Chapter 36. Anesthesia

Perioperative management of the older adult is complex. Anesthetic care itself is challenging, and both preoperative preparation and postoperative care for older adults assume greater importance than for young, healthy adults. Ideally, care is based on a comprehensive plan that integrates the roles of the anesthesiologist, surgeon, geriatrician, primary caregivers, and medical specialists. This chapter provides an overview of the role of the anesthesiologist and anesthesia techniques in pre-, peri-, and postoperative care of the older adult, in order to improve communication with other professionals who provide care to this vulnerable population.

Anesthesia is the art and science of controlling physiologic processes in order to permit interventions on the body that would be intolerable owing to normal compensatory mechanisms. Surgery, which involves a direct assault on body tissues, is the standard paradigm for understanding anesthesia. Anesthesiologists perform other valuable services, including sedation for less invasive procedures, management of pain syndromes, and the provision of critical care. European anesthesiologists are also actively involved in emergency care. Anesthesia is based on an understanding of homeostatic mechanisms and their manipulation. Homeostenosis, the restricted range and capacity of homeostatic mechanisms associated with aging, provides a challenge to the anesthesiologist. In order to tailor appropriate anesthesia care for an older adult, the anesthesia team must be familiar with the effects of aging on multiple organ systems, particularly the heart, lungs, and brain, and must be familiar with physiologic changes of age, such as increase in body fat, decreased glomerular filtration, and reduced hepatic blood flow, which affect anesthetic drug action and duration.

Preoperative care can take place in multiple settings and can occur briefly or over a period of time. The most important goal of preoperative assessment is not the risk assessment, but the improvement of the patient's medical status prior to surgery and planning for the recovery process. The anesthesiologist provides care in a different time frame and uses a different approach than is typical for other specialties. The chief complaint usually takes the form of a request for anesthesia services for a specific operation. The anesthesiologist uses a vertical or systems-based approach, most clearly articulated by Stanley Muravchick (Figure 36-1). As part of this approach, the anesthesiologist assigns an American Society of Anesthesiologists physical status classification (Table 36-1). The American Society of Anesthesiologists score is a clinician's tool, not a predictor of perioperative risk, because, while the American Society of Anesthesiologists score correlates broadly with outcomes, it does not incorporate age or type of operation, both significant influences on outcomes.

In terms of ability to tolerate anesthesia and surgery, physiologic status is more important than chronological age. Aging is a highly variable process; young adults are more similar to each other physiologically than are older adults. While chronological age is insufficient to explain surgical outcomes, it has repeatedly been found to be associated with an increased risk of morbidity, mortality, and poor surgical outcomes. Age appears to be an important modifier of disease load, and the number of diseases (disease burden) appears to be the primary determinant of outcomes. Older adults with limited or low disease load have relatively lower risk of postoperative complications (Figure 36-2A), and the combination of high disease load and age is associated with extremely high rates of morbidity and mortality (Figure 36-2B). In addition, aging is not uniform across organ systems, and some organ systems may have aged or have been affected by disease more than others. Since aging is so heterogeneous, a careful preoperative evaluation is extremely important in preparing the older adult for surgery.

How and when the preoperative evaluation occurs is highly variable. Many modern surgical practices admit patients for major surgery on the day of the operation. In this scenario, the anesthesiologist must perform the preoperative evaluation and discuss the perioperative experience and anesthetic options with the patient in a tight time frame. Alternative approaches include preoperative clinics and telephone calls, which can provide important opportunities to develop an understanding of the patient's physical condition prior to the day of surgery.

The preoperative evaluation is an essential process in developing an anesthetic plan. Just as importantly, it provides an opportunity to answer questions for the patient and family. The most constructive of preoperative evaluations provide plans for minimizing risk in the perioperative period (i.e., before, during, and after surgery). As part of the evaluation, a general internist maybe consulted. Institutional policy and practices vary as to whether such input is required or even routine. A generalist's statement that the patient has been “cleared for surgery” is of limited value. Much more useful to the anesthesiologist and surgeon is a complete overview of the patient's medical condition, recent evaluations, and medications. The more the preoperative evaluator knows and understands, the less often the need for additional consultation. The best plan is based on good communication between the generalist and specialist physicians, the surgeon, the anesthesiologist, and the patient.

Preoperative management of significant medical problems should be determined. Medication regimens (such as anticoagulants, antidiabetic agents, and antiparkinsonian medications) require modifications and advance planning prior to surgery. Several members of the medical team can lead this effort, but the individual responsible should have a good understanding of the issues that need to be addressed and how they should be handled. Medication management protocols that have been jointly set by surgeons, anesthesiologists, and internists are often useful. When additional consultation is sought, the request should clearly define the issues that need to be addressed and the expected role of the consultant. The example shown in Figure 36-3 illustrates a reasonable delineation of the problem to be addressed. Anecdotally, an organized preoperative clinic can improve patient satisfaction and decrease the number of perioperative consults.

Risk assessment is a part of the preoperative process. Most of the time the objective of risk assessment is to decide how aggressive one should be in preparing the patient for surgery, rather than deciding that the surgical risk is too high. Risk assessment is most formalized in the form of the AHA/ACC guidelines for the evaluation of cardiac disease. The guidelines attempt to integrate severity of comorbid disease, exercise tolerance, and surgical severity into a rational plan, in order to determine who needs to enter an advanced evaluation process or to avoid diagnostic procedures with a risk greater than the proposed surgery itself. In fact, the AHA/ACC guidelines were originally developed specifically to limit the frequency of noninvasive cardiac testing, which has the potential to be used unnecessarily. It is important to remember that the AHA/ACC guidelines have never been formally tested, nor are they meant to supplant clinical judgment. Many medical problems in the older adult cannot be addressed using established guidelines so that decisions are based on clinical judgment. The impact on postoperative outcomes of optimizing medical conditions prior to surgery is usually unknown. Typical conditions that might be optimized preoperatively, even if not seriously unstable, include high blood pressure, diabetes, reactive airway disease, and coronary artery disease. One of the best examples of medical optimization leading to improved outcome is the use of beta-blockade for patients with known or suspected coronary heart disease.

