Obstructive Sleep Apnea

Obstructive sleep apnea (OSA) occurs in the general population but is much more common in obesity. Nonetheless, most cases can be managed safely and effectively on an ambulatory basis. Perioperative problems such as difficult tracheal intubation and airway obstruction should be anticipated. However, patients undergoing more invasive surgery, especially involving the chest or airway, or those requiring large doses of perioperative opioids, may be less suitable. Frequently, OSA is suspected but has not yet been definitively diagnosed and treated. A simple questionnaire, supplemented with some basic measurements (e.g., STOP-Bang), can identify most patients at high risk for OSA, but there is insufficient evidence to recommend delaying surgery until the diagnosis is confirmed. In children, OSA is now one of the main indications for tonsillectomy and this has been seen as a relative contraindication to ambulatory surgery. However, one recent study has shown that same-day discharge is still safe in the absence of other comorbidities.

Age

Medical and social problems increase with age and should be evaluated and managed individually rather than applying an arbitrary upper age limit for ambulatory surgery. The risk for death or readmission within 7 days of surgery in patients over 65 years of age are 41/100,000 and 2.53%, respectively. Although somewhat more frequent than values from younger patients, the major risk factors appear to be very advanced age (older than 85 years of age), more invasive surgery, and recent inpatient hospital care. The frequency of perioperative adverse cardiovascular events also increases with age. Overall, older patients had a twofold increase in the risk for intraoperative adverse cardiovascular events; however, this was seen not as a contraindication to ambulatory surgery but rather as indicating the need for more careful intraoperative management. In contrast, the incidence of postoperative complications seems to be reduced in older patients. In particular, older patients appear to experience far less postoperative pain, dizziness, nausea, and vomiting than younger patients, and do not require a higher rate of unplanned admission or readmission. One study suggests a reduced incidence of postoperative cognitive dysfunction in older patients after ambulatory surgery compared to similar procedures performed on an inpatient basis, presumably because of the use of short-acting anesthetic techniques and reduced separation from their familiar home environment. Reducing length of hospital stay after hip and knee surgery appears to confer similar benefits.

At the other age extreme, the lower limit for ambulatory surgery may vary depending on the expertise and specialization of the individual facility. Premature babies have a higher risk for postoperative apnea, and hence ambulatory surgery may be inadvisable until they have reached an appropriate postconceptual age (PCA). Evaluation of several historic and retrospective studies suggests the risk for postoperative apnea is less than 5% once PCA exceeds 48 weeks, provided gestational age was at least 35 weeks, the infant was not anemic, and no apnea occurred in the recovery room. However, because considerable variability exists in the incidence of apnea and the relatively small sample size of these studies, the PCA at which risk becomes acceptably low is controversial, with a value of 60 weeks most commonly taken as the cutoff for ambulatory management. Caffeine administration appears to markedly reduce the incidence of postoperative apnea in babies who were premature, but this is not seen as a substitute for careful patient selection. Spinal anesthesia may offer advantages to infants who were premature undergoing abdominal surgery in the first week of life, but it is distressing to the infant and has a high (28%) failure rate. Compared with older studies involving more soluble anesthetics, sevoflurane and desflurane were associated with lower rates of postoperative apnea after hernia surgery in infants born at less than 37 weeks gestation and under 47 weeks PCA. Although no airway interventions were required to manage apnea, episodes still occurred throughout the 12-hour postoperative observation period and were equally common with either anesthetic.

Cardiovascular Disease

Hypertension is the most common cardiovascular disease and has been a frequent cause for delay and cancellation of ambulatory surgery. Although hypertension is an important risk factor for long-term health, a meta-analysis of nearly 13,000 patients showed that it increased the risk for perioperative complications by only 1.35 times, an amount that may be clinically insignificant. In ambulatory surgery patients, hypertension resulted in an approximate 2.5-fold increase in the risk for perioperative cardiovascular events, but these were relatively minor. Hypertension is not an independent risk factor for perioperative cardiovascular complications if the diastolic pressure is less than 110 mm Hg. Higher arterial blood pressures may predispose to perioperative ischemia, arrhythmias, and cardiovascular lability, but no clear evidence indicates that deferring surgery reduces perioperative risk. In the United Kingdom, patients with documented blood pressures below 160 mm Hg systolic and 100 mm Hg diastolic in the community can be accepted for elective surgery without further measurement. In practice, poorly controlled hypertension is usually identified at the preoperative evaluation and can be treated before surgery is scheduled. Delaying surgery until hypertension is controlled is unlikely to be beneficial.

Patients with known hypertension should continue to take their chronic medication, especially β-adrenergic blockers. β-Adrenergic blockers should not be abruptly stopped, and it may be more likely that they will be unintentionally omitted if patients are advised to take all of their cardiac medications up to and including the day of surgery. Controversy surrounds angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs). It has been considered “reasonable” to continue these, as any hypotension after induction of anesthesia is usually transient and responds to intravenously administered fluids and repeated vasopressor administration. However, evidence is accumulating of an increase in mortality, stroke, and myocardial injury if ACEIs and ARBs are continued, and current practice is to hold these on the day of surgery.

Patients are generally unsuitable for ambulatory surgery if they have severe unstable angina that causes marked limitation of activity or pain at rest. In the absence of complications, such as arrhythmias or ventricular dysfunction, cardiac risk after myocardial infarction or revascularization procedures returns to baseline after 3 months. Exercise tolerance is a major determinant of perioperative risk and the inability to climb a flight of stairs (∼4 metabolic equivalent tasks of activity) is highly predictive (89%) of a postoperative cardiopulmonary complication.

Patients taking anticoagulants and antiplatelet drugs require careful evaluation to balance the risk for perioperative bleeding against the risk in withholding therapy. For minimally invasive procedures, the international normalized ratio can be briefly decreased to the low or subtherapeutic range, with the usual dose of oral anticoagulation resumed immediately after the procedure. , If the risk for bleeding or thromboembolism is high, low-molecular-weight heparin can be used as bridging therapy. Patients are increasingly being managed with oral direct thrombin or factor Xa inhibitors (e.g., dabigatran, apixaban, rivaroxaban, and edoxaban). The effects of these agents are not reliably measured by commonly used coagulation tests, such as prothrombin time and activated partial thromboplastin time, but their shortened half-lives may make bridging therapy unnecessary. Antiplatelet therapy should be continued for the recommended intervals in patients with bare metal and drug-eluting coronary artery stents, because premature discontinuation is associated with a 25% to 30% risk for stent occlusion, which in turn has a more than 60% risk for myocardial infarction and 20% to 45% risk for death.

Management of Anxiety

The use of anxiolytic premedication is now uncommon in ambulatory anesthesia, probably because of concerns that these medications will delay recovery. In fact, one meta-analysis found no evidence that anxiolytic premedication delayed the discharge of ambulatory patients, although impaired performance in some tests of psychomotor function was found and the authors questioned the relevance of some of the older studies to modern ambulatory practice now that short-acting anesthetics are the norm.

Anxiety is nevertheless common in ambulatory surgery patients, with up to twothirds showing symptoms. A preoperative consultation more than 2 weeks before surgery can reduce patient anxiety and improve satisfaction, especially if the anesthesiologist was perceived as empathic. Patient satisfaction was further improved when their intraoperative care was provided by the same anesthesiologist who performed the preoperative consultation.

