitals recommends against revascularization in patients with stable cardiac symptoms.39 Preoperative PCI does not decrease the risk of future MI or mortality in patients with stable coronary disease, and only targets stenotic lesions, rather than those most likely to rupture. One retrospective study found no reduction in morbidity or perioperative MI after percutaneous transluminal coronary angioplasty, and the authors proposed that surgery within 90 days of balloon angioplasty increased the risk of thrombosis.40 However, PCI done more than 90 days before surgery did provide benefit when compared to those who had no intervention at all. Another retrospective study found that patients who have surgery within 2 weeks of stenting had a high incidence of perioperative MI, major bleeding, or death.41 Although a retrospective review from the Coronary Artery Surgery Study registry showed a lower mortality rate in patients with coronary artery disease who were post-CABG than those without CABG (0.09 vs 2.4%), this benefit did not include the morbidity associated with CABG itself. Unfortunately, the benefit was overwhelmed by the 2.3% morbidity rate seen with CABG in this cohort.42 Survival benefit of CABG over medical management is realized for 2 years after surgery43 so that preoperative mortality may decrease overall short-term survival. Revascularization and bypass grafting should be restricted to patients who would benefit from the procedure independent of their need for noncardiac surgery. One of the disadvantages of PCI in the preoperative setting is the need for anticoagulation to prevent early stent occlusion. The use of platelet inhibitors to prevent stent occlusion must be included in the overall risk assessment, especially for surgery of the central nervous system.

Catecholamine surges can cause tachycardia, which may alter the tensile strength of coronary plaques and incite plaque rupture.44,45 Catecholamine surges can also increase blood pressure and contractility, contributing to platelet aggregation and thrombosis after plaque rupture and increasing the possibility of complete occlusion of the arterial lumen.46 Perioperative beta-blockade mitigates these effects and has been shown to reduce MI and mortality from MI by over 30% in vascular surgical patients with reversible ischemia.44 Patients at highest risk still have a cardiac event rate of 10%, even with adequate perioperative beta-blockade.37

In 1998, a landmark study47 demonstrated a 55% reduction in mortality in non-cardiac surgical patients with known coronary disease who were given atenolol perioperatively. This was followed by the DECREASE trial48, which showed a 10-fold reduction in perioperative MI and death compared to placebo. Thereafter, perioperative beta-blockade was widely adopted as a quality measure. However, more recent investigations have shown that while perioperative beta-blockers benefit patients with known ischemia, low-risk patients may in fact be harmed.49 Tight rate control has been associated with increased risk of hypotension and bradycardia requiring intervention, and stroke without any significant decrease in mortality.50–53 Furthermore, critical analysis of the literature shows that studies have been inconsistent in the type of medication administered, the duration and timing of administration, or the target for heart rate control.54 Consequently, results are difficult to interpret. Thus, prophyalactic perioperative beta-blockade should be restricted to patients with cardiac ischemia and has a limited role in patients with low or moderate risk of postoperative cardiac events.

Congestive Heart Failure and Arrhythmia

Congestive heart failure (CHF) is associated with coronary disease, valvular disease, ventricular dysfunction, and all types of cardiomyopathy. These are all independent risk factors that should be identified prior to surgery. Even compensated heart failure may be aggravated by fluid shifts associated with anesthesia and abdominal surgery and deserves serious consideration. Perioperative mortality increases with higher New York Heart Association class and preoperative pulmonary congestion. CHF should be treated to lower filling pressures and improve cardiac output before elective surgery. Beta-blockers, angiotensin-converting enzyme inhibitors, and diuretics can be employed to this end. The patient should be stable for 1 week before surgery.55

Arrhythmias and conduction abnormalities elicited in the history, on examination, or on ECG, should prompt investigation into metabolic derangements, drug toxicities, or coronary disease. In the presence of symptoms or hemodynamic changes, the underlying condition should be reversed and then medication given to treat the arrhythmia. Indications for anti-arrhythmic medication and cardiac pacemakers are the same as in the nonoperative setting. Nonsustained ventricular tachycardia and premature ventricular contractions have not been associated with increased perioperative risk and do not require further intervention.56,57

Valvular Disease

Valvular disease should be considered in patients with symptoms of CHF, syncope, and a history of rheumatic heart disease (RHD). Aortic stenosis (AS) is a fixed obstruction to the LV outflow tract, limiting cardiac reserve and an appropriate response to stress. History should elicit symptoms of dyspnea, angina, and syncope; examination may reveal a soft S2, a late peaking murmur, or a right-sided crescendo-decrescendo murmur radiating to the carotids. AS is usually caused by progressive calcification or congenital bicuspid valve. Critical stenosis exists when the valve area is less than 0.7 cm2 or transvalvular gradients are greater than 50 mm Hg, and is associated with an inability to increase cardiac output with demand. If uncorrected, AS is associated with a 13% risk of perioperative death. Valve replacement is indicated prior to elective surgery in patients with symptomatic stenosis.58 Myocardial ischemia may occur in the absence of significant coronary artery occlusion in the presence of aortic valve disease. Perioperative management should include optimizing the heart rate to between 60 and 90 beats per minute and avoiding atrial fibrillation if possible. Because of the outflow obstruction, stroke volume may be fixed and bradycardia will lower cardiac output. Similarly, hypotension is also poorly tolerated.

Aortic regurgitation (AR) is associated with backward flow into the left ventricle during diastole and reduced forward stroke volume. Bradycardia facilitates regurgitation by increased diastolic time. Chronic AR causes massive LV dilatation (cor bovinum) and hypertrophy, which is associated with decreased LV function at later stages. AR is most often caused by rheumatic disease or congenital bicuspid valve. Medical treatment includes rate control and afterload reduction. Without valve replacement, survival is approximately 5 years once patients become symptomatic. This is an obvious consideration when planning any other surgical procedures.

Tricuspid regurgitation is usually caused by pulmonary hypertension secondary to severe left-sided failure. Other causes include endocarditis, carcinoid syndrome, and primary pulmonary hypertension. Hypovolemia, hypoxia, and acidosis can increase right ventricular afterload and should be avoided in the perioperative period.

Mitral stenosis is an inflow obstruction that prevents adequate LV filling. The transvalvular pressure gradient depends on atrial kick, heart rate, and diastolic filling time. Tachycardia decreases filling time and contributes to pulmonary congestion. Mitral regurgitation is also associated with pulmonary hypertension with congestion, as the pathologic valve prevents forward flow, causing left atrial dilatation, and subsequent atrial arrhythmias. History and physical examination should focus on signs of CHF such as orthopnea, pedal edema, dyspnea, reduced exercise tolerance, and auscultatory findings such as murmurs and an S3 gallop. Neurological deficits may signify embolic sequelae of valve disease. Perioperative rate control is essential for maintaining adequate cardiac output. ECG findings will reflect related arrhythmias and medications, but will not be specific for valve disease. Laboratory studies should identify secondary hepatic dysfunction or pulmonary compromise. Left ventricular hypertrophy is an adaptive response, which may cause subsequent pulmonary hypertension and diastolic dysfunction.

Prosthetics in the mitral position pose the greatest risk for thromboembolism, and the risk increases with valve area and low flow. Mechanical valves pose a higher risk than tissue valves in patients with a history of valve replacement. Diuretics and afterload-reducing agents will enhance forward flow and minimize cardiopulmonary congestion. Patients with mitral valve prolapse (MVP) should receive antibiotics.

Mitral regurgitation may also impair LV function and lead to pulmonary hypertension. Stroke volume is reduced by backward flow into the atrium during systole. The left ventricle dilates to handle increasing end-systolic volume, eventually causing concentric hypertrophy and decreased contractility. The end result may be decreased ejection fraction and CHF. A decrease in systemic vascular resistance and an increase in atrial contribution to the ejection fraction can both improve forward flow and reduce the amount of regurgitation. Echocardiography can clarify the degree of valvular impairment. Medical treatment centers on afterload reduction with vasodilators and diuretics. MVP is present in up to 15% of women, and is usually associated with a midsystolic click and late systolic murmur on physical examination. Murmur is indicative of prolapse. Although MVP is associated with connective tissue disorders, it usually occurs in otherwise healthy, asymptomatic patients. Echocardiography is used to confirm the diagnosis and evaluate the degree of prolapse. Chronically, MVP may be associated with mitral regurgitation, emboli, and increased risk of endocarditis. Prolapse may be aggravated by decreased preload, which should be minimized in the perioperative period. Patients with MVP are at risk of ventricular arrhythmias with sympathetic stimulation and endocarditis, which can be addressed with pain control and antibiotic prophylaxis, respectively.