Sometimes internists and anesthesiologists disagree on preoperative management, often about whether or not planned diagnostic tests can be delayed until after the surgery. The consult shown in Figure 36-3 provides an example. In this case, atrial flutter is not a contraindication to surgery. Why should its evaluation delay surgery? Some would argue that if the tests provide useful information or lead to an intervention that improves the patient's status, then the patient will be better off having the tests done prior to elective surgery. In the example in Figure 36-3, an echocardiogram provides useful information, even if normal. If a test should be performed, it is reasonable to do it before surgery, unless surgical urgency dictates otherwise. Furthermore, if elective cardioversion is being considered, it should be done before surgery because heart rate control in the intra- and postoperative periods is more easily achieved when the patient is in sinus rhythm. As stated above, there is frequently little evidence upon which to base such judgment but coherent preoperative planning in an older adult can avoid confusion, surgical delay, and requests for emergent consultations that are rarely as complete as elective evaluations.

Geriatrician involvement improves perioperative care but is undervalued in most usual care settings. Comprehensive geriatric assessment (discussed in detail elsewhere) has a place in perioperative care because it is the best way to prevent geriatric-specific postoperative issues like delirium and pressure ulcers.

A unique part of the preoperative assessment of the older adult arises when a patient has a “do not resuscitate” order or other advanced care directives. Many patients have, in consultation with their primary-care physicians, elected to forgo resuscitative measures, because of fears that resuscitation will lead to poor outcome, or because they feel that resuscitation measures will be futile. These fears are supported by numerous studies that demonstrate that cardiac arrest in most settings carries a very poor prognosis, because arrests are usually unwitnessed and may be caused by advanced disease states. However, the operating room is a special setting in this regard. In contrast with other in-hospital cardiac arrests, cardiac arrests in the operating room carry a very favorable prognosis, because arrests are often because of reversible causes such as hemorrhage or drug effects, and because they are witnessed events where resuscitation is instituted within seconds. This means that, in the setting of surgery, a patient's DNR order must be reconsidered in light of the different prognostic implications of cardiac arrest during surgery. Some organizations have standing policies that DNR orders are rescinded in the operating room. The DNR may be reinstituted either upon a physician's order or automatically upon return to the ward. Regardless of the management plan, DNR orders require review and discussion with the patient. While the anesthesiologist should have such discussions with the patient, input from the patient's other physicians can be invaluable.

Sometimes a primary-care physician or consultant offers advice on anesthetic management as a part of the anesthesia preoperative referral. These recommendations are usually not helpful. Requests to avoid hypotension and hypoxemia are so obvious as to be unnecessary. Surgical clearance limited to local or regional anesthesia can create problems. These restrictions may not be based on clear evidence. Regional techniques have limitations (see section on “Regional Anesthesia”). Regional anesthetic techniques occasionally fail in the operating room, requiring an immediate alternative. In some cases, local anesthesia may be feasible, but not preferable, as a result of patient characteristics such as apprehension or positioning issues, or because adequate anesthesia would be difficult to achieve with local approaches alone.

Anesthesiologists have developed two major approaches to the provision of anesthesia. The first is general anesthesia, based on the systemic administration of medications whose principal effects occur in the central nervous system. The provision of various levels of sedation can be thought of as a subtype of general anesthesia. The second major approach is regional anesthesia, based on the use of one of a family of local anesthetics to block nerves. Spinal or epidural anesthesia applies these agents either in or around the spinal canal. Peripheral nerve blocks provide them around nerves subserving specific areas of the body. Local anesthetic agents injected into areas of skin to block the surrounding nerve endings can be thought of as a form of regional anesthesia. The two main forms of anesthesia are not mutually exclusive. A patient undergoing general anesthesia may have a peripheral nerve block to control postoperative pain, and a patient undergoing a procedure with spinal anesthesia is likely to receive hypnotic medications in concentrations that produce sedation. General anesthesia and regional anesthesia work via different mechanisms, so their effects in the older adults are discussed separately below.

General Anesthesia

This section will review the goals of general anesthesia, principles of intraoperative care, types of anesthetic agents, and age-related special management issues. While unconsciousness is the most apparent distinguishing feature of general anesthesia, it has many other important effects. The cardinal features of general anesthesia are (1) lack of consciousness, including amnesia for events that occur under anesthesia, (2) analgesia or absence of pain, (3) lack of movement in response to painful stimuli, and (4) control of the reflex responses to painful stimuli. In the early years of anesthesiology, ether was found to provide all of these properties, although it did cause prolific vomiting following emergence and had a tendency to cause explosions in the operating theater. The observation that ether, a relatively simple molecule, produced an adequate anesthetic state, along with compelling physicochemical evidence suggesting that potency was linked to the length of the carbon chain of the anesthetic, led anesthesiologists to search for a single mechanism or receptor that was responsible for anesthesia. In subsequent years, it has become clear that anesthesia does not result from alterations of a single receptor, but that all available drugs with anesthetic properties have specific targets. General anesthesia results from alterations at multiple receptors and multiple drugs are almost always used to produce general anesthesia.

Intraoperative Anesthesia Care

Standard procedure for anesthetic care is to monitor the electrocardiogram, blood pressure, and oxygen saturation. When general anesthesia is undertaken, end-tidal carbon dioxide monitoring is required and temperature should be monitored for most procedures. Intravenous access with large-bore (18 gauge or larger for major surgery) catheters inserted into peripheral veins is almost universally required as part of anesthetic care. The induction of anesthesia is generally accomplished with intravenously administered hypnotic agents. The appropriate dosages of these agents are all decreased with aging. Maintenance of a patent airway and adequate ventilation are primary goals, usually attained via an endotracheal tube and positive pressure ventilation. In recent years, anesthesiologists have developed a variety of new airway devices, such as the laryngeal mask airway, as alternatives to endotracheal intubation. The choice of device and backup procedures varies with procedure and practitioner. Spontaneous breathing is also a feasible and potentially desirable alternative to positive pressure ventilation.

Adjustment of the depth of anesthesia consumes a good deal of the anesthetic manipulations during surgery. How “deep” or “light” the patient is at any moment depends on the balance of drug effects and the surgical stimulus. Titration of the depth of anesthesia is typically a matter of balancing the tendency of the surgical stimulus to activate the brain and stimulate the sympathetic nervous system against the anesthetic's ability to depress brain activity and limit the stimulation of the sympathetic nervous system. Thus, a patient might be fully anesthetized one moment and then be inadequately anesthetized a moment later if the surgical (painful) stimulus suddenly increases. The effective depth of anesthesia can be monitored through a variety of methods. The primary approach for years has been to monitor blood pressure and heart rate. Hemodynamic signs of light anesthesia generally appear long before a patient has any chance of “waking up” or of having postoperative recall. A rise in blood pressure and/or heart rate is typically the first sign of light anesthesia and triggers an increase in the concentration of volatile gas (currently isoflurane, sevoflurane, or desflurane) concentration or administration of a small bolus of a fast-acting opioid such as fentanyl. A safe strategy is to keep blood pressure at or slightly lower than the patient's usual baseline, although this can be quite difficult to accomplish in older patients. The risk of awareness/recall is relatively remote under such circumstances.

