Published: January 2009
Operative treatment of disease has a tremendous yet unrecognized impact on modern medical systems. An estimated 33 million patients undergo surgery in the United States yearly. Serious adverse events occur in more than 1 million of these patients at an estimated cost of $25 billion annually. With the aging population, it is anticipated that surgical referrals will increase by 25%, costs by 50%, and costs of perioperative complications by 100%. Given these staggering numbers, it is imperative that clinicians involved with patients undergoing surgery know the basics of perioperative diagnosis and management.
Cardiovascular complications are one of the most common perioperative adverse events in patients undergoing noncardiac surgery. Although in absolute numbers they are rare, they are associated with a mortality rate as high as 70%. It is essential for clinicians to be familiar with current cardiac risk evaluation and preventive strategies for patients undergoing noncardiac surgery.
The core goals of preoperative cardiac assessment are to determine the status of the patient's cardiac conditions, to provide an estimate of risk, to determine if further testing is warranted, and to determine if interventions are warranted to reduce perioperative cardiac risk.
The prior cardiovascular history of the patient is the foundation of the perioperative assessment. Because the incidence of perioperative cardiovascular events varies according to the patient risk profile, risk of the proposed surgery, and the patient's functional capacity, all of these elements should be parts of the preoperative history.
The clinician should inquire about prior myocardial infarction (MI), congestive heart failure, valvular disease, angina, or arrhythmia. If the patient has had prior diagnostic testing or therapeutic interventions, inquire about when and where these were done and the results of such procedures. Traditional risk factors, such as hypertension, dyslipidemia, tobacco use, and diabetes, are essential elements, as well as comorbid conditions that might limit functional capacity such as peripheral vascular disease, chronic obstructive pulmonary disease, cerebrovascular disease, and renal insufficiency. Current symptoms such as chest pain at rest or on exertion, shortness of breath, claudication, syncope or presyncope, or anginal equivalent symptoms should be noted.
Functional capacity is vital information, as exercise capacity is a reliable predictor of future cardiac events. This is usually expressed in metabolic equivalents (METs), where one MET is defined as the oxygen consumption of a 70-kg man at rest. Greater than 7 METs of activity tolerance is considered excellent, whereas less than 4 METs is considered poor activity tolerance. The Duke Activity Status Index suggests questions that correlate with MET levels; for example, walking on level ground at about 4 miles per hour or carrying a bag of groceries up a flight of stairs expends approximately 4 METs of activity. Patients limited in their activity from noncardiac causes, such as severe osteoarthritis or general debility, are categorized as having poor functional capacity, because one cannot discern if significant cardiac conditions exist without the benefit of a functional study (noninvasive testing).
The degree of surgical risk contributes to a patient's risk for cardiac complications. In general, procedures that are longer and have greater potential for blood loss, hemodynamic instability, and intravascular fluid shifts carry greater risk. Procedural risk is often stratified into high (estimated mortality >5%), intermediate (mortality 1%-5%), and low (<1%) risk categories.
The physical examination serves to confirm this information and can reveal information of importance unknown to the patient. Vital signs can detect hypertension or hypotension, tachycardia or bradycardia, significant arrhythmias, or hypoxia if pulse oximetry is used. Jugular venous distention, the presence of a S3 gallop, or rales suggest decompensated heart failure. Cardiac murmurs should be noted, especially if aortic stenosis is suspected. Carotid, femoral, or abdominal bruits suggest peripheral vascular or cerebrovascular disease.
The electrocardiogram is a commonly used tool in traditional preoperative cardiac assessment, although its role in the asymptomatic patient is unclear. Incidental findings that might be significant include evidence of prior MI, conduction abnormalities such as second- or third-degree heart block, bundle branch block, and left ventricular hypertrophy suggesting hypertensive heart disease. Although the current literature notes no evidence that asymptomatic findings on the preoperative electrocardiogram affect postoperative cardiac risk, clinicians often obtain this test as a preoperative baseline for comparison in the patient with prior heart disease or with intermediate to high clinical predictors for cardiovascular events.