Individuals with underlying structural cardiac defects are at increased risk for developing endocarditis after invasive procedures. Surgical procedures involving mucosal surfaces or infected tissues may cause transient bacteremia that is usually short-lived. Certain procedures are associated with a greater risk of endocarditis and warrant prophylaxis (Table 2-4). Abnormal valves, endocardium, or endothelium can harbor the blood-borne bacteria for a longer period of time, and infection and inflammation can ensue. While there are no randomized trials regarding endocarditis prophylaxis, the American Heart Association recommends prophylaxis for those59 at high and moderate risk for developing the condition. Highest-risk patients have prosthetic heart valves, cyanotic congenital heart disease, or a history of endocarditis (even without structural abnormality).60 Conditions associated with moderate risk include congenital septal defects, patent ductus arteriosus, coarctation of the aorta, and bicuspid aortic valve. Hypertrophic cardiomyopathy and acquired valvular disease also fall into this category. Mitral valve prolapse is a prevalent and often situational condition. Normal valves may prolapse in the event of tachycardia or hypovolemia, and may reflect normal growth patterns in young people. Prolapse without leak or regurgitation seen on Doppler studies is not associated with risk greater than that of the general population and no antibiotic prophylaxis is necessary.61,62 However, the jet caused by the prolapsed valve increases the risk of bacteria sticking to the valve and subsequent endocarditis. Leaky valves detected by physical examination or Doppler warrant prophylactic antibiotics.62 Those with significant regurgitation are more likely to be older and men, and other studies have shown that older men are more likely to develop endocarditis.63–65 Some advocate prophylaxis for men older than 45 years with MVP even in the absence of audible regurgitation.65 Prolapse secondary to myxomatous valve degeneration also warrants prophylactic antibiotics.66,67 Prophylaxis is indicated in cases in whom mitral regurgitation cannot be determined.68

For patients at risk the goal should be administration of antibiotics in time to attain adequate serum levels during and after the procedure. For most operations, a single intravenous dose given 1 hour prior to incision will achieve this goal. Antibiotics should generally not be continued for more than 6–8 hours after the procedure to minimize the chance of bacterial resistance. In the case of oral, upper respiratory, and esophageal procedures, alpha-hemolytic streptococcus is the most common cause of endocarditis, and antibiotics should be targeted accordingly. Oral amoxicillin, parenteral ampicillin, and clindamycin for penicillin-allergic patients are suitable medications. Erythromycin is no longer recommended for penicillin-allergic patients because of gastrointestinal side effects and variable absorption.69 Antibiotics given to those having genitourinary and nonesophageal gastrointestinal procedures should target enterococci.69 While gram-negative bacteremia can occur, it rarely causes endocarditis. Parenteral ampicillin and gentamicin are recommended for highest-risk patients. Moderate-risk patients may receive amoxicillin or ampicillin. Vancomycin may be substituted in patients allergic to penicillin.

Estimates suggest that 250,000 patients receiving chronic anticoagulation require surgery in the United States each year. Operative bleeding risk must be balanced against thromboembolic risk for the patient off of anticoagulation, and requires careful judgment. Factors that influence the risk of thromboembolism include the condition requiring chronic anticoagulation, the duration of the procedure, time expected off of anticoagulation, and the duration of perioperative immobility. Thromboembolic risk increases with the amount of time that the patient's anticoagulation is subtherapeutic.

Primary indications for chronic anticoagulation include arterial embolism associated with mechanical valves and atrial fibrillation and venous thromboembolism (VTE). Arterial events precipitate stroke, and valvular and atrial clot and systemic emboli are higher risk for morbidity and mortality than venous events. According to the American College of Chest Physicians Practice Guidelines for the Perioperative Management of Antithromotic Therapy,70 patients at highest risk for perioperative embolism have mechanical mitral valves, aortic caged-ball and tilted valves, RHD, or history of stroke or transient ischemic attacks (TIA) in the past 3 months. The risk of thromboembolism without anticoagulation is higher than 10% per year in high-risk patients.

Patients at moderate risk of thromboembolism without anticoagulation, 4–10% per year, have atrial fibrillation, a bileaflet valve, or history of stroke or TIA. The congestive heart failure-hypertension-age-diabetes-stroke score (CHADS2) further stratifies embolic risk for patients with atrial fibrillation based on comorbidities. One point is assigned for hypertension, diabetes, congestive heart failure, and age >75 years; two points are assigned for history of stroke or TIA. Patients with a cumulative score of 5–6 are highest risk; 3–4 are moderate risk, and 0–2 without history of stroke or TIA are low risk.

Chronic anticoagulation is indicated for VTE. Patients with VTE within 3 months of surgery and severe thrombophillia are at highest risk for perioperative events and should receive bridging anticoagulation with therapeutic doses of low-molecular weight heparin (LMWH) or intravenous unfractionated heparin (IV UFH). Patients at moderate risk include those with thromboembolic event 3–12 months before surgery and less severe thrombophillias. They can receive therapeutic or subtherapeutic doses of anticoagulation depending on the risk of bleeding associated with the procedure. Patients with a remote event are at lowest risk, and do not require bridging anticoagulation. It is generally recommended to stop Warfarin 5 days prior to surgery if a normal International Normalized Ratio (INR) is desired. Vitamin K may be administered in days leading up to the event if the INR is not coming down quickly enough.

LMWH should be held 24 hours before surgery, and IV UFH should be held 4 hours before surgery. Oral anticoagulants may be started 12–24 hours postoperatively because they take at least 48 hours to effect coagulation. The timing of resuming IV and SC anticoagulants should be determined on a case-by-case basis.

Low-risk patients receiving Clopidogrel or aspirin should have it held 5–10 days before surgery. Patients with coronary stents are chronically treated with Clopidogrel and aspirin to mitigate the risk of stent thrombosis. Interruptions in therapy are associated with high risk of thrombosis and infarct. Patients with bare metal stents placed within 6 weeks of surgery, and drug-eluting stents within 12 months of surgery, should continue Clopiogrel and aspirin in the perioperative period.

Pulmonary complications are common after surgery, and can prolong hospital stays for 1–2 weeks.71 Complications include atelectasis, pneumonia, exacerbations of chronic pulmonary disorders, and respiratory failure requiring mechanical ventilation. Smoking, underlying chronic obstructive pulmonary disease (COPD), and poor exercise tolerance are the greatest risk factors for postoperative pulmonary complications. Physicians should ask about a history of smoking, decreased exercise capacity, dyspnea, and chronic cough. Examination should note pursed lip breathing, clubbing, and chest wall anatomy that could impair pulmonary function. Pulmonary testing is unnecessary in patients without a clear history of smoking or pulmonary disease. The predictive value of screening spirometry is unclear, and no threshold value has been identified to guide surgical decision making. Forced expiratory volume in 1 second less than 50% of predicted is indicative of exertional dyspnea and may herald the need for further testing. Preoperative chest x-ray abnormalities are associated with postoperative pulmonary complications,71 but to this point there are no recommendations for screening radiographs in patients without pulmonary disease. Any preoperative chest x-ray must be examined for signs of hyperinflation consistent with COPD. While compensated hypercapnia has not been shown to be an independent predictor for postoperative ventilatory insufficiency in patients undergoing lung resection, preoperative arterial blood gas analysis provides useful baseline information for perioperative management of patients with chronic CO2 retention. Transverse and upper abdominal incisions are associated with a higher rate of postoperative pulmonary complications than longitudinal midline incisions and lower abdominal incisions.72 Surgery longer than 3 hours is also associated with higher risk.73 General anesthesia is also associated with a higher risk of pulmonary complications than spinal, epidural, or regional anesthesia.74

Physiologic changes can be seen in the postoperative period, especially after thoracic and upper abdominal procedures. Vital capacity may decrease by 50–60%, and is accompanied by an increased respiratory rate to maintain tidal volumes. Normally, functional residual capacity usually exceeds the closing capacity of the alveoli, so they remain open throughout the respiratory cycle. Prolonged effects of anesthetics and narcotics reduce functional reserve capacity postoperatively, causing alveolar collapse. These changes can last for weeks to months. A distended abdomen can impair diaphragmatic excursion; painful incisions around the diaphragm and other respiratory muscles contribute to splinting and inadequate pulmonary toilet. Narcotics can inhibit sighing and coughing reflexes, which normally prevent alveolar collapse during periods of sleep and recumbency. Analgesics must be titrated carefully to permit deep breathing and avoid impairing respiratory effort.