Anxiolytic Premedication

Given that anxiety was commonly reported, doubtless some patients will benefit from anxiolytic premedication, but what is the optimal regimen? Oral midazolam provided more anxiolysis than temazepam but also produced more sedation and amnesia, resulted in more oversedated patients, and delayed recovery. Oral alprazolam produced comparable anxiety reduction to midazolam without causing amnesia, but it also caused greater impairment of psychomotor function in the early postoperative period. Neither anxiolytic premedication delayed clinical recovery, but this may be a relatively crude assessment tool, because recovery primarily depends on other factors. Intravenous midazolam given shortly before induction of anesthesia reduced both anxiety and postoperative nausea. Such relatively late administration will not reduce anxiety while the patient waits preoperatively, but may be useful in a patient who needs to have an uncomfortable procedure before surgery, such as mammographic needle localization, in which satisfaction with both the procedure and subsequent breast biopsy was improved.

Premedication is more commonly used in children. In one investigation, premedication with oral midazolam 0.2 mg/kg reduced emergence agitation associated with sevoflurane anesthesia without significant delay in recovery, and doses up to 0.5 mg/kg did not delay recovery. However, other patients have experienced delayed recovery at this dose while not always having a reduction in anxiety. The less stringent recovery endpoints needed for children compared with the independent functioning needed for adults should be kept in mind. Premedication with oral midazolam may itself provoke anxiety in children, but giving the child a small toy to play with first has been shown to be effective in reducing this anxiety. Unfortunately, the effect of the toy as an alternative to midazolam was not evaluated. Play therapy and distraction can be more effective than midazolam in reducing anxiety in children, although this may require an elaborate approach. A preoperative educational program portrayed as a “Saturday morning club” also reduced anxiety, although the authors questioned whether the benefits achieved were worth the time and resources required. A simpler and highly effective approach is to allow children to watch age-appropriate video clips or movies during anesthetic induction with intravenous (South Korea) or inhaled (Canada) techniques. A preoperative educational DVD also encouraged more positive parental involvement during the recovery period and reduced children’s postoperative pain.

Because of variable efficacy and the potential for delayed recovery, alternatives to benzodiazepine premedication have been sought. In children, oral transmucosal fentanyl reduced preoperative anxiety and postoperative agitation, but a high incidence of predictable side effects, such as PONV and delayed discharge, limited its usefulness. Selective α ₂ -adrenergic agonists have potentially useful sedative and analgesic effects, yet despite years of study they still have no clearly defined role in ambulatory surgery in which their advantages outweigh the risks for adverse events. Clonidine is commonly used in pediatric practice, particularly to reduce agitation during emergence, but the promising results of clinical trials do not always translate well into routine clinical practice and its effects on induction anxiety are less persuasive.

Analgesic Premedication

Ambulatory patients are commonly premedicated with prophylactic oral analgesia in the expectation of achieving analgesic levels in the early postoperative period. Because the duration of action of acetaminophen (paracetamol) is relatively short (4-6 hours), it is unlikely to provide worthwhile postoperative analgesia, other than for an extremely short procedure. When patients received oral acetaminophen 1 g an hour before arthroscopic knee surgery, barely a third had plasma levels in the therapeutic analgesic range 30 minutes after arriving in recovery, whereas therapeutic levels were consistently achieved with intraoperative intravenous administration.

The nonsteroidal antiinflammatory drugs (NSAIDs) are consistently more effective as analgesic premedicants. Some evidence indicates that NSAIDs have a small preemptive analgesic effect (i.e., they achieve a superior effect if given before rather than after surgery), although one of the studies most strongly supporting this conclusion has been called into question. Preoperative parecoxib followed by postoperative valdecoxib significantly reduced the opioid analgesic requirements and the incidence of opioid-related adverse effects after laparoscopic cholecystectomy. As well as providing effective postoperative analgesia, etoricoxib premedication had an anesthetic-sparing effect during ambulatory ankle surgery. Using a common dental model of postoperative pain, preoperative ibuprofen, diclofenac, and acetaminophen with codeine effectively controlled early postoperative pain. Similarly, rofecoxib and ketorolac were equally effective in controlling postoperative pain after ambulatory surgery, as were ibuprofen and ketorolac. Sustained-release preparations of NSAIDs improve convenience by allowing earlier preoperative administration and a prolonged postoperative effect. Modified-release ibuprofen 1.6 g delayed the time until rescue analgesia was required after ambulatory third molar surgery in contrast to standard ibuprofen. Several of these NSAIDs (and most cyclooxygenase-2 [COX-2] inhibitors are unavailable in the United States.

Selective COX-2 inhibitors are no more effective and offer few advantages over traditional NSAIDs in ambulatory surgery. Despite any inhibition of platelet function, the amount of blood loss seen during high-risk surgery, such as tonsillectomy, was not reduced in comparison with nonselective NSAIDs. Nonetheless, COX-2 inhibitors are preferred by surgeons, so age-appropriate doses of oral celecoxib are used as premedication in the United States. Attempts to limit GI adverse effects, which are rare in acute use, has introduced other side effects and led to some drugs being withdrawn. Selective COX-2 inhibitors may be better tolerated in patients with aspirin-sensitive asthma.

Several other drugs have been evaluated for premedication. Controlled-release oxycodone did not improve pain scores or reduce opioid requirements within 24 hours of ambulatory gynecologic laparoscopic surgery. In combination with ibuprofen, pregabalin 150 mg reduced median pain scores at rest and in motion after gynecologic laparoscopic surgery, but did not reduce postoperative analgesic requirements. Perioperative administration of pregabalin 75 mg provided short-lived pain reduction after laparoscopic cholecystectomy, and a meta-analysis confirmed limited analgesic benefit, no reduction in opioid side effects, and increased sedation from its use. Even less effective was pretreatment with magnesium sulfate 4 g, which had no impact on postoperative pain or analgesic consumption in patients undergoing ambulatory ilioinguinal hernia repair or varicose vein surgery.

Prophylactic Antiemetics

Patients at moderate-to-frequent risk for PONV should receive prophylactic antiemetics. Preoperative steroids can provide both analgesic and antiemetic prophylaxis. Some investigators have given dexamethasone as premedication, but it is more usually given after induction of anesthesia to reduce administration side effects.

Antacid and Gastrokinetic Premedication

The incidence of aspiration of gastric contents is infrequent in fasted elective surgical patients. Despite improvements in several surrogate measures, insufficient evidence exists of clinical benefit (i.e., a reduction in morbidity or mortality from aspiration) to recommend the routine use of antacids, metoclopramide, H ₂ -receptor antagonists, or proton pump inhibitors before elective ambulatory surgery. Patients who are receiving these medications chronically should take them before surgery. Patients who regularly suffer from significant acid reflux in the fasted state will also benefit from the head-up tilt position during induction of anesthesia. This approach also may be useful in the super-morbidly obese patient undergoing bariatric surgery, in whom the use of prophylactic proton pump inhibitors and sodium citrate may be considered.