Since Goldman and colleagues1 created the first risk stratification tool in the late 1970s, several risk indices have been published, each with their own benefits and limitations. The most prominent are the guidelines published jointly by the American College of Cardiology and the American Heart Association in 1996, subsequently updated in 2002 (the ACC/AHA guidelines, Figure 1a and Figure 1b); the American College of Physicians guideline (ACP), and the Revised Cardiac Risk Index (RCRI). The ACC/AHA guidelines incorporate available evidence and expert consensus if no evidence is available, and they use the specific risks of surgeries; the ACP guidelines cite the available evidence only and do not make recommendations in the absence of evidence. The RCRI is a simple tool that discerns the presence of six independent predictors of major cardiovascular complications (Box 1). The authors of this tool did not make recommendations for risk reduction, but subsequent studies suggest the use of beta blockers based on RCRI score results.
|Box 1: Revised Cardiac Risk Index Criteria|
|History of myocardial infarction|
|History of or current angina|
|Use of sublingual nitroglycerine|
|Positive exercise test results|
|Q waves on electrocardiogram|
|Patients who have undergone percutaneous transluminal coronary angioplasty or coronary artery bypass graft surgery and who have chest pain presumed to be of ischemic origin|
|History of transient ischemic attack|
|History of cerebrovascular accident|
|Diabetes mellitus requiring insulin therapy|
|Chronic renal insufficiency, defined as a baseline creatinine level of at least 2.0 mg/dL (177 μmol/L)|
In patients with suspected occult coronary artery disease or with risk factors and limited functional capacity, noninvasive cardiac testing can unveil the presence of significant coronary artery disease and assess the patient's functional capacity. Treadmill stress testing, with and without thallium imaging, has been assessed in the literature and found to have good negative predictive value for perioperative cardiac events. Dipyridamole or adenosine thallium imaging can be used in patients unable to reach an adequate heart rate with physical activity and also has a comparable negative predictive value. Dobutamine echocardiography has similar risk-stratification usefulness, with the added advantage of lower cost; this test is more limited in patients with preexisting wall motion abnormalities or in the presence of bundle branch blocks. Transthoracic echocardiography may be helpful if congestive heart failure or significant valvular disease is suspected (ejection fraction <35% has been shown to predict postoperative congestive heart failure), but this should not be pursued for screening purposes. Although these tests have good negative predictive value, they have very poor positive predictive values for perioperative cardiac events; thus, a positive test is more limited in its value.
If significant coronary artery disease is seen by noninvasive testing and/or by cardiac catheterization, options for management include medical optimization or revascularization. Asymptomatic patients with prior coronary artery bypass graft surgery or percutaneous coronary interventions have lower rates of perioperative mortality and nonfatal MI compared with historical controls; this protective effect lasts approximately 4 to 6 years. However, more recent literature suggests that prophylactic revascularization, even in high-risk surgeries, does not reduce risk in patients without unstable symptoms; this might partly be due to the risks of the revascularization itself (coronary artery bypass graft surgery complications or stent thrombosis). If patients have an independent indication for revascularization, then this should be pursued and elective surgery postponed; otherwise, medical optimization should be considered as the primary means of risk reduction.
Medical means of perioperative cardiac risk reduction in recent times have predominantly focused on two classes of medications: selective beta1 antagonists (beta blockers) and 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors (statins).
Beta blockers are proved to reduce cardiac mortality, MI, and ischemia in ambulatory settings; four studies between 1996 and 2001 found that perioperative beta blockade reduced the risk of death or MI between 50% and 90%, at postoperative intervals from 28 days to up to 2 years. Subsequent randomized controlled trials of vascular surgery patients showed no clear benefit from metoprolol on major cardiac events; a meta-analysis of the trials, by Devereaux and colleagues, 2 showed a small benefit on 30-day adverse outcomes that failed to reach statistical significance.