Inspired nonhumidified oxygen and halogenated anesthetics are cytotoxic and interfere with surfactant production and mucociliary clearance. Depressed respiratory reflexes, diaphragm dysfunction, and decreased functional reserve capacity all contribute to alveolar collapse and pooling of secretions. Aspiration risk is also increased. Excess secretions cause further alveolar collapse and create a milieu ripe for bacterial infection and pneumonia. Intubated patients should receive antacid prophylaxis and gastric drainage to minimize the risk of aspiration.

Multiple analyses have found that poor exercise tolerance is the greatest predictor of postoperative pulmonary impairment. The ASA risk classification is a gauge of general status, and is highly predictive of both cardiac and pulmonary complications.75,76 Although advanced age is associated with increased incidence of chronic pulmonary disease and underlying impairment, it is not an independent risk factor for pulmonary complications.

Clearly, all smokers should be urged to stop before surgery. Even in the absence of coexisting pulmonary disease, smoking increases the risk of perioperative complications. Smoking confers a relative risk of 1.4–4.3, but a reduced risk of pulmonary complications has been shown in patients who stop smoking at least 8 weeks before cardiac surgery.77 Even 48 hours of abstinence can improve mucociliary clearance, decrease carboxyhemoglobin levels to those of nonsmokers, and reduce the cardiovascular effects of nicotine. A nicotine patch may help some patients with postoperative nicotine withdrawal, but may not be advisable in patients at risk for poor wound healing.

COPD confers a relative risk of 2.7–4.7 in various studies. Symptoms of bronchospasm and obstruction should be addressed before surgery and elective procedures should be deferred in patients having an acute exacerbation. Preoperative treatment may include bronchodilators, antibiotics, steroids, and physical therapy to increase exercise capacity. Patients with active pulmonary infections should have surgery delayed if possible. Asthmatics should have peak flow equivalent to their personal best or 80% of predicted, and should be medically optimized to achieve this goal. Pulse corticosteroids may be used without an increased risk of postoperative infection or other complication.78,79

Malnourished patients may not be able to meet the demands of the increased work of breathing, increasing their risk for respiratory failure. Obese patients have higher rates of oxygen consumption and carbon dioxide production, which increases their work of breathing. They may also exhibit restrictive physiology due to a large, stiff chest wall. A complete history should inquire about sleeping difficulty and snoring. Obesity increases the amount of soft tissue in the oropharynx, which can cause upper airway obstruction during sleep. Fifty-five percent of morbidly obese patients may have sleep-related breathing disorders such as obstructive sleep apnea and obesity-hypoventilation syndrome.80 Symptoms include snoring and daytime sleepiness, and formal sleep studies are employed for definitive diagnosis. Sleep-disordered breathing is associated with hypoxia, hypercapnia, changes in blood pressure, nocturnal angina, and increased cardiac morbidity and mortality including stroke and sudden death.81 Arterial blood gas with partial arterial oxygen pressure less than 55 mm Hg or partial arterial carbon dioxide pressure greater than 47 mm Hg confirms the diagnosis. An increased incidence of pulmonary hypertension and right-sided heart failure is seen in patients with obesity hypoventilation syndrome and these patients should have an echocardiogram before surgery. In severe cases, intraoperative monitoring with a pulmonary artery catheter may be prudent.

In the patient who is awake, postoperative care should include coughing and deep breathing exercises, and in nonambulatory patients, early mobilization should include turning every 2 hours. Early ambulation prevents atelectasis and pooling of secretions, and increases the ventilatory drive. Upright position distributes blood flow and minimizes shunting. Preoperative medications should be resumed expeditiously. Incentive spirometry and pulmonary toilet are pulmonary expansion maneuvers, which reduce the relative risk of pulmonary complications by 50%.82 Patients should receive preoperative education about these techniques. Inhaled ipratropium and beta-agonists, used together, may prevent postoperative wheezing and bronchospasm, and should be prescribed in patients at risk. Intermittent positive pressure ventilation and nasal bi-level positive airway pressure may be enlisted for secondary prevention. Epidural analgesia is superior to parenteral narcotics in abdominal and thoracic procedures for preventing pulmonary complications.

Stress ulceration has been a well-recognized complication of surgery and trauma since 1932, when Cushing reported gastric bleeding accompanying head injury. With later research in gastric physiology and shock, it has been recognized that the appearance of gastric erosion results from failure of the protective function of gastric mucosa and back diffusion of hydrogen ion, enabling gastric acid to injure the mucosa. Once the mucosa is injured, the defenses are further weakened, leading to further injury in a vicious cycle. The protective functions of the mucosa rely on the stomach's rich blood flow to maintain high oxygen saturation. The most critical factor in the development of erosive ulceration now appears to be mucosal ischemia. Once the rich blood supply of the mucosa is compromised, the protective mechanisms are impaired, and gastric acid causes erosion, bleeding, and perforation.

In the late 1970s, the incidence of gastric bleeding in critically ill patients was 15%. Recognition of the importance of organ perfusion has resulted in decreased rates of erosive stress gastritis. Factors often cited for this observation are: improvement in resuscitation and monitoring technology, nutritional support, and effective agents for medical prophylaxis. The prophylactic medicines are targeted to reduce gastric acidity. Antacids have been shown to provide effective protection against erosive ulceration; however, there is some risk of aspiration pneumonia. Antagonists of the histamine-2 (H2) receptors of the parietal cells impair gastric acid secretion and also are effective prophylaxis for erosive ulceration. Due to ease of use, H2-blockers have become the mainstay of stress ulcer prophylaxis in abdominal surgery.83

In the setting of elective operations when the patients are not critically ill, the incidence of stress ulceration is now very low and routine use of ulcer prophylaxis medication has been questioned. In addition, the routine use of H2-antagonists in this setting may lead to increased risk of pneumonia because of failure of the gastric juices to kill bacteria.

Postoperative Ileus

Ileus is a condition of generalized bowel dysmotility that frequently impairs feeding in the postoperative setting. Ileus typically occurs after abdominal surgery, even if the bowel itself is not altered. It has been shown that laparotomy alone, without intestinal manipulation, leads to impaired gastrointestinal motility. The small bowel is typically affected the least, and can maintain organized peristaltic contractions throughout the perioperative period. The stomach usually regains a normal pattern of emptying in 24 hours, and the colon is last to regain motility, usually in 48–72 hours.

The exact mechanism that causes postoperative ileus is not known; however, physiologic studies have demonstrated the significant contribution of both inhibitory neural reflexes and local mediators within the intestinal wall. Inhibitory neural reflexes have been shown to be present within the neural plexuses of the intestinal wall itself, and in the reflex arcs traveling back and forth from the intestine to the spinal cord. These neural pathways may account for the development of ileus during laparotomy without bowel manipulation. In addition, inflammatory mediators such as nitric oxide are present in manipulated bowel and in peritonitis and may play a role in the development of ileus.

Ileus can be recognized from clinical signs, such as abdominal distension, nausea, and the absence of bowel sounds and flatus, which should prompt the diagnosis. Abdominal x-ray imaging typically shows dilated loops of small bowel and colon. Bowel obstruction must also be considered with these clinical findings, however, and CT or other contrast imaging may be required to rule out obstruction.

Ileus can also appear following nonabdominal surgery, and can result from the effects of medications (most often narcotics), electrolyte abnormalities (especially hypokalemia), and a wide variety of other factors.

Occasionally, the patient sustains a prolonged period of postoperative ileus. This can be due to a large number of contributing factors, such as intra-abdominal infection, hematoma, effects of narcotics and other medications, electrolyte abnormalities, and pain. In addition, there can be prolonged dysmotility from certain bowel operations, such as intestinal bypass.