Propofol

Propofol has many of the properties of an ideal anesthetic induction drug and is popular because it rapidly and smoothly induces anesthesia without airway irritation and results in a rapid recovery with an infrequent incidence of early PONV and a clear head. Propofol has some disadvantages, however, including pain on injection, involuntary movements, transient apnea, and hypotension after induction of anesthesia. Numerous strategies have been proposed to reduce pain on injection, of which the use of a large antecubital vein or pretreatment with lidocaine in conjunction with venous occlusion are the most effective, and superior to altering the formulation of propofol.

Some of the adverse effects of propofol can be minimized by reducing the propofol dose by coinduction with adjuvant drugs, the most common being midazolam. Pretreatment with midazolam 0.1 mg/kg allowed anesthesia to be induced with a decreased propofol dose and also attenuated the resultant hemodynamic changes. A substantial decrease in the propofol dose can be achieved even when midazolam is administered up to 10 minutes before induction of anesthesia, a technique that also can make the induction experience for the patient more pleasant. However, this approach may delay recovery. Midazolam 0.03 mg/kg halved the required propofol dose but significantly impaired psychomotor recovery, even though awakening times were not delayed. Coinduction with short-acting opioids, such as alfentanil, can improve the quality of induction and ease of laryngeal mask airway (LMA) insertion, but at the expense of an increased incidence of hypotension and prolonged apnea. Similarly, fentanyl reduced propofol dose and improved conditions for LMA insertion, but also prolonged respiratory depression. Furthermore, the use of 1 μg/kg or fixed doses of 75 μg to 100 μg of perioperative fentanyl increase the incidence of PONV. Giving an initial bolus of 30 mg of propofol reduced the total propofol induction dose to a degree similar to that achieved by 2 mg of midazolam. This technique, which has been termed propofol autocoinduction, can reduce propofol dose and hypotension to a degree comparable to that with midazolam pretreatment but without delaying recovery. Following the concepts of opioid minimization and enhanced recovery after surgery, propofol coinduction can also be performed with intravenous lidocaine.

The pharmacokinetics of propofol allow it to be used as a variable rate infusion to maintain anesthesia, either in association with N ₂ O or combined with alfentanil or remifentanil as total intravenous anesthesia (TIVA). Recovery after anesthesia maintained with propofol is typically no faster compared to volatile anesthetics after an induction dose of propofol, and initial awakening may be delayed compared to that with the shorter-acting inhaled anethetics. These differences in emergence times are no more than 2 to 3 minutes and so not of clinical significance. Recovery to home readiness may occur up to 15 minutes sooner after propofol than isoflurane but is no faster than sevoflurane or desflurane. There are conflicting reports on whether propofol may be more or less likely than inhaled anesthetics to impair cognitive dysfunction after ambulatory anesthesia in older patients. One consistent feature of propofol anesthesia is a less frequent incidence of early PONV in contrast to the use of inhaled anesthetics. However, even this benefit has been described as of doubtful clinical relevance, except when the baseline incidence of PONV is very high. The reduction in PONV achieved by TIVA, specifically the use of propofol and the omission of N ₂ O, was similar to that produced by the use of single-drug antiemetic prophylaxis. Furthermore, any reduction in PONV appears to be confined to the early recovery phase as propofol use for ambulatory surgery does not alter the rate of unanticipated hospital admission and increases the incidence of nausea following discharge, probably because long-acting antiemetics (e.g., dexamethasone) had not been given.

Outside of the United States, delivery of propofol is increasingly performed by TCI, which makes the delivery of propofol easier, with fewer interventions required by the anesthesiologist, in contrast to manual infusions. However, it does not improve the quality of anesthesia, shorten recovery time, or reduce adverse events, although the quality of trials performed to date is poor. A further issue is that TCI systems calculate the predicted plasma propofol concentration and considerable variation exists between this and the actual measured value. Further sources of error are introduced when the effect site, rather than the plasma concentration, is targeted. Currently two different pharmacokinetic models are in common use, both derived from healthy subjects. The choice of model may make relatively little difference in the young, fit patient, but performance differs considerably in older patients and neither model is reliable in the morbidly obese. Even newer pharmacokinetic models developed specifically with data from obese patients have a tendency to underestimate plasma propofol concentrations.

Sevoflurane

The low solubility and minimal airway irritation of sevoflurane make it a readily controllable, short-acting anesthetic for ambulatory surgery. In contrast to isoflurane, sevoflurane resulted in significantly earlier awakening and orientation, caused less postoperative drowsiness, and facilitated earlier discharge by an average of 25 minutes. Compared to propofol, orientation occurred earlier but the time to home readiness was similar. The lack of airway irritation with sevoflurane means that rapid increases in inspired concentration are well tolerated, facilitating control of anesthetic depth. This is in contrast to the cough reflex and transient increases in heart rate and blood pressure, which sudden changes in isoflurane or desflurane concentration may provoke.

Lack of airway irritation makes sevoflurane almost ideal for inhaled induction, which may be especially desirable in children and needle-phobic adults. An 8% bolus induction of sevoflurane in adult patients can be faster than with propofol, with a similar rate of different side effects. Sevoflurane induction also caused significantly less depression of mean arterial pressure than propofol in older patients and was equally as well accepted as an induction choice. Using sevoflurane for volatile induction and maintenance of anesthesia (VIMA) has several benefits for ambulatory patients, but is associated with a greater incidence of PONV than when propofol is used for induction, maintenance, or both, of anesthesia. This appears to be in part due to coadministered opioids, which are rarely needed with VIMA, because omitting the opioids minimizes the problem. In children, the rapid emergence from sevoflurane (and desflurane) anesthesia is associated with a high incidence of emergence delirium, especially if inadequate measures are in place to control postoperative pain. Numerous strategies have been evaluated to reduce emergence delirium. Midazolam premedication is ineffective, but supplemental use of fentanyl 1 μg/kg, propofol, ketamine, and α ₂ -adrenergic agonists all reduce agitation to some degree. Although emergence delirium is undesirable, it is not associated with any long-term consequences and does not delay recovery room discharge. There is insufficient evidence of adequate quality to show whether propofol TIVA reduces the risk of behavioral disturbances (or PONV) compared with inhaled anesthesia.

Desflurane

With its low blood solubility, desflurane should be an ideal anesthetic for ambulatory anesthesia. However, metaanalysis of 25 randomized studies showed that patients receiving desflurane followed commands, had their trachea extubated, and were oriented only 1.0 to 1.2 minutes earlier than those receiving sevoflurane for up to 3.1 hours. No differences were found in recovery room stay or incidence of PONV. Concerns about the more frequent incidence of airway irritation with desflurane may limit the rapidity with which anesthetic depth can be altered, although relatively few problems were encountered in patients spontaneously breathing desflurane through an LMA when coadministered with fentanyl. Although desflurane has lower fat solubility than sevoflurane, the rate of diffusion into the fat compartment is extremely slow with either inhaled anesthetic; therefore, lower fat solubility does not make the case for use of desflurane as the ideal agent for morbidly obese patients unless procedure length is very long. In practice, some studies of ambulatory surgery in the morbidly obese have found more rapid recovery after desflurane than sevoflurane, whereas others found emergence and recovery to be similar with either drug.