More recently, a retrospective review by Lindenauer and colleagues 3 of more than 600,000 patients extracted from Medicare databases used the RCRI score and found that low-risk patients (RCRI = 0) did not benefit and might actually have been harmed by beta blockade, whereas higher-risk patients (RCRI ≥ 3) had significant risk reductions for in-hospital death. Auerbach and Goldman 4 suggest an algorithm for selecting patients for beta blockade (Figure 2).
Statin therapy has been evaluated in the perioperative setting by observational studies by Poldermans and colleagues 5 and by Kertai and colleagues, 6 and by a small randomized trial by Durazzo and colleagues. 7 These studies consistently reveal a 60% to 70% reduction in mortality in patients taking statins. Because patients at risk for postoperative MI and cardiac events often have indications for statins, the perioperative period may be an opportune time to consider long-term statin candidacy in these patients, regardless of the perioperative period.
In contrast to cardiovascular risk assessment and management, the literature on perioperative pulmonary assessment and intervention is less robust. However, the American College of Physicians has published a summary of the literature and guidelines for evaluation and management.
Postoperative pulmonary complications (PPCs) are equally prevalent compared with cardiac complications and contribute similarly to morbidity, mortality, and length of postoperative hospital stay. Patient factors increasing the risk for PPCs include chronic obstructive pulmonary disease, age older than 60 years, American Society of Anesthesiologists (ASA) class II or greater, functional dependence, and congestive heart failure. Obesity and mild to moderate asthma have not been consistently shown to predict PPCs. A low serum albumin level (<35 g/dL) has been found to be a powerful predictor of PPCs, likely as a reflection of impaired general health or immune compromise.
Surgical factors increasing PPC risk include thoracic or abdominal surgical site (which can lead to splinting due to pain and impaired diaphragmatic excursion); neurosurgery; head and neck procedures; vascular procedures, especially abdominal aortic aneurysm repair; any emergent procedures; use of general anesthesia; and prolonged (>3 hr) procedures.
Routine pulmonary function testing and chest radiography are not indicated preoperatively because they do not predict PPCs; obtain these only if the patient is symptomatic, has unexplained dyspnea, or is undergoing lung volume reduction surgery or other intrathoracic procedures.
Interventions that successfully reduce PPCs in high-risk patients include incentive spirometry or deep-breathing exercises and selective use of nasogastric tube decompression. Right heart catheterization and total enteral or parenteral nutrition also have been studied in this vein, and neither intervention has proven benefit in reducing PPCs.
Until recently, no scoring systems existed for predicting PPCs akin to those used for cardiovascular risk stratification. Re-searchers with the Veterans' Administration National Surgical Quality Improvement Project (NSQIP) developed and prospectively validated scoring systems for predicting postoperative pneumonia and respiratory failure that include many of these predictors in numeric scoring systems. Although these tools have prognostic value, guidance of preventive therapy based on these tools is limited.
Other ongoing avenues of research include the effect of obstructive sleep apnea on PPCs and the role of continuous positive airway pressure and bilevel positive airway pressure in preventing or treating postoperative respiratory failure and PPCs.
Venous thromboembolism (VTE), which includes deep venous thrombosis (DVT) and pulmonary embolism (PE), are quite common causes of morbidity and mortality that are largely preventable in the postoperative patient. Several national quality-improvement organizations have cited VTE prophylaxis for patients at risk as a priority for both individual physicians and for hospitals, because this intervention reduces both adverse patient outcomes and hospital costs.
Surgical patients in particular have significantly increased risks for VTE due to advanced age, multiple medical comorbidities, and prolonged procedure times, in addition to the hypercoagulable state of surgery and immobility. Thus, clinicians must consider VTE risk and risk-reduction strategies in all patients undergoing surgery.
Postoperative DVT is typically asymptomatic, and fatal PE can often be the first sign of VTE; screening modalities (such as with venous duplex imaging) in asymptomatic patients have low sensitivity to detect clot, so it is not appropriate to use these unless clinical suspicion is present. Therefore, the approach of choice is to systematically apply prevention strategies to all patients undergoing surgery, with treatment choices based on patient-related and procedure-related risks.