Since depth of anesthesia is such a critical issue in perioperative care, and because it can change moment to moment, anesthesiologists have attempted for years to use electroencephalograms to define a brain-based measure of depth of anesthesia. Recently, devices have been developed that use advanced signal processing and computer miniaturization to provide a single number between 0 and 100 that indicates depth of anesthesia. When these devices have been applied to older adults, the findings suggest that the amount of anesthetic typically used to control blood pressure and heart rate frequently results in extremely deep anesthesia for the brain. Excessive depth of anesthesia has been proposed to contribute to poor outcomes. This important area requires further study and might be valuable in the clinical arena in the future.

Patient movement in response to a surgical stimulus, even if very slight, suggests an inadequate depth of anesthesia. Too light anesthesia can result in patient awareness and failure of amnesia. The amount of movement is usually subtle. Most slight movement does not imply such a light level of anesthesia as to result in awareness or recall. Movement cannot be used as an indicator of depth of anesthesia if a patient is paralyzed by neuromuscular blocking agents and common adjuvants for abdominal surgical cases or in cases where even minimal movement is surgically unacceptable. Unintentional excessively light anesthesia is most likely to occur when neuromuscular blocking agents produce muscle relaxation in the presence of lighter general anesthetic agents, such as nitrous oxide and opioids, with little or no potent gases. The electroencephalogram monitors described above have an unclear role in preventing awareness during surgery.

Anesthetic Agents

The pharmacokinetics and pharmacodynamics of the intravenous and volatile anesthetic agents are altered in the older adult, so that administration must be adjusted. A general rule of the thumb is that all drugs will have more dramatic and longer-lasting effects in older patients, although this is not universally true. The mechanisms vary with the drug but include changes in protein binding, decreases in initial volume of distribution, alterations in receptor sensitivity, and slowed renal or hepatic metabolism. The therapeutic effect of many intravenous anesthetic agents is not dissipated by metabolism, but by redistribution of the drug from the brain to the fat. These drugs are highly lipid soluble and cross membranes rapidly. On initial injection, the blood levels are very high but will fall as the drug is taken up into fat. The high initial blood levels will promote rapid transfer of the drug into highly perfused organs such as the brain. Even though fat serves as an almost limitless sink for lipid-soluble drugs, blood flow to fat is low and transfer of drug into fat takes time. Transfer time largely dictates the duration of a drug's clinical effect. As blood levels decrease, drug returns to the blood from the more vascular organs and is then transferred to the fat, dissipating the therapeutic effect. Substantial and/or prolonged administration of many drugs results in drug levels in fat that become high enough to sustain a (residual) blood level that has a therapeutic effect. From an anesthetic management viewpoint, the pharmacokinetic and pharmacodynamic changes of aging rarely present a major problem. Patient response to drugs is variable at any age, so, with the exception of induction, drugs are given in small doses and titrated to effect. The trend today in anesthesia practice is to use drugs that are short acting. These drugs either have shorter metabolic half-lives or redistribute to yield very low blood levels with minimal residual effects, or do both. For example, propofol and thiopental are equally effective induction agents and both are “short-acting” because these redistribute rapidly into fat. The residual blood levels of propofol, however, produce less residual sedation than occurs with thiopental, and propofol also has a shorter metabolic half-life. These principles of drug pharmacokinetics apply to patients of all ages, but the ability to use drugs with rapid fat uptake, but minimal residual effects, is particularly helpful in older patients.

Hemodynamic Stability during Anesthesia

Hemodynamic stability is a core goal of anesthesia and one of the more challenging aspects of anesthetic care of the older adult. Many characteristics of aging contribute to the increased vulnerability of older adults to hemodynamic instability (Table 36-2). General anesthesia lowers the blood pressure in everyone. This effect is owing to decreased sympathetic tone, which, in turn, may lower heart rate, decrease systemic vascular resistance, and cause peripheral pooling of blood, which will lower cardiac preload. There is also some direct myocardial depression. These effects are exaggerated in older adults. Sympathetic nervous system activity commonly increases with age and so may be more reactive to stimuli than in young people. Chronic hypertension, common in older patients, further exaggerates the vascular response to changes in sympathetic nervous system activity. In particular, the aorta stiffens with age and typically results in left ventricular hypertrophy. Hypertrophy leads to a reduction in the rate of diastolic relaxation, which, in turn, diminishes early diastolic filling. Adequate ventricular preload becomes dependent on passive filling related to left atrial pressure and active filling related to atrial contraction. Both mechanisms must overcome the increased ventricular stiffness and both require a full atrium. Yet atrial blood volume is difficult to keep constant. Veins also stiffen with age. Venous stiffness reduces the capacity to buffer changes in blood volume, making atrial filling more sensitive to overall volume status as well as to the effects of the sympathetic nervous system on blood distribution. In consequence, older patients are more prone than young patients to both hypotension and hypertension during surgery.

Management of the inevitable swings in blood pressure can be a challenge. Alterations of blood pressure are particularly common during induction of anesthesia and the laryngoscopy/intubation sequence. Hypertension is usually managed by increasing the depth of anesthesia with more opioid and/or increasing the anesthetic gas concentration. Unfortunately, the surgical stimulus can increase or decrease faster than the depth of the anesthetic can be manipulated. One approach is to complement the anesthetic with beta-blockade to minimize the hemodynamic response to the surgical stimulus. Another approach is to use relatively high doses of opioid and relatively less volatile anesthetic. Opioids produce less intrinsic depression of blood pressure than do volatile anesthetics while preventing significant rise in blood pressure during surgery. If high-dose opioids are used to prevent the hemodynamic response to the surgical stimulus, the primary role of the volatile anesthetic is to produce unconsciousness, requiring approximately half the concentration of gas needed without opioids. An ultra-short-acting opioid, remifentanil, can be employed primarily for short cases that are accompanied by profound simulation in sensitive areas. High doses of moderate-duration opioids (e.g., fentanyl or morphine) are limited by residual respiratory depression, which can unacceptably depress ventilatory drive at the end of the surgery. Continuous use of ultrashort agents is not currently encouraged because of cost. Another approach is to use a limited amount of opioids compatible with the surgical procedure and then maintain the volatile anesthetic concentration high enough to block the hemodynamic response to the highest level of surgical stimulus. Such a strategy will minimize hypertension but will also make the patient more prone to hypotension at other, less stimulating, times of the surgery. Sometimes even the minimum level of anesthesia causes an unacceptable depression of blood pressure. In these circumstances, blood pressure must be actively supported. Volume administration can support blood pressure but is of limited utility in older patients. Volume may restore stroke volume but will not compensate for hypotension caused by severe bradycardia or a low vascular resistance. Bradycardia generally responds to glycopyrrolate, an anticholinergic drug that crosses the blood–brain barrier poorly and is presumably less likely to promote postoperative delirium than atropine. Ephedrine stimulates both alpha and beta-receptors and effectively raises pressure, but tachyphylaxis eventually develops and so it is not used for extended time periods. Phenylephrine, a pure alpha agonist, can be conveniently given as either a bolus or a low-dose infusion and effectively maintains blood pressure. On occasion, inotropic infusions (e.g., dopamine) are used, especially if bradycardia does not respond to anticholinergics or there is concern over baseline ventricular function. All pressors have potential adverse effects, of course, but serious problems from their use are infrequent.