Nitrous Oxide

Although N ₂ O is by far the oldest anesthetic drug still in use, its role is regularly questioned. N ₂ O is weakly emetogenic, yet its omission only significantly decreases vomiting when the baseline incidence is high and has no effect on nausea or the complete control of PONV. Others have found the effects of omitting N ₂ O to be “modest.” This may be because the alternatives, specifically a higher inhaled anesthetic concentration or the use of supplemental opioids, also induce PONV. The effect of N ₂ O on PONV appears to be time-dependent. There was no clinically significant effect from at least an hour’s exposure, allowing N ₂ O to retain a place in modern ambulatory anesthesia where its use also improves the quality, speed, and safety of induction of anesthesia, facilitates faster recovery, and reduces overall costs.

Opioid Analgesics

The indiscriminate use of opioid analgesics should be discouraged in ambulatory surgery to avoid PONV and unplanned admission. Although long-lasting drugs such as morphine are especially detrimental, even the ultra–short-acting opioid remifentanil resulted in a 35% incidence of PONV when used with desflurane and multimodal prophylaxis, compared with only 4% when all opioids were avoided. For minor to intermediate ambulatory surgery procedures, the routine administration of as little as 1 μg/kg of fentanyl serves only to increase the incidence of PONV, and its omission does not worsen postoperative pain, provided prophylactic analgesia is ensured with local anesthetic infiltration and preoperative NSAIDs. However, small doses of fentanyl may be useful to provide additional analgesia toward the end of more painful procedures.

Opioids are an essential component of TIVA, but the antiemetic effect of propofol minimizes the incidence of PONV in contrast to the use of the same opioid with inhaled anesthetics. Compared to alfentanil, remifentanil provides more effective suppression of intraoperative responses without prolonging awakening. However, there is increasing evidence emerging that remifentanil is associated with both acute opioid tolerance and opioid-induced hyperalgesia, which can increase postoperative analgesic requirements. This hyperalgesia is partially reduced by NSAIDs and by concurrent N ₂ O administration, although remifentanil is usually used as an N ₂ O replacement.

Nonopioid Analgesia

For longer or more invasive procedures, intravenous acetaminophen given toward the end of surgery also can have a beneficial effect, comparable to that of the opioid tramadol. For shorter procedures, acetaminophen can be given orally before the anesthetic, at much lower cost.

Cardiovascular Drugs

Although intraoperative hemodynamic disturbances are usually managed by increasing the primary anesthetic concentration, administering an opioid analgesic, or both, it may be more appropriate to treat these responses with cardiovascular drugs. Using an infusion of esmolol as an alternative to alfentanil to control heart rate reduced emergence times after arthroscopy, and substituting esmolol for remifentanil reduced nausea after gynecologic laparoscopy. Using a combination of esmolol and nicardipine to blunt increases in heart rate and arterial blood pressure, respectively, prevented the need for increases in inspired anesthetic concentration and consequently shortened emergence and recovery times. Ambulatory patients who received esmolol intraoperatively for hemodynamic control also had significantly smaller requirements for postoperative opioid analgesia. Administration of a continuous infusion of esmolol during laparoscopic gynecologic surgery reduced sevoflurane consumption by 18%, shortened recovery room stay, reduced postoperative pain scores, and reduced fentanyl requirements. In a similar study, an esmolol infusion decreased intraoperative remifentanil requirements and again resulted in lower postoperative pain scores and halved the dose of rescue fentanyl. Labetalol may be a cost-effective alternative to esmolol for longer cases, especially in older patients in whom it is less likely to cause inadvertent hypotension.

Neuromuscular Blocking Drugs

Neuromuscular blocking drugs (NMBDs) may be used in ambulatory anesthesia to facilitate endotracheal intubation or to provide profound surgical relaxation. Although several of the available compounds are suitable, muscle pain after administration of succinylcholine and concerns about residual neuromuscular block from intermediate-duration NMBDs after brief procedures have fueled the search for alternatives. The search for a nondepolarizing equivalent to succinylcholine has been unsuccessful to date. The most promising candidate, rapacuronium, was withdrawn from clinical use because of a frequent incidence of severe bronchospasm, although other factors such as inconvenient administration, declining use of endotracheal intubation, and high costs also contributed to its lack of commercial success. The search for a rapid-onset, short-acting nondepolarizing NMBD continues with investigation of several fumarate compounds, of which gantacurium was found to have rapid onset and brief duration, but was also associated with histamine release. Subsequent attention has moved toward designing a molecule of intermediate duration, but which can be rapidly reversed at any time by administering L-cysteine.

An alternative to short-acting NMBDs is terminating the neuromuscular blockade with sugammadex, which can rapidly and completely reverse rocuronium (or vecuronium) independent of the degree of residual neuromuscular block. Sugammadex is very expensive compared to neostigmine, although in the United States the cost of neostigmine has risen dramatically. Currently no reliable evaluations have been done on the cost-effectiveness of sugammadex in routine clinical practice. Although improved recovery times resulting from the elimination of residual neuromuscular block by sugammadex may potentially be cost-effective, this depends on the productivity of staff during the time saved and would probably require major changes to the workflow for actual benefits to be realized.

Endotracheal intubation also can be achieved without the use of NMBDs, thereby avoiding all of their adverse effects. This practice is most common in children, for procedures in which airway protection, but not prolonged muscle relaxation, is required. The optimum technique depends on personal experience and preference, but most commonly involves deep sevoflurane anesthesia or propofol supplemented by remifentanil or alfentanil. For adults, a dose of remifentanil 3 μg/kg with propofol 2 mg/kg is recommended, but satisfactory conditions can be obtained with remifentanil 2 μg/kg, thereby reducing the side effects of bradycardia and hypotension.

Spinal Anesthesia

The use of spinal anesthesia makes ambulatory surgery accessible to some patients for whom the risks of general anesthesia are excessive. It also increases patient choice, allows participation in intraoperative decision making (e.g., in sports injuries), and provides beneficial postoperative analgesia. Various ambulatory surgical procedures, such as transurethral prostatectomy, female incontinence surgery, and ankle and foot surgery, are well suited to spinal anesthesia.

The ready availability of fine-gauge, pencil-point spinal needles has reduced the incidence of significant postdural puncture headache to 0.5% to 1%. The main challenge now is to prevent prolonged motor block or impairment of joint position sense from delaying discharge. Although lidocaine has an appropriate duration for ambulatory surgery, its use in spinal anesthesia has virtually ceased as a result of a high incidence of transient neurologic syndrome (TNS). Until fairly recently, bupivacaine was the most obvious alternative, being devoid of TNS, but causing unacceptable delays in home discharge if used in standard doses.

For its acceptance in ambulatory surgery, bupivacaine spinal anesthesia needs to be “modified.” Reducing the dose of bupivacaine shortens recovery, whereas the use of patient positioning or the addition of adjuvants, such as fentanyl, ensures adequate analgesia for surgery. These techniques can be summarized as selective spinal anesthesia (SSA), defined as “the practice of employing minimal doses of intrathecal agents so that only the nerve roots supplying a specific area and only the modalities that require to be anaesthetized are affected.” SSA provides analgesia suitable for surgery, and light touch, temperature, proprioception, motor, and sympathetic function are all preserved. This results in remarkable cardiovascular stability but can make the block difficult to test, and patient cooperation is important to the success of this technique.