Patient-related risk factors for VTE include age older than 40 years, malignancy, immobilization, varicose veins, severe cardiopulmonary disease (prior MI, congestive heart failure, chronic obstructive pulmonary disease), prior stroke, paralysis or spinal cord injury, prior VTE events, hyperviscosity syndromes (polycythemia vera or malignancy related), and major vascular injury.
VTE risks also vary with the type of procedure; orthopedic and neurosurgical procedures have the highest reported rates in the literature in the absence of prophylaxis. Total knee replacement procedures have up to 65% rates of postoperative DVT; hip surgery (fracture or elective replacement) is associated with a 50% rate. These patients require the most aggressive approach, often combining pharmacologic and nonpharmacologic means. In contrast, outpatient procedures such as cataract surgery and laparoscopic procedures have quite a low risk for perioperative VTE and do not require preventive means other than early ambulation unless other VTE risks are present. Other surgery types, including general, vascular, gynecologic, and neurosurgical, have similar risks for VTE and may be stratified based on patient age, preexisting risk factors, and length of the operation.
Modalities to prevent VTE events are categorized into nonpharmacologic and pharmacologic means. Nonpharmacologic methods include early ambulation, graduated compression stockings, and intermittent pneumatic compression devices. Pharmacologic methods routinely evaluated include aspirin, low-dose unfractionated heparin (LDUH), low-molecular-weight heparin (LMWH), warfarin, and factor Xa inhibitors such as fondaparinux. Another agent recently showing promise was ximelagatran, an oral direct thrombin inhibitor, but this was withdrawn from the market due to risks of severe hepatotoxicity noted during clinical trials.
The American College of Chest Physicians (ACCP) recommendations for antithrombotic recommendations for preventing perioperative VTE is highlighted in Table 1 based on the risk of thromboembolism. Note that aspirin alone is not recommended by the ACCP guidelines because data show limited effectiveness compared with other modalities and increased bleeding risk, mostly gastrointestinal in origin.
|DVT (%)||PE (%)|
|Level of Risk||Calf||Proximal||Clinical||Fatal||Successful Prevention Strategies|
|Low risk: Minor surgery in patients <40 yr with no additional risk factors||2||0.4||0.2||<0.01||No specific prophylaxis; early and aggressive mobilization|
|Moderate risk: Minor surgery in patients with additional risk factors; surgery in patients 40-60 yr with no additional risk factors||10-20||2-4||1-2||0.1-0.4||LDUH q12h, LMWH <3400 U daily, GCS, or IPC|
|High risk: Surgery in patients >60 yr, or 40-60 yr with additional risk factors (prior VTE, cancer, molecular hypercoagulability)||20-40||4-8||2-4||0.4-1.0||LDUH q8h, LMWH >3400 U daily, or IPC|
|Highest risk: Surgery in patients with multiple risk factors (age >40 yr, cancer, prior VTE); hip or knee arthroplasty, HFS; major trauma; SCI||40-80||10-20||4-10||0.2-5||LMWH (>3400 U daily), fondaparinux, oral VKAs (INR, 2-3), or IPC or GCS + LDUH or LMWH|
DVT, deep venous thrombosis; GCS, graduated compression stockings; HFS, hip-fracture surgery; INR, international normalized ratio; IPC, intermittent pneumatic compression; LDUH, low-dose unfractionated heparin; LMWH, low-molecular-weight heparin; PE, pulmonary embolism; SCI, spinal cord injury; VKA, vitamin K antagonist; VTE, venous thromboembolism./p>
Adapted from Geerts WH, Pineo GF, Heit JA. Prevention of venous thromboembolism: The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004;126:338S-400S.