The role of laparoscopic surgery in prevention of ileus is controversial. In theory, with less handling of the bowel laparoscopically and with smaller incisions, there should be less stimulation of the local mediators and neural reflexes. Animal studies comparing open and laparoscopic colon surgery indicate earlier resumption of normal motility studies and bowel movements with the laparoscopic approach. Human trials have not been conclusive. Several series demonstrate earlier tolerance of postoperative feeding with the laparoscopic approach to colon resection; however, these have been criticized for selection bias, and such studies are impossible to conduct in a blind fashion.

Early mobilization has long been held to be useful in prevention of postoperative ileus. While standing and walking in the early postoperative period have been proven to have major benefits in pulmonary function and prevention of pneumonia, mobilization has no demonstrable effect on postoperative ileus.

In the expected course of uncomplicated abdominal surgery, the stomach is frequently drained by a nasogastric tube for the first 24 hours after surgery, and the patient is not allowed oral intake until there is evidence that colonic motility has returned, usually best evidenced by the passage of flatus. Earlier feeding and no gastric drainage after bowel surgery can be attempted for healthy patients undergoing elective abdominal surgery, and has a high rate of success provided clinical symptoms of ileus are not present. In such patients, the use of effective preventive strategies is highly effective. These include maintenance of normal serum electrolytes, use of epidural analgesia, and avoidance of complications such as infection and bleeding. The routine use of nasogastric tubes for drainage in the postoperative period after abdominal surgery has come into question since the mid-1990s.

The most effective strategy for management of postoperative ileus following abdominal surgery has been the development of epidural analgesia. Randomized trials have shown that the use of nonnarcotic (local anesthetic–based) epidural analgesia at the thoracic level in the postoperative period results in a decreased period of postoperative ileus in elective abdominal surgery. Ileus reduction is not seen in lumbar level epidural analgesia, suggesting that inhibitory reflex arcs involving the thoracic spinal cord may play a major role in postoperative ileus.

Narcotic analgesia, while effective for postoperative pain, has been shown to lengthen the duration of postoperative ileus, especially when used as a continuous infusion or as PCA. Patients report better control of postoperative pain with continuous infusion or PCA as compared to intermittent parenteral dosing. Many studies have been done comparing various types of opioid analgesics, in attempts to find a type that does not prolong ileus. There has been no clearly superior drug identified; all currently available opioids cause ileus. Opioid antagonists such as naloxone have been used in trials to decrease ileus in chronic narcotic use, and there is evidence that antagonists are effective in that setting; however, in postoperative ileus, the antagonists have not been shown to be clinically useful, again suggesting that other mechanisms are contributing to postoperative ileus.

Early Postoperative Bowel Obstruction

Early postoperative bowel obstruction refers to mechanical bowel obstruction, primarily involving the small bowel, which occurs in the first 30 days following abdominal surgery. The clinical picture may frequently be mistaken for ileus, and these conditions can overlap. The clinical presentation of early postoperative bowel obstruction is similar to that of bowel obstruction arising de novo: crampy abdominal pain, vomiting, abdominal distention, and obstipation. The incidence of early postoperative bowel obstruction has been variable in published series, due to difficulty in differentiating ileus from early postoperative bowel obstruction, but the reported range is from 0.7% to 9.5% of abdominal operations.

Retrospective large series show that about 90% of early postoperative bowel obstruction is caused by inflammatory adhesions. These occur as a result of injury to the surfaces of the bowel and peritoneum during surgical manipulation. The injury prompts the release of inflammatory mediators that lead to formation of fibrinous adhesions between the serosal and peritoneal surfaces. As the inflammatory mediators are cleared and the injury subsides, these adhesions eventually mature into fibrous, firm, and bandlike structures. In the early postoperative period, the adhesions are in their inflammatory, fibrinous form, and as such do not usually cause complete mechanical obstruction.

Internal hernia is the next most common cause of early postoperative bowel obstruction, and can be difficult to diagnose short of repeat laparotomy. Internal hernia occurs when gaps or defects are left in the mesentery or omentum, or blind gutters or sacs are left in place during abdominal surgery. The typical scenario is colon resection involving extensive resection of the mesentery for lymph node clearance. If the resulting gap in the mesentery is not securely closed, small bowel loops may go through the opening and not be able to slide back out. A blind gutter may be constructed inadvertently during the creation of a colostomy. When the colostomy is brought up to the anterior abdominal wall, there is a space between the colon and the lateral abdominal wall, which may also trap the mobile loops of small bowel. Defects in the closure of the fascia during open or laparoscopic surgery can cause obstruction from incarcerated early postoperative abdominal wall hernia. Fortunately, internal hernia is a rare occurrence in the early postoperative period; however, it must be suspected in cases where bowel anastomoses or colostomies have been constructed. Unlike adhesive obstruction, internal hernia requires operative intervention due to the high potential for complete obstruction and strangulation of the bowel.

Intussusception is a rare cause of early postoperative bowel obstruction in adults, but occurs more frequently in children. Intussusception occurs when peristalsis carries a segment of the bowel (called the lead point) up inside the distal bowel like a rolled up stocking. The lead point is usually abnormal in some way, and typically has some intraluminal mass, such as a tumor or the stump of an appendix after appendectomy. Other rare causes for early postoperative bowel obstruction include missed causes of primary obstruction at the index laparotomy, peritoneal carcinomatosis, obstructing hematoma, and ischemic stricture.

Management of early postoperative bowel obstruction depends on differentiation of adhesive bowel obstruction (the majority) from internal hernia and the other causes, and from ileus. Clinicians generally rely on radiographic imaging to discern ileus from obstruction. For many years, plain x-ray of the abdomen was used: if the abdominal plain film showed air-distended loops of bowel and air-fluid levels on upright views, the diagnosis of obstruction was favored. However, plain radiographs can be misleading in the postoperative setting, and the overlap of ileus and obstruction can be confusing. Upper GI contrast studies using a water-soluble agent has better accuracy, and abdominal CT using oral contrast has been shown to have 100% sensitivity and specificity in differentiating early postoperative bowel obstruction from postoperative ileus.

Once the diagnosis is made, management is tailored to the specific needs of the patient. Decompression via nasogastric tube is usually indicated, and ileus can be treated as discussed. Adhesive bowel obstruction warrants a period of expectant management and supportive care, as the majority of these problems will resolve spontaneously. Most surgical texts recommend that the waiting period can be extended to 14 days. If the early bowel obstruction lasts longer than 14 days, less than 10% resolve spontaneously, and exploratory laparotomy is indicated. The uncommon causes of early postoperative bowel obstruction, such as internal hernia, require more early surgical correction, and should be suspected in the setting of complete obstipation, or when abdominal CT suggests internal hernia or complete bowel obstruction.

Renal Evaluation

Patients without a clinical history suggesting renal disease have a low incidence of significant electrolyte disturbances on routine preoperative screening.84 However, those patients with renal or cardiac disease who are taking digitalis or diuretics, or those with ongoing fluid losses (ie, diarrhea, vomiting, fistula, and bleeding) do have an increased risk of significant abnormalities and should have electrolytes measured and replaced preoperatively.

Preoperative urinalysis can be a useful screen for renal disease. Proteinuria marks intrinsic renal disease or congestive heart failure. Urinary glucose and ketones are suggestive of diabetes and starvation in the ketotic state, respectively. In the absence of recent genitourinary instrumentation, microscopic hematuria suggests calculi, vascular disease, or infection. A few leukocytes may be normal in female patients, but an increased number signifies infection. Epithelial cells are present in poorly collected specimens.

Patients with renal insufficiency or end-stage renal disease often have comorbidities that increase their overall risk in the perioperative period. Hypertension and diabetes correlate with increased risk of coronary artery disease and postoperative MI, impaired wound healing, wound infection, platelet dysfunction, and bleeding. Preoperative history should note the etiology of renal impairment, preoperative weight as a marker of volume status, and timing of last dialysis and the amount of fluid removed routinely. Evaluation should include a cardiac risk assessment. Physical examination should focus on signs of volume overload such as jugular venous distention and pulmonary crackles. In patients with clinically evident renal insufficiency, a full electrolyte panel (calcium, phosphorus, magnesium, sodium, and potassium) should be checked preoperatively, along with blood urea nitrogen and creatinine levels. Progressive renal failure is associated with catabolism and anorexia. Such patients need aggressive nutritional support during the perioperative periods to minimize the risk of infection and poor healing.