A variety of SSA regimens have been described, which typically permit discharge within a little over 3 hours after surgery. The addition of fentanyl marginally delays recovery and is associated with pruritus (although most cases do not require treatment), but also reduces postoperative pain and analgesic requirements. For unilateral knee arthroscopy, 4 to 5 mg of hyperbaric bupivacaine appears to be sufficient without adjuvants, provided the patient is kept in the lateral position. Clonidine has also been used to supplement low-dose spinal anesthesia, but it may prolong motor block, exacerbate hypotension, and delay voiding.

Spinal anesthesia may contribute to postoperative urinary retention. This is uncommon in low-risk patients, but may be more likely in older patients, with certain procedures and when more than 7 mg of bupivacaine is used. The risk for urinary retention is particularly high after inguinal hernia surgery, but simple infiltration anesthesia is often sufficient for this procedure and may be a more appropriate choice.

New Drugs for Ambulatory Spinal Anesthesia

With the increasing popularity of ambulatory spinal anesthesia, some older local anesthetics have been reevaluated and, in some countries, marketed for spinal use. Of these, prilocaine and 2-chloroprocaine have been evaluated the most. Hyperbaric prilocaine has a faster onset and shorter duration than the plain preparation, with patients receiving 40 mg of 2% hyperbaric prilocaine being home ready in 208 ± 68 minutes. This is comparable to discharge times reported with low-dose bupivacaine plus fentanyl, although 13% of the patients receiving hyperbaric prilocaine required supplementation for inadequate analgesia toward the end of surgery, which was scheduled to last for less than an hour. Doses of 40 to 60 mg achieve analgesia of about 90 minutes duration for lower abdominal or limb procedures, while 10 to 30 mg is adequate for perineal surgery of up to 40 minutes, with home discharge usually achievable within 4 hours. The duration of 2-chloroprocaine is even shorter, with recovery being significantly faster than that of low-dose bupivacaine or articaine and comparable to that of lidocaine. The addition of fentanyl appears to extend the block duration of both 2-chloroprocaine and prilocaine with little delay in recovery in either case. TNS is much less common than with lidocaine, but still occasionally reported after both prilocaine and 2-chloroprocaine.

Ambulatory spinal anesthesia that relies on short-acting local anesthetics should be compared with the cardiovascular stability of SSA. Prilocaine 20 mg combined with 20 μg fentanyl was associated with a lower incidence of clinically significant hypotension compared with bupivacaine 7.5 mg with fentanyl 20 μg. However, the doses of bupivacaine and fentanyl used in both groups were somewhat higher than usual.

Epidural Anesthesia

Epidural analgesia is seldom used in adult ambulatory anesthesia. Although it allows the block duration to be extended by using a catheter technique, this is offset by the time required in establishing the block and less certainty of success, as well as the risk for inadvertent intravascular injection or dural puncture. In knee arthroscopy, recovery after 15 to 20 mL of epidural 3% 2-chloroprocaine was more rapid than spinal procaine with 20 μg fentanyl and also had a decreased incidence of pruritus. Recovery after epidural 3% 2-chloroprocaine required fewer top-up injections and permitted discharge an hour sooner after ambulatory knee arthroscopy than with 1.5% epidural lidocaine. However, the use of epidural 2-chloroprocaine has a more frequent incidence of back pain.

In children, caudal epidural analgesia—for example, using levobupivacaine 0.5 to 1 mL/kg of 0.25%—is popular for postoperative pain relief. The technique is most useful for bilateral surgery or other situations in which the maximum safe dose would not permit an adequate volume for wound infiltration. After circumcision, caudal analgesia was no better than parenteral or systemic analgesia or dorsal nerve block in reducing analgesic requirements, nausea, or vomiting. However, motor block and leg weakness were significantly more common in boys receiving a caudal block. The addition of clonidine or dexmedetomidine augments caudal analgesia, but concerns remain about the frequency of sedative and hemodynamic side effects and the risk for neurotoxicity.

Intravenous Regional Anesthesia

Intravenous regional anesthesia (IVRA) (Bier block) is a simple, reliable method of analgesia most commonly used in the upper limb, but sometimes also used effectively in the lower extremity. In the United Kingdom and Europe, prilocaine is the preferred local anesthetic because of its high therapeutic index. Lidocaine has also been used for many years, and appears to be a safe alternative. Ropivacaine has also been extensively evaluated for IVRA. Compared to lidocaine, 0.2% to 0.375% ropivacaine resulted in prolonged and improved postoperative analgesia. However, compared to prilocaine, ropivacaine’s onset was slower, no useful prolongation of postoperative analgesia occurred, and ropivacaine plasma concentrations were more than double those of prilocaine, despite a 60% lower dose.

IVRA was a cost-effective alternative to general anesthesia for outpatient hand surgery, being equally rapid to administer and with faster recovery and fewer postoperative complications. However, analgesia was inadequate in approximately 11% of cases, which required supplemental local anesthesia, repeat of the block, and even conversion to general anesthesia. IVRA was also cost effective, and quicker to perform, compared to brachial plexus block, but with a 4.4% failure rate due to tourniquet pain.

A variety of adjuvants have been used in IVRA to decrease tourniquet pain, improve block quality, and prolong analgesia after cuff deflation. Opioids are relatively ineffective and cause nausea, vomiting, and dizziness after tourniquet deflation, but several NSAIDs have been shown to be beneficial. Lornoxicam decreased tourniquet pain and improved postoperative analgesia and tenoxicam improved postoperative pain, but the effects of ketorolac are unclear after most of the evidence showing its efficacy in IVRA was withdrawn. Dexamethasone improved both block quality and postoperative analgesia. Similar results have been observed with the α ₂ -adrenoceptor agonist dexmedetomidine, whereas clonidine delayed tourniquet pain without improving postoperative analgesia.

Other Local and Regional Anesthetic Techniques

A wide range of other regional anesthetic techniques can be used to facilitate ambulatory surgery or provide postoperative analgesia ( Table 72.3 ). The applicability of these techniques depends on the nature of the planned procedure, the patient, the preferences of the surgeon and anesthesiologist, as well as the skill and experience of the anesthesiologist performing the block. Benefits include excellent postoperative analgesia and reduced PONV, but this must be offset against pain and discomfort while performing the block, the difficulty of transitioning to satisfactory analgesia when the block wears off, and the risk for perioperative nerve injuries. Regional anesthesia may be contraindicated in patients with bleeding problems or those taking anticoagulants, as well as in patients with local infections. Regional anesthesia has a high failure rate in inexperienced hands, especially in the morbidly obese patient, although ultrasound guidance appears to improve the success rate of many blocks and can reduce pain during block placement. While there is no conclusive evidence that ultrasound guidance reduces the incidence of peripheral nerve injury, it does appear to reduce local anesthetic systemic toxicity and lowers the incidence of hemidiaphragmatic paresis and pneumothorax associated with certain blocks.