Intravascular fluid therapy in older adults is a controversial topic. Much has been written, little has been proven, and clinical practice tends to be based more on habit than on science. Much of the science underlying the most standard anesthetic and surgical approach to volume replacement is based on studies undertaken in the 1960s and 1970s. As noted above, cardiovascular aging makes volume administration more difficult. The stiffness of the venous system exaggerates the response to a given volume excess or deficit. Well-anesthetized patients may appear hypovolemic during surgery as a result of the suppression of sympathetic nervous system activity. A common concern of the geriatric anesthesiologist is that large amounts of crystalloid used for blood pressure control may achieve what appears to be normovolemia during surgery, but, when surgery is over and postoperative pain restores sympathetic activity, that volume may shift back to the central circulation and create a volume overload. Recently, the basis for fluid administration has been questioned, and an alternative strategy that uses less fluid has been proposed. To date, this approach to fluid management has not been extensively tested or adopted in the United States. The recent literature on reduced fluid volume has not attended to the special issues of cardiovascular aging. Strategies that limit fluid administration frequently are attended by increased use of vasopressors (phenylephrine, norepinephrine, and vasopressin). Since traditional anesthesia care reserves vasopressor use for when fluid administration has failed to maintain perfusion, vasopressor use can be considered an ominous clinical sign that is frequently cited as an adverse consequence of limited fluid administration. In many cases, volume administration is unavoidable and there are no extant guidelines for managing fluid, particularly in older adults. There is also an extensive literature about the proper content of fluids; particularly comparing crystalloid to colloid solutions. There are no clear-cut guidelines for use in older adults.

Anesthesia and the Pulmonary System

Aging and general anesthesia have important effects on the respiratory system (Table 36-3). Closing capacity increases with age because of diminished airway tethering by connective tissue. General anesthesia enhances expiratory muscle tone and diminishes inspiratory muscle tone, thereby decreasing functional residual capacity. Sighs are suppressed. Atelectasis develops more easily because of the increased closing capacity, decreased functional residual capacity, and the loss of sighs. Atelectasis occurs in the great majority of patients undergoing general anesthesia. Mechanical ventilation disturbs the normal pattern of diaphragmatic breathing in the supine patient, altering normal ventilation/perfusion matching. Volatile anesthetics reduce the effectiveness of hypoxic pulmonary vasoconstriction. Both volatile anesthetics and the placement of an endotracheal tube diminish ciliary action of the respiratory tract, an effect that extends into the postoperative period. Opioids diminish the carbon dioxide ventilatory drive. Volatile anesthetics suppress the hypoxic ventilatory drive, even at the low concentrations present postoperatively. Particularly very frail patients are at increased risk of ventilatory failure postoperatively. The increase in body oxygen consumption during recovery from surgery may exceed the ventilatory reserve that is compromised by increased chest wall stiffness and decreased skeletal muscle mass. Finally, silent regurgitation and aspiration is more common in older patients. At-risk patients will be challenged postoperatively by drugs that may further depress the airway protective reflexes and impair gastric emptying. Aspiration is much more likely to occur postoperatively than intraoperatively. It is important to understand, however, that patients who suffer aspiration to a degree that causes clinical signs or symptoms will present with some evidence of dysfunction (e.g., hypoxia) within a few hours of the aspiration. It is unlikely that a patient will aspirate, have no consequences for many hours, and then develop a clinical problem that could be attributed to much earlier aspiration.

Intraoperative respiratory management of older patients is straightforward. Bronchospasm is probably the most common intraoperative untoward event. Bronchospasm usually resolves promptly with easily administered treatments. Volatile anesthetic gases themselves produce bronchodilation, and inhaled bronchodilator medications can be administered through the endotracheal tube. Atelectasis prevention strategies during the intraoperative period have been variously successful. Supplemental oxygen is freely used in the operating room, and all patients are monitored by pulse oximetry and frequently with expired gas analysis. This has allowed the anesthesiologist to ensure that all patients are well oxygenated and has markedly decreased the incidence of clinically significant hypoxemia in all age groups.

Most pulmonary problems do not manifest until after surgery. Some degree of hypercarbia is common for at least the immediate postoperative period, especially after considerable opioid administration. This finding usually means nothing more than the presence of good analgesia, so long as supplemental oxygen is administered and the patient can be watched carefully. Supplemental oxygen will likely be necessary for a longer period of time after surgery in older patients. Patients with tenuous respiratory function may need ventilatory support until they are past the acute recovery period. Pneumonia remains an important complication and the ability to prevent its development is limited. Deep breathing and vigorous coughing by the patient are thought to help prevent pneumonia, but postoperative pain may impair these maneuvers. Improved pain control may be one mechanism by which epidural analgesia improves perioperative outcome (see section on “Postoperative Analgesia”). The role of silent aspiration in the development of postoperative pneumonia is unclear and warrants closer examination.