Dialysis-dependent patients should have dialysis within 24 hours before surgery, and may benefit from monitoring of intravascular volume status during surgery. Blood samples obtained immediately after dialysis, before equilibration occurs, should only be used in comparison to predialysis values to determine the efficacy of dialysis.85

Postoperatively, patients with chronic renal insufficiency or end-stage renal disease will need to have surgical volume losses replaced, but care should be taken to avoid excess. Replacement fluids should not contain potassium, and early dialysis should be employed to address volume overload and electrolyte derangements. Patients with impaired creatinine clearance should have their medications adjusted accordingly. For example, meperidine should be avoided because its metabolites accumulate in renal impairment and can lead to seizures.

The choice of postoperative fluid therapy depends on the patient's comorbidities, the type of surgery, and conditions that affect the patient's fluid balance. There is no evidence that colloid is better than crystalloid in the postoperative period, and it is considerably more expensive.86 Sepsis and bowel obstruction will require ongoing volume replacement rather than maintenance. Ringer's solution provides six times the intravascular volume as an equivalent amount of hypotonic solution. In patients with normal renal function, clinical signs such as urine output, heart rate, and blood pressure should guide fluid management. Once the stress response subsides, fluid retention subsides and fluid is mobilized from the periphery, and fluid supplementation is unnecessary. This fluid mobilization is evident by decreased peripheral edema and increased urine output. Diuretics given in the period of fluid sequestration may cause intravascular volume depletion and symptomatic hypovolemia.

Postoperative management includes close monitoring of urine output and electrolytes, daily weight, elimination of nephrotoxic medications, and adjustment of all medications that are cleared by the kidney. Hyperkalemia, hyperphosphatemia, and metabolic acidosis may be seen and should be addressed accordingly. Indications for renal replacement therapy include severe intravascular overload, symptomatic hyperkalemia, metabolic acidosis, and complicated uremia (pericarditis and encephalopathy) (Table 2-5).

Postoperative renal failure increases perioperative mortality. Risk factors for postoperative renal failure include intraoperative hypotension, advanced age, congestive heart failure, aortic cross-clamping, administration of nephrotoxic drugs or radiocontrast, and preoperative elevation in renal insufficiency. Up to 10% of patients may experience acute renal failure after aortic cross-clamping. Postoperative renal failure rates are higher in hypovolemic patients, so preoperative dehydration should be avoided. Contrast nephropathy is a common cause of hospital-acquired renal failure, and manifests as a 25% increase in serum creatinine within 48 hours of contrast administration.

Nephropathy is caused by ischemia and direct toxicity to the renal tubules. Diabetes and chronic renal insufficiency are the greatest risk factors for dye nephropathy. Recent trials87 have shown that patients receiving contrast have a lower incidence of contrast-induced nephropathy when treated with a sodium bicarbonate infusion. N-acetylcysteine given orally on the day prior to contrast exposure also decreases the incidence of radiocontrast nephropathy.88

Rising blood urea nitrogen and creatinine and postoperative oliguria (<500 mL/d) herald the onset of postoperative renal failure. Management is determined by the cause of renal insufficiency. Acute renal failure is classified into three categories: prerenal, intrarenal, and postrenal. Prerenal azotemia is common in the postoperative period. It is caused by decreased renal perfusion seen with hypotension and intravascular volume contraction. Intrarenal causes of oliguric renal failure include acute tubular necrosis (from aortic cross-clamping, shock, or renal ischemia), and less commonly, acute interstitial nephritis from nephrotoxic medication. Postrenal causes include obstruction in the collecting system (from bilateral ureteral injury, Foley catheter occlusion, or urethral obstruction). Workup should include urinalysis, serum chemistries, and measurement of the fractional excretion of sodium. Invasive monitoring and cardiac echocardiogram may be employed to evaluate volume status. Ultrasound is indicated if obstruction is suspected.

Initial management of oliguria in adults includes placement of a bladder catheter, and a challenge with isotonic fluids (500 mL of normal saline or Ringer's lactate). If a bladder catheter is already present, it should be checked to ensure that it is draining properly. A urinalysis should be obtained with special attention to specific gravity, casts, and evidence of infection. Hematocrit should be evaluated to exclude bleeding and blood pressure measured to rule out hypotension as causes. The fractional excretion of sodium can help determine the etiology of the renal failure (Table 2-6). Serum creatinine is used to follow the course of acute renal failure. Patients who have been adequately resuscitated or who are in CHF require evaluation to rule out cardiogenic shock. Urinary retention can be treated with a Foley catheter, and ureteral obstruction can be addressed with percutaneous nephrostomy.

Intravascular volume depletion adversely affects cardiac output, tissue perfusion, and oxygen delivery. Monitoring includes total body weight, urine output, vital signs, and mental status. However, body weight should not be used alone because total volume overload can be seen in the setting of intravascular volume depletion. Most cases of postoperative renal failure are associated with an episode of hemodynamic instability,47 and perioperative hemodynamic optimization has been shown to decrease acute kidney injury and mortality.89 Invasive monitoring to measure cardiac filling pressures may be utilized when clinical assessment is unreliable.

Fluid overload may be seen in patients with renal, hepatic, and cardiac disease, and is associated with increased morbidity.90 Critically ill patients may develop anasarca. It is difficult to determine volume status by observation alone, and invasive monitoring may be required.

Electrolyte abnormalities are common in the perioperative period. Serum sodium reflects intravascular volume status. Hyponatremia signifies excess free water in the intravascular space, and is caused by excess antidiuretic hormone in the postoperative period. It occurs in the setting of normo-, hypo-, or hypervolemia. It may be avoided by judicious use of isotonic fluids. Conversely, hypernatremia suggests a relative deficit of intravascular free water. Patients who are unable to drink, or those with large insensible losses, are most at risk. Treatment includes free water replacement.

Diuretics, malnutrition, and gastrointestinal losses may cause postoperative hypokalemia. Metabolic alkalosis shifts potassium into the intracellular compartment. Serum potassium levels less than 3 mEq/L warrant ECG monitoring and replacement in patients who are not anuric. Replacement in patients with renal insufficiency may be complex. Hyperkalemia is more commonly seen in renal patients. It may also be seen in myonecrosis, hemolysis, and acidosis. Cardiac arrhythmias are seen at levels above 6.5 mEq/L and death is associated with levels greater than 8 mEq/L. These patients should have cardiac monitoring until their levels normalize. ECG will show widened QRS interval, peaked T waves, and absent P waves. Hyperkalemia should be treated with sodium bicarbonate to stimulate acidosis, as well as intravenous calcium and insulin with glucose to drive potassium into the intracellular compartment. Cation exchange resins can be administered orally or per rectum to bind ions in the gastrointestinal tract, but care should be taken for the patient who is post–GI surgery or has underlying gastrointestinal problems. Dialysis can by employed if other measures fail.

Hyperglycemia is a risk factor for postoperative infection and perioperative mortality. Intensive insulin therapy (IIT) has been associated with improved outcomes for intensive care unit patients, and after cardiac surgery, in brain injury and after acute MI. However, early enthusiasm for IIT has waned as more recent studies have shown that it is not as beneficial in medical patients as it is in surgical patients, and has been associated with severe hypoglycemia.91 More recently, a meta-analysis of 29 randomized trials of IIT in adult ICU patients showed no difference in mortality in patients receiving IIT versus conventional insulin therapy with goal blood sugar <200 mg/dL.92

Although there does appear to be consensus that controlling glucose is a worthwhile therapeutic goal in surgical patients in particular, appropriate targets for control remain controversial. While hyperglycemia is associated with increased infection and mortality,93 IIT is associated with hypoglycemia and increased mortality. Results from a recent international, randomized controlled trial in ICU patients demonstrated a 2.6% increase in absolute risk of death in ICU patients with blood glucose target of 81–108 mg/dL versus 180 mg/dL.94 Others suggest that the variability in glucose level may affect morbidity and mortality more than blood glucose levels alone.95 More investigation is needed to determine the optimal way to manage blood glucose levels in the postoperative patient.