Simpler locoregional techniques may be more appropriate for some procedures. Intraarticular local anesthesia produces a moderate and relatively brief reduction in postoperative pain after arthroscopic knee surgery, but this is still considered to be of clinical significance in ambulatory surgery. In many cases, simple wound infiltration may be as effective as central and proximal peripheral blocks and allows patients to mobilize more quickly. Concerns about the possible risk for infection, or systemic toxicity resulting from large volumes of local anesthetic, appear unfounded in clinical practice.

Infiltration anesthesia offers advantages for many ambulatory procedures and is the method of choice for inguinal hernia repair. Several large series have confirmed excellent results with this highly cost-effective technique, with 79% of patients requiring oral analgesia for 7 days or less after surgery and 91% of patients returning to normal activities within 5 days. Repair under local anesthesia is not an independent factor for hernia recurrence, although this is influenced by the type of hernia and the level of surgical experience. Infiltration anesthesia resulted in fewer medical and urological complications after groin hernia repair than either general or regional anesthesia. If spinal anesthesia is chosen for groin hernia repair, the anesthesiologist should be aware of the increased risk for urinary retention and other medical complications, especially in older patients.

The concept of simple wound infiltration has been further developed into local infiltration analgesia (LIA), used for major orthopedic ambulatory surgery. This multimodal technique was developed for the control of pain after inpatient knee and hip surgery in Australia by Kerr and Kohan. Their technique used a mixture of 300 mg ropivacaine with 30 mg ketorolac and 1.5 mg epinephrine diluted to 150 to 200 mL with saline, infiltrated into all the tissue planes of the surgical field over approximately an hour as the operation progressed. A catheter was placed into the wound to permit later repeat dosing. In the original series of 325 patients, mainly undergoing elective hip resurfacing but including some primary hip and knee replacements, pain was well controlled (numerical pain score 0-3) without morphine in twothirds of patients, most of whom could walk with assistance 5 to 6 hours after surgery; 71% were discharged, walking independently after a single night stay ( Fig. 72.2 ). These promising results have been confirmed in a blinded, randomized trial in patients undergoing unicompartmental knee arthroplasty, where a similar LIA regimen significantly reduced pain and opioid consumption and decreased the median length of stay by 2 days, with 68% of patients discharged after a 1-night stay. A recent review suggests that LIA is beneficial after knee surgery, even when combined with multimodal systemic analgesia, but adds little additional analgesia after hip surgery when a multimodal regimen is also used.

Numerical pain scores at various time points after hip resurfacing arthroplasty (A; n = 185), total hip replacement (B; n = 54), and total knee replacement (C; n = 86) in patients receiving local infiltration anesthesia.

Data from Kerr DR, Kohan L. Local infiltration analgesia: a technique for the control of acute postoperative pain following knee and hip surgery: a case study of 325 patients. Acta Orthop. 2008;79[2]:174–183.

A limitation of infiltration anesthesia in general is the relatively short duration of pain relief it provides, even with long-acting drugs such as bupivacaine. The encapsulation of bupivacaine into a biodegradable carrier offers the prospect of extended-duration local anesthesia, which has demonstrated significant analgesic effects for at least 96 hours in pilot studies. Several potential problems need to be solved before widespread application, however, including ensuring that the large doses of local anesthetic contained within the carrier are not rapidly released, leading to toxicity, or that the encapsulating materials do not break down into harmful products. One extended-duration preparation of bupivacaine, Exparel, which uses an established liposomal-based drug delivery system (Lipo Foam, Contour ^MD ), is currently approved by the US Food and Drug Administration (FDA). Early trial results report improved pain scores and reduced opioid analgesic consumption compared to plain bupivacaine during the first 24 to 48 hours or more after bunion surgery, knee replacement, and breast augmentation. Myalgia was observed after administration of liposomal bupivacaine in six patients in one study, but a 2-year follow up from the same group reported no long-term complications. Several studies appear to support the efficacy and patient acceptability of liposomal bupivacaine after abdominal wall reconstruction, mastectomy, and mammoplasty, although the quality of comparative studies to date is poor. However, systematic reviews suggest no improvement over conventional analgesia for shoulder arthroplasty or bilateral knee surgery and only marginal benefits of questionable clinical significance after unilateral knee surgery. Catheter techniques also can extend the duration of effective local or regional analgesia, reducing pain scores, opioid-related side effects, and length of hospital stay, as well as improving patient satisfaction.

Several naturally occurring alkaloid toxins, which may have the potential to provide longer lasting and safer local anesthesia, have been considered. A preliminary study with neosaxitoxin, an extracellular sodium channel blocker, described superior analgesia compared to that with bupivacaine at 12 hours after laparoscopic cholecystectomy. Subcutaneous injection of neosaxitoxin in combination with bupivacaine and epinephrine produced a five-fold increase in analgesic duration compared with bupivacaine alone, without an apparent increase in toxicity.

Choice of Sedatives

The specific requirements for supplemental sedative and analgesic drugs depend on the procedure undertaken, the skill of the operator in using local anesthesia, and the experiences and expectations of the individual patient. The use of pharmacologic agents can be reduced by compassionate patient management at every stage, preoperative explanation, and the use of distractions, such as deep breathing, conversation, and listening to music. Adjuvant drugs should be chosen to deliver those specific aspects that are required, such as reduction of anxiety, sedation to relieve boredom or aid immobility, and analgesia for pain that is not amenable to further infiltration of local anesthetic.

Midazolam

Midazolam produces anxiolysis and dose-related hypnosis and is a popular component of sedation. It also produces marked antegrade amnesia. At times this may be beneficial, but it is not always welcomed by patients. Midazolam is preferred to other benzodiazepines because it is water-soluble, does not cause venoirritation or pain on injection, is rapid-acting, and has a relatively short-elimination half-life of about 2 to 4 hours. As sole anesthetic, a single dose of 0.05 to 0.1 mg/kg allows reasonably predictable recovery after brief procedures, but individual responses are highly variable. Recovery is also considerably slower if repeat or prolonged administration is required.

Propofol

Propofol is an excellent sedative-hypnotic because its pharmacokinetic properties result in a rapid recovery from the effects of a single bolus dose and a continuous infusion. Propofol is a relatively pure hypnotic, providing no analgesia and only moderate amnesia. Infusion rates of 25 to 75 μg/kg/min are typically used, but the short duration of action facilitates titration to effect.

Because of the excellent conditions that can be achieved and the superiority of recovery compared with that of midazolam, demand for propofol sedation is growing rapidly in areas such as endoscopy in which anesthesiologists have not traditionally been involved. Propofol can rapidly cause apnea, before loss of consciousness, even in sedative doses, and unintended transition to general anesthesia is a constant risk. The safety of nonanesthesiologist administration of propofol is therefore controversial. In the United Kingdom, the Royal College of Anaesthetists and the British Society of Gastroenterology jointly state that propofol sedation for complex upper GI endoscopic procedures should be administered only by an “appropriately trained anaesthetist.” In the United States, the FDA package insert states that propofol may be given “only by persons trained in the administration of general anesthesia,” which the ASA supports. However, the American Society for Gastrointestinal Endoscopy considers propofol use safe by anyone “proficient in the management of upper and lower airway complications, including manual techniques for reestablishing airway patency” and holding at least basic life support certification.