The use of inferior vena cava (IVC) filters has increased significantly over the past several decades, with an increasing percentage being placed for prophylaxis in high-risk patients. More recently, retrievable IVC filters have been developed that can be removed up to 180 days after placement. However, placing retrievable IVC filters appears to be cost prohibitive in multisystem trauma patients. Because it is not clear that they improve outcomes, are costly, and have a fair rate of complications (29%, including improper placement, migration, caval occlusion or wall penetration, and venous stasis), current ACCP guidelines do not support prophylactic placement of IVC filters. However, patients who cannot tolerate pharmacologic prophylaxis or who have a complication of anticoagulation and who have a temporary contraindication are reasonable candidates for temporary IVC filter placement for protection from fatal or disabling PE.
More than 2 million Americans currently take anticoagulant agents to prevent or treat thromboembolic events. As the population ages, so will the frequency of surgical procedures, and consequently clinicians will need to manage perioperative anticoagulation more often. Discontinuation of anticoagulation leaves patients unprotected from thromboembolic risk for several days around the time of surgery. However, aggressive anticoagulation in the postoperative period can increase bleeding risk. Thus, clinicians must consider the indication for long-term anticoagulation and extrapolate the risk for thrombotic events compared with the risk for bleeding events (Box 2).
|Box 2: When Patients Taking Warfarin Need Surgery|
|High Risk of Thomboembolism: Bridging Advised|
|Known hypercoagulable state as documented by a thromboembolic event and one of the following:
|Hypercoagulable state suggested by recurrent (two or more) arterial or idiopathic venous thromboembolic events (not including primary atherosclerotic events, such as stroke or myocardial infarction due to intrinsic cerebrovascular or coronary disease)|
|Venous or arterial thromboembolism within the preceding 1 to 3 months|
|Acute intracardiac thrombus visualized by echocardiogram|
|Atrial fibrillation plus mechanical heart valve in any position|
|Older mechanical valve model (single-disk or ball-in-cage) in mitral position|
|Recently placed mechanical valve (<3 mo)|
|Atrial fibrillation with history of cardioembolism|
|Intermediate Risk of Thromboembolism: Bridging on a Case-by-Case Basis|
|Cerebrovascular disease with two or more strokes or transient ischemic attacks without risk factors for cardiac embolism|
|Newer mechanical valve model (e.g., St. Jude) in mitral position|
|Older mechanical valve model in aortic position|
|Atrial fibrillation without a history of cardiac embolism but with multiple risks for cardiac embolism (e.g., ejection fraction <40%, diabetes, hypertension, nonrheumatic valvular heart disease, transmural myocardial infarction within preceding month)|
|Venous thromboembolism >3-6 months ago*|
|Low Risk of Thromboembolism: Bridging Not Advised|
|One remote venous thromboembolism (>6 months ago)|
|Intrinsic cerebrovascular disease (such as carotid atherosclerosis) without recurrent strokes or transient ischemic attacks|
|Atrial fibrillation without multiple risks for cardiac embolism|
|Newer model prosthetic valve in aortic position|
*For patients with a history of venous thromboembolism undergoing major surgery, postoperative bridging therapy only (without preoperative bridging) may be considered.
High-risk patients have up to a 10% rate of thromboembolism per year and are typically managed with bridging therapy, because the concern for a thromboembolic event outweighs the perceived bleeding risk. Bridging involves discontinuing warfarin 5 to 6 days before surgery and initiating therapeutic-dose subcutaneous LMWH or intravenous unfractionated heparin when the patient's international normalized ratio (INR) falls below the therapeutic range. The bridging agent is stopped 6 to 24 hours before surgery, and anticoagulation is resumed as soon as possible postoperatively.
Low-risk patients have an annual arterial thromboembolic risk of less than 5% or a monthly VTE risk of less than 2%. Because the perioperative bleeding risk with bridging therapy outweighs the thromboembolic risk in these patients, bridging therapy is not indicated. Warfarin can be discontinued approximately 5 days before surgery and resumed postoperatively when the bleeding risk related to surgery is minimal.
Intermediate-risk patients have an annual arterial event risk of between 5% and 10% and a monthly VTE risk of between 2% and 10%. These patients can have comparable risks of bleeding and thromboembolism, so individual patient and procedure factors must be assessed on a case-by-case basis.