Our current recommendations for glucose control in noncardiac surgery patients is to maintain blood glucose less than 180 mg/dL.

A complete preoperative evaluation should include assessment of hematological disorders, which can increase the risk for postoperative bleeding or thromboembolism. Patients should be asked about a family history of bleeding disorders and personal history of bleeding problems, especially after procedures. Excessive bleeding after dental procedures and menorrhagia in women can alert the physician to undiagnosed hematologic disease. Risk factors for postoperative hemorrhage include known coagulopathy, trauma, hemorrhage, or potential factor deficiency.96 Factor deficiencies can be seen with a history of liver disease, malabsorption, malnutrition, or chronic antibiotic use. Even high-risk patients have only a 1.7% risk of postoperative hemorrhage and a 0.21% risk of death related to postoperative hemorrhage.96,97

Routine tests may include a complete blood count, prothrombin time (PT), activated partial thromboplastin time (PTT), and INR, but are not required in the asymptomatic patient with no associated history. The complete blood count will reveal leukocytosis, anemia, and thrombocytopenia or thrombocytosis. A baseline hematocrit is useful for postoperative management when anemia is suspected. Platelet count also provides a useful baseline, but does not provide information about platelet function. A bleeding time may be required to provide more information in select patients. However, bleeding time results are operator dependent and highly variable, making it a poor screening tool for identifying high-risk patients.98,99 An abnormal bleeding time is not associated with increased postoperative bleeding,100 nor has it proven useful in identifying patients taking nonsteroidal anti-inflammatory medication or aspirin.98 None of the aforementioned tests can be used to diagnose hereditary bleeding disorders. However, an elevated PTT may be seen in factor XI deficiency, and should be obtained in patients at risk for this deficiency. Low-risk patients are very unlikely to have bleeding complications even if the PTT is abnormal,99 and have an increased risk of false-positives that can lead to unnecessary testing. PTT is not a reliable predictor of postoperative bleeding,101 and should not be used to screen for bleeding abnormalities in patients without symptoms or risk factors.102,103

A platelet count of 20,000 or more is usually adequate for normal clotting. Aspirin causes irreversible impairment of platelet aggregation, and is commonly prescribed in patients at risk of cardiovascular and cerebrovascular disease. The clinical effect of aspirin lasts 10 days, and it is for this reason that patients are asked to stop taking aspirin 1 week before elective surgery. Desmopressin can be used to partially reverse platelet dysfunction caused by aspirin and uremia. Other NSAIDs cause reversible platelet dysfunction and should also be held before surgery. Glycoprotein IIb-IIIa inhibitors prevent platelet-fibrin binding and platelet aggregation, and are used for 2–4 weeks after coronary angioplasty. Elective surgery should be avoided during these 2–4 weeks, as stopping treatment increases the risk of thrombosis. Patients who do not receive 4 weeks of antiplatelet therapy are at risk of stent thrombosis.104

Indications for red blood cell transfusion remain somewhat controversial and are often empirical in practice. Transfusing one unit of red blood cells or whole blood can increase the hematocrit by approximately 3% or hemoglobin by 1 g/dL. Multiple studies have demonstrated that overutilizing transfusion may adversely affect patient outcome and increase risk of infection. ASA guidelines105 suggest that transfusion should be based on risks of inadequate oxygenation, rather than a threshold hemoglobin level. Generally, transfusion is rarely indicated when the hemoglobin level exceeds 10 g/dL, but is almost always indicated when it is less than 6 g/dL, especially in the setting of acute anemia. Healthy individuals can usually tolerate up to 40% of blood loss without requiring blood cell transfusion, and blood products should not be used solely to expand volume or to improve wound healing. The decision to transfuse red cells or whole blood should be based on the patient's risk of complications associated with impaired oxygen delivery, including hemodynamic indices, history of cardiopulmonary disease, rate of blood loss, and preexisting anemia.

Conditions associated with abnormal platelets and low platelet counts can be treated with platelet transfusions. The usual dose, one unit of platelet concentrate per 10 kg body weight, can be expected to increase the platelet count by approximately 5,000–10,000 in an average adult. In patients without increased risk of bleeding, prophylactic platelet administration is not indicated until counts fall below 20,000. Higher thresholds may be indicated for patients at increased risk of bleeding, known platelet dysfunction, and microvascular bleeding. Desmopressin can augment platelet function in uremia and incite release of Von Willebrand's factor (VWF) from the endothelium, which can improve platelet function. The decision to transfuse platelets should be based on the amount of bleeding expected, the ability to control bleeding, and the presence of platelet dysfunction or destruction.

Transfusion of fresh frozen plasma (FFP) is indicated to reverse warfarin before procedures or in the presence of active bleeding, for inherited or acquired coagulopathy that can be treated with FFP, and for massive transfusion of more than one whole blood volume. Microvascular bleeding can be seen if the PT/PTT is greater than 1.5 times normal, and FFP can be used to reverse bleeding in this setting. Warfarin reversal can be achieved with doses of 5–8 mL/kg, and 30% factor concentration can be achieved with 10–15 mL/kg. FFP should not be used to address volume depletion alone. Cryoprecipitate contains factors VIII, VWF, XIII, fibrinogen, and fibronectin, and can be used preventively in patients with these factor deficiencies and uremia.

Endothelial injury and venous stasis are the greatest risk factors for VTE. The patient with hereditary thrombophilia, or a personal history of VTE, cancer, or recent surgery (within 4 weeks) has an increased risk of VTE.106 Preventive measures include external pneumatic leg compression, early mobilization after surgery, and anticoagulation. Compression devices are contraindicated in patients with severe peripheral vascular disease, venous stasis, or risk of tissue necrosis. Inferior vena cava (IVC) filters are indicated in patients who cannot take anticoagulation or who have failed anticoagulation therapy. Patients with a history of VTE benefit from IVC filter placement in the short term, but IVC filter placement is accompanied by an increased incidence of deep venous thrombosis over the long term.107 Systemic anticoagulation is the preferred long-term option. LMWH and UFH are equally effective for prevention of pulmonary embolism in patients with deep venous thrombosis.107 Recent VTE, atrial fibrillation, and mechanical heart valves are common indications for warfarin treatment.

Clinically, UFH activity is measured by PTT and the therapeutic goal is usually 2.0–2.5 times normal. LMWH is a relatively stronger inhibitor of factor Xa and does not have the same effect on the PTT. The anticoagulant effect of LMWH is measured by factor Xa activity. Protamine can reverse the effects of heparin, but may cause allergic reactions and induce hypercoagulability, and should be used cautiously. FFP will not reverse heparin, and can actually increase heparin activity because it contains antithrombin III. Direct thrombin inhibitors can also prolong the PTT. Direct thrombin inhibitors are not reversible with protamine and may require large amounts of FFP for reversal.

Heparin can be used for the prevention and treatment of VTE. Surgical patients over age 40 or those at increased risk for VTE should receive 5,000 U SC every 8–12 hours, depending on their weight. High-risk patients with a history of VTE, cancer, morbid obesity, or those having orthopedic procedures should either receive SC heparin with a goal of high range of normal or LMWH. In the event of acute VTE intravenous heparin should be started promptly with a therapeutic goal of PTT 1.5–2.0 times normal. Oral anticoagulation should be started within 24 hours and continued for 3–6 months.106

Heparin-induced thrombocytopenia (HIT) is a potentially lethal complication of heparin therapy. HIT is caused by an IgG mediated hypersensitivity reaction between the heparin moiety and platelet factor 4 (PF4). Patients with previous heparin exposure, such as orthopedic and cardiac surgical patients, are at greatest risk. The incidence of HIT is 0.5–5.0% in patients receiving UFH. HIT occurs with UFH or low molecular weight heparin; the risk is highest with UFH.

Platelet counts usually drop 40–50% from baseline. Thrombosis can be venous or arterial leading to deep vein thrombosis, extremity ischemia, and mesenteric ischemia of stroke. Digital ischemia and skin necrosis can also be seen. HIT remains a clinical syndrome which can be diagnosed by a decrease in platelet count <405 of baseline in 4–14 days of heparin administration once other causes of thrombocytopenia have been ruled out. The diagnosis can be supported by the ELISA assay for antiplatelet antibodies.