Regional Anesthesia

Regional anesthesia uses a family of pharmacologic agents with unique properties and a range of techniques designed to meet the needs of the surgical plan. This section will review agents first and then techniques. Local anesthetic agents reversibly inhibit the propagation of signals along nerves. When applied to a specific nerve or pathway in sufficient concentration, these agents produce anesthesia (loss of sensation) and paralysis (loss of muscle power). The currently used local anesthetics belong to one of two classes: aminoamide and aminoester local anesthetics. All of these medications are structurally related to cocaine but do not stimulate the sympathoadrenal system and, thus, do not produce hypertension or local vasoconstriction, nor do they have any abuse potential. Local anesthetics vary in their pharmacological properties. Field and nerve blocks are frequently performed using long-acting drugs such as bupivacaine. This often provides exceptional postoperative analgesia for up to 24 hours. The local anesthetic lidocaine is also used as a Class Ib antiarrhythmic drug. At higher doses, lidocaine and all local anesthetics are arrythmogenic. All agents have limited therapeutic windows, and overdosage can result in cardiac arrest and death. There are large variations in how aging alters the pharmacokinetics of local anesthetics, and aging per se accounts for less than 20% of variability. There is little suggestion that sensitivity to local anesthetics is altered by aging. Some forms of regional anesthesia can be combined with a light general anesthetic.

Regional anesthesia techniques include topical anesthesia (surface), infiltration, plexus block, epidural (extradural) block, and spinal anesthesia. As a rule, field and nerve blocks have minimal systemic effects unless there is an untoward event during placement, such as intravascular injection of the local anesthetic, or there is an allergic reaction to the drug. Many procedures, such as cataract surgery, can be accomplished with local anesthetic drops and no additional sedation. Blocks of specific nerves or nerve plexi are associated with anatomically specific side effects and complications. An interscalene block of the cervical plexus provides good anesthesia for operations on the shoulder. It may also result in anesthesia of the ipsilateral phrenic nerve, resulting in hemidiaphragmatic paralysis. An ipsilateral Horner syndrome may also be observed. Field blocks are often placed by the surgeon at the operative site prior to or at the conclusion of surgery.

Neuraxial blocks, including spinal and epidural anesthesia, are considered complete anesthetics; no other supplemental treatment is necessary, although sedation is often provided at the request of the patient or the surgeon. These blocks produce the greatest physiological effects and are presented here in more detail than other forms of regional anesthesia because their use in older adults is common. Spinal anesthesia differs from epidural anesthesia in the location of anesthetic administration, in the consequences of location of administration on type of nerve affected, and on the resulting clinical manifestations of treatment. With spinal anesthesia, the local anesthetic is injected into the cerebrospinal fluid, inside the dura and arachnoid, where it quickly diffuses into the spinal nerves. Inside the dura and arachnoid, spinal nerves have minimal connective tissue surrounding them so that the local anesthetic effectively blocks all nerve fibers, including motor, touch, and sympathetic nerves. The patient is, thus, provided not only anesthesia but also paralysis and sympathetic blockade. In contrast, an epidural anesthetic is typically administered through a catheter placed just outside the dura. In the epidural space, spinal nerves are covered with a thick connective tissue sheath, which slows the onset of anesthetic effect. By varying the concentration of the injected local anesthetic, nerve types can be differentially blocked. Differential blockade is important for postoperative analgesia, where partial blockade of pain fibers are desired, but touch and motor nerves are to be spared. Unfortunately, it is not possible to fully anesthetize pain fibers without also affecting the sympathetic nerves. When high concentrations of the local anesthetic will be used to achieve full anesthesia, all nerve types will be blocked.

The hemodynamic effects of neuraxial anesthesia stem from the blockade of sympathetic nerves. The hemodynamic response depends on many factors, including how many thoracic dermatomes of sympathetic fibers are affected and how much of a reflex response can be mounted via the vagus and any unblocked sympathetic fibers. In the case of spinal anesthesia, the sympathetic fibers may be at least partially blocked for many dermatomes above the level of sensory blockade, as defined by sharp–dull discrimination. It is not uncommon, therefore, to have a near-total sympathectomy with spinal anesthesia. The hemodynamic consequences can be dramatic. If one accepts the hypothesis that vascular resistance is increased with age because of increased sympathetic tone, then removal of that tone often results in a large decrease in vascular resistance (Figure 36-4). Pharmacologic sympathectomy in young adults results in a much smaller decrease in vascular resistance but a similar decrease in cardiac output. In Figure 36-4, the sympathectomy also decreased left ventricular filling as blood volume shifted to the legs and mesentery. However, the increase in ejection fraction ameliorated much of what the decrease in preload would have otherwise been expected to do to stroke volume and cardiac output.

Many patients request sedation during surgery performed under a regional anesthetic, usually because they are not interested in being aware of the activities in the operating room. Small quantities of midazolam and fentanyl often suffice, but if higher levels of sedation are desired, typically a propofol infusion will be used. Interestingly, older patients commonly fall asleep during spinal anesthesia, even in the absence of any sedative medications. The mechanism behind this phenomenon is unknown but may involve the loss of sensory input to the brainstem. If sedative and analgesic drugs are used, the risk of airway obstruction increases. The combination of sedation plus spinal anesthesia, for example, produces more postoperative hypoxia than sedation or spinal anesthesia alone. With careful monitoring and skill in airway management, the risk of airway obstruction should be minimal. In unskilled hands, however, airway obstruction can rapidly lead to hypoxia and patient injury. This risk to the patient forms the basis for the recent JCAHO requirement that trained personnel must be present at procedures performed anywhere in the hospital whenever the patient is likely to be sedated to the point of drowsiness that requires verbal stimuli or worse to arouse. The requirements include assigning a person to the sole task of monitoring the patient.

Regional anesthetic techniques can be combined with general anesthesia. Combined general/regional anesthesia involves the provision of primarily a high thoracic epidural anesthetic with subsequent induction of general anesthesia with endotracheal intubation. This approach has been most extensively studied in Denmark in association with a fast-track approach to colonic surgery, which also includes rapid ambulation and food consumption (day of surgery) and early discharge (postoperative day 2). The advantages of the technique combine the ability to use considerably less general anesthetic with resulting rapid emergence, with the complete absence of pain, because the epidural is still active and will be used to manage postoperative pain. The main limitation of the technique is the added difficulty in maintaining blood pressure in the operating room. In addition, it is common to place an epidural catheter exclusively for postoperative pain control, which will only be activated at the end of surgery, i.e., it is not used as part of the anesthetic technique.