Assessment of thromboembolic risk perioperatively is not simply achieved by taking the yearly risks and dividing by 365 to obtain a daily risk. Surgery creates a prothrombotic milleu that can increase VTE risk by 100-fold, and discontinuation of warfarin has been associated with biochemical evidence of rebound hypercoagulability. Therefore, even in low-risk and intermediate-risk patients, appropriate VTE prophylaxis measures should still be applied, even if bridging therapy is not indicated. In addition, consider the consequences of the thromboembolic event being averted by bridging. Although arterial events are less common, they cause significantly more death and disability when they do occur: 20% to 30% mortality and 30% to 40% rate of disability, compared with combined rates of 5% to 10% for VTE and 3% to 13% for bleeding events.
Intravenous unfractionated heparin and LMWH are typically used for bridging therapy, although the use of LMWH in patients with mechanical heart valves is controversial. The prescribing information for enoxaparin states that its use for “thromboprophylaxis in pregnant women with mechanical prosthetic heart valves has not been adequately studied.” This refers to the study where two of eight pregnant women receiving enoxaparin developed clots resulting in blockage of the valve and leading to maternal and fetal deaths, whereas none of four patients receiving unfractionated heparin developed valve thrombosis. In the two deaths, anti-Xa levels were subtherapeutic at some points during treatment; subsequent studies note that the physiology of pregnancy can affect the pharmacokinetics of enoxaparin, leading to lower anti-Xa levels. Therefore, if enoxaparin is to be used in pregnant patients for any reason, anti-Xa levels should be checked frequently and kept between 0.5 and 1.2 anti-Xa units. Nevertheless, unfractionated heparin should be used for any pregnant patients with mechanical heart valves requiring bridging therapy. Several studies document the safety of use of LMWH for bridging in patients with mechanical heart valves; Ansell and colleagues reviewed data on 461 patients from 10 studies; three patients had TIAs and no patients had strokes or valve thromboses. 8
Stopping and restarting medications in the perioperative period is an essential component to perioperative care. Appropriate medication management helps to maintain stability of chronic conditions, prevent medication withdrawal, avoid interactions with anesthetic agents, and facilitate transition to discharge. Although published clinical trial data in this area are limited, management strategies are extrapolated based on case reports, expert consensus, in vitro studies, and pharmaceutical manufacturer recommendations.
Postoperatively, patients present with a significant stress response, including an increase in pituitary, adrenal, thyroid, and hypothalamic activity, which leads to heightened sympathetic nervous system activity. This can affect the activity and metabolism of medications for chronic conditions. Also, gut motility and absorption may be diminished by factors such as villous atrophy, splanchnic blood flow changes, ileus, and narcotics use. In one study of general surgery patients, patients taking chronic medications had 2.5 times the likelihood of developing postoperative complications.
Several general principles can be applied in managing perioperative medications. First, medications with significant withdrawal potential that do not negatively affect the procedure or anesthesia administration should be continued during the perioperative period. Examples of this include beta blockers and alpha blockers such as clonidine. Second, medications that increase surgical risk and are not essential for short-term quality of life should be discontinued during the perioperative period. If a medication does not fall clearly into one of these categories, then one must rely on physician judgment, based on the stability of the condition being treated and anesthetic and surgical concerns.
Cardiovascular medications in general should be continued throughout the perioperative period, because they treat and stabilize conditions such as coronary artery disease, congestive heart failure, and cardiac arrhythmias. Some notable exceptions include anticoagulants and antiplatelet agents (discussed later), diuretics, and angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs), which are associated with hypotension with induction of anesthesia, often requiring intraoperative pressor therapy. The literature does support a benefit of perioperative ACE inhibition toward mortality, but these studies do not note whether these agents were stopped before surgery, how far in advance they were stopped, and how quickly they were restarted postoperatively if they were stopped. More studies are needed in this area, but given the available data, it seems prudent to discontinue ACE inhibitors or ARBs on the morning of surgery.