Because HIT can be life-threatening, heparin should be stopped as soon as HIT is suspected, and treatment with an alternative anticoagulant, such as the thrombin inhibitor bivalrudin, should be started immediately. Platelets should return to baseline after therapy is initiated. If thrombosis is present, patients should be anticoagulated for 6 months with Coumadin. Coumadin should not be started until platelet counts have recovered.

Warfarin inhibits synthesis of vitamin K–dependent clotting factors (II, VII, IX, X, and proteins C and S). Poor diet, prolonged antibiotic use, and fat malabsorption can also cause vitamin K deficiency and cause abnormal coagulation. Liver disease can lead to multiple coagulation abnormalities including factor deficiencies, vitamin K deficiency, fibrinolysis, and elevated levels of fibrin degradation products. All patients with known or suspected liver disease should be tested for coagulopathy. Vitamin K can be administered subcutaneously or intravenously in deficient patients. The initiation of warfarin therapy is associated with a transient thrombotic state because plasma concentrations of protein C fall approximately 24 hours before concentrations of other clotting factors.

Heparin is the drug of choice for VTE during pregnancy because it does not cross the placenta. Adverse effects of heparin therapy may include hemorrhage, thrombocytopenia, and osteoporosis. HIT is an immune disorder seen in patients with prior exposure to heparin, which may cause thrombosis. Treatment includes cessation of heparin and utilization of alternative anticoagulants such as lepirudin, danaparoid, or argatroban. These should be given until platelet counts recover.

For patients on long-term anticoagulation therapy, the INR should be 1.5 or lower before elective surgery. After warfarin is discontinued, it takes about 4 days for an INR in the range of 2.0–3.0 to spontaneously reach 1.5, and about 3 days for the INR to reach 2.0 after it is restarted. If therapy is withheld preoperatively, most patients will have a window of 2–4 days when they are not anticoagulated and at risk for venous thrombosis. This risk is compounded by the increased risk of thromboembolism associated with surgery.108,109 It has been estimated that surgery increases the risk of VTE by 100-fold in patients with recurrent disease.110 Without anticoagulation, there is a 50% chance of recurrence within the 3 months after the first episode of venous thrombosis. Warfarin therapy reduces the risk to 10% after 1 month and 5% after 3 months. It is not advisable to interrupt anticoagulation within 1 month after an event of VTE, and if possible, surgery should be deferred until the patient has completed 3 months of therapy.110 Chronic anticoagulation lowers the risk of thromboembolism in patients with atrial fibrillation and mechanical heart valves by 66% and 75%, respectively.110

Patients with prior embolic episodes are at increased risk for recurrence. Six percent of episodes of VTE and 20% of arterial thromboembolism may be fatal,110 and a significant percentage cause disability. Alternatively, the risk of death after postoperative hemorrhage is less than 1%,111 so the judicious use of postoperative anticoagulation can be relatively protective. Preoperative heparinization is not required during the second and third months of warfarin treatment for deep vein thrombosis (DVT) because the risk is sufficiently low. Such patients have increased VTE risk after surgery and should receive postoperative anticoagulation. Patients who are at risk for recurrent DVT, and are within 2 weeks of the first episode, or who cannot tolerate anticoagulation are candidates for an IVC filter.107

Elective surgery should be deferred for the first month after arterial embolism because of the high risk of recurrence during this period. If necessary, patients should receive perioperative heparin while oral anticoagulation is held. Patients on long-term anticoagulation to prevent arterial thromboembolism do not need perioperative heparin because the risk of bleeding outweighs the risk of arterial embolism during this period.

Heparin should be titrated to a goal PTT of 1.5–2.0 times normal and given as a continuous intravenous infusion. It should be stopped 6 hours prior to a procedure, and can be restarted 12 hours after surgery if there was no evidence of bleeding at the end of the case. Heparin can be restarted without a bolus at the anticipated maintenance infusion rate.110,111

Infectious complications can be most unwelcome and difficult to control after major abdominal surgery, yet they are surprisingly frequent despite all modern prophylactic measures. Reported surgical wound infection rates in elective operations vary from 2% for inguinal hernia repair112 to 26% for colectomy,113 and are even higher for emergency surgery.114 Surgical site infections (SSIs) increase overall mortality and morbidity, and increase hospital length of stay and overall costs. Therefore prevention and treatment of infectious complications should be included in surgical decision making for all abdominal procedures.

Prevention of SSIs begins with preoperative evaluation and identification of patients at high risk for SSI. Patient factors implicated in risk of SSI include age, diabetes mellitus, smoking, steroid use, malnutrition, obesity, active distant infection, prolonged hospital stay, and nasal colonization with Staphylococcus aureus.115–118

Standard basic surgical rules should be followed with every patient. These were codified as formal guidelines by the Centers for Disease Control and Prevention (CDC) in 1999119 and include recommendations for skin preparation with alcohol or iodophor, surgical barriers such as drapes and gowns, careful hand scrubbing, and appropriate selection of prophylactic antibiotics. Preoperative hair removal and antiseptic shower have not been shown to decrease SSI rates, and shaving and clipping of hair can increase SSIs. The CDC recommendations are summarized in Table 2-7. (See Table 2-3 for extended recommendations.)

Antibiotic prophylaxis may be indicated for patients at high risk, or in contaminated surgical procedures, but antibiotics should not be used indiscriminately. Overuse of antibiotics is associated with emergence of multidrug-resistant bacteria and increased rates of hospital-acquired infections. Selection of patients for antimicrobial prophylaxis requires stratification of patient risk factors as discussed above and procedure-specific risk factors. The degree of contamination in the surgical site has long been recognized as an independent risk factor for SSI,120 leading to the wound classification system (Table 2-8) in use since 1983.

Patients undergoing class I (clean) procedures have a very low infection rate and generally do not benefit from prophylactic antibiotics, unless there is some suspicion at the start of the procedure that some contamination may occur, such as unplanned enterotomy in a patient with many previous abdominal procedures. In addition, many surgeons prefer to use antibiotic prophylaxis in class I procedures when a prosthesis is implanted; examples include hernia repair and vascular bypass. In this setting, the risk of SSI is low, but the morbidity and mortality of an infected prosthesis are great, and prophylaxis may decrease the risk. To date, large prospective trials have not shown benefit of antibiotic prophylaxis in preventing prosthetic infections,121,122 but smaller trials have suggested a decrease in site infection without change in implant infection rate.123,124 Therefore, there is no strict guideline for the use of systemic antibiotics for implant surgery, and the surgeon must tailor the use of antibiotics to the individual patient's risk.

Patients with class II (clean/contaminated) surgical wounds do benefit from systemic antibiotic prophylaxis. The most studied example of this class of wound is elective colon resection. Most current guidelines recommend systemic broad-spectrum antibiotic coverage using a second-generation cephalosporin plus metronidazole if the parenteral route is used, and neomycin plus metronidazole or erythromycin base (both as nonabsorbable antibiotics), if the oral route is used.125 Published evidence supports administration of antibiotics preoperatively in order to achieve maximum therapeutic levels at the time of incision, and repeat dosing to maintain therapeutic levels during a long procedure. There is no documented study showing benefit to additional doses of antibiotics after the procedure is over and the skin is closed, and prolonged use of prophylactic antibiotics contributes to emergence of resistant bacteria.126,127

Patients with class III (contaminated) wounds are a mixed population. Some of these wounds are the result of inadvertent entry into a contaminated field, some result from traumatic injury, and some are planned operations for débridement of infected tissue. In the latter case, antibiotic therapy is indicated for specific therapy rather than prophylaxis. In the case of penetrating traumatic injury to the colon, there is strong evidence to support single-dose antibiotic prophylaxis at the time of laparotomy, similar to elective colon resection.128,129 Surgical judgment must be individualized in these cases as to whether the risk of skin closure can be justified due to the high rate of wound infection despite antibiotic prophylaxis.

Patients with class IV (dirty) wounds are generally undergoing débridement of already infected and necrotic tissue, and should be receiving antibiotic therapy targeted to the relevant organisms. Skin wound closure is generally not advised in these patients.