Regional versus General Anesthesia

Many anesthesiologists prefer regional anesthesia over general anesthesia. There is something attractive to providing an anesthetic that just blocks nerves but spares direct effects to the brain and other vital organs. Markers of stress such as cortisol, catecholamines, and cytokines become elevated during and after surgery with a general anesthetic, while spinal and epidural anesthesia markedly attenuate these changes during surgery, and much of the attenuation will continue after surgery if epidural analgesia is continued postoperatively. The presumption has been that the reduction in stress would translate into a reduction in morbidity and mortality, but it has been surprisingly difficult to prove this hypothesis. Early studies centered around the benefits of postoperative analgesia via an epidural catheter typically infused with a dilute local anesthetic plus opioid. These studies employed random group assignment but enrolled small numbers of patients (40–60 each). Perhaps because a high-risk population was recruited, several of the studies demonstrated lower mortality and morbidity with respect to cardiac and pulmonary complications. Other studies have demonstrated additional benefits of regional anesthesia, including less blood loss during hip replacement surgery, a decreased incidence of deep vein thrombosis and pulmonary emboli, and a reduction in early graft thrombosis in peripheral vascular surgery. Not all studies demonstrate consistent benefits, however.

In order to circumvent the problem of small patient enrollment, a meta-analysis has been performed on 141 studies that had employed randomized assignment to general anesthesia versus spinal or epidural anesthesia. In approximately half of the studies, the regional anesthetic was accompanied by a general anesthetic. When an epidural catheter was employed (75% of the neuraxial blocks), postoperative epidural analgesia was frequently employed. Thus, group comparison was somewhat muddled. Nevertheless, the results were strongly in favor of the regional technique with deaths from all causes reduced by 30%. Decreases in rates of myocardial infarction, pneumonia, and pulmonary embolism (but not stroke) were also demonstrated.

These encouraging results must be tempered by two prospective studies, each involving roughly 1000 patients randomly assigned to general anesthesia alone with parenteral opioids for postoperative analgesia versus combined general plus epidural anesthesia with postoperative epidural analgesia. In one study, the general plus epidural group demonstrated a decreased incidence of respiratory failure. The other study only found a lower incidence of adverse events in the subset of patients within the epidural group who were undergoing repair of an abdominal aortic aneurysm. Therefore, neither study found overwhelming evidence of improved outcome from the addition of epidural anesthesia/analgesia.

It is apparent, then, that the current evidence of the benefit of postoperative epidural analgesia is weak, as is the evidence of the superiority of neuraxial anesthesia over general anesthesia as the primary anesthetic technique. Nevertheless, better pain relief is usually achieved by regional techniques, and it is still possible that some medical benefits exist. Therefore, these techniques continue to be used, at least in selected circumstances. Given the difficulty in proving a difference in outcome between regional and general anesthesia, it would be surprising if any study could demonstrate a difference in mortality or major morbidity when comparing, for example, different types of general anesthesia. In the absence of compelling outcome data, whether to perform a regional or a general anesthetic depends on many factors. Carotid endarterectomies, for example, may be performed under either a local block with an awake patient or a general anesthesia. Experienced teams of surgeons and anesthesiologists can perform either of these techniques with excellent and equivalent results. Usually, a center chooses one or the other technique. Both the surgeon and the anesthesiologist must be comfortable working with the chosen technique. Patient physical limitations, including uncomfortable patient positioning for the surgery, may prevent the use of a pure regional technique. For example, total hip replacement surgery requires the patient to lie in a fixed position on their side for the entire procedure. Even with careful positioning and padding, the position can be uncomfortable if only the lower half of the body is anesthetized. Sometimes sedation is adequate to make the procedure tolerable, whereas other patients require sedation bordering on general anesthesia that requires a secured airway. The presence of coagulopathies or the use of preoperative anticoagulation may preclude the use or regional techniques, as a result of the risk of bleeding and, in the case of spinal or epidural anesthesia, the risk of epidural hematoma (see the section below on “Postoperative Analgesia”). Also, the patient must be willing to undergo the regional technique, which will invariably involve placing a needle somewhere into the body. Some patients will not accept needles under any circumstance unless they are totally asleep, but, for safety reasons, most blocks will not be performed under such conditions. Other patients may refuse regional techniques owing to prior bad experiences. There is also a common fear of paralysis following spinal anesthesia, which is, for the most part unfounded, representing an extremely unusual complication with multiple potential causes. On the other hand, many patients welcome the opportunity to avoid a general anesthetic and the attendant postoperative sedation and risk of nausea. There is little evidence to suggest that the choice of anesthesia has a significant impact on either cognitive abilities or the incidence of postoperative delirium. All anesthetics should be preceded by a thorough discussion with the patient about anesthetic options, including their benefits, disadvantages, and associated risks. This discussion is best carried out by an anesthesiologist, who is best able to recognize the characteristics of the patient and procedure that will determine which anesthetic options are appropriate, and to explain to the patient the important risks and benefits of each option. In many cases, the choice of anesthetic technique can include patient preference; therefore, it is important to provide patients with comprehensive information and involve them in the decision-making process.

The options for postoperative analgesia have expanded considerably in the past 15 years. Highlighted here are patient-controlled analgesia and peripheral nerve blocks.

Patient-Controlled Analgesia

Patient-controlled analgesia devices are programmable infusion pumps that can provide either basal infusion or controlled access on request (usually pushing a button) to an opioid preparation, which is maintained in a locked compartment. The device offers the patient the opportunity to obtain pain medication without having to wait for a busy nurse. The device also provides more smooth analgesia compared to intermittent administration, which produces a cycle of heavy sedation with the initial peak effect, followed by a period of alertness and reasonable analgesia, and ending with increasing pain and anxiety until the next dose comes due. Patient-controlled analgesia allows the patient to finely titrate the analgesia against the opioid side effects, including nausea, sedation, dysphoria, and itching. In recent years, patient-controlled analgesia has become a standard part of postoperative pain management, in which orders follow specific templates. Since there is a tremendous variability in pain tolerance and response among individuals in general, and older adults in particular, effective postoperative pain management programs require a knowledgeable individual to observe the patient on a regular basis, in order to adjust the device properly. While elderly patients are sensitive to the respiratory depressant effects of narcotic agents, it is impossible to prescribe an age-specific approach to the use of postoperative opioids. Each patient is highly individual, and the therapeutic window for many analgesics is narrowed in the elderly.

Peripheral Nerve Blocks

Peripheral nerve blocks performed with long-acting local anesthetics provide up to 24 hours of postoperative analgesia. These blocks provide both significant advantages and risks. These are typically performed in a sedated, but awake and cooperative, patient prior to surgery. The most elaborate technique for postoperative analgesia is provided through indwelling catheters. Given the time involvement for catheter placement and management and the uncertainty of enhanced outcome, this technique is used selectively. Initially, epidural analgesia was nothing more than boluses of morphine through a lumbar catheter. Now the typical formula is low concentrations of both a local anesthetic and an opioid, which are infused continuously through a catheter placed at the dermatome most central to the incision, be it thoracic or lumbar in location. The quality of analgesia afforded by a well-managed epidural catheter is excellent and associated with high levels of satisfaction.