Advocates of endoscopist-administered propofol claim a very low rate of airway complications (0.1%) and a mortality (4 in 646,080) less than that associated with endoscopist-administered midazolam-opioid sedation. Their data also reveal a wide range for the frequency with which rescue airway support was required, suggesting considerable variation, either in technique or willingness to intervene to protect the airway. When anesthesiologists administered propofol for patients undergoing endoscopic retrograde cholangiopancreatography (ERCP), deeper levels of sedation or general anesthesia were often intended and sedation-related adverse events were far more common (21%), especially in patients of higher ASA physical status classification, but were managed without adverse consequence or delay, and high levels of patient and operator satisfaction were achieved. A more recent study of anesthesiologist-administered propofol sedation has revealed a very similar (23%) incidence of significant unplanned events.

Potential New Sedatives

Fospropofol was a water-soluble phosphate ester prodrug of propofol, with a slower onset of action and longer duration. Although approved in the United States for sedation in adults in 2009, restriction of its use to anesthesiologists and a lack of any clear advantage over propofol prevented its commercial success.

Remimazolam is a new ester-based benzodiazepine, designed to be rapidly broken down by ubiquitous tissue esterases into an inactive metabolite. Preliminary findings suggest a faster onset and shorter duration than midazolam after a single dose, although not all procedures could be completed without “rescue” sedation. Remimazolam may have a role as a sedative for brief (<10 minute) procedures, but its behavior after repeated dosing and its consequent role as a sedative have yet to be evaluated.

Analgesic Adjuvants

Opioids are useful for procedures in which pain cannot reliably be blocked with local anesthesia alone. Compared to alfentanil in women undergoing breast biopsy, remifentanil resulted in lower pain scores during deep dissection and reduced the need for supplemental local anesthesia. Remifentanil was a useful adjuvant to propofol sedation during transvaginal tape insertion and in achieving optimal conditions for awake craniotomy, surgery that is increasingly being performed as a day and short-stay procedure. Remifentanil produced lower pain scores than fentanyl in patients receiving propofol sedation during hysteroscopy, but it did not improve recovery or patient satisfaction. For hysteroscopic surgery, paracervical block supplemented by a remifentanil infusion permitted earlier mobilization and discharge compared to TIVA, and the sedation technique was also preferred by a higher proportion of patients. Remifentanil is usually delivered by infusion, but intermittent bolus doses may be more effective for some procedures. Hyperalgesia after remifentanil administration remains a concern.

Both clonidine and dexmedetomidine have potentially useful analgesic, anxiolytic, and sedative properties. The slow onset and offset of clonidine and frequent reports of cardiovascular instability mean that neither drug has established a routine place in sedation, although dexmedetomidine may be a useful analgesic adjuvant in awake craniotomy. Ketamine improved analgesia when used to supplement propofol sedation, but an increase in PONV, psychomimetic side effects, and delayed discharge were observed at higher doses.

Delivery of Sedation

Because of the varying needs to provide hypnosis, anxiolysis, analgesia, and amnesia, use of drug combinations is common during anesthesiologist-administered sedation. Preceding a propofol infusion with a 2 mg dose of midazolam improved anxiety, sedation, and amnesia for early intraoperative events with no detrimental effect on postoperative sedation, amnesia, or recovery times. However, using drug combinations increases the risk for interactions, leading to adverse effects. At sedative doses, propofol and remifentanil each have only a modest effect on heart rate and arterial blood pressure, but their effect on respiration is strikingly synergistic, with the potential for severe respiratory depression. Respiratory depression is also enhanced by the interaction of midazolam and remifentanil. Antagonists of benzodiazepines and opioids may be useful in cases of unintended overdose, but should not be routinely relied on to reverse deep sedation because of their short duration of action, which may permit resedation before or after discharge. It is also possible to achieve high rates of patient satisfaction without procedural recall when using intermittent bolus doses of propofol alone for endoscopic procedures.

As with TIVA, the stability and control of sedation regimens can be improved by using TCI systems. Target concentrations of 0.5 to 2 μg/mL for propofol and 0.5 to 1 ng/mL for remifentanil are typical, but should be titrated to individual effect. Effective sedation is usually judged by clinical endpoints or scores. Bispectral index (BIS) changes with increasing sedation, but is too variable to be useful routinely and was not effective as an endpoint for midazolam sedation. BIS monitoring did not improve the quality of sedation or reduce the propofol dose or rates of hypoxemic, bradycardic, or hypotensive complications during ERCP. However, BIS-guided TCI propofol may be useful in the management of patients with intellectual disability, in whom clinical signs may be less reliable. Allowing patients to adjust their own sedation level by patient-controlled sedation resulted in less propofol being used compared with a continuous infusion and was significantly more popular with patients. However, this approach necessarily requires comparatively light levels of sedation and appears to be an undesirable option for a significant proportion of patients.

Inhaled sedation by a low dose of sevoflurane is an alternative technique that provides good control of sedation and rapid recovery. However, it can be complicated by a frequent incidence of perioperative excitement and an increased risk for transition to general anesthesia.

Managing Opioid Complications

Although multimodal techniques aim to minimize opioid use, strong opioids can be necessary as a rescue option after more invasive surgery. Analgesics such as morphine and oxycodone provide more intense and prolonged effects, but are associated with more intense and prolonged typical opioid side effects. Novel delivery techniques, such as iontophoretic or nasal fentanyl, or sublingual sufentanil, may improve patient convenience but do nothing to reduce adverse effects. In addition to multimodal analgesia, attempts to limit these adverse effects have included the development of drugs such as tramadol, which combine opioid and nonopioid mechanisms of action. Tramadol is effective after ambulatory surgery, but it is still associated with a frequent incidence of unwanted effects, including sedation, dizziness, and particularly PONV. Tapentadol is licensed in both the United States and the United Kingdom. It has a dual action similar to that of tramadol, with comparable efficacy to oxycodone but fewer GI adverse effects, such as nausea, vomiting, and constipation. Unlike tramadol, it does not require metabolic activation or suffer from isomer-dependent pharmacodynamics.

Combining oral oxycodone with naloxone (Targinact) antagonizes the GI effects, in particular preventing opioid-induced constipation, but has little effect on analgesia because first-pass metabolism prevents much naloxone from reaching the central nervous system. Alvimopan is another peripherally acting μ-opioid (MOP) receptor antagonist intended to reduce opioid-induced constipation. However, peripheral MOP receptors also may partially mediate opioid-associated PONV, delayed gastric emptying, and urinary retention, offering the hope that opioid analgesia without serious adverse effects may yet be a possibility.

Local Anesthetic Administration at Home

Sending patients home with perineural, incisional, and intraarticular catheters is a new and evolving area of postoperative pain management. Patient-controlled regional anesthesia (PCRA) with an elastomeric balloon pump has provided effective analgesia and reduced pain intensity during mobilization when ropivacaine is delivered into a perioperatively placed wound catheter after subacromial decompression. Although patients were kept in hospital for study observations and assessments, based on pain scores, the authors concluded that all patients in the active group could have been discharged within 2 hours of surgery. Another small pilot study showed excellent analgesia with continuous interscalene block, continued at home, allowing same-day discharge after a variety of shoulder operations, including open rotator cuff repair, subacromial decompression, and joint replacement.