The wound classification system does not take into account patient risk factors or site-specific risk factors. Various physiologic scoring systems including the Acute Physiology Score and the Acute Physiology, Age, and Chronic Health Evaluation index have been used to predict perioperative infection risk with some success. In an effort to provide more accurate risk stratification, the CDC's National Nosocomial Infection Surveillance project has developed a risk index that accounts for patient risk factors such as malnutrition and chronic medical conditions, and operative factors including duration and site of procedure.130 Enlightened risk assessment of perioperative infections should be included in the discussion for informed surgical consent.

The importance of proper nutritional assessment and management cannot be overstressed. In surgical patients, malnutrition increases risk for major morbidity,131,132 including wound infection, sepsis, pneumonia, delayed wound healing, and anastomotic complications. Careful preoperative clinical assessment can identify those patients at increased nutritional risk. The assessment should include a thorough history and physical examination with attention paid to usual weight, recent weight loss, changes in eating and bowel habits, changes in abdominal girth, loss of muscle bulk, and the presence of diseases that carry a risk of malnutrition such as COPD, diabetes mellitus, inflammatory bowel disease, and psychiatric conditions such as bulimia and anorexia nervosa. The history and physical examination should identify those patients with nutritional risk; that risk can be stratified by calculation of the Nutritional Risk Index (NRI). The NRI is a simple calculation (15.19 × serum albumin (g/dL) + 41.7 × present weight/usual weight), which has been shown in prospective studies to correlate with increased rates of mortality and complications from major abdominal surgery.133,134 NRI less than 83 indicates a significantly increased rate of mortality and complications, especially wound dehiscence and infection. Severely malnourished patients have been shown to benefit from preoperative nutritional support.135,136

Malnutrition can be classified into protein deficiency (kwashiorkor), calorie deficiency (marasmus), or mixed protein calorie deficiency. In order to complete the nutritional assessment and to guide nutritional support, it is useful to classify the patient's specific nutritional state (Table 2-9). Malnutrition states are much more common than is generally acknowledged, with 30–55% of hospital inpatients meeting criteria for one of the diagnoses.137

Some interval of deficient nutritional intake is expected after an abdominal operation. In uncomplicated cases, this is usually the result of postoperative adynamic ileus and resolves promptly, in less than 7 days. Traditional surgical management includes provision of dextrose-containing intravenous fluids. The goal of this therapy is not to provide sufficient calories for complete nutritional support, but simply to provide enough carbohydrate to prevent breakdown of lean body mass. Certain organs, including the heart and brain, have an obligate requirement for carbohydrate as a primary energy source, and do not store energy in the form of fat or glycogen. If intake is insufficient to meet this requirement, the body breaks down hepatic glycogen to provide glucose to the circulation, and ultimately the brain and heart. Once hepatic glycogen stores have been depleted (after about 1 day of no intake), lean muscle mass is converted to glucose via gluconeogenesis to produce carbohydrate. Provision of only 100 g of exogenous glucose per day is sufficient to prevent breakdown of lean muscle mass in otherwise healthy subjects.

In already malnourished patients, or in patients who do not return to normal bowel function promptly, nutritional support is indicated. As in the preoperative setting, a thorough evaluation of the patient's nutritional status is necessary, as is the identification of the cause of bowel dysfunction. In the postoperative setting, there are many potential causes of bowel dysfunction (Table 2-10), and nutritional support should be individualized for each patient's needs. Some patients may respond to enteral support and some may require parenteral support. Whenever available, the enteral route is the preferred route of support, as it has been shown to cause less morbidity and mortality.138

Enteral nutritional support is effective in patients that have functional small bowel; examples include esophageal or gastric resection, patients with postoperative delirium or dysphagia, and patients who have gastroparesis. In the short term, if the dysfunction is expected to respond to treatment, nasogastric tubes can be used effectively to deliver full support. Patients that need long-term enteral support are best served with gastrostomy or jejunostomy tubes, which may be placed operatively or percutaneously. With good preoperative nutritional assessment and sound surgical judgment, these patients' needs for long-term postoperative support can often be anticipated, and long-term feeding access can be included in the operative plan. Enteral support may not be suitable for some patients; examples include early postoperative bowel obstruction, fistula, or intestinal insufficiency (short-gut syndrome). In such patients, parenteral support is indicated, and should be initiated without delay, and futile attempts to use the enteral route should be avoided.

To establish the diagnoses of mild or moderate protein calorie malnutrition, two of the five criteria shown must be met; to establish the diagnosis of severe protein calorie malnutrition, three of the seven criteria must be met.

Irrespective of the route of support, every patient on nutritional support should have his or her nutritional needs assessed and provided. The assessment begins with the calorie requirement. There are several formulas and nomograms that estimate basal energy expenditure, taking into account height, weight, age, gender, stress factors, and activity factors.139 All of these methods are estimations, and may underfeed or overfeed certain subgroups, especially the obese. The method in most common clinical use bases basal energy expenditure on adjusted body weight (ABW). Using this method, ABW is defined as the patient's ideal body weight (IBW) plus the difference between actual body weight (BW) and the IBW divided by two:

ABW = IBW + 0.5(BW – IBW)

The baseline caloric requirement for weight maintenance based on ABW is 25 kcal/kg/d. This target may be adjusted upward in patients with extreme metabolic demands, as is the case in burns or head injury.126 Furthermore, the ABW can be used to establish the protein requirement. In unstressed normal subjects, the minimum daily protein requirement is 0.8 g protein/kg/d. In postoperative patients with healing wounds, this target is adjusted to 1.0–1.5 g/kg/d, and in severely ill patients to 2.0 g/kg/d. The highest requirements are seen in severe burn and bone marrow transplant patients.

Essential nutritional components must be provided, again irrespective of the route of support. These include water- and lipid-soluble vitamins, trace elements such as zinc and selenium, essential fatty acids such as linoleic and linolenic acids, and the eight essential amino acids. These trace elements are provided in abundance in all enteral feeds, and are part of the standard additives in parenteral formula.

Once nutritional support has been initiated, the patient's response to support must be followed closely, especially in parenteral support and in patients with preexisting metabolic conditions such as diabetes. Blood glucose should be monitored regularly during the first few days of support. Recent evidence has linked hyperglycemia in the postoperative setting, especially in critically ill patients, with increased risk of death and infection.140,141 In addition, electrolyte abnormalities (especially those of potassium, magnesium, and phosphate) are often seen in the early period of nutritional support, and should be corrected.

It is also important to follow the markers of nutrition repletion to ensure that the calories and protein provided (based on the initial estimate) are sufficient, and the patient is not mobilizing lean body mass due to inadequate support. Serum markers such as prealbumin, retinol binding protein, and transferrin can be useful in this regard. They are serum proteins with short (2–7 days) turnover times that reflect the body's ability to synthesize new protein.139 Unfortunately, the serum concentrations of these proteins are also affected by acute disease states and renal and hepatic failure, and can be difficult to interpret in postoperative patients. Nitrogen balance can also be used to monitor nutritional support and reflects the ability to synthesize new protein. Nitrogen balance is calculated by subtracting nitrogen excretion from nitrogen intake. Nitrogen intake is calculated from the protein intake, where each gram of protein divided by 6.25 is equal to the number of grams of nitrogen. Nitrogen excretion has two components: urinary urea nitrogen (UUN) and insensible loss. UUN can be measured in a 24-hour urine collection; insensible loss is generally accepted to be 4 g/d, unless there is another source of loss, such as abdominal drainage of proteinaceous ascites, enterocutaneous fistula, or nephrotic syndrome. Thus, in most cases, nitrogen balance can be simplified to:

Nitrogen balance = protein intake/6.25 – 24 hour UUN   – 4g (insensible loss)

A patient that takes in more nitrogen than he or she excretes in the urine and feces is in positive nitrogen balance and is synthesizing new protein. On the other hand, a patient that is excreting more nitrogen than he or she is receiving in nutritional support is in negative nitrogen balance, and is therefore losing lean body mass, becoming more malnourished. These patients should be reevaluated for nutritional needs and for sources of nutritional depletion, such as uncontrolled diabetes mellitus, sepsis, and organ failure.

By itself, uncontrolled diabetes mellitus can be viewed as a perioperative nutritional complication, as it results in nutritional depletion, interferes with delivery of parenteral and enteral nutrition, and is associated with increased infectious morbidity.140,141