There are significant risks to epidural catheter use. The most common problem is failed placement. The next most common problem is inadvertent dural puncture. As epidural catheters are placed with large-bore needles, the risk of a spinal headache is relatively high. Fortunately, the risk of postdural puncture headache decreases with age. If a thoracic catheter is placed, dural puncture could also lead to direct spinal cord trauma. This complication is rare. The most feared complication is the development of an epidural hematoma. The epidural space has many veins that are occasionally punctured. Generally, any bleeding stops spontaneously and quickly, but if it does not, epidural hematoma may cause cord compression. Permanent paraplegia becomes likely, since the success of emergency laminectomy is poor once symptoms appear. Retrospective analysis of spinal and epidural anesthetics suggests that the intrinsic risk of paralysis is less than 1:150 000. Patients who have abnormal coagulation, however, are thought to be at greater risk. Most reports of epidural hematoma have been in patients who were anticoagulated when the block was performed or when heparin therapy was instituted less than an hour after the needle placement. The relative risk owing to anticoagulation is unknown, as all reports are anecdotal and do not include a denominator. In the case of epidural catheters, epidural hematomas develop almost as often on catheter removal as on insertion, so the catheter removal must be considered an at-risk event as well. There are case series reports on spinal or epidural block performed in anticoagulated patients, but the largest series involved 1000 patients and is too small to be meaningful despite a zero incidence of hematoma. Even subcutaneous heparin can provide enough anticoagulation to increase risk at its peak effect at around 2 hours after administration. Particularly distressing has been the high number of epidural hematomas reported in patients receiving low-molecular-weight heparin. Many of the hematomas developed after catheter removal. In response to these reports of epidural hematoma, it is recommended that any patient with an epidural catheter be given frequent neurologic checks if they receive anticoagulants (warfarin, heparin, and low-molecular-weight heparin). Guidelines for catheter placement and removal are given in Table 36-4.

The most common surgical indicator of complications is morbidity and mortality within a defined period following a procedure, frequently 30 days. Hamel reported on 26 648 patients aged ≥80 years (median age 82 years) and 568 263 patients <80 years (median age 62 years) from the Veterans Administrations National Surgical Quality Improvement Program (NSQIP) database. Mortality is low (<2%) for many common procedures (transurethral prostatectomy, hernia repair, knee replacement, carotid endarterectomy, vertebral disc surgery, laryngectomy, and radical prostatectomy). The incidence of complications increases and the impact of complications on mortality and functional recovery increases with age. The presence of a complication increased mortality from 4% to 26%. Respiratory and urinary tract complications were most common.

Anesthetic risk is difficult to quantify separately from the risk of surgery, because few people receive an anesthetic in the absence of a surgical procedure. For healthy patients, the risk of death purely caused by the anesthetic has been estimated to be as low as 1 in 250 000. Such a figure is trivial in comparison to the 1 in 500 risk of death associated with surgery and anesthesia overall. Mortality is dependent on patient age and strongly influenced by the type of surgery. Even in patients older than 90 years of age, minor surgeries carry a near-zero 30-day mortality, but major abdominal, thoracic, and vascular surgeries are associated with a 20% to 30% mortality. Clearly, comorbid disease, the effects of surgical trauma, and presumably the anesthetic must interact in some way to increase risk. The mechanisms underlying increased risk are poorly understood but are presumed to be influenced by the interaction of age and comorbid disease on reduced physiological reserve. This reduced reserve is presumed to make the body less able to withstand the various stresses associated with surgery such as pain, a hypermetabolic state, and altered neuroendocrine hormones.

Course of Recovery and Common Perioperative Complications in Older Patients

The influence of age on the course of recovery after surgery has not been a focus of research but a recent study by Lawrence et al. provides valuable insights. Among geriatric patients who had undergone abdominal surgery, recovery of independence in basic activities of daily living took almost 3 months on average. Some measures, such as hand grip strength, had, on average, not returned to baseline even at 6 months. Since recovery from surgery involves patient effort, it is important for the patients to understand their own role in the recovery process and to have a realistic idea of how long it will be before they will feel like themselves again.

An exhaustive discussion of complications that may occur in association with anesthesia and surgery is beyond the scope of this chapter. Nevertheless, it is worth commenting on a few complications. The most feared complication from anesthesia is the loss of the ability to ventilate a patient with subsequent hypoxic injury. This type of complication has become rare, and there is nothing to suggest that these events are more common in elderly patients.

Pulmonary complications are most common in elderly surgical patients, with postoperative pneumonia carrying a 20% mortality rate. Gag reflexes are decreased in the elderly and easily impaired by sedative and analgesic medications. Unless an older adult is specifically assessed as needing prolonged mechanical ventilation following surgery, there is nothing to suggest an advantage delaying extubation based on age. The current practice is to return to spontaneous ventilation and avoid prolonged intubation. Programs to prevent the development of pneumonia have had mixed results to date.

Cardiac events have received intensive focus in the perioperative arena. The nature of perioperative risk has changed during the years, with full Q-wave infarctions becoming less common. The risk of myocardial infarction increases with the prevalence of coronary disease in the patient population examined. In a relatively high-risk group such as patients undergoing peripheral vascular surgery, the incidence of myocardial infarction is approximately 4% and congestive heart failure around 9%. Among patients with a prior history of myocardial infarction, a perioperative myocardial infarction is still associated with approximately 30% mortality. The most common time when myocardial infarctions and congestive heart failure are detected is 2 days after surgery. The use of perioperative beta-blockers for patients at risk of myocardial infarction is rapidly becoming a standard of care, and there is no reason to avoid beta-blockers solely on the basis of age.