Statins (HMG-CoA reductase inhibitors) have clear associations with perioperative mortality benefit by randomized trial and epidemiologic data. They might prevent vascular events by other mechanisms besides cholesterol lowering, such as plaque stabilization, reduction in inflammation, and decreased thrombogenesis. Previous concerns for increased risk of rhabdomyolysis are not well founded, because these were based on scant individual case reports with marked confounding.
Antiplatelet agents include aspirin, nonsteroidal anti-inflammatory drugs (NSAIDs) and thienopyridines, such as clopidogrel. Although aspirin and clopidogrel are typically used for patients with preexisting coronary, cerebral, or peripheral vascular disease for maintaining vessel patency and reducing event risk, they also increase the risk of postoperative bleeding. For most patients, short-term discontinuation of these agents perioperatively does not lead to increased adverse outcomes; therefore, it is reasonable to discontinue them 7 to 10 days before elective surgery. One important exception to the risk equation is when patients have a recently placed drug-eluting stent in a coronary artery or arteries. Patients with paclitaxel-eluting stents should continue to take combination aspirin and clopidogrel for a minimum of 3 months, whereas patients with sirolimus-eluting stents should continue taking combination therapy for no less than 6 months. Premature discontinuation of antiplatelet agents tremendously increases the risk for early and late stent thrombosis, which has a case fatality rate of 45% and a 70% rate of myocardial infarction. Therefore, elective surgery should be deferred at least 6 months in this patient population if at all possible, until more data surrounding the perioperative management of drug-eluting stents become available.
Tight glycemic control in the perioperative period is clearly associated with reductions in mortality and length of critical care unit stay, as well as reductions in wound infections and complications in cardiac surgery patients. However, many clinicians fear the possibility of perioperative hypoglycemic events given the variable caloric intake and disruption of anabolic and catabolic processes surrounding major surgery.
Patients taking intermediate-acting insulin should take at least one half to two thirds of their evening dose the night before and on the morning of surgery, because approximately one half of insulin is used for non-nutrient metabolic needs. Full-dose intermediate or long-acting insulin (such as glargine) should be considered. Insulin coverage should be anticipatory and dosed for basal coverage with long-acting and intermediate-acting agents and mealtime doses with additional units for coverage as needed with short-acting or ultrashort-acting insulin. Sliding scale insulin alone is insufficient and has been shown to lead to unacceptable rates of both hyper- and hypoglycemia. Insulin administration should also mirror the route and frequency of nutrient intake; continuous feedings require more continuous insulin administration (such as with an insulin drip or long-acting subcutaneous agent), whereas intermittent feedings require intermittent insulin doses for mealtimes or bolus feedings.
Herbal medications are used by up to one third of patients undergoing surgery. These agents are often perceived by the public as being natural and therefore completely safe; however, they have no FDA regulation because they are considered food supplements, and they can contain varying amounts of the active ingredient, among other compounds. On the other hand, prescription medicines are often perceived as artificial and therefore less safe, despite rigorous standards from the FDA for dosing and safety. Clinicians must specifically inquire about herbal preparations and over-the-counter medications, because many patients do not even consider these to be medications. These agents should be discontinued at least 1 week before surgery, preferably 2 weeks.
Psychiatric medications should be continued perioperatively, because decompensation of psychiatric conditions should be avoided if possible. Agents such as selective serotonin reuptake inhibitors, the newer serotonin-norepinephrine reuptake inhibitors, and benzodiazepines are safe to continue. Some concern exists for perioperative arrhythmias in conjunction with tricyclic antidepressants, but the literature does not support this concern. Monoamine oxidase inhibitors are used much less commonly now, but they still are used for refractory depressive disorders. These agents lead to an accumulation of biogenic amines in the central nervous system, which can lead to a hypertensive crisis if used with indirect sympathomimetics or can lead to a serotonin-like syndrome when used with meperidine or dextromethorphan. However, anesthesia may be performed safely if meperidine is avoided and only direct-acting sympathomimetics such as phenylephrine are used.