Recently, several institutions have described cases of chondrolysis that appear to be related to the use of postoperative intraarticular local anesthetic pain pumps. In a series of 375 patients from one individual orthopedic surgeon, chondrolysis was observed only in cases in which either bupivacaine or lidocaine had been infused into the joint during the postoperative period. Most local anesthetics, including bupivacaine, lidocaine, and ropivacaine, are chondrotoxic to human articular cartilage in vitro, and the risk may be increased with longer exposure to higher concentrations of local anesthetic, such as with a pain pump, compared to that associated with a single injection. Consequently, the use of local anesthetic infusions after shoulder surgery has declined or stopped in many institutions.

A variety of infusion pump designs are available for local anesthesia infusions or PCRA after discharge. Technical problems have been found with electronic pumps, which seem to be unreliable for home use. Patient satisfaction has been reported with disposable elastomeric pumps, which are more reliable, although not all types function equally well and in vivo performance does not necessarily match results obtained in vitro.

First-Generation Drugs

Metoclopramide is a dopaminergic (D ₂ ) and serotonergic (5-HT ₃ , peripheral 5-HT ₄ at higher doses) antagonist with gastric prokinetic properties that was first described in 1964. A meta-analysis of a standard clinical dose of 10 mg showed little evidence of pharmacologic benefit for PONV, but after a number of studies by Fujii were withdrawn, reanalysis suggests that there may be some benefit. However, current guidelines still do not list metoclopramide for first-line treatment. Higher doses, such as 20 to 25 mg, are more effective but are associated with more akathisia.

Droperidol is a butyrophenone with an antiemetic effect resulting from dopaminergic receptor (D ₂ ) antagonism that received a black box warning for the FDA-approved dose of 2.5 mg or greater because of potential QT interval prolongation. This and other side effects of droperidol, such as sedation and akathisia, were significant but at higher doses than are commonly used for PONV prophylaxis, which are typically 1.25 mg or less. A recent meta-analysis has confirmed the antiemetic action of low-dose droperidol. Droperidol use is less in the United States since the black box warning was added but primarily for medicolegal reasons rather than because of concerns with efficacy or side effects. It is also little used in the ambulatory environment in the United Kingdom due to perceptions of adverse extrapyramidal effects, particularly akathisia, even at 0.5-mg doses. An Australian study evaluating the incidence of akathisia in 228 women undergoing ambulatory gynecologic laparoscopy reported a rate of 29% with administration of 10 μg/kg.

Antagonists of the histamine H ₁ receptor have a particular efficacy in the management of nausea and vomiting precipitated by vestibular pathways, with a documented benefit in the management of travel sickness and surgical procedures for strabismus or involving the middle ear. Dimenhydrinate (a combination of diphenhydramine and 8-chlorotheophylline that was added to reduce drowsiness) has an antiemetic effectiveness deemed similar to that of both droperidol and 5-HT ₃ antagonists, but potential adverse reactions from these antihistamines include marked sedation, dry mouth, urinary retention, and blurred vision as a result of associated muscarinic receptor antagonism. Meclizine, an H ₁ -receptor antagonist and anti–motion sickness medication, is minimally sedating, long-acting, and effective for treatment of PONV and for prevention of PDNV. In the United States, it is inexpensive and available without prescription, making it an attractive postdischarge choice.

A transdermal scopolamine delivery system is also available. This patch was designed to deliver a total dose of 1 mg of scopolamine at a sustained, constant rate over a period of 3 days. Numerous studies have shown transdermal scopolamine to be effective in reducing the incidence and severity of PONV and PDNV, with efficacy comparable to that of ondansetron or droperidol. It has a prolonged duration but the onset is delayed, becoming effective within 2 to 4 hours of applying the patch. The slow onset can be overcome by applying the patch the night before surgery. Used in this way, the incidence of nausea and vomiting was reduced from 62.5% and 37.5% to 20.8% and 8.3%, respectively, compared to placebo after gynecologic laparoscopy. In the United States, clinicians apply the patch preoperatively, knowing that onset of PDNV efficacy will be in the early recovery period for shorter surgery. Adverse effects, particularly dry mouth but also somnolence, dizziness, and blurred vision, are relatively common but generally described as mild.

Serotonin 5-HT3 Receptor Antagonists

The 5-HT ₃ antagonists have played a substantial role in the management of PONV since their introduction in the 1980s, because of their relatively benign adverse reaction profile compared with those of the then available drugs. Their prophylactic use is more effective when given just before the end of surgery. They are effective drugs for emetic rescue therapy, with ondansetron seeming to have a more preferential effect on vomiting (number needed to treat [NNT] = 4) rather than nausea (NNT = 7). Although relatively well tolerated, their side-effect profile includes an increased risk for headache (number needed to harm [NNH] = 36) and elevated liver enzymes (NNH = 31). All serotonin subtype 3 antagonists have also been associated with QT interval prolongation. More recently released serotonin 5-HT ₃ receptor antagonists, such as dolasetron, granisetron, and palonosetron, seem to exhibit similar characteristics with equal reduction of PONV rates in high-risk patients when used for prophylaxis, though the longer half-life of these drugs (8, 10, and 40 hours, respectively) may afford better protection against symptoms after discharge. In particular, palonosetron, by nature of its unique binding properties leading to internalization of the 5-HT ₃ receptor and a long half-life, may have a role in the management of postdischarge symptoms. A large cost remains a significant barrier to use of the newer, nongeneric serotonin antagonists.

Steroids

Dexamethasone is an effective antiemetic at intravenous doses of 4 to 5 mg (depending on local formulation). This steroid may act centrally to either modulate the release of endorphins or inhibit prostaglandin synthesis. Given its delayed onset of action, dexamethasone should be administered as early as possible after induction of anesthesia. Prophylactic dexamethasone is also effective in reducing postoperative pain and improving the quality of recovery, although at somewhat higher doses (typically 8 mg) than those required for an antiemetic effect. The long-term side effect profile of this dose of dexamethasone has not yet been evaluated.

Neurokinin-1 Antagonists

The role of tachykinins in the emetic pathway was first elucidated by immunohistologic studies identifying substance P in the dorsal vagal complex of the ferret, an area of the brain deemed essential in the vomiting reflex. Subsequent research identified the potential value of specific antagonists to the neurokinin-1 receptor, at which substance P and neurokinins A and B interact both centrally and peripherally in the gut to suppress vomiting. Aprepitant was the first commercially released drug of this class. A preoperative oral dose of 40 mg has efficacy similar to that of ondansetron for the reduction of nausea, with a potentially superior effect in the suppression of vomiting for 48 hours after administration. An intravenous prodrug formulation, fosaprepitant, was more effective than ondansetron in preventing vomiting, but not nausea, and with no effect on complete response rates for PONV in small numbers of patients undergoing gynecological or lower limb surgery. A multicenter study evaluating the use of oral rolapitant, a competitive NK-1 antagonist with an exceptionally long half-life of 180 hours, suggested similar effectiveness in comparison with placebo and ondansetron (administered at induction) with early control of symptoms, but use of rolapitant similarly seemed to confer prolonged protection from postoperative emetic symptoms. Rolapitant does not seem to have been further evaluated in PONV and cost remains a barrier for this class of drugs as well.