Stroke is a less frequent, but feared, complication of surgery. Perioperative stroke occurs much more frequently than would be expected for the same population in the absence of surgery. The most common cause of stroke in medical patients is thromboembolism, and there is no indication that it is any different perioperatively. In all likelihood, the same changes that increase the risk of myocardial infarction are responsible for the increased risk of stroke. While perhaps widely assumed to be a contributor to surgical stroke risk, hypotension is not a likely cause of stroke in the setting of surgery and anesthesia. There are two main sequences of evidence for this assertion. The first rests on the timing of onset of stroke in the perioperative period, and the second rests on the lack of relationship between duration of hypotension and stroke. The vast majority of strokes occur after the patient recovers from anesthesia and has demonstrated an immediate postoperative unchanged neurologic examination. Intraoperative hypotension is certainly not uncommon in older patients, although it rarely is allowed to persist for long periods of time. It is unlikely that brief periods of intraoperative hypotension are associated with stroke. If brief periods of hypotension caused strokes, then extended periods should, too. Yet no study involving deliberate hypotension for control of blood loss has reported an associated stroke. The best such study, involving 235 patients with an average age of 72 years who were randomly assigned to have mean arterial pressure maintained between 45 and 55 mm Hg or 55 and 70 mm Hg during surgery, demonstrated no strokes. Of much greater concern than hypotension as a mechanism for stroke is atrial fibrillation and cardiac emboli, perhaps accentuated by hypercoagulability. The degree to which atrial fibrillation is a risk factor is not well understood, but it raises the issue of how to manage patients who take warfarin chronically for atrial fibrillation or artificial valves. Based on the low risk of stroke in the immediate few days surrounding surgery in the absence of anticoagulation, it seems statistically safe to stop warfarin 4 days before surgery. A heparin infusion does not need to be used unless the patient has a history of arterial embolus or deep vein thrombosis within the previous 3 months. Warfarin therapy should be reinstituted postoperatively as soon as feasible.

There are two neurocognitive syndromes, delirium and postoperative cognitive dysfunction (POCD), that emerge in the postoperative period most commonly in older adults. Postoperative delirium is a change in mental status that consists of the inability to focus, sustain, and shift attention that is accompanied by other cognitive symptoms (e.g., disorientation and episodic memory dysfunction), and/or perceptual disturbances (misinterpretations, illusions, or hallucinations). Delirium as an entity is discussed elsewhere in this text. There are two forms of delirium after anesthesia. Emergence delirium occurs early, during emergence from general anesthesia. Postoperative delirium on the other hand develops later, usually 24 to 48 hours after an otherwise uneventful recovery from surgery. The risk of postoperative delirium is higher in older adults. For any major general surgery, the incidence is approximately 9% but may be as high as 60% in select surgical populations (major orthopedic and cardiac procedures). In contrast to medical patients, surgical patients rarely have delirium on hospital admission. Approximately a third to a half of the cases of delirium in medical patients are present at admission to the hospital, while only 7% of patients with delirium associated with hip fracture were delirious on admission. The etiology of delirium in general, and in postoperative delirium in particular, remains unknown. Drugs such as meperidine and anticholinergics are potential contributors and should be avoided in the elderly, but the risk from opioids is not clear, especially when failure to use such drugs could lead to increased pain and decreased sleep. One might expect general anesthesia to be a major risk factor for delirium, but multiple studies have all failed to demonstrate any differences in rates of delirium between general and regional anesthesia, even when careful monitoring of attention and cognitive function is performed. Delirium, if present at any time in the hospitalization, is associated with increased rates of complications, significantly prolonged hospital stay, and a decrease in function (e.g., ADLs) on hospital discharge. Aggressive management of geriatric patients can result in significant decreases in the incidence of postoperative delirium, as shown primarily by Edward Marcantonio in hip fracture patients (see section below on “Models of Collaborative Care for the Surgical Patient”). Both the regular confusion assessment method (CAM) and the CAM-ICU can be used in the perioperative period to detect delirium. As with medical patients, some patients are hyperactive and aggressive, but many delirium patients are hypoactive and quiet, so monitoring programs are required to reliably detect postoperative delirium. Management resembles delirium prevention and care in other settings (see Chapter 53). POCD is defined as a measurable decline in performance on a battery of neurocognitive tests. POCD appears to be distinct from delirium but may prove to be related. For years, there have been anecdotal reports of cognitive decline following surgery and anesthesia. In 1998, a prospective study tested surgical patients before and at 1 week and 3 months after surgery. A matched set of control subjects not having surgery was selected and tested as well. At 3 months, 10% of the surgical patients had demonstrable cognitive decline, in comparison to only 3% of the control subjects. Subsequent follow-up of this cohort suggested that approximately 1% had a persistent deficit after 2 years. As neurocognitive testing is not routinely done prior to surgery, POCD is primarily a research finding. It appears to be primarily a problem of the elderly, in that incidence is minimal below 60 years of age. POCD is uncommon following minor surgery. POCD following cardiac surgery occurs at a higher rate, approaching 50% at discharge from the hospital. POCD following cardiac surgery had traditionally been attributed to the influence of cardiopulmonary bypass; however, a recent study showed similar levels of cognitive decline among patients undergoing coronary artery bypass with and without cardiopulmonary bypass. Thus, the cause of POCD for both cardiac and noncardiac surgery remains unknown. There are neither strategies to prevent POCD nor treatments once cognitive decline occurs.

The most mature literature regarding the role of geriatricians in the care of surgical patients has been focused on patients with fractures of the upper femur. Heyburn and colleagues described four models that have been applied to hip fracture patients: the traditional model in which care is directed by the orthopedic surgeon and medical queries are directed to a consultant, the second is a variation in which multidisciplinary rounds with geriatricians and surgeons increase awareness of cross specialty issues, the third involves early postoperative transfer to a geriatric rehabilitation unit, and the fourth is combined orthogeriatric care in which the patient is admitted to a specialized ward where care is coordinated by geriatricians and orthopedic surgeons. There are clear advantages to patients from these models, the key factor being the implementation of the geriatric parts of the care plan. Marcantonio demonstrated the potential for a marked improvement in outcome in a randomized trial of proactive geriatric consultation based on a structured protocol for patients with hip fractures. The intervention reduced delirium by more than one-third. Some perioperative comprehensive geriatric assessment programs have cut down on length of stay. More complex programs such as the Hospital Elder Life Program are being evaluated for their impact on surgical patients.

Since there are more and more older adults, and since surgical procedures are increasingly feasible in complex older adults owing to technical and medical advances, the demand for anesthetic care of older patients is increasing. This care is intricate and carries real risks. Better preoperative preparation and postoperative care are likely to have a greater impact on outcomes than further advances in intraoperative anesthetic management, at least in the near future. The role of postoperative analgesia in improving outcomes and reducing the costs of medical care is still in its infancy. Optimal care of the older adult in the perioperative setting depends on the combined expertise of many specialties and an integrated, comprehensive